[This Transcript is Unedited]
DEPARTMENT OF HEALTH AND HUMAN SERVICES
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
Meeting of Subcommittee on Standards and Security
August 17, 2004
Hubert H. Humphrey Building
200 Independence Avenue, S.W.
Washington, D.C. 20201
CASET Associates, Ltd.
10201 Lee Highway, Suite 160
Fairfax, Virginia 22030
Simon Cohn, M.D., Chair
TABLE OF CONTENTS
- Call to Order, Welcome and Introductions – Simon Cohn, M.D.
- Messaging Standards
- Scott Robertson, HL7
- Lisa Miller, X-12
- Lynne Gilbertson, NCPDP
- Richard Peters, M.D., ASTM
- HL7/NCPDP Harmonization
- Karen Eckert
- Ross Martin, M.D.
- Terminology Standards: Readiness for Adoption and Demo; Mapping to Drug Knowledge Bases
- Randy Levin, M.D., FDA
- Steve Brown, M.D., VA
- Stuart Nelson, M.D., NLM
- Identifier Standards
- Karen Trudel, CMS
- Lee Ann Stember (NCPDP-HCIdea)
- Standards for Formularies: Messages, Structure and Prior Authorization
- Teri Byrne, RxHub
- Brian Bamberger, MediMedia Information Technologies
- Open Microphone
P R O C E E D I N G S [9:11 a.m.]
DR. COHN: Well, good morning. I want to call this meeting to
order. This is the first day of three days of hearings and sessions of
the Subcommittee on Standards and Security of the National Committee on Vital
and Health Statistics. The committee is the main public advisory
committee to the U.S. Department of Health and Human Services on national
health information policy.
I am Simon Cohn, a physician and chairman of the Subcommittee. I am
the national director for health information policy for Kaiser Permanente.
I want to welcome fellow subcommittee members, our staff, who has been so
great in this process so far and others here in person. I also want to
welcome those listening in on the Internet and I just want to remind everyone,
both testifiers, as well as subcommittee members and others to speak clearly
and into the microphone just because these microphones don’t catch you if you
get too far away from the mouthpiece. So, just a warning and wisdom
Lots of work over the next three days. Now, this morning and really
today we continue our hearings regarding e-prescribing standards.
As all of you know, the Medicare Modernization Act of 2003 calls on the
Secretary to adopt standards for e-prescribing and the NCVHS has been directed
to develop such standards and recommendations. Today, we hear from
e-prescribing standards development organizations, code set developers and
others, how they plan to address the gaps and issues that have been identified
by users and other parties during our previous testimony and hearings.
Additionally, testimony will be given on the plans to harmonize prescribing
standards used in the ambulatory environment, in other words, the NCPDP script
and the acute care environment, HL7, just to update us about what is going on
there, as well as the issues involved with that. For each of the
topic areas, we will have a panel followed by questions and discussion.
I obviously want to thank Jeff Blair, our vice chair, who has been great
support and help to the subcommittee and bull committee as we move through this
relatively accelerated process. I also want to thank Maria Friedman, our
lead staff, for once again, I think, phenomenal effort and support in all of
We also want to thank Margaret Amatayakul for her help and support as a
consultant on this project.
I want to emphasize that this is an open session. Those in attendance
are welcome to make brief comments, as time permits, if you have information
pertinent to the subject being discussed. We will also have time during
the day for open sessions, for brief comments by those in attendance.
Finally, for those on the Internet, we welcome e-mail and letter comments as
issues come before the — on the issues coming before the subcommittee.
Tomorrow and Thursday, the committee will deliberate in open session about
observations and draft recommendations we may wish to communicate to the full
committee for the September full committee meeting. This, obviously, is
on e-prescribing. Our intent is to come forward with a first set of
recommendations that is commented this is just a first set of recommendations.
There will be additional hearings and testimony taken as you move into the
fall and early winter in preparation also for a second set of recommendations
that we expect to have the full committee in March of 2005.
With that, let’s have introductions around the table and then around the
room. For those on the committee, I would ask as always if you have any
conflicts of interest related to any issues coming forth today, would you
please so publicly indicate during your introduction.
MR. BLAIR: Thank you, Simon. I am Jeff Blair, vice president of
Medical Records Institute, vice chair of the Subcommittee on Standards and
Security. I am a member of ASTM, IMIA(?), HL7, HIMS(?) and there might be
an apparent conflict of interest with anything related to ASTM because my
employer is chair of ASTM. So, I would recuse myself from any votes or
discussions on that.
MR. STEINDEL: Steve Steindel, Centers for Disease Control and
Prevention, staff to the subcommittee and liaison to the full committee.
MR. HUFF: Stan Huff with Intermountain Health care and the University
of Utah in Salt Lake City. I have a potential conflict of interest with
HL7. I am a former chair of HL7 and vocabulary co-chair within HL7 at the
current time. So, I would recuse myself from those votes related to HL7.
DR. FITZMAURICE: Michael Fitzmaurice, Agency for HealthCare Research
and Quality, liaison to the National Committee and staff to the subcommittee.
MR. HETHITER: Hart Hethiter(?), VA, staff to the subcommittee.
MS. AMATAYAKUL: Margaret Amatayakul, Margaret A. Independent
consultant advising the committee, helping the committee.
MS. BEBEE: Suzie Bebee, ASPI, staff to the subcommittee.
MR. ROBERTSON: Scott Robertson with Kaiser Permanente, representing
MS. MILLER: Lisa Miller with Washington Publishing Company,
MS. GILBERTSON: Lynne Gilbertson, National Council for Prescription
MR. LEVIN: Randy Levin, Food and Drug Administration, staff to the
MS. AULD: Vivian Auld, National Library of Medicine, staff to the
MS. TRUDEL: Karen Trudel, Centers for Medicare and Medicaid Services,
staff to the subcommittee.
MR. REYNOLDS: Harry Reynolds, Blue Cross and Blue Shield, North
Carolina, member of the committee.
MS. WARREN: Judith Warren, University of Kansas, School of Nursing,
member of the subcommittee. I may also have a potential for a conflict
with HL7 in that I am a co-chair of the HL7 Patient Care Technical Committee.
MS. FRIEDMAN: Maria Friedman, CMS, lead staff of the subcommittee.
MS. JACKSON: Debbie Jackson, National Center for Health Statistics,
staff to the committee.
MR. ROBINICK: Bill Robinick(?) ExpressScripts.
MR. DE CARLO: Michael DeCarlo, Blue Cross/Blue Shield Association.
MR. SINCO: Mike Sinco(?), Walgreens.
MR. TAVE: Agee(?) Tave(?), SureScripts.
MR. WHITTEMORE: Ken Whittemore, SureScripts.
MR. SCOTT: Philip Scott, NCPDP.
MS. STEMBER: Lee Ann Stember, NCPDP.
MS. RIPPON: Helga Rippon, assistant secretary for planning and
MS. WILLIAMSON: Michelle Williamson, National Center for Health
MR. SILK: David Silk(?), health strategies consultancy.
MR. KEROWSKI: Lee Kerowski(?), FDC Reports.
MR. KASPER: Andrew Kasper(?), FDC Reports.
MS. HIRSCHORN: Rebecca Hirschorn(?), Centers for Medicare and Medicaid
MS. ECKERT: Karen Eckert with Medi-Span, part of WoltersKluwer Health.
MS. BAVARA: Patrice Bavara(?), Metger(?) Health Solutions.
MS. WOOLEY: Shelly Wooley, RXM.
MS. BYRNE: Teri Byrne, RXM.
MR. BROWN: Jeff Brown.
MS. BLOM: Lola(?) Blom(?), the joint commission.
MR. SHEATH: Tony Sheath, Point of Care Partners.
MS. HELM: Jo Helm, All SCRIPTs.
DR. NELSON: Stuart Nelson, National Library of Medicine.
MR. ROBINSON: George Robinson, First Data Bank.
MS. HOWARD: Cynthia Howard, CMS.
MR. BAMBERGER: Brian Bamberger, MediMedia.
DR. COHN: Thank you all for joining us.
DR. PETERS: I was out of line. I apologize. I wasn’t out
at my assigned — I am Rick Peters, emergency physician with Kaiser Permanente,
here representing the American Academy of Family Physicians.
DR. COHN: Well, before I hand this over to Jeff for a couple of
introductory comments, I guess I should publicly disclose, one, that I work for
Kaiser Permanente and it seems like that we are filled with panelists
presenting from Kaiser Permanente. So, the good news is I believe that
Scott Robertson, who is really testifying for HL7 is not formally testifying
for Kaiser Permanente, as I understand. I would also expect that Rick
Peters, who is working for Kaiser Permanente, is testifying specifically for
ASTM and not representing Kaiser Permanente in this testimony.
DR. PETERS: And the American Academy of Family Physicians.
DR. COHN: And the American Academy of Family Physicians, but just to
note that issue.
The other thing, of course, is I think as you all know if there are issues
coming before the committee today related to CPT, I sit on the AMA CPT
Editorial Panel. I suspect that that is not going to be an issue for
discussion significantly today, but just to note it for the record.
Jeff, would you like to make some introductory comments before we start on
our first panel?
MR. BLAIR: Sure. Thank you. Let me extend my welcome to
everyone for being here today, especially those that have had to come on short
notice. You are aware of the Medicare Prescription Drug Improvement and
Modernization Act, which was passed November of last year winds up calling on
the NCVHS to study, evaluate and recommend standards for e-prescribing.
As we have done that, it is very clear that there is a sense of urgency for
us to move forward as quickly as we can. For that reason, the NCVHS has
tried to accelerate these hearings to provide information to Health and Human
Services and specifically CMS as quickly as we can. That is the reason
that the first set of recommendations will be offered, or at least we are going
to try to offer them in September, if we manage to stay awake and work through
the weekends and nights and everything, and the second set in March.
Now, for anyone who wants to try to understand the plan for how we try to
achieve this accelerated path, you could go to the NCVHS web site and the 10th
version of our plan, I believe, if it is not posted there, the 9th one is, and
then the 10th one would be available shortly. That will lay out
essentially our strategy, which was to try to hear from all of the users of
these strategies, the vendors of these — that use these standards, the
networks, the providers, the state accrediting agencies, the health care
providers, the pharmacists and pharmacies. In short, all stakeholders
that were identified in the MMA law, but in addition to those stakeholders
identified in the law, we have also tried to hear from consumer groups and from
other users that were not identified.
Those hearings were — okay, I will use the word — “jammed” into
March, May and July hearings. We tried to hear as many folks as we
could. Maria Friedman gets tremendous amount of credit for really working
to try to make sure that everyone gets included and has a voice in this.
During that process, our consultant, Margaret Amatayakul, has done a fantastic
job capturing all of the comments, criticisms, the standards, the potential
standards, the limitations to the standards, the gaps that may not be addressed
by the standards and the related issues.
That was disseminated back to not only the standard developers and
identifier developers and terminology developers on August the 4th, with a
request that they give their responses to whether they can address this in the
near term or long term, but also to other interested parties that also made
comments on that. Today, we have selected those that really appear to be
the most pertinent to wind up telling us how they could address the limitations
and gaps in the e-prescribing standards that have been identified in all the
In addition, some harmonization work effort so that we don’t wind up
recommending standards that are isolated exclusively to those involved with
Medicare only, although that is the driving force. That is the primary
objective of what we are doing. Okay?
So, given that introduction to try to put things in perspective, let me turn
it back to Simon and thank you all for being here.
DR. COHN: Okay, Jeff. Thank you for your comments.
Our first panel is on messaging standards. I believe we have Scott
Robertson leading off with your views from HL7. Thank you for joining us.
MR. ROBERTSON: Thank you. On behalf of the Board of Directors
and membership of Health Level 7, thank you for the opportunity to testify
before this subcommittee today on the subject of e-prescribing HL7’s activities
in this realm.
My name is Scott Robertson. I am co-chair of the HL7 Medication and
Information Special Interest Group. I am here in my role as co-chair and
as an active participant in the development of health care messages and as a
pharmacist with a background in community and clinical pharmacy.
As mentioned before, I am employed by Kaiser Permanente as a standards
architect, but my testimony today is from HL7, no perspective from Kaiser
HL7 currently has a number of established standards and standards in
development. Of these standards, five are particularly noteworthy in the
context of e-prescribing; Version 2 and Version 3 messaging, clinical document
architecture, structured product labeling and the electronic health record
system functional model.
In the United States, there is a general perception that HL7 messaging is
used only in inpatient or clinical settings. This perception is not
entirely accurate. HL7 Version 2 supports a complete range of
health care factors and is implemented in the inpatient clinical and
outpatient environments, particularly in integrated health care
organizations. Offering considerable optionality and, thus,
flexibility, the Version 2.X series of messages have been widely implemented
and are very successful.
These messages have evolved over several years using a bottom-up approach
that has addressed individual needs through an evolving ad hoc
methodology. However, this optionality and ad hoc development throws some
difficulties in implementation. There is neither a consistent view of the
data that HL7 moves nor that data’s relationship to other data.
While providing great flexibility to Version 2 optionality also forces
implementers to spend more time analyzing and planning their interfaces to
ensure that both parties are using the same optional features in the same
manner. This need to align usage implies that interoperability between
applications is achievable, but not inherently guaranteed in Version 2 of
In the 1990s, it became apparent that this ad hoc methodology and
optionality were limiting factors in the continued evolution of the
standards. In response to this, the HL7 Board endorsed the development of
a rigorous methodology for the next generation of HL7 standards. As a
fundamental shift, this has been termed HL7 Version 3.
The Version 3 methodology brings modern software engineering practices, such
as object oriented analysis and design, along with formal modeling to the
standards development process. This methodology incorporates a concept of
substance and behavior modeling in a rigorous set of work products and
expresses the standard in a widely accepted UML notation.
The goals of V3 and the methodology are that you provide a framework for
coupling events, data elements and messages to improve clarity and precision of
the specifications and to improve the adaptability of the standards to change.
Core to the development of Version 3 was the establishment of a reference
information model for health care. Based upon input from several academic
institutions, provider organizations and vendors, the first was introduced in
1997. An early diagram is included in the testimony and illustrates the
detailed complexity of the early versions of the RIM(?).
Application of a structural vocabulary facilitating their abstraction in the
model and continued harmonization with health care demands, there has also been
a visually simpler, highly abstract RIM, which was approved as a standard in
2003. A similar diagram is presented for comparison purposes. But
even this current RIM diagram is overwhelming at first. During the
development of the RIM, a set of backbone classes has emerged that provides a
means to conceptualize this model. Those backbone classes are entity,
role, participation and act.
Entities are things, such as persons, organizations, manufactured materials,
places. The role class applies context to an entity. For example, a
person may play the role of a patient or a person may play the role of a
licensed entity or a physician. The participation class represents the
performance of a particular role. The role of a physician may participate
as an author or a performer.
Finally, the act class represents what is being done, has been done and can
be done or is requested to be done. Common acts include observation,
substance of administration, financial transactions, patient encounters.
Taken together these backbone classes can be placed into a narrative form, such
as entities play roles whereby they participate in acts.
From my example fragments, we can state one portion of a medication order
model as a person in the role of a physician is the author of a medication
order. By replicating, cloning and specializing these backbone classes,
we can represent all of the information objects and attributes in any health
care interaction. In this highly abstract model, a well-defined
structural vocabulary is required for the core attributes.
The Version 3 vocabulary is a tightly controlled terminology, which allows
the potential of the RIM to be realized, but the RIM and vocabulary could still
produce divergent models and messages without strict process methodology.
Briefly, the Version 3 methodology is based on conceptual modeling of actors,
triggers, interactions and data concepts.
The resulting conceptual model then derives the derivation of the refined
message information model or RMIM from the RIM. Message specifications
are then derived from the RMIM to support the specific interactions.
In April 2003, the HL7 Board of Directors approved the development of a
model, functions, that may be represented within an electronic health record
system. This project was supported by a combination of public and private
funds and has produced the HL7 electronic health record system functional model
draft standard for trial use.
It is important to note that EHRS is a functional model and is not a
messaging standard. EHRS is a tool that is used not by engineers to build
EHR systems, but by clinicians, vendors and others, who seek a common language
to describe the functional capabilities needed to support clinical care.
Eventually there will be associations between EHRS functions and the
messages and APIs that support or implement those functions. While EHRS
is certainly not limited to e-prescribing, e-prescribing is a component of EHRS
and may be included in an EHRS profile. The HL7 EHR Committee has
designed a task group to review the EHRS functional specification, identifying
those functions that relate to e-prescribing and to propose changes to existing
functions or additional functions needed to fully support the requirements of
An early application of HL7 Version 3 concepts and methodologies was the
establishment of the clinical document architecture standard in November of
2000. The CDA provides an exchange model for clinical documents, such as
discharge summaries and progress notes. By leveraging the use of XML, the
RIM, current vocabularies, the CDA mix, documents, both machine readable, so
they are easily parsed, and electronic — processed electronically and human
readable, so they could be easily retrieved and used by the people who need
CDA was used as a basis for the structured document and the representation
for drug product labeling or structured product labeling. The HL7
Regulated Clinical Research Information Management Committee worked with the
FDA to apply CDA in the specific context of the regulated product information
that pharmaceutical manufacturers must supply to the FDA and health care
The SPL(?) specification has been approved as an HL7 standard. Moving
away from the standards, there are several activities at HL7, which relate to
the development and implementation of e-prescribing. First among these is
the relationship between HL7 and NCPDP. HL7 and NCPDP have maintained a
cooperative relationship for a number of years. The basis for this
relationship has been an acknowledgement between our two SDOs of an overlap of
domains that can be used to our mutual advantage.
Recent developments have brought greater focus on the NCPDP, HL7 mapping
effort. You will receive additional testimony on this collaboration, but
I will leave this topic by mentioning three points. First, from my
personal involvement with the collaboration team and interaction with other
members of both organizations, I have noted a great deal of excitement and a
general attitude that not only is this mapping feasible, but there is a strong
motivation for the mapping to become a reality in the near term.
My second point is that the mapping is essential to the success of
e-prescribing between clinical provider applications and community pharmacy
applications. Without such bridging technology, either the clinical or
community applications would have to re-engineer their systems to provide
continuity between these realms.
For this reason, HL7 strongly recommends that the committee support this
mapping effort with funding and resources. The final point on the
collaboration is that any extended effort to map between our standards should
employ a structured methodology and an information model. HL7 will be
introducing the V3 methodology and RIM to the joint collaboration team as a
suggestion for the ongoing mapping effort.
We anticipate that this mapping could be beyond current e-prescribing
concepts to include clinical, administrative and other financial data. I
also briefly want to mention that there is a great deal of international
interest in HL7 to the NCPDP HL7 mapping effort. While NCPDP does not
directly impact the HL7 community, there is a strong international interest in
this effort, its implications and pharmacy messaging and e-prescribing in
Moving to a few committee specific activities that I want to mention, the
Medication Information Committee and its parent, the Order of an Observation
Committee, are the primary owners of the pharmacy content in the HL7
standards. The primary concern is that the business processes,
interactions and data elements are supported in 2.X and clearly defined under
the HL7 RIM and specific pharmacy models that derive from the RIM.
The Medication Information Committee also works closely with the HL7
vocabulary committee to address the identification or development of
appropriate terminologies. There are several recognized terminologies
employed in HL7 pharmacy messages, such as NOMED(?), LOINC, NBC(?)
numbers. There are a number of other terminologies that remain
incomplete, unidentified or need to be developed. Among the terminologies
that are being addressed are consideration of the identification of ordered
medication and associated attributes, such as dose form.
And, two, the standardization of directions for use. The Patient Care
Committee is moving forward with the continuity of care record as a CDA
implementation. The CCR can include a medication history when appropriate
and would closely be tied to e-prescribing messages in some situations; for
example, when a patient is discharged from an inpatient facility.
While not always recognized, e-prescribing is possible is possible today in
HL7 standards. U.S. and international implementations use the Version 2
messages for pharmacy orders and dispense reporting. Also, there are
example projects requiring large scale integration, which employ V3
messaging. United Kingdom’s NHSETP project and The Netherlands National
Medication Dosier are two examples of pharmacy specific projects, either in
development or in early demonstration.
HL7’s near term focus for promoting e-prescribing is the joint mapping
effort with NCPDP to bridge the barrier previously discussed. A
successful demonstration in the immediate future would go a long way to showing
that e-prescribing across the health care sector is a viable concept.
Isolating inpatient clinical e-prescribing for the community e-prescribing will
exclude a significant population and reduce the overall effectiveness of
e-prescribing and providing for continuity of care.
In the mid-term, we need to come to a consensus on various terminology
issues. If we cannot come to a consensus, then these issues must at least
be detailed such that a resolution path can be established. In the mid to
long term, HL7 suggests that a comprehensive mapping between HL7, NCPDP and
other related standards should be considered how we have adjusted the existence
of mapping, priorities, processes, participant’s funding, should be qualified
and quantified as mid-term objectives.
In closing, I would like to restate our recommendations for the
committee. The first, that NCVHS acknowledge and endorse existing
implementations of both HL7 and NCPDP standards for the use in
e-prescribing. Existing implementations that can support e-prescribing
using either HL7 or NCDPD messaging should be allowed to continue present
operations. There does not appear to be any imperative for existing
installations at either standard to be re-engineered simply to replace existing
functional interfaces and messages.
The second recommendation, that NCVHS endorse and facilitate where possible
the joint NCPDP HL7 mapping effort. This project is on track, as will be
discussed in other testimony, to show essential interoperability between our
standards. However, this essential interoperability is a first
step. Additional work will be needed to map the complete range of
The committee’s support and funding in this ongoing effort will be an
important driver to the success of the mapping effort and the success of
e-prescribing across the institutional community boundary.
Third, that NCVHS endorse the HL7 V3 RIM and associated methodology as an
overarching information model for health care messaging and its process
standards. The point of this recommendations is not for HL7 Version 3 to
supplant any existing standards, but to provide a single concept model and
methodology that can be used for mapping and alignment and thus support
interoperability between these existing standards.
HL7 looks forward to presenting the merits of this approach to the other
Fourth, that NCVHS endorse the functional definitions of the HL7 EHRS in
particular as it relates to the functional aspects of e-prescribing.
While e-prescribing may exist as an independent application, in the long term
e-prescribing must be considered in the overall context of the electronic
In conclusion, I want to reiterate and reinforce HL7’s support for
e-prescribing in the efforts of this committee in making e-prescribing a
DR. COHN: Scott, thank you very much.
Lisa, you are on next and we will obviously have discussion after you have
MS. MILLER: Very good.
On behalf of the Steering Committee and the membership of X-12, specifically
X-12 and our insurance subcommittee, I thank you for the opportunity to testify
before this subcommittee today on the subject of e-prescribing.
My name is Lisa Miller. I am the chair of the Convergence and Outreach
Task Group and serve on the Steering Committee for X-12. I am also in my
role of X-12 the former task group chair of the Modeling Task Group for X-12
and we are the insurance subcommittee.
I also serve on X-12C as the chair of TG-4; that is, our functional
acknowledgement reference task group, where we handle those things like
acknowledgements and how we say accept and reject. So, needless to say,
HIPAA has been my role for awhile.
In my active participant in the standards process, I have been involved in
health care since 1976. I am a clinician as well. In the interest
of full disclosure, I must inform the committee that I am also employed by
Washington Publishing Company, the publisher for the HIPAA implementation
guides on behalf of X-12N. I will not be speaking in my role as publisher
or employed by the publisher for those guides.
This is our first time at the table for e-prescribing as X-12 and when I was
given the spreadsheet on August 5th by Maria Friedman, we were a little bit
taken aback that we have not been at the table and that there were questions
around our transactions, but more than that, I have to state that I am a little
bit concerned that as we look at this new initiative and MMA, that as we
struggle with the HIPAA implementations and X-12 is so ingrained in that at
this point in time, as well as NCPDP, that perhaps we had our head in the sand
just a little bit and didn’t notice what we going on around us. So, for
that, please forgive us. We have been rather involved.
So, as we looked at e-prescribing and the spreadsheet that Margaret gave us,
there were specific transactions that currently live within X-12 that were
identified, specifically the 270, 271 or the eligibility request and response,
the 278 prior authorization, but as we looked at it from our industry
expertise, from administrative simplification and the administrative process,
we identified perhaps there were more transactions that really were impacted as
well and that our continuing the administrative process continuum had more of
an impact that needed to be evaluated and that we did not want to see blinders
put on only looking at serve our transaction sets, rather than the entire
continuum and our administrative process model.
For example, we may want to use the 275 transaction to transmit clinical
information in an HL7 form, just like the claims attachment for a prior
authorization that needs additional supporting documentation. That would
be one example of a transaction that we not mentioned. X-12 has a history
of many transactions. Unfortunately, HIPAA only looked at 12. We
have many more than that. In fact, our transaction numbers are well into
the 200 for X-12N and that we may have those transactions that aren’t mandated
by HIPAA that would be useful. One that was mentioned earlier to me today
was the 274 or the provider roster that may be beneficial in the e-prescribing
So, as we look at these transactions and their use today, specifically the
270, 271, some of the concerns that we would have is that in its implementation
today, it is not implemented fully. An example would be that a provider
asks some very verbose questions. It sends to the payer, hello, this is
who I am. This is my patient today. This is my procedure and the
answer from that payer is yes or no, but nothing further about that procedure
and not to the detail that we have noticed in the spreadsheet and the questions
that were asked by the subcommittee.
So, as you move forward with e-prescribing, perhaps this is our opportunity
to fully implement and push the implementation so that our provider communities
have the benefit of those verbose answers back. Yes, provider. We
can answer that question and we can answer it much better.
We also noticed that there were things that were asked for the 270, 271
transactions that were outside of our scope, such as clinical information, we
can answer things like does the patient have a reaction? Is there an
allergy? Those are things outside of the administrative process that we
would not believe are appropriate at this time.
For the prior authorization transactions, the request and the response
provides limited support for the prior authorization of drugs. We end up
with a work-around. This work-around may not be to the best benefit of
our community nor would it be cost effective. We look for that return on
investment. Why would people want to implement rather than be forced to
implement? Why would vendors want to do this?
One of the concerns we have is something called our message
segment. Let’s just put it in there and we will just have a string of
text and you parse it and decide how to use it. That may not be the best
way to implement something. Maybe a short term work-around, but we do not
want to see a work-around become a de facto standard and harm the industry.
So, we do have work ahead of us. We do have changes that need to occur
to both of the transactions that I have mentioned. We must go
forward. We need to codify. We may need an additional
segment. So, for the long term X-12 is committed to those changes and
assisting the community.
Our recommendations — and I will be brief today since we haven’t been at
the table and we are certainly not going to speak to clinical information, but
we would like to make some recommendations. One, that X-12 is included in
the 2006 demo and that we are considered a very integral part of these
discussions moving forward; that we look at the business process and how the
administrative process is fed by both the clinical system and by the
e-prescribing systems through NCPDP and that there is a direct correlation to
that information as you go through the continuum.
A clinical event will spawn an administrative process event. Every
time I go into see a physician and every time you go to see a physician, at the
end of the day that physician would like to be paid and it ends up in an
administrative transaction. So, that data does flow from one area to
another. With that said, you have heard a lot about the NCPDP and HL7
interoperability. I would suggest to you that we also need X-12 at that
table as well.
For your information, a recent document, a memo of understanding document,
and inquiry from our chair and to revitalize the relationship between X-12 and
HL7 was delivered to the chair of HL7 in the past month. X-12 is
recognizing the fact that we must interoperate with other organizations because
our data is coming from those clinical systems, as well as the pharmacy
In that effort, I would like to bring forward to the committee a new X-12
effort called CICA or Context Inspired Component Architecture. This is
X-12’s XML effort, but more than that, it is a process in which we divorce
ourselves in the standards process from our business process and syntax,
protecting the effort that we put into formulating our business messages.
We would propose that this is a perfect point of harmonization in
interoperability, not only for health care but for all standards and for all
industry. CICA is a continuum that can cover all industries, not just
In addition, we would like to prevent the work-arounds that become from the
de facto, we did it this way for years. By using the message segments or
finding those work-arounds for the short term are we harming ourselves in the
long term, allowing us to have the time and to work and move forward and in
that demo perhaps use a draft standard and one that is not HIPAA mandated so
that we can move those standards forward.
Perhaps this is a way for X-12 HIPAA to move forward to the next
version. We look forward to working with the committee and look forward
to working with NCPDP and HL7 in this effort. We would like to be more
aware, included and involved in this process. We believe that if you
begin with a new transaction around a new service, i.e., e-prescribing, that we
will have perhaps a cascading effect through our entire administrative process,
not only including eligibility, prior authorization, but perhaps even our
claim, our remittance advice, our claim status and that this will have an
impact larger than we are aware of sitting at the table.
We need to be aware and we need to be looking at that. Thank you very
DR. COHN: Yes, Lisa. Thank you very much. Will we have a
copy of your testimony?
MS. MILLER: You will. We worked more on the spreadsheet than we
did on the testimony. We figured you would like that data more.
DR. COHN: Thank you very much.
MS. GILBERTSON: Lynne Gilbertson from NCPDP.
NCPDP offers the following testimony on
e-prescribing. Included in your testimony is a document,
“E-Prescribing Guidance and Implementations,” which offers more
detail regarding electronic prescribing. The document discusses the
electronic prescribing process in general ambulatory settings, which can apply
to the acute care environment for prescribing, discharge, medications
electronically in long term care settings.
However, it does not address inpatient care and other specialized
settings. We did not have that level of expertise. While there are
obstacles to overcome, the document and testimony will address the current
implementation, the gaps that can be filled short term, long term and some
The document was prepared by NCPDP members and industry participants.
We would like to thank them for their vast input and participation on the
paper. The purpose of the document is twofold, to articulate existing
electronic prescribing standards and practices and assess their completeness
and/or gaps relative to the requirements and recommendations of the Medicare
Prescription Drug Improvement and Modernization Act, MMA, to recommend a model
electronic prescribing environment, inclusive of challenges and obstacles that
represent all stakeholders and contemplate the requirements and recommendations
of the MMA, while accounting for the current status of the marketplace and
anticipating future advances in e-prescribing capabilities and market adoption.
The paper summarizes sections of the MMA that are applicable to electronic
prescribing, describes current functionality, identifies relevant obstacles for
adoption, explains the NCPDP SCRIPT Standard and its current usage and
capabilities, reports other industry standards and processes used in the
implementation of e-prescribing applications, reports on standards that have
been developed by the industry, but not yet proposed to SDOs and their current
implementations, analyzes gaps and recommends actions and identifies obstacles.
The recommendations for standards: The NCPDP Committee recommended
that the following actions be taken to enable the model electronic prescribing
system to be widely implemented; the transfer of new prescriptions, changes,
renewal, prescription cancel and fill status notification.
Recommendation 1, to continue to support the NCPDP SCRIPT Standard in
applicable domains. Two, to continue to support relevant prescribing HL7
messaging standards in all applicable domains. Three, support the
collaborative effort between HL7 and NCPDP to map the two standards in
appropriate business functions. As you will hear later, this has already
begun. We can mark that as a short term.
Under eligibility and benefit information, continue to utilize the X-12 270,
271 eligibility request response between the prescribing system and the PBM or
payer system. NCPDP will create a guidance document to map the pharmacy
ID card, as you have noted before, to the appropriate field in the 270, 271, to
give guidance for implementation. This work will begin by November 2004.
NCPDP will facilitate bringing together the industry to approve a new
standard that allows the transfer of formulary and benefit coverage information
to the prescribing application to be downloaded and then utilized in a real
time mode during the prescribing process. This process is estimated to
complete in approximately two years from start of project. This was
discussed at length during the NCPDP meetings last week and the work is
Regarding medication history, NCPDP will facilitate bringing together the
industry to approve modifications to the NCPDP SCRIPT Standard, to support the
transfer of medication history from payers for the treatment of patients.
This process is estimated to complete in approximately one year from the start
of the project. Once again, this was also discussed during the NCPDP
meeting last week and the work is beginning.
Continue to support the use of HL7 messages for transferring medication
history when applicable. On prior authorization we recommend that CMS to
undertake a pilot demonstration project that will test the use of existing or
new standards and identify gaps in the e-prescribing process to request, obtain
and communicate prior authorization. There is much to analyze as we
have discussed in previous sessions on the whole functions regarding prior
Regarding provider directory, NCPDP will facilitate bringing the industry
together to modify the SCRIPT Standard to support the communication of
prescribers, nursing facilities and pharmacies that are available
electronically. This process is estimated to complete in approximately
one year from start of project and once again, it started last week. We
have got models already and practices as you have already heard from previous
testifiers of eligibility information actively going on, medication history
actively going on, provider directory actively going on. So, we are
not starting from scratch. We are starting from those efforts and moving
forward. Hence, the short time frame to get the stuff done.
MR. BLAIR: Just one second. Since the word “provider”
is used somewhat differently within NCPDP, are you talking about the dispensers
when you talk about providers or are you talking about the prescribers?
MS. GILBERTSON: We are talking about the entities that will send and
receive electronic prescribing transactions or all, if there are tailors(?) and
other things involved. Right.
Regarding provider directory as well, as Lisa mentioned, there is the X-12
274 health care provider directory transaction. We don’t have a lot of
experience with that. We think it bears some further look.
Obviously, the NCPDP Task Group will be looking at it, but also we recommend
CMS should also be looking at that to see functionality.
Regarding standard identifiers, we recommend CMS look into the utilization
of HCIdea as a standard identifier for prescribers. You will hear more
testimony on that later. We recommend CMS further analyze the feasibility
of the NPI to meet the needs of the prescribing industry. NCPDP will also
facilitate bringing the industry participants together to standardize the
SIG(?) as much to determine there are many aspects of this statement and there
is a lot of work going on. So, we are going to try to see what low
hanging fruit, what longer term and what things can be done.
HL7 has been doing a lot of work in this environment. There are other
organizations that have also been working on standardizing SIG and there are
probably five or six different perspectives of this effort. So, we are
going to try to bring people together and there are some aspects that the SDOs
can do, meaning, for example, HL7 and NCPDP. There are some aspects
that will be outside of our scope, such as the actual terminology.
When you talk about SIG, depending on the perspective, it may be from the
time the doctor actually chooses or writes the SIG, all the way through to what
kind of decision support, what kind of number crunching can I do with the data
that might make up a SIG. So, it is a very vast topic that needs to be
honed in on and chunked up into smaller pieces that we can get some actions
We recommend the current use of drug name, dosage, strength, form, et
cetera, the fields that are on the SCRIPT from the prescriber with the use of
the NDC HRI UPC identifiers as appropriate for the pharmacies. NLM and
the industry should determine RxNorm readiness to be used in the industry for
the pilots in 2006.
We recommend studies on the use of the RxNorm in the current e-prescribing
environment to determine effectiveness, ROI, et cetera. Regarding long
term care, NCPDP will engage industry participation to address the special
needs of long term care settings, to apply the standards and where needed to
suggest extensions of the standards.
The assumption in this testimony is that current and proposed standards will
be utilized in demonstration projects in 2006. Regarding the spreadsheet,
we had some other thoughts that I just wanted to highlight. There is much
more in the spreadsheet, the particular points. I am referring to the
actual rows so you can have a basis for reviewing the testimony when it is in
Regarding the row line message/benefit DUR inquiry from prescriber to
payer/PBM and response from payer/PBM to prescriber, please see our
comments for more information, but we are not sure that the DUR noted in this
statement is part of the eligibility or benefit information. Perhaps this
might be part of a patient profile, which might be different than the
eligibility or benefit information that is shared.
Regarding the row, messages/fill status notification, an NCPDP Task Group
will create further guidance about the use of the fill status notification
transactions. This was discussed last week. A task group is in
place to start.
Regarding terminologies, medication history row, one thought that hit us
when we were reviewing this is any new databases or terminologies that will be
brought forth would not exist in a historical reference. So, we have to
bear that in mind when we are going back in medication history and we have a
different set of data than we might have going forward when we knew what we
were doing better.
Regarding the terminologies tab, we added some of the field’s drug use
evaluation information, such as a drug status — drug coverage status, a drug
utilization, reason for service code, a professional service code, the results
of service code, the co-agent ID and a qualifier. These are just fields,
which have been submitted in previous NCPDP testimony, but they are fields
related to the drug utilization review process we thought would be important.
On the related issues tab, we added a statement for controlled substance
prescriptions and just a thought, that perhaps controlled substance
prescriptions should not be technically processed a different way than other
prescriptions. Perhaps the digital encryption/stringent environment that
has been discussed quite at length the last couple of years, is not required
when the active notification of that prescription gives greater assurance of
the prescriber’s intent and that that was captured in the current paper
environment of today.
Regarding demonstration pilots and studies, this section does not
describe all the possible demonstration pilots but some that we would like to
highlight. CMS may wish to fund studies with existing industry
participants of existing e-prescribing functionality to show the return on
investment of current transaction implementations for prescribers, pharmacies
and the payer themselves. The ROI information could be used in
educational material in the future.
CMS may wish to fund studies with existing industry participants of existing
e-prescribing functionality to show the return on investment of transactions,
code sets, et cetera, that are available but not widely used today for
prescribers, pharmacies and the payer themselves. This ROI information
could once again be used in educational material for the future.
CMS may wish to fund studies to verify operationally that Medicare as a
payer will provide sufficient eligibility and benefit information for the
electronic prescriber’s needs. The ROI information could be used in the
future as well. There is a theme here.
To provide — we have talked many times that the ROI is really
important. Why should I purchase this system? Why should I do
this? Yes, it makes sense, but. And maybe we can use a
demonstration project time frame to also have some studies going on in
Regarding RxNorm, we had a recent discussion with the NLM and preliminary
tasks were discussed to take examples of medications that are currently
prescribed today and walk them through the process. What level or
functionality of RxNorm would support what the prescriber needs to
request? How would that information need to be supported in the SCRIPT
Standard? Is it one field, multiple fields, the code set or what?
Then further map that RxNorm code or however it is represented to what the
pharmacy needs for their pharmacy system and tie into the drug databases that
the pharmacy supports. With enough examples to show really going through,
here is what the prescriber wanted, here is what it translated to and whatever
is appropriate at the RxNorm level, to here is what the pharmacy is going to do
against their drug databases and DUR and all that and make sure it flows all
the way through and provide a mapping. That should show the industry a
guidance of how to go forward and what is needed, how many pieces and parts,
where changes need to happen, where gaps need to happen.
So, we just talked about that and I think we are ready to roll on
that. Then the last comment is how should the demonstration or pilot
programs be measured. What is success? NCPDP is committed to
support the industry in e-prescribing. So much has been accomplished thus
far and many opportunities await.
Thank you very much.
DR. COHN: I presume you have additional written testimony that you
will be providing to us since it is somewhat different than what we have.
MS. GILBERTSON: We had it last week but it didn’t show up on your
DR. COHN: Okay. Because I was looking at 22 pages here and you
went beyond that in your testimony. So, thank you.
Rick Peters, please.
DR. PETERS: Simon, thanks to you and to the committee for inviting us
at the 11th hour, the ASTM CCR committee and all of the supporting vendors and
organizations, particularly the specialty societies and consortia that are
involved in the CCR, thank you for this opportunity.
One of the things in deference to my colleagues, I would like to start out
with is we are up here on a panel of messaging standards and we want to make
very clear that we are not a messaging standard. What I want to start out
with is just a quick overview of our coordination efforts and mapping that is
critical to understand where we fit within this whole discussion, particularly
in regards to e-prescribing.
With HL7, we have a memorandum of understanding underway to allow us to
integrate directly with the CDA and with the architecture, particularly the RIM
and their approach to messaging. With NCPDP SCRIPT, SureScripts and
RxHub, one of our lead authors, Dr. Allen Zuckerman, who testified recently to
NCVHS, with the CCR group, his coordinating efforts to make sure that we were
fully mapped with those standards.
All of our administrative data is mapped to X-12, both to the traditional
paper form of HIPAA 1500 and UB-92 claims, as well as the 271, 272, 278
preauthorization, as well as the utilization review. As we go through
things, we will talk in detail about what effects us as providers, the
utilization review, particularly especially pharmacy and disease management by
the PBMs. It is a critical issue more and more so to us on a daily
basis. So, therefore, we reflect those in the CCR in the way we manage
The last is WC3. We feel very strongly in putting together a standard
that we can’t not look at the junk computer industry. So, things such as
date, time, things such as SIG and things like that, we turn to the general
computer industry to look for XML standards in that regard. The CCR is
not a messaging standard. It is a patient summary. It is an XML map
that allows us to do a detailed summary of pertinent and relevant patient data
for decision-making on a real time basis for doctors, nurses and long term care
facilities and inpatient facilities, so we can support the decisions needed to
make a critical clinical decision on the care of a patient.
Our focus, which is driven by the specialty societies primarily and by the
EHR and e-prescribing vendors, who participate, is really on a whole wide array
of things, but in e-prescribing, which I will limit our discussion to today, it
is really in four key areas. A detailed and comprehensive medication
history, taking into account so many of the things that Lynne brought up that
we don’t have vocabulary and nomenclature that is comparable. Therefore,
we have to be very careful about how we put medication histories together and
make sure that you can take a look at that, either as an electronic message or
from a human perspective and understand what the patient is actually on or
Also, adverse reaction allergy history and a link to medication history,
there are often adverse reactions that are not allergies that we need to track
or we need to fix with information relative to this to make clinical
decisions. Also, medical history only has pertinence in terms of drug
utilization review, regulation audit and the other things like that,
particularly in prescribing and also what we call medication management.
Now, we don’t talk as much in the specialty societies and the EHR world at
e-prescribing per se and I know that we are focused on e-prescribing but I want
to define for us what medication management really is. If we look at it
as an integrated whole, particularly from the practice setting, e-prescribing
to us is retail and mail order prescriptions and refills. So, for us that
would be NCPDP, SureScripts, RxHub, drug utilization review and interaction
with the PBMs, eligibility, pharmacy and other issues like that.
We also look at in-office administration and we like the same systems to be
able to be used to allow us to order medication to be either prescription of
these in the office to be administered and so if we look at that as the EHR
vendors, are very interested in having a standard and they are looking at XML
as an ideal way to do that. So you can have uniformity of order entry in
the EHR world so when we generate one of these, it can be passed from one
system to the other or at least you will have some standardization in its
We look at inpatient order entry from the office setting and we look at HL7
as a key to this because most of the systems will be interacting with the HL7
message relative to the inpatient order. We would like to pass that from
the practice or most EHR systems that we use in the practice to the inpatient
setting. If we look at continuing care and long term care, we look at
those specifically, some of those are HL7, but regrettably most of those are
paper. Long term care organizations we are working with are dealing with
us in terms of making sure we can do paper support for printouts that would be
consistent with state requirements or things like that for information.
They need to administer drugs for patients.
The same thing for home health, which tends to be paper, although we would
like to see those systems more automated and then referral. Referral for
us is EHR to EHR. Primarily, when you look at chemotherapy, when you look
at referring people for specific drug therapy on disease management or things
like that. We see that as a need as well under medication management.
I don’t want us to overlook billing as well. We need to look at
preauthorization. We need to look at generation of claims, particularly
in the subspecialties that deal with administration of medication to the
inpatient setting — I mean, in the outpatient setting and also patient
medication lists, we need a way to generate out of EHRs some type of uniformity
in a paper or electronic personal health record, for example, medication lists
for patients. So, if we look at e-prescribing and we look at it as an
element of medication management, but when we in the CCR and EHR need
prescribing, then if they are involved in this effort, look at it — we look at
it as we want the same systems and the same interoperability between all of
these so that when a doctor, nurse or someone else interacts with your
prescription or a refill or an administration order or inpatient order, we use
the interaction on the screen and we use the same messaging format, at least
internally within the system.
Our core focus under that relative to the CCR work today in the XML has been
on the SIG segment that Lynne brought up. We are very concerned that the
SIG segment reflect the complexity of the difference between a prescription
that is written to be filled by a pharmacy versus an order for administration
of a drug use as an IV or other add mixture in the inpatient or home health,
long term care or outpatient setting.
So, we want to make sure that the system and the tagging that we use allows
for both the prescription refill, as well as the administration of
medications. We think we have done a pretty thorough job of that.
We also have been very focused on medication history and medical history,
linking drug-drug, drug allergy, drug adverse event, drug procedure and drug
diagnosis or drug problem within that medication segment, the indication for
the medication or restrictions on its use or things that define when and how
you would administer it.
We have also been very focused on pediatrics because of the AFP and the
AAP’s participation in this on dose calculation on a real time basis, which
doesn’t appear as much in prescriptions, but appears a lot in orders for home
health or for administration by parents or others.
We have basically focused on clinical care and real time decision support in
this regard. So, when we look at who is really driving this, the EHR van
is the physician specialty society, e-prescribing vendors and long term care
organizations. That focus is not on messaging. That focus is
on content. So, we really are very strongly focused on that and work
integrally to generate messages that would go through the correct messaging
standards to the correct sources, inpatient, outpatient, the prescribing
pharmacy, the TBM, i.e., through RxHub or through the switch vendors or others.
We are trying to focus on the data that they need and make sure that that is
complete and uniform, rather than on the mechanism, i.e., the messaging
standard that we use to get that data to them. One other key point and I
would like to finish with this.
The feedback we have had from the vendors is explicit. We had a trial
at HIMS in January. We had another trial at the Tepper(?) Convention in
May. We will have a final standard out the first week of September for a
ballot. It will be under final ballot in early October or at least
October 15th for IHE to do another demo at HIMS and then we will do another
demo probably in the spring at Tepper again.
The vendors have been adamant that the CCR standard be locked down, that we
not allow variability, that we not allow anything other than a very explicit
implementation. So, we have been under pressure from them to have an
implementation guide and approach and a tagging approach that is absolutely
uniform so that when they take one of these messages, they will be able to use
it implicitly in their system and not have to worry about local extensions of
other things like that.
This has helped us considerably in being able to give something to the other
messaging organizations and SDOs so that we can tell them exactly what we are
doing and get an exact map back from them. So, again, we are not a
messaging standard, but we are here with messaging standard
organizations. We think we work integrally with them and, again, our
focus has been primarily on content and we think that that brings a lot to the
table. Our vendors are very that we begin as quickly as possible and if
we can help in any way in the e-prescribing situation with NCVHS, we would be
more than glad to.
DR. COHN: Thank you very much. I realize that you didn’t have a
lot of time to prepare the testimony. Will we have a copy of it that we
DR. PETERS: If you will allow me the time to type this up and send it
to you, I will be glad to.
DR. COHN: That is fine. Hopefully, relatively, the next couple
of days if we could.
DR. PETERS: Not a problem.
DR. COHN: I guess I should ask the subcommittee. We obviously
are going to be going into discussion. We have a break planned.
Would you like to do a break now and then come back? Let’s do a break
first. Okay. Why don’t we take about 15 minutes and then we will go
into questions and all that and then we will just go until noon or a little
after with questions and discussions.
Thank you very much.
DR. COHN: I really want to first of all thank our testifiers for what
I think has been some very useful testimony and information. I guess I
will ask the subcommittee if you have questions and who would like to
start. Steve is raising his hand first. Jeff, are you on second?
MR. BLAIR: I am going to wait for awhile.
DR. COHN: Okay.
DR. COHN: I guess I should warn both staff as well as subcommittee and
guests that we do have to eventually get out of here by the end of the
day. So, we will try to hold you —
MR. STEINDEL: Thank you, Simon. I suspected that. I would
like to thank the presenters for the very interesting and useful presentations
that — like I mentioned very quickly, probably if I had an hour or two, I
might get through all my questions, both on the spreadsheet and what was
presented today. But as Simon indicated, I probably just want to hit a
couple of the high points.
So, I have maybe three, four questions. The first one I would like to
pose to everyone who presented and get your thoughts on this because all of you
indicated at one point specific standards and another point multiple standards
that could be used in this process. There was a lot of talk about
harmonization and mapping and the use of this standard and this environment or
that standard and that environment. I think we can give specific examples
of each of those. And I am talking in generality.
The question that I have is if we focus on one standard in the
recommendations, would we stall or promote progress? Who wants to start?
MS. GILBERTSON: You are saying one standard —
MR. STEINDEL: No, not for everything, for a specific aspect. You
know, for instance, we were talking about eligibility. And, you know, we
heard that the 270, 271 is useful, but needs some work, et cetera. If we
had other standards for eligibility and we said, okay, we want to try multiple
standards, we were talking about HL7 and NCPDP and how they had focus in
different environments and that we should work on harmonizing but moving
forward with both.
My question is if we made a recommendation that we should use a
standard for a specific task, would we stall or promote progress because it
gives industry one thing to focus on, which could promote progress, but it also
could prevent, you know, new thoughts, new developments, progress in
areas that the standard doesn’t apply with and we have seen this in the HIPAA
process, you know, where we have focused on a standard and we have seen
problems with that. So, I am posing this question to the group.
MR. ROBERTSON: I think that what we have been putting forward, at
least my intention in putting forward the idea of different standards is that
in a particular situation, yes, there should be a particular standard. We
don’t want multiple things going on, but the problem with saying that one
standard and focusing on one standard might mean that you are focusing on one
aspect of the overall picture.
The pint of mapping and harmonizing the standards is not to allow multiple
things to happen between two trading partners, but to make sure that those two
trading partners can communicate to each other if they aren’t using the same
standard to begin with. So, the X-12 eligibility might be the one thing
to do and, yes, that might be something to focus on, but if you are talking
about the order message from a doctor in his office to a pharmacy, NCPDP SCRIPT
is probably going to be pretty clear as the recommendation.
I would say that HL7 would not dispute that at all. Once you get to a
provider and an inpatient environment talking to a pharmacy down the street,
then you have to worry about how you get from one center to the other, but
saying one standard is what you are going to do is only talking about one
aspect of the whole picture.
MS. MILLER: Since you have picked on the X-12 transaction, I suppose I
should speak. It was the easy one to do.
I would like to take this from maybe perhaps a different perspective, which
is that of an implementer and pose to you how difficult it would be if there is
a mixed message coming from NCVHS saying, gosh, I could implement the 270, 271,
but perhaps down the street is — I will just carry this out — eligibility XML
or eligibility this where the data content is different.
Now you are throwing this back into perhaps a Tower of Babel effect.
That is not really what we want to do. We do want progress and I think
that our communities that are out there that are implementing are looking for
tell us the one method we are going to use to do this, especially in messaging
when you are going from trading partner to trading partner. You would add
confusion and I think we would find that we would not have the progress that
you are looking for. So, my recommendation would be that sticking with
one standard, although it may not be perfect and staying with that standard,
again, although it may not be perfect, the progress is perfecting that
standard, learning from our mistakes and moving forward.
HIPAA is a good example. Our standards body, although we are
implementing 4D-10 and we know that there are issues, we are not
stopping. We are still moving forward to collect those mistakes so that
the next version of whatever it is is bigger, better, when it comes along and
that the industry can see that the standards body has been reactive and that we
have made progress moving forward.
So, from an X-12 perspective, I would say let’s not create that discontent
and the misdirection to the constituents out there, but rather let’s give them
a clear pact for implementation and this I look at as an opportunity to help
perfect and perhaps really give stronger wording around some of those
transactions, like the eligibility transaction and really push them to
utilizing the full extent.
One of the members of the audience here approached me in between and said,
gosh, we implemented eligibility. We did a good job. Well, not
everybody has. So, I look at this for the opportunity of increased
progress and increased return on investment, especially to the provider
MR. STEINDEL: Lynne, Rick, do you have anything in addition to add
because if you don’t we could move on to the next question?
DR. PETERS: I will just answer quickly. >From the experience
we have at the CCR, any standard that you state is the standard would be a
benefit to the industry, as long as you put two caveats in that. One, if
you are explicitly clear from NCVHS perspective and probably sequentially from
HHS, exactly what that standard applies for an put a box around what it does
and what it is supposed to do. So, it is for transactions, certain
interaction and that is the standard there.
Also, that the SDO provide to all the players and that NCVHS and ultimately
HHS require that everybody in the food chain implemented exactly the same with
no variation. That would be a service to the industry. That also
relates to some of the lessons we have learned from HIPAA as well.
MR. STEINDEL: Thank you. That helped me a lot with your
It also segues into my second question very well. We heard the talk
from multiple presenters about how important it is to pilot this process, which
really is something that is new that is happening in HIPAA, but we learned from
experience was very important. That will help us refine whatever is
proposed as the final steps for e-prescribing when it comes out.
Now, my question is how closed — and you can use “closed” in
parentheses regulated — should e-prescribing be when it is implemented?
We have seen lots of problems with the HIPAA process because when we want to
make changes there, it now has to go into the regulatory process. I would
like some comments on that.
MS. GILBERTSON: I don’t know that I can speak for the industry, but
his has got to be more open than HIPAA. I mean, as someone who has been
trying for the last couple of years to figure out how we move forward in the
standards development organization, in the HIPAA world, for example, the
Telecommunications Standard Version 5.1 was named. We have gone through
all the numbers, 6, 7, 8, 9, and are already balloting Version A’s and
B’s. The industry is moving forward and to even think of trying to name
next Version A and thinking that through the regulatory process it may take two
years before that is recognized and by that point we will be on Version C and
then what do we do?
It has been a real trial and as the pharmacy industry has moved
forward so quickly into the real time on-line environment, it has been a real
stifling of, you know, now we can only play based on what is being
regulated. So, from that personal perspective, from what I recall of the
MMA guidance, the committee is to recommend and then HHS could then promulgate.
You know, I would think a basic guidance as Lisa and Scott mentioned
earlier, of these are the standards being used for these functions and let the
industry keep moving forward. We don’t move forward at a pace that kills
the industry. We do not move forward at a snail’s pace. For
example, the SCRIPT Standard has balloted as many as four revisions in a
year. So, things move forward if the industry wants to see changes.
Things can also be — you know, you can also ballot only one or two a year
because that is all that was brought forward for that time frame. So, I
think the industry needs to step up and say this is what we want to do
and we really don’t need to be regulated about it. That is just a kind of
MS. MILLER: Well, from an X-12 perspective, here we go again.
And I will speak more to perhaps living through the standards process from the
HIPAA perspective. It is very difficult to respond. You try very
hard to write recommendations and as a government it is very difficult to do
that, where you are meeting all of the needs. I will bring out two words
that I will tell you are probably the hardest things within HIPAA to
define. They are very simple words; “accept” and
“reject.” Just those two simple words alone have created a huge
conundrum for HIPAA and the people that implement HIPAA.
So, I would also have to recommend perhaps not so much an open standard but
a more gentle, flexible approach to the standards development where people can
come forward to the standards body, get prescriptive advice. Their
implementations are truly usable. Again, those terms that I used,
“accept” or “reject,” my example will be a world that is
very close to me, which is the acknowledgement model. Although we have
our transaction sets as mandated by HIPAA, in the appendices, we had the
functional acknowledgement or that transaction that said, yes, I got it, thank
you for sharing, basically, in layman’s terms.
That wasn’t specifically mandated. So, what we have found is that
people didn’t look to the standards body. They weren’t guided back there
to that open ability for us to react and support the industry’s implementation
but rather it wasn’t mandated, so don’t use it. Now, people are creating
their own — and they are going outside of the realm of the standards.
So, as you look in that question, how closed, regulated, e-prescribing
should be? A suggestion would be closed enough that you give them
specific guidance as to this is the standard to use, but open enough for the
standards bodies to react to the industry and to support the implementation,
whether it be with transactions or messages that are not mandated but still
support their implementation and are handled in a way that it puts everyone on
an even playing field. But when you turn around and say, oh, no, it is
not mandated, you don’t have to do it that way, well, if you mandate a
standard, our standards, you are mandating an implementation of the
standard. So, our standard is the base and we have rules and regulations
that we have implemented and we have said to be a standard, this is what you
have to do.
So, close enough to support that. Open enough that we respond to the
industry. That is probably a response that is difficult to do from a
MR. STEINDEL: Again, to move on, Scott and Rick, do you have anything
that is significantly different from what we have heard?
DR. PETERS: I guess it is required from the subspecialty perspective
to say that the general feeling of the subspecialty societies that we have been
working with in the AFP that I represent is against tight regulation
historically relative to the concern that that will stifle innovation.
However, there has been a concern particularly reflected in some of the HIPAA
things particularly the practice management situation with X-12 and the
variability allowed under the X-12 implementation that has not given doctors in
the office setting the efficiencies that they expected from it.
There is a feeling that too tight a regulation even on the standard side may
be dangerous. However, we have not to date seen the vendors move
aggressively towards this, although we are getting hopeful particularly at our
space any prescribing of the vendors themselves, particularly with the input of
SureScripts, RxHub, NCPDP, the harmonization efforts between the various SDOs
that there finally is in one space a vendor drive that is really making
something happen. So, we might be at an interesting juncture where we
might prefer regulation to move things forward. We would be opposed to it
on principle. We may not need it because industry is finally getting its
MR. STEINDEL: Okay. Thank you.
My last question is actually directed to Lynne Gilbertson because she
brought it up, but if anybody else wishes to chime in, they can. This
concerns ROI, which you mentioned multiple times during your testimony and
Karen Trudel has sitting in front of us the encyclopedia that was just
published with the e-prescribing regulation. For those of you who have
taken a look at it, there is a proposal in the regulation for payments for
prescriptions that are submitted electronically.
Would that make a difference in the ROI equation?
MS. GILBERTSON: I would have to punt that to the industry. I
don’t think from NCPDP’s perspective I can really address incentives and where
they should or shouldn’t go. I think the participants would have a much
better feel. What I was trying to mention in the testimony is that we
have between now and January 2006 to prepare for the pilot and I think we need
to have what the goals and objectives are and to start preparing for what do we
anticipate this particular demonstration project will show. How do we
measure it and how can we use that information going forward so that when it
goes outside of the Medicare prescription drug benefit realm, that there is
information that can be shared with other payers, other providers of all sorts
of care, that they can point to and say, see what was shown by this study.
We can build on that or that doesn’t have any bearing on me or — we need to
something of educational materials based on the demonstration projects.
MS. MILLER: From more of a personal perspective, I think when you look
at return on investment and you talk about ROI, it is very difficult in a
provider setting — and I will speak from my history as being a provider and
not speak on behalf of X-12 — it is very difficult to warrant doing something
like e-prescribing if you are making a decision between I need this piece of
equipment or I am going to spend $3 million again on a new initiative as
mandated by the government.
So that return on investment is what is the benefit to me? Where am I
going to have a return on investment? But more than that, the education
on how to obtain that return on investment based on implementation. My
example would go back to that transaction we were talking about earlier, which
was eligibility. In HIPAA so many people started with the claim rather
than with the eligibility and if you look at our business model — and I think
our friends over at HL7 really focus on the modeling, but if you look at our
business model, the eligibility is the beginning of that encounter and
everything is driven off of that particular transaction where as a provider if
I had done that eligibility transaction first, I would not have incurred the
expense of creating the claim, but because I started my implementation with the
claim, I am still not reaping my benefit and my return on investment.
So, a poor implementation practice can create the lack of return on an
investment on something that would have had exponential return on
investment. So, from that piloting and from what Lynne was speaking to
that prescriptive advice, how to do this, the how-to guide. Do your
eligibility first and here are the returns on investment that you will
immediately be — will be benefit to you. That type of information,
especially when HIPAA was implemented, just didn’t exist to empower the
communities, whether it is payer, provider, PBM, when they go to implement,
when they go forward to do this effort, that they have that so that when they
talk to vendors, they can say, look, I want to do it this way because I will
pay for it in the long run.
So, that was a missing aspect when we implemented and did the wide sweeping
HIPAA changes. So, the studies, the pilots that are going to be so
important will lend us to that return on investment prescriptive advice, that
how-to, that I think we really should have available for the community.
MR. STEINDEL: Thank you.
DR. COHN: Okay. Stan, Mike and then Jeff.
MR. HUFF: Lynne, this is a question. In your testimony, there
were a number of transactions that were specified, the new prescription change,
but those were a subset of those that are allowed in SCRIPT and I didn’t know
whether you intentionally left — you know, there are a bunch of responses, for
instance. Each one of those messages that you list has a response.
Did you leave those out assuming that the responses are included with those or
are those transactions not used and we shouldn’t — what is your recommendation
about those other —
MS. GILBERTSON: No, those are the functions, basically the function of
a new prescription, the function of a refill. We did not leave the
responses out. They are part and parcel of each other.
MR. HUFF: Okay. What about the other things that are used for
mailbox stuff like, you know, get message and those things, are those in use
and you are recommending the use of those transactions as well?
MS. GILBERTSON: There is some usage, although not as prevalent as the
others. We didn’t go to that level of detail, I think, mainly because it
is how you wish to connect. So, if you are on a network-based
environment, you may not ever need to use any of the mailbox
transactions. If you choose to be in a dial-up situation, for example,
you would need to use those. It is the only way you are going to
function. We just didn’t go down to the technical level of
recommendations, like in the — you know, there is a password change if you are
using the network and mailbox and things like that.
So, it was just — we didn’t go that level of detail for this testimony.
MR. HUFF: But those transactions are in use and as you point out — I
mean, it would seem to me that if that is the way you are situated, if you are
doing mailbox stuff, then those would be the transactions you would want
to use. Obviously, if you are doing direct connects, you don’t need
that, but —
MS. GILBERTSON: Right.
MR. HUFF: Okay.
The second question, I mean, one of the things that would be desirable is
that we would have — we could specify standard codes for the drugs and SCRIPT
rather than leaving that open. If the terminology for the ordered drug
was specified, how is the best way to do that? I mean, would we want some
new version of SCRIPT that specified that or could the committee just make that
recommendation and there wouldn’t be any need to change the SCRIPT Standard in
that regard or how is the best way to approach that if the committee had a
desire to tighten the coding of the drugs?
MS. GILBERTSON: I am guessing that it would be a new version of the
SCRIPT Standard, but it would be what everybody would be pointing to because
there would the benefit of moving to that coding or whatever. I hesitate
a bit mainly because I am not quite sure what its impact would be. I
mean, it may be as simple, if that is possible, to say add a qualifier to this
field and what shows up the qualifier is new drug classification or whatever
you are going to call it and what follows it is the actual code. If it is
that simple, you add a code and off you go.
But I would venture that the implementation guide would need some guidance
in it that says if you are using this new pointed to code value, you would not
need to send these other fields that could be sent currently today that are
used to supplement. So, yes, it can be changed. I would want to
see, as I mentioned from the NLM discussion, if we went that route, how it
plays out in the whole industry sending it back and forth because what may be
good at a prescriber level, may just kill the pharmacy trying to dispense
So, I am not trying to tap dance your answer. I think we can work it
out as long as we know what it is we are working toward.
MR. HUFF: Sure. Would there be changes — assuming the work to
codify the SIG happened as well, which is actually a combination of
terminology, as well as some formatting and structure, some information model
around that, would that require a change to the standard as well or how would
you see that evolving?
MS. GILBERTSON: Once we had what the business cases would be, I would
envision that this would be, you know, just another version upgrade that would
incorporate whatever those new fields were and the guidance for how to use them
or where to point to if we are appointing to external code sets or
whatever. That is really not the difficult point. It is more the
implementation. Where are you going to get the files of what all the SIG
values are now and how you are going to load them in and how they are
going to correlate with your DUR module. Those are much more
difficult than putting some fields in the standard with some guidance of how to
DR. PETERS: One great example is to look at SureScripts and what they
have done. They discretely map the SIG, but then they convert to a string
and it can generate NCPDP SCRIPT out of that. So, there is also the
intermediary when people out in the real world are looking at it, they may not
go directly to the pharmacy and connect. They may use intermediaries
RxNorm to the PBMs or SureScripts to the pharmacy. So, that is somewhat
standardized in that somebody can use something that is an industry standard,
rather than an SDO standard to get connectivity and to get, you know, finite
data aggregation and still use a standard that gives the pharmacist what they
need to do.
There is some disconnect in there.
MR. HUFF: Yes. Well, I think there is more than one source for
structured SIGs that just haven’t been generally used yet. Is the Sure
SCRIPT Standard, is that available and is that a standard that is published
that would be available publicly for free for use?
Who is here from SureScripts?
DR. PETERS: It is proprietary as far as I know. You are XML
MR. WHITTEMORE: I think there might be a little confusion because of
the proprietary provider ID —
DR. COHN: Please introduce yourself.
MR. WHITTEMORE: Ken Whittemore from SureScripts.
I think there might be a little confusion. We have a proprietary
subscriber provider ID and we have an algorithm for handling the SIG codes, but
I wouldn’t call it a standard. I will also give the caveat that I am not
one of the technical, highly technical people that should be explaining that,
but we can get additional information to the committee.
DR. PETERS: What I am referring to is there is the XML guide that they
have for people who are going to send their messages, that actually lays out
all the components of the SIG, you know, route, units, values, all the various
components of the frequency and that is then converted into a string and a
NCPDP scut(?) message is generated from that.
DR. COHN: Stan, I guess that also reminds you of the NCPDP in their
testimony talked about bringing the industry together to standardize
MR. HUFF: Yes.
And then to clarify, Scott, what I understand you to say is, you know,
basically use standards in the circumstance where they are used.
So, we are not proposing that inpatient medication orders would start using the
SCRIPT Standard, but we continue to use HL7 there, but from a physician’s
office to a pharmacy, we would use SCRIPT. So, in the case, which
admittedly is the more rare case where you are transmitting discharge
medications from a hospital to a pharmacy, is your working model of that
that — I am wondering which end of the line the conversion would happen
on. Are you thinking that hospitals, I mean, would order and they would
convert from there HL7 message to the SCRIPT message before they transmitted to
the pharmacy or that they would transmit HL7 to the pharmacy and the system
would have the responsibility to translate it into SCRIPT or you are agnostic
on the possibilities?
MR. ROBERTSON: Actually, agnostic is not a bad term in this
case. It can happen at any of three points, depending on the level of
sophistication of the systems involved. The sending application
could convert into — from HL7 to NCPDP. It could be translated on the
receiving side or if they are using an intermediary, which is quite common or
fairly standard in NCPDP transactions, it could be that intermediary processor,
like RxHub or Sure SCRIPT, could accept HL7 from one partner and then translate
it to NCPDP for the receiving partner.
So, it could happen in any of three points, depending on
the level of sophistication of the system.
MR. HUFF: It would seem to be advantageous to either have the sending
system or the intermediary do that translation so that the pharmacies didn’t
have to worry about —
MR. ROBERTSON: I would tend to favor either the sender or the
intermediary over the pharmacy application.
MR. HUFF: And then, Lisa, if you could tell me, make sure I understood
what you said, that eligibility seems to be working but you seem to imply that
we could do better? Could you mention things specifically that we could
do better at or what is needed?
MS. MILLER: Well, certainly. One of the issues around
eligibility today, if you look at the entire medical and understand that
pharmacy did a much better job in HIPAA implementation than the medical
community did at large. If you look at the way eligibility was
implemented, if it was implemented — and I say that strongly. There are
many places today that have not implemented the 270, 271 transaction, even
though it was mandated. They are still in contingency.
They are still trying to get their claim working. So, they haven’t even
started to think about it and if they did, they more than likely did not do
what I call the verbose answers. So, if you look at the transaction and
what it is capable of, its capability is not fully utilized today.
So, as a provider, I could ask a very complex question down to the procedure
code and this is what I am going to do or in the case of e-prescribing, this is
the drug that I want to give to my patient. The answer coming back from
that payer is totally owned by that answering, that responder, the
When they come back, they don’t have to give the verbose answer. So, I
could take all the time in the world so you talk about return on
investment. This is another good one. I took all the time as the
person that has formulated the question and what I get back is uh huh.
So, we have lost the return on investment. We have lost the ability of
that transaction to answer that question saying, yes, indeed, here you
go. That patient can have that drug. They haven’t met their
deductible. They have a co-pay, that we could get very verbose or it is
not available. That was one of the — in the spreadsheet if you noticed
we can’t give that answer today, but we can certainly add that into the
transaction. We need to have a code that says, oh, it is not in our
formulary. So, it is not one that this benefit would talk to.
But are we using as an industry the eligibility transaction to its
fullest? No, we are not. We are working up to that still. So,
the e-prescribing initiative would help reinforce the use of that transaction
and, again, show another return on investment, why the industry should really
I hope that answered you. One other thing if I could, I am going to go
back to your question what about other transactions. I would really
strongly hope that if you move forward recommending the eligibility, which it
sounded like there was consensus on this panel, that you took the time to
also mention the ancillary transactions that are stated in that implementation
guide, the TA-1, the 997. So, you talked about the mailbag and the other
things around NCPDP, I think that was something that was missed in HIPAA, that
that implementation guide specifically says if I send the request and I don’t
get the response, that then I expect some type of response. Here is what
I want you to do, which was the TA-1, which is, gosh, I couldn’t even open
it. Maybe you sent me a word document and I couldn’t figure it out, and
the 997, which says, you know what, syntactically it is messed up and you have
got errors in this message. So, I couldn’t give you the expected
response, which was the 271 or the response transaction.
So, if we are going to look at what other transactions, let’s dig a little
bit further even on the X-12 side of things and specifically say in
implementation there is acknowledgement type transactions that are mentioned
and if you read the forefront of that implementation guide, it specifically
says you must do this, but we didn’t do a very good job, I think, in the first
process of ferreting out all of those implementation issues.
So, I think there are other transactions to mention as well for X-12.
DR. COHN: Okay.
DR. FITZMAURICE: Scott, on the NCPDP HL7 mapping, sometimes it helps
just to know more about the specifics of it. What is being mapped?
Is it the data elements, the information flows or the functions? Are you
looking for consistency of vocabulary? Is it a matter of making sure
there are three text spaces in HL7 so you can push in the NCPDP specified
elements and codes for them? What is the mapping doing?
MR. ROBERTSON: You will probably hear more about that in subsequent
testimony, but generally for its mapping of functions, first of all, so that we
know which messages go with which messages. It is mapping as data
elements so that an NCPDP field maps to a specific HL7 field in their
Something that must go along with that is the terminology over — although
there are issues about terminology that we are using the same terminology that
appropriate terminologies exist in particular in terms of identifying the
medication. There is no full imbedding of one message in the other.
I mean, like we are not going to just serialize the HL7 message inside a SCRIPT
because they would still have to understand HL7 and that is what we are trying
Basically, it comes down to a field by field mapping or identified
DR. FITZMAURICE: A second question for Scott and then one for Rick.
Scott, you mentioned the U.K. and The Netherlands and internationally there
are over 80 clearly Version 3 adopters. Are there harmonization problems
that would affect e-prescribing across the health care spectrum in the United
States; that is, do they have same information flows as we do, the same
concerns with DUR, with coverage and with payment or is what they are doing not
really what we are doing in terms of e-prescribing?
MR. ROBERTSON: Well, there are some differences in overall
flows. Actually I should say, while there are differences in specific
interactions, the overall flows are very similar and one of the things that
even — I am a pharmacist and I deal with the pharmacy realm in
particular. Seeing how the U.K. has to deal with aspects of their
prescription and what their prescription is as opposed to what our prescription
is, has been somewhat enlightening to me to understand more completely what we
do in the United States.
So, I don’t believe there are complications from the international
cooperation that we have and actually that in some respects it might help us to
understand some aspects of our own processes that we take for granted.
DR. FITZMAURICE: Thank you. You make it sound like there is a
good benefit for the international collaboration.
MR. ROBERTSON: We think that it is beneficial to both the
international community and to the U.S.
DR. FITZMAURICE: And the question for Rick on the CCR deals with
element definitions, the elements are then represented on the CCR. Whose
representations are you using for the recommended representations? Are
you designing new representations, like dates, names and things that go into
the CCR? That is the first part of it.
The second part is on the SIG, you see the development of the SIG as an
opportunity to work together with NCPDP, HL7 and others, who would be good for
convening a group. I know NCPDP has volunteered to convene such a
group. What I was going to ask is this work already being done, but I —
from the industry and from your comments earlier in response to Stan’s
question, it is kind of obvious that SIG has not been standardized yet.
So, whose representations for the element definitions of CCR are you using
or are you using your own? Then secondly, the potential of the SIG or
working together with other standard developing organizations to get a good
DR. PETERS: Well, Michael, the answer to the first is wherever we can
we are trying to standardize not necessarily with health care but with the
Internet as a whole. So, date and time, date and signature and any naming
conventions we try to do consistent with WC3. Most of our professional
feedback on XML comes from either the vendor community, you know, the vendors
who are doing XML or from the financial services, which is most of our
professionals come from, which is very standards oriented towards general
There are health care particularities. So, for example, in X-12, all
our administrative data does do almost a field mapping to all of the data that
you would have on a claim or in either electronic or paper format, things such
as name or other things like that. We are generally consistent. We
call a name one thing. We are trying to harmonize that, for example, with
the way HL7 would internationally look at a name, the way the State of
California would look at a unique identifying someone.
So, where we can get standards, we are doing it. Where we can’t, we
are getting wide feedback in voting from the vendors as to how they would like
to do it or how they have done it. In terms of content, the vendors are
contributing a lot and we are looking at standardized vocabularies,
nomenclatures, coding sets, SNOMED, ICD, CPT, LOINC, other things like that
that will be sort of recommended or required under the implementation guide.
In terms of the second part of your question, would you refresh that for my
DR. FITZMAURICE: Yes. It was on the SIG. Do you see the
development of the SIG as a chance to work together with NCPDP, HL7 and others
to get a really good SIG and consensus on it and who would be for convening
such a group and I was going to ask is the work already being done, but it is
not being done or hasn’t been completed certainly.
DR. PETERS: We think it is an ideal area to cooperate on. I
don’t think anyone has done it to a level of detail that at least the provider
community would like to see it. I think that a lot of people have taken a
stab at it. Most of the EHR vendors can do a relatively discretely tagged
SIG, although some can’t. There is a feeling among that group that it
needs to be there because of the interaction with some of the things you will
talk about this afternoon, the drug databases. Once you start to get into
dosing and things like that and once you take data out of one medication list
from one EHR and pass it to another EHR and then try to run decision support
against it and one happens to be running Multim(?) and the other half is
running First Data Bank.
There you get into issues of naming, into issues of coding, but you also get
into issues in the SIG. Can you interpret from the SIG whether or not
that patient is on a correct dose or that dose needs to be varied now because
they have renal failure or hepatic disease or their LFTs are elevated, others
things like that. So, we think it would be a great area of harmonization.
A lot of work has been done and probably more in industry than in the SDOs.
DR. FITZMAURICE: Is there a natural focal point for this
DR. PETERS: You mean NCVHS?
DR. FITZMAURICE: No, I wasn’t trying to tout anything.
DR. PETERS: No, I am pulling your leg.
DR. FITZMAURICE: NCPDP has offered. Other people could also
offer it. Is there an industry group? Is NCPDP the place to do it?
DR. PETERS: Well, traditionally, I would have thought something like
NCVHS, where you could get the standards organizations together to do it, but I
will tell you where the drive from the CCR is coming from. It is from the
EHR and e-prescribing vendors, who are saying we have to have this
standardized. If you will lay it out for us, then we will go back and
redo our architectures to reflect even in our database these segments.
So, I think it is a unique opportunity. I would like to see all the
people here get together because you need it in an inpatient message for
HL7. You need it in SCRIPT to do detailed DUR and dosing and what we are
depending on the pharmacies to do, which is to look over our shoulder and make
sure we are doing things correctly as providers.
DR. FITZMAURICE: Do you support NCPDP as being the convener of such a
DR. PETERS: Absolutely.
MR. ROBERTSON: If I may add one point to that. In the NCPDP
working group meeting last week, there was discussion about SIG standard
division and the work they want to do and I was present. We were actually
sort of somewhat joking about, okay, who is going to actually be the focal
point and it was like, okay, you do it, no, you do it. There is an
agreement that — well, there is an agreement that we will continue moving
forward and we want to — and between HL7 and NCPDP at least, we are trying to
get our two dialogues, internal dialogues, working together on the one topic
and NCPDP at the moment is the focal point that we are going to be using.
DR. FITZMAURICE: I think the issue would be if they extended those —
develop into something that there is interesting consensus on. It would
be nice to have NCVHS recommend it. But work has to be done pretty
DR. COHN: Margaret, did you have a clarification on this or another
MS. AMATAYAKUL: I have a follow-up question.
Scott, I know that the structured product label is something that HL7 has
been working on and some of the responses to the worksheet indicated that the
SPL could be a source for the SIG, standardizing the SIG. Could you speak
to that? Is that — can you clarify what that means?
MR. ROBERTSON: First, I will not claim to be an expert on SPL in
particular, but the content of the SPL would include directions for use.
The specific representation, though, is still — if you are talking about a
coded representation, it is still as nebulous as it is in any of the
messages. We are still limited by the available terminologies. So,
yes, it can be a source for that information. Is that source going to be
uniformly recognized? Probably not at the moment.
Any efforts that we can do collectively to standardizing the SIG would apply
directly to the SPL and its representation.
DR. PETERS: Margaret, we have another issue from provider perspective,
too, and that is that when we write the prescription relative to getting
it filled by the pharmacy and we write 1POBID times ten days, we know that on a
product label it is going to go on the actual bottle that is going to go to the
patient. There is a conversion of that text into something that logically
places a patient. We know then as well that when the patient then enters
it in their personal health record, they are going to put the label
in there that won’t then again match up to what we have done.
So, the product label relative to how the product is, i.e., labeled NDC,
what the format is it comes in versus the product label that would go out with
the bottle versus what is actually in the SIG from the provider to the
pharmacist and that they communicate with are probably three different
MS. GILBERTSON: That is one of the things that was mentioned, spent a
lot of time discussing. I think we were probably an hour’s discussion on
just the codification of SIG or the standardization of SIG and Scott had made a
point last week that was fairly important about the SDOs can only go so
far. We can present here are the fields you would use, the code sets, the
whatever you use if you were going to do something with crunching that data
with the SIG. But at some point we are out of our league and that is why
it is outreach to the rest of the industry. During testimony we heard
that JCHO is working on cleaning up abbreviations or whatever. I haven’t
gotten into that aspect yet.
But it is from the point of when the prescriber writes it that may need to
be standardized as well all the way through so that the data formats can then
take it and consistently translate it. So, NCPDP will help
facilitate. We have got obviously a lot of opinions and knowledge and
information but we are going to look at what HL7 has already done. People
have been thinking about this for years. So, it is time to just if
nothing else throw us all in the room and figure out what are the different low
hanging fruits and then the heavy work that has to go on afterwards.
DR. PETERS: One of the things we found from the vendors in the
EHR and the e-prescribing space is when they look at the SIG, they would like,
for example, route to be codified or enumerated. In the XML world,
enumeration would mean that you can only put these values in there. So,
PO, SL, all the various things, abbreviations because those are common and that
is the way the doctors who use these interfaces and nurses would like to
interact with it.
They also believe that it is an enumerated and then through XML, for
example, you can write an LT SCRIPT that they could generate then a patient
medication list. They can write PO orally or by mouth, but they would
like enumerated as well. So, the conversion to a human readable or lay
readable SIG would be uniform.
The next thing they would like and try for a very structured SIG was a very
known bucket, even for units and other things like that is that then you can
very much control that translation and make it so it is automated. There
is no ability for somebody to change that.
We think this would be a benefit to the pharmacies as well, although they do
that pretty well because most of that is the manual. They will type it
out or they have pre-canned statements that they put onto the labels. So,
they are much more automated than we are.
We do think that from the EHR and e-prescribing perspective, it allows you
to convert between a prescription in medical terminology automatically to lay
terminology and potentially back and that that is one of the areas where the
really detailed XMLs, they would have a huge benefit.
DR. COHN: Yes, structured no matter how you structure it.
Okay. Jeff, I think you are up.
MR. BLAIR: I have two questions. They are really rather specific
but I need to explain a very poor choice of words that I used earlier when I
wound up saying trying to jam things in. For the record, people ought to
be aware of the fact that when we began this process of our work plan and our
target dates for e-prescribing, we have extended the number of days of
testimony twice or three times since then because we have discovered the
complexity and diversity of this e-prescribing space.
While, yes, we do feel some, you know, sense of urgency here and I wound up
using the word “jammed.” That was not the right word because
whatever, we are not able to properly recommend by September 1st and 2nd, we do
have that second, you know, date of March and even if beyond that, there are
things we can’t clarify to the extent that we have historically clarified and
made our recommendations in NCVHS, you know, that can also be articulated.
So, I wanted to indicate that that was an inappropriate word that I
I have two questions. Thank you to all of the testifiers. I
really appreciated your comments. Lynne, you wound up giving us very much
in particular what I was looking for in that you wound up indicating how NCPDP
SCRIPT would be able to address the gaps and limitations and in what time frame
and there were two areas within that that maybe you could clarify a little bit
more. One is in the area of the messages for communicating the
formularies from the PBMs or payers, in particular to prescribers, as well as I
guess to dispensers as well.
We have gotten tremendous cooperation and I think there are members of RxHub
who are here also, who have developed message solutions and there has been
discussion of, you know, the possibility and/or the intent to be able to have
them become — go through the standards development process, the accredited
standards development process in consensus fashion in NCPDP’s
environment. So, I was just wondering when — is the date that you gave
us with the understanding or assumption that our RxHub would be part of that
development for those formulary messages?
MS. GILBERTSON: That is correct. The medication history, the
formulary, the provider directory, those are all transactions or functions that
are being performed now, some with modified SCRIPT, proprietary SCRIPT and
others with proprietary formats and those will be brought forward through
NCPDP, yes. I am sorry I wasn’t clear about that.
MR. BLAIR: No, no, no. That is okay.
MS. GILBERTSON: They will have a little thunder this afternoon when
they get to testify.
MR. BLAIR: I don’t want to take their thunder. I love thunder.
MS. GILBERTSON: All the different ones — all the different standards,
yes, that is totally the intent.
MR. STEINDEL: Lynne, the dates you gave us, were they for final
approved standards roughly?
MR. BLAIR: And that was two years, wasn’t it?
MR. STEINDEL: It depends. Some were one year, some were —
MS. GILBERTSON: Some were one year, some were two years.
I don’t remember if we said actually, finally, completely approved,
Teri, or whether we said —
DR. COHN: Teri, do you want to come up and introduce yourself and —
MS. GILBERTSON: — as far as we can go. Like in the formulary,
it will be anybody who is involved and we heard from MediMedia things like —
MR. STEINDEL: Let me clarify with the extent to why I was asking the
question. They will be at a stage where the SDOs feel that they are
comfortable for use in a pilot.
MS. GILBERTSON: Yes.
MR. STEINDEL: Thank you. That is enough of a clarification for
what I need right now.
MS. GILBERTSON: I just can’t guarantee they will be completely through
the ballot process, but, yes, the intent is they are ready for prime time.
DR. COHN: Did you want Teri to comment also?
MS. BYRNE: Yes, I agree with that. We are hoping that we will be
able to use the standards as they exist and as they are proposed, as they exist
today in the pilots in January of 2006. But while we are taking them
through the process of NCPDP.
MR. BLAIR: For those of you who have shared the last few months with
the NCVHS, you know this is one of my areas of concern and that is trying to
make sure that there is some type of a balance with preauthorization and the
balance can’t be achieved unless the messages are electronically sent to a
prescriber or let’s put it this way, a prescriber can electronically send a
request for preauthorization and can receive one so that a lot of time is not
The actual criteria for the preauthorization, you know, remains in the hands
of the PBMs and the payers. That is not what I am talking about. I
am just talking about the fact that without the message, the prescriber is at a
distinct disadvantage to express their concern, in particular, possibly, about
patient safety. So, I was concerned when you — if I heard you correctly,
Lynne, I had the impression that in terms of the messages for preauthorization,
that is an area that needs study and that may be the right answer. You
know, that may be the right answer. But even if it is the right answer,
could you give me some additional information so that I could understand that
this would not be an area, which would either have a low priority or could be
deferred because as long as it goes on, I perceive a disparity here, a lack of
So, could you elaborate a little bit more on the area of what needs to be
done with the messages for preauthorization?
MS. GILBERTSON: I don’t purport to be an expert at all on the whole
prior authorization process. There are already prior authorization
transactions in use between the pharmacy and the PBM or the payer. That
exists and is in use. But as we talked in previous testimony, that is
down the food chain basically. There are, if I am not mistaken, X-12
transactions for prior authorizations from what we would in this entity call
the prescriber to the payer. I don’t know the usage of that.
Hence, one of the question marks in the testimony. We really just
don’t know that aspect. Another question we discussed during the last
testimony was how — or whether it is low hanging fruit or not is how much of
the prior authorization process would be invoked in the Medicare prescription
drug program. If there is no functionality for that, then it might be a
longer term study if Medicare prescription drugs will come out with prior
authorization requirements, then we really need to get that whole flow under
control because if it is transactions bouncing all over the place, but we
aren’t really getting the answer to whoever needs it, then that is where we
thought the further study could be started immediately and gone forward.
Because some of the transactions are there, but we are not really sure they
all flow together or whatever the business case is that MMA might be
introducing are part of those flows that do exist for today in the transaction.
MR. BLAIR: Thank you.
MS. MILLER: I actually wanted to relate to you some of what our
authorization workgroup sent with me. So, although I am not the expert on
authorization, they are. So, I will read to you their concern and where
they see authorization with prescription drugs.
The 278 request response provides limited support for request for prior
authorization of drugs. The transaction provides the ability to identify
the drug using NDC or other HIPAA standard code sets. However, most
physicians request drugs by name or formulary and have resorted to using the
MSG or message segment. It is a text strain to request the drug.
The following URL — and I do have a URL that I will give to the committee
to go look at — provides the work-arounds recommended for the limitations and
the HIPAA implementation of the transaction. So, X-12, the work group
specifically has identified these limitations for those transactions in use
today. I am not sure how much RxHub has found out, but I am sure Teri
will address that when she speaks.
The 278 supports code sets that identify the patient condition, rationale
for the drug, as well as the ability to attach medical history
information. The 278 does not support all the identifiers noted under the
identifier’s tab and specifically does not currently support the DEA
identifier. The 278 HFD segment does support the ability to express the
dosage. In addition to the use of the 278 for prior authorization, one
might use the 278 to enable the supplier to request permission to fill an order
from an identified provider. The response conveys permission from the
payer to the supplier to fill the order.
The 278 notification could be used to send a copy of the prescription
request script from the provider to the pharmacy. It could also be used
by the pharmacy back to the provider to notify the provider that the
prescription has, indeed, been filled. The pharmacy and supplier could
use the 278 inquiry to determine if authorization has already been granted for
the prescription. In addition, the 278 provides the ability for the
prescriber to request, cancel, renew, extend, revise an authorization or order.
While the traditional transaction exchange request and response between a
provider and the payer, UML, the 278 standard does not limit the business
entities that participate. For example, the request may be from the
dispenser to the payer between the prescriber and the dispenser. The
dispenser response can indicate the prescription or order has been modified by
the dispenser or payer.
Now, the most important thing in this statement from this is that they
mention other 278 transactions that are outside of the HIPAA mandate. So,
you have the 278 inquiry and the notification, but certainly that workgroup
within X-12 has the capability and can move forward. Short term, they
have the work-around to keep the industry moving today. Long term, it
would take us greater than a year and I can’t give you much more information
than that to move these forward or to get adoption. However, the other
278 transactions do exist and you can look at the X-12N registry. I
believe you can find that at WPC, WPC-EDI.COM, and you can look at the current
transaction that would meet this business need.
So, X-12 does have the authorization, transactions that can be utilized for
MR. BLAIR: Thank you very much.
DR. COHN: Margaret, did you have something right now or can I —
okay. Harry and then I have questions and Margaret, do you have
MR. REYNOLDS: Excellent testimony. It made me think about a lot
of this a little bit different than when I came in the room. Anytime you
look at anything we are trying to do, you try to look at the deal
breakers. One of the reasons everybody started with claims and HIPAA is
it was the deal breaker. If you didn’t at least get that automation level
that you had before, it was a deal breaker. So, I think the year of 2005
will be when your other X-12 transactions will blossom.
When you look at e-prescribing, one of the deal breakers is will the
physician use it and is it really going to do that business process in a way
that it really becomes real? I have no affiliation with any of the
standards organizations. So, nothing I say here was pushing anything one
way or another, but I guess my question is when you look at HL7 and X-12, they
are looking at the entire end-to-end health care process in dealing with many,
whether it is lab or pharmacy or claims or just procedures or anything else
that is going on pretty much setting that — when you look at NCPDP and you
look at continuity of care and you look at some of those, those almost appear
to me to be subsets of the industry.
When you talk about who should maybe lead each of these — and the word
“ROI” has been thrown around here a number of times — there is an
industry out there that has spent a large fortune on implementing the ability
to communicate health information back and forth in many, many different ways,
in many transaction levels and ideas.
So, I would really like to understand if a lot of the group now, for
example, Lynne is mentioning, the NCPDP would head it, you know, a lot of your
recommendations, which I think, thank you, were very precise and very helpful,
but if those divert from the investments that have already been made, including
the doctor, which — and I will just forget the payers, forget everybody else
because everybody is going to — the big players are going to do what they need
to do, but at the doctor’s office that can do the 270s and the 276s and the
278s and I won’t bore you with the rest of the numbers, they have already
invested in those in many ways.
You guys have come close to hugging, but I still sense a real — I still
sense that it would b e one of those that we are not going to see each other
again. We will hug today but we are not going to see each other
again. So, I would really like you to honestly — who should have the
ball? Because when you think of ROI and you think of that doctor’s
office, who should really have the ball? Should it be a segment because
this session happens to be e-prescribing and then the next session we will have
as a committee is something like EHR and then the next one would have as the
next update X-12 HIPAA transactions.
So, that sets a framework and then obviously within that, one of the issues
is in HIPAA there was a — I called it an interpretation free-for-all and I
would hope that whatever you as standards organizations would do
in e-prescribing and others would not leave it quite as open because
everybody took their own stance and didn’t quite get there.
So, I would like a comment maybe from all four of you.
MS. MILLER: The HIPAA and you mentioned some of those transactions, a
couple of thoughts. X-12 participates in HISB, the Health Information
Standards Board and that might be a very good forum for the convergence for
everyone who sits at this table to bring some of those issues forward for the
benefit of our industry. So, perhaps one recommendation would be the
involvement at HISB and the continued cooperation there at HISB.
You also have the Designated Standards Maintenance Organization, which is
another forum for all of us to come together and at least three of us at this
table are participating at that level. But more than that, when you
look at, as you said, the interpretation free-for-all, it is very difficult to
put a standard out that covers the entire health care continuum and not have
the issue of interpretation free-for-all, making standards broad enough that
everybody can use them, yet specific enough that they have the intent of
that business content message is not an easy task. There are lots of
compromise that goes on.
So, you are back to that closed and open thought process around the
standards. How closed do you make them? How tight? How many
non-extensions? How specific does the business messaging get? And
how hard do you make it to implement versus do we run into that risk of the
interpretation? So, I am not sure I have a clear direction for that, but
we have learned. We have learned how to make things better through the
trials and tribulations that we faced with the HIPAA implementation.
But from the perspective of who should lead it, I certainly wouldn’t
recommend that X-12 would lead e-prescribing. That is certainly not our
area of expertise and we would look to our fellow standards body, NCPDP, to be
the people that lead it. But I do think we need a forum where we all meet
and we work together towards the industry and perhaps HISB is the place for
MR. REYNOLDS: A follow-up, I guess, to your comment. There is no
question that in some of the open interpretation in HIPAA, you had
entities dealing with entities. Very seldom did it go down to the
MS. MILLER: That is correct.
MR. REYNOLDS: The concern I have in e-prescribing is if it is not
enough for a doctor to consider it helpful in e-prescribing, I mean,
consistently across the country, then I just don’t think it is going to carry
nearly the oomph. It is not going to carry — it is not going to be a
reason to do it and it is going to be very difficult.
So, I think, again, a different model because entities dealt with entities
in HIPAA and now you are dealing with these individuals. So, that is a
little different, at least the way I am seeing it.
MS. MILLER: I will comment from a personal perspective. I have
spent most of the last five years traveling around the country doing an awful
lot of education on HIPAA and I found myself in front of a lot of organizations
that were the payers, large organizations and they would stand there and say,
yes, we got HIPAA. We don’t really need to hear this. But the
people at the end of the day that would listen to the presentation would be the
providers and they would walk up and say why didn’t anybody tell me HIPAA would
do that for me.
My example that I would give was actually the eligibility transaction.
If I may, I will share it with you. I think it is more in the message to
the provider community. HIPAA did, in those transactions, really did give
the opportunity to the provider. They just weren’t informed
and they didn’t realize the empowerment they had from those transactions,
whether they were used entity to entity or from little provider to entity.
So, my example was the eligibility transaction and in the example that I
give is that a nice elderly lady walks into a doctor’s office and she is there
to be seen and he has been writing prescriptions for her for over six months
and he doesn’t have the ability of the eligibility transaction, but now
we are going to move that same example into today with HIPAA and the
eligibility transaction and he fires off the question. I have this
lady. She is in my office. I am seeing her. Tell me about
We get the verbose answer back that says, gosh, here is her deductible,
which she hasn’t met. Here is the information I can tell you today, yes,
she is covered, but I am going to tell you that you can see her and
that her deductible isn’t met. No, that doesn’t sound like a lot until as
a provider you start looking in that and go, well, gosh, you know, I have
written prescriptions for her for six months.
She would have met her deductible had she filled them. So, obviously,
I know now that my patient isn’t getting her prescriptions filled. So,
those administrative transactions had a clinical impact that was met and a
helper impact that was met. More than that, that provider, if they found
out that there wasn’t a benefit could have asked for payment at time of service
and it had an impact to their business and their cash flow, as well as not
taking the time and the cost to move forward with that claim and to prepare the
claim. That message never gets out to the individual. So perhaps
one of the recommendations through ROI and through these pilots is we get that
message to that provider. We tell them what their benefit is going to be.
We really didn’t get that message out well with HIPAA. We were
concentrating on those entities and e-prescribing really does focus down to
that provider. So, from that personal perspective and from being a
provider myself and living in that realm, I wish I would have at the time I was
practicing had the eligibility transaction so that we could have seen a bigger
picture of what that patient looked like and what information that benefit
package told us about that patient.
So, I think part of it was HIPAA and in defense of those transactions, they
have the capability. We just missed getting the message out.
MR. ROBERTSON: First, going back to what I heard as one of your
primary questions in terms of is there a problem if one industry subset leads a
particular effort and at this point in the ongoing cooperation that
we have between the different organizations, different SDOs, I don’t know that
it is really significant that any one organization take the lead providing that
we are all at the table and working together for the common goal. It
really becomes a matter of who has the resources, who is available to keep the
ball moving forward.
Somewhat in response to what Lisa was just talking about in your questions
about getting the prescribers to buy into it because it really — I mean, we
can say a whole lot but if the doctors aren’t going to use it, it is not going
to work no matter what we try.
There is something I try to keep in mind. I deal with messages and I
can say a whole lot about, you know, you use this structure, you use this
terminology, you use this code set, but there is an entirely different world in
terms of what the user sees. So, the user interface has to be useful to
them. It has to be something that they can deal with. They are not
going to want to deal with the terminologies. They really won’t care if
behind the scenes it is NCPDP codes or RxNorm codes or however it is
represented. They will care a great deal if they are expected to put an
NDC number in because they won’t know what it is.
So, the user interface really gets divorced from some of the standards talk
that we are discussing, but there is a direct — there is a relationship
between the two that needs to be addressed. Again, also, as Lisa was
mentioning, you know, when the providers didn’t understand about what HIPAA
would do for them, they didn’t know what they needed to like or dislike about
it, except that it was new regulations they had to follow, which generally
means they didn’t like it because I don’t like being told what to do.
Along the same lines, it is important that the prescriber community, the
worker bees that are going to have to deal with this understand what
e-prescribing will do for them and why it is really a very good thing that is
coming down the road. I could see some people, some organizations and a
lot of individuals that at first glance would think, you know, e-prescribing,
here is another set of regulations coming out of the government that just means
it is going to complicate my life further.
Well, there might be a little bit of complication but effectively
introduced, it could provide a whole lot more benefit to their work flow and
ability to provide better care for the patient than what they might perceive as
issues about regulations coming down from the government.
MS. GILBERTSON: A couple of years ago, NCPDP had an electronic
prescribing educational forum and one of the most powerful speakers of that day
was a doctor, who stood up and said here is what my practice was, here is what
my practice is now because I have introduced some e-prescribing
functionality. That speech was — it was right onto what is important to
me as a provider. So, I think that once again is the message.
Did I have to hire more staff? Did I put in new phone lines? Was
I spending from 4:00 to 6:00 p.m. everyday just dealing with refill requests
and things like that? As Scott said, the standards, we can make them do
just about anything we are asked to do from the industry. But we really
need to focus in on what the benefit is and how we can improve the flow of
information. I don’t mean that to be just a trite statement because I
think it is very important that we look at — and when NCPDP facilitates — you
notice I did not say NCPDP lead anywhere — it is NCPDP will facilitate.
Just because I have heard the testimony since March saying we need and we have
some bodies who are very interested in this. So, if we don’t do it, who
will and that is not to show any negative to anybody else, who has been working
on this. It just means we want to come to the plate, too. If it
takes conference calls to get people together, we could at least get started
because if we don’t do that, six months from now, a year from now, we will have
the same discussion and will have gotten nowhere.
DR. PETERS: I think Scott said it very eloquently, the doctors don’t
care about what is under the covers. They care about the fact of what
they can interact with and that that is easy to use. The vendors in
e-prescribing and electronic health records, they think they can do a good job
of that, what they are looking for since they are between the doctors and the
actual standards and such like that is they want some uniformity, but they have
to deal with variability anyway. They deal with X-12 routinely because
most people don’t want to buy an EHR that is a stand-alone anymore. They
want a part of a package management system and they want it to do the billing
automated and relatively successful in that.
In labs, they have to deal with HL7 and they do, but, for example, AFP has
been running a pilot project for six months on a dedicated EHR implementation
in family practices and it started out as just a documentation implementation
employing EHR for documentation alone. The doctors said don’t really want
that. What we want is we want e-prescribing. We want connectivity
to lab. We want order entry. We want electronic billing to go
directly through our products management system, a various array of practice
management systems for e-prescribing is relatively
We tend to overlook as we did under HIPAA that pharmacies are the segment —
retail pharmacies and mail order pharmacy are the segment in our industry that
is automated. Ninety-five to 98 percent, I don’t know what the current
numbers are, but they do use the NCPDP standard for all their connectivity on
billing and claims. They are used to it. They are comfortable with
it. They are comfortable with the organization and they make up that
organization and they represent it.
We know and the vendors know that when it comes to actually sending a
prescription to a pharmacy or getting a verification or validation of
eligibility, I can do it easier, instead of trying to check the formula on the
Internet or doing things like that, I can call my pharmacist and say would you
check the eligibility on this. Instantly back through an NCPDP or X-12
So, I don’t think the doctors or the vendors are worried about different
standards. They are worried about siloing by just having a prescribing
application. They would just as soon from the end user perspective, as
Scott mentioned, have a uniform approach with the EHR e-prescribing practice
management integrated. They don’t even care if that is from one vendor,
just so that the vendors work out the interoperability. But what they do
want is uniformity in terms of the back end. The vendors are asking for
Just to give one last example on the pilot project that we are running, we
have three labs represented. Hospital labs and Siemans is one of them;
LabCorps and Quest, the two largest providers of lab service in the
country. We can get HL7 results back from them. Not a
problem. Now, they are different results for each one. So, there
are three separate interfaces. In fact, with Siemans, which is multiple
hospitals, there are multiple interfaces, but they can be converted back into a
management that the EHR vendor can deal with and then just play on the screen.
Cannot for any of the three of those generate an HL7 order. This is
not HL7’s problem. HL7 fully supports the entire order entry process, but
the hospitals and the commercial lab vendors do not support HL7 as a vehicle to
get orders. So, what happens? The EHR system generates a fax.
This fax is entered by hand into the hospital or into the lab, to LabCorps or
Quest system and that is the work-around that has to occur there.
Again, so we talking about these standards being universal across our
industry. The only area that is really automated is the pharmacy.
So, when we talk about e-prescribing, we are right in their camp. So, we
need to make sure that when we talk about harmonization, there is a reason we
want to do this.
Hospital pharmacies don’t use NCPDP generally unless they are dealing with
multiple payers. So, look at community hospitals, they have it.
Does Kaiser do it? No, because we have a controlled formulary and we have
our own pharmacies. So, all this harmonization that the SDOs are doing
and the industry is doing is critical for the customers, but if what you can do
is help us, guide us and help the SDOs focus and each do the things that they
need to do and then we can all work together with the vendors to make it
happen, I think that would be the ideal thing.
I wouldn’t worry about the vendors or the doctors asking for one standard
just overarching or one SDO to manage the whole thing.
DR. COHN: Okay.
Let me try to ask a couple of very specific questions. I realize we
have got about, oh, 15, 17 minutes. I want to give Margaret a chance to
ask also some specific questions.
Obviously, we have been hearing a lot of testimony about standards and a lot
of it is sort of we will work together to do things in x years or y years or
whatever. I think as you will remember part of the MMA actually suggests
that if there are actually standards that exist now that are commonly accepted,
that they don’t need to go through piloting and the Secretary can advise the
industry earlier than 2009 about their applicability and use for
e-prescribing. So, we have listened to a lot of things. I am
actually just sort of curious on are there standards that the NCVHS should be
recommending to the Secretary that are maybe not perfect, I mean, because we
are all hearing about, well, you want to do this, you want to have a bunch of
draft standards for trial use in 2006 to improve things, but let me just ask, I
mean, I think what I am hearing is is that NCPDP SCRIPT, while not having all
the functionality needed, is probably a usable vehicle.
I guess I would ask that question. I think I am also hearing the 270,
271, while not perfect and needing a lot of improvements for eligibility and
all of this, has a lot of work-arounds that mean that it actually could be used
now, but these are not sleek. This is not as nice as people would like to
have it, but it is right now — am I right and what else am I missing here in
this sort of very high level — once again, I am not talking about all the
identifiers, but just the message formats.
What else am I missing?
MS. GILBERTSON: I think in the testimony that I provided, that
assumption, which is current and proposed standards will be utilized in the
demonstration projects in 2006, I think that was an important point that
standards that are currently being used today and that different testifiers
have presented before the committee of active in-use can jump immediately to
the forefront of the studies by CMS, for example, for that, quote, unquote, ROI
There are a lot of models of things being done with new prescriptions, with
refill requests, with formulary, with eligibility checking that is the low
hanging fruit. It is out there. It is active. Now, while — I
mean, it still has to be demonstrated as Medicare, as a payer, because that is
the unknown here. You know, you have a lot of transactions that are going
on with known payers, but Medicare is not — they are not deeply involved in
the prescription benefit environment today. So, therefore, they are a new
element and, therefore, should be part of a demonstration in the 2006 time
When it comes to new prescriptions and refill requests and things like that,
those are between the doctors and the pharmacies. That can immediately
jump to the educational materials that say based on what is currently going on
in the industry today, here is where the investment — here is what is good
about those practices that are being done today.
So, there are a couple of different levels that say based on the testifiers,
here is the stuff that is going on now that could immediately — if you want to
make it part of a demonstration, it is fine, but you have also got the model of
what is going on. So, you could cut to the chase and say, you know, there
are millions of transactions going on for new prescriptions while what do we
want to glean out of that and let’s go forward.
When you introduce the payer environment, that is where I think it is
important that they be part of the demonstration because Medicare is the
unknown entity here when it comes to some of the prescription drug program
DR. COHN: Okay. Well, let me just push you on that one a little
bit only because I am just really curious. I haven’t read the entire MMA,
but I have read enough of it. My understanding is is that
they are not the payers, that basically they are contracting with prescription
drug plans and Medicare advantage organizations, those groups that look
like you have been dealing with probably for years already. Obviously,
the benefits are different. Is it the benefits issue or is it that — I
mean, it isn’t leading with new payers.
MS. GILBERTSON: I wouldn’t want to make the sweeping generalization
that says that if PBMs today who are currently active in a lot of the
prescription benefit information, if they can continue to function as they do
today and we have models of entities that are exchanging transactions
with prescribers with pharmacies and those same PBMs, then we have got — that
part is already the low hanging fruit.
What I don’t know is what is Medicare introducing into the PBM’s
environments that may be different than what they are doing today. That I
would have to be a little careful about, just assuming its drugs are — drug
claims are drug claims are drug claims, just like they are today. I do
not know. So, that is the hesitancy.
DR. COHN: Okay.
DR. PETERS: Simon, the other issue with the PBMs and Medicare is that
the PBMs aren’t clamoring right now to get that business, due to the
concern over the drug spend and the expense of managing that population.
The other issue, too, is that we don’t know yet from a provider perspective
what the PBMs are going to go back and sell to the Federal Government or to
Medicare in terms of services. There is a lot of specialty pharmacy with
that and drug utilization review and drug utilization audit. That would
change somewhat the requirements for messaging, even from the pharmacy side of
So, again, for the PBMs to get into that business, they may ask for some
things that allow them a greater level of control. That will sort of flow
down hill in the food chain to more information that needs to be provided from
the provider and the pharmacy on a given prescription.
MS. GILBERTSON: One of the things we run across and I am going to get
out of my league very quickly, but in the claims processing, we have 4 billion
pharmacy drug claims being submitted every year, but we did not have in the
on-line real time environment Medicare as a payer until HIPAA. We are
still struggling to meet Medicare’s needs in the standards because they need
different data elements than any other payer in the United States.
I am just a little concerned that I may see that same thing show up in the
DR. COHN: Okay. Lisa and I think Steve has a comment and then I
want to give it to Margaret.
MS. MILLER: One concern I have, you mentioned the eligibility
transaction. I do want to bring one thing out. We are kind of
constrained as X-12 right at the moment because we do have a mandated
implementation of that transaction and speaking to this testimony, we were
constrained that we are talking about the 4010 transaction. Now, if I had
the latitude and the standards body had the latitude to speak at maybe the 4050
implementation guide, which is published or the new transaction, the 5010, that
we are working on, we could be more eloquent and do those things.
So, I do want to bring forward that, yes, the eligibility is out
there. We have it install base within the industry, but it is on that
4010. Again, just as Lynne pointed out, as you bring forward these new
requirements, there is going to be an impact on that transaction and even into
our prior authorization transaction, that they weren’t really looking at
e-prescribing at the time that they were designed, developed, mandated,
implemented. So, as you are looking at this, I would strongly urge you,
how do we move forward and I know this is a big topic anyway, but as you are
moving forward with e-prescribing, how are you going to enable us, the
standards body, to move forward to the next versioning and meet the needs of
the environment and meet the needs of these new needs as they are coming down
So, rather than looking at just our install base, yes, we want to leverage
that, but wouldn’t it be nice to be able to use a newer transaction that could
better meet the needs as well? So, just to put that out. I was a
little bit concerned when I heard you say, you know, it is there. It is
not quite meeting the needs. It does meet the needs, but you are now
presenting new requirements to us in this spreadsheet and saying, look, here
are the requirements to utilize eligibility now in this new realm.
So, as you introduce new things, we need to be able to react to them and
move forward with new versions of our transactions.
DR. COHN: Well, Lisa, I agree with you, but I think I did hear you say
that you had a list of work-arounds.
MS. MILLER: We do. We do have a list of — absolutely and it is
in the spreadsheet where we address — but we also have areas where we say
there is no work around and the spread sheet, I think we did a very good
job. Margaret and I had already discussed —
DR. COHN: Okay. Testing the waters here.
Scott, did you have a comment?
MR. ROBERTSON: Just quickly that you are looking to — your statement
was there are standards in existence or are there standards in existence and
through the range of standards that exist today, you can — the committee
should be able to say, yes, these are standards that exist today that cover
these aspects of e-prescribing. That would then help to identify the
moving forward issues that we have all been talking about. It may help
some organizations to realize some things that are available they didn’t really
think about before.
There might be some provider organizations that just never thought about the
fact that, oh, you mean we could do prior authorization. That is a good
thing and also it could be part of the overall education of the health care
community to this is what you can do if you are e-prescribing today and this is
where we are going forward.
DR. COHN: Okay.
MR. STEINDEL: Lynne, I have a very specific question for you since
Simon drilled down very low and I feel like I can ask this question. In
the set of the NCPDP SCRIPT transactions, I think we have heard numerous
testimony that there are certain areas of things like a new prescription, a
changed prescription, that we do need to work on, like, for instance, the SIG,
the drug name and stuff like that, which we will be covering later today.
But I think in your responses to the spreadsheet and what my sense is is the
refill transaction can go through relatively unmodified, that it doesn’t
require anything new and we have also heard repeatedly in testimony before this
group over the last few months, that one of the things that e-prescribing can
help tremendously in a physician’s office is the refill. Is that a real
low hanging fruit?
MS. GILBERTSON: Yes. Definitely.
DR. COHN: Steve, thank you.
MS. AMATAYAKUL: I want to ask a question about acknowledgement of
receipt. I know that X-12 has the capabilities of the TA1 and the
997. I know NCPDP has a request for receipt, which was a verify component
to it. But in response to the worksheet questions, we did get concern
raised about the cost of these transactions and whether they should be trading
partner agreement issues or whether they should be required for every single
transaction or what. We did hear some testimony earlier that prescribers
want an acknowledgement, but if there is an additional cost, what does that
DR. PETERS: From the proprietor’s perspective, I am not sure who said
that university providers want receipts. One of the things that we found
is that the actual doctors themselves on the street aren’t interested in the
receipt. What they want is fulfillment data. They want when they
refill a prescription the ability which someone thinks that RxNorm, for
example, can provide to the PBMs, they would like to see how many times that
patient has filled that or what they filled or what else they may be on from a
The notion of receipts, which the systems could handle, the doctors don’t
want to see that on their desktop necessarily because just the fact that the
pharmacy got it, they are more concerned, we find, among our members, of what
the patient has actually taken or what they are filling so that they can look
not only at what they are they are prescribing but what other doctors are
prescribing to get a better picture overall of what the patient is taking.
MS. GILBERTSON: In the NCPDP SCRIPT transactions, there is —
everything is request and response based. So, you send a new prescription
request. You get a status or an error message back with, you know, good
or bad basically. Refill requests have refill responses. So, all of
those have real time bases to them.
The verified transaction, which you were mentioning, that is more of a —
basically built on the USPS model of there is something special about
this transaction and I want to know a specific response to it, so I can tag it
as I want a special notification. So, it is not used very widely.
It was just when we were blue skying, what are all the different kinds of
functionality someone may want to know, rather than the request and the
response, I may want to tag this one with return receipt. So, that is
what that one is for.
I don’t think any of the testifiers really said that transaction was
something they thought important to go forward. It is just out there if
you need it for a specialty situation, but the real time request and response
handles what they needed today.
MS. AMATAYAKUL: I think the testimony I heard mixed the fill status
notification and the acknowledgement of receipt. Thank you.
MS. MILLER: And, Margaret, for X-12, you mentioned the TA1 and the
997. You need to add to that list the 999, which is actually the
acknowledgement to an implementation; 997 is to the base standard and a TA1 is
the acknowledgement that says whether or not I understood it or not and it is
at the request of the sender, not of the receiver.
One other point, we do expect, when you send an eligibility, that the proper
response in real time is the 271, which is actually the informative
answer. We have the other transactions in there in case of a catastrophic
failure. So, in the case of real time, where you have a timeout, you
would expect something to come back to say what that timed out. I didn’t
get all of the — so, just from a technical point of view, there are
acknowledgements and that model that is core to the X-12 standard and how we
function with an X-12 and that crosses all industries. This is not
health care specific. This is across all of our industries,
finance, government. I can go on and on and on, that that is just core to
the use of the X-12 standard and at times is implied.
Unfortunately, in health care when we implemented, many people just looked
at the implementation guide and missed the fact that there was an entire rich
standard behind it. So, there was some, again, misinterpretation, as
someone said earlier, but please add the 999 because that would be — to the
implementation guide if you wanted to report back to what that implementation
guide said, the transaction that would be appropriate to be used.
DR. COHN: One more question? Okay.
Well, you know, I want to thank everyone for what has been a very
interesting set of testimonies. I actually appreciate the fact that you
all are working together. I observe that the NCVHS works best when we can
all work together to help drive industry consensus. I think it was
intended to make you all be here all summer with us, but the end result, I
think, has been some very good conversations going on, but I see there are
people beginning to work together and, obviously, this will make all
recommendations all that much easier. So, I want to thank you.
Now, we are going to adjourn for lunch for an hour. We will come back
at 1:15 and then we will talk about NCPDP, HL7 harmonization.
So, thank you.
[Whereupon, at 12:20 p.m., the meeting was recessed, to reconvene at 1:30
p.m., the same afternoon, Tuesday, August 17, 2004.]
A F T E R N
O O N S E S S
I O N [1:30 p.m.]
DR. COHN: Okay. The first set of presentations and discussions
on HL7/NCPDP harmonization, we have Karen Eckert and Ross Martin. Thank
you for joining us, both updating us with what is going on.
Karen, are you on first or is it a joint presentation?
DR. MARTIN: I will be starting. Karen will be filling in the
middle part of the Oreo cookie and then I will be doing my part again.
So, substance will come from Karen. I will just either — the rest.
Good afternoon. I hope I can keep you all awake and in your
postprandial state after lunch. My name is Ross Martin. I am a
senior manager of business technology at Pfizer and have been involved in some
of the coordination efforts between HL7 and NCPDP, as have some of my
colleagues, one of whom, Jim McCain, could not be here today, with the Veterans
Health Administration, and Karen Eckert from Medi-Span WoltersKluwer.
First, I would like to thank this committee in particular, the subcommittee
in particular, for inviting us here. As you mentioned earlier on about
your perceived role, the fact that you invited us to the table to testify has
done more to advance the actual efforts of collaboration than probably any
single thing because we realized when you put us on the schedule, we had to
tell you that we had done something. So, we have now something to show
you, which I hope we can go through very briefly.
This need that we are about to discuss has been acknowledged for some time
between the organizations, this need to have some degree of if you want to call
it harmonization. We have been calling it coordination at this point
because harmonization has some different connotations to some of us, but we
have noted that this has been a necessary issue for some time, but there hasn’t
been a compelling business need that has driven the community to actually put
concerted efforts together to make this a possibility.
Of course, one of those has been the path to the MMA and probably one of the
primary drivers because we know that there are many business cases where people
using HL7 centric systems are having to communicate with other systems that are
NCPDP centric. The NHII efforts and the appointment of Dr. Brailer as the
national coordinator for health information technology has certainly been a
part of that as well.
Along those lines some time ago, NCPDP and HL7 selected some liaisons to
help facilitate the communication between these two organizations. And
Karen Eckert and Jim were among the first two to do that. I joined them
this year. Our conversations around these efforts really quickly focused
on e-prescribing and the need for more activity around this and I think it was
back in April of this year, I circulated a brief draft about a proposal about
how we could go about accomplishing some of these tasks and send it to Kelly
Cronin and Michael Fitzmaurice from AHRQ and said would this be something worth
funding and one of my problems was I didn’t ask for enough money. So, it
didn’t really get on their radar screen.
If I had asked for like a serious amount of $120,000, maybe we could have
done it, but it was more like 30 or 40 because it really didn’t seem like this
was a project that needed an incredible amount of resources. We needed
facilitation more than anything else. So, we didn’t get that initial
round of funding but it certainly spurred enough conversation to make this
So, Pfizer went ahead and we supported the creation of a brief stakeholder
meeting that we held the day before the last series of NCVHS hearings on July
27th of this year. We did it in our offices and we provided breakfast and
lunch and that was pretty much the only offering that was there that was, you
know, a financial offering. We just made it possible for these people to
meet. So, here is a picture of us at the beginning of the meeting and I
can say that it pretty much quickly disintegrated into a little bit more chaos
as soon as we started talking about how we were going to put logos on one tee
shirt to show how we were collaborating. We did eventually get through
These are the folks that participated in the meeting. I briefed Jeff
that there would be a humorous picture in there. So, he was on cue to
laugh. So, at that kick-off meeting, you can see the name — for Jeff’s
benefit, I will read them off. We had Bob Beckley from SureScripts, Tom
Bizarro from First DataBank, Roy Bussewitz from NACDS, Karen from Medi-Span,
Lynne Gilbertson from NCPDP, Chuck Hauris from Cleveland Clinic, Ed Larsen, who
is an independent consultant for HL7 and HIMS, who at the NHII meeting
indicated an interest in joining us and was extremely helpful in facilitating a
lot of activity; myself from Pfizer, Jim McCain, again, from VA, who is a
consultant through EDS, Livea Reich, who is a consultant for Pfizer that I have
kind of put on loan for this project to help facilitate the coordination of
activities, things like teleconferences and such; Scott Robertson, whom you
have heard from this morning from Kaiser, who represented HL7, Phillip Scott
from NCPDP, Terrie Shepherd from Walgreens, Marge Simos from ExpressScripts,
Mark Singleton from Rx America and Lee Ann Stember, who is president of NCPDP.
So, you can see we had a bit of a heavier mix from the NCPDP side.
However, there was a reason for that. The HL7 Board was meeting at the
same time, but they urged us to go ahead and start off with this mapping
meeting to get things going. At this point, Karen is going to talk about
some of the major discussion points and I will just run this for you.
MS. ECKERT: When we met, we as a group were trying to identify what
were the major discussion points we were going to talk to and what we had
worked to was scope, resources, deliverables and inversion. Specifically
on the scope, what should the goals of the projects first phase and subsequent
phases relative to the total potential scope of the needs be? We knew the
previous collaboration attempts in the last several years between the two
organizations have not succeeded, sometimes due to too broad a scope or too
vague of goals or there wasn’t an actual business need among the participants
to include the ongoing resources needed to make the activity happen.
So, what this group did was define a very specific, succinct and achievable
goal that we could accomplish as sort of a starting point of where we needed to
go from there. We needed to achieve something first before we could
continue. Resources, we are trying to identify where will we get the
expertise and the project coordination and whatever financial needs for
the coordinations that might be necessary.
Deliverables, we need to make sure that the expectations is internally
within the group, as well as with the outside stakeholders regarding what will
the finished product be from the project. Then we also had a long
discussion regarding what versions of the two standards should we target for
the mapping and I will go into that in more detail in an upcoming slide.
What we identified for the project goal and the initial scope is to, one,
produce a guidance document outlining the typical mapping requirements between
HL7 standard messages and NCPDP SCRIPTS standard messages. Two, conduct a
working demonstration project of these two mappings that will be a result from
this guidance document. Three, we realize the standards would go along
and continue to live and evolve while we were doing this mapping and after this
mapping. So, we changed the management process from both organizations to
make sure that this guidance document has continued to be updated as time goes
Four, as we started talking we realized there were so many other things that
is collaboration could be doing, but we were fearful — we kept bringing them
into the scope here — we would never get our original goal done. So, we
are also creating a future phase document of additional work that is currently
out of scope. There could be further collaborations as time goes
So, specifically, what we are including in this project is given this
scenario, a prescriber sees a patient and utilizing their CPOE system to
generate a prescription request that will be delivered to a pharmacy for
dispensing, an example is if you are upon hospital discharge being able to send
the prescription electronically to the retail pharmacy.
We have heard in previous testimony that this is something that we thought
in the first phase would be the greatest benefit to the area was on the new
prescriptions and renewals. So, that was what we were trying to target in
this project. For prescribers that means from this scenario, a prescriber
would need to initiate a request for a new prescription and, two, initiate a
response to a refill or renewal request from a pharmacy and then from the
pharmacy side, for the same scenario, they would be able to accept the request
for a new prescription and initiate a request for a refill.
We had a long discussion regarding which versions of the standards to
use. What we settled on was HL7 Version 2.3 and NCPDP SCRIPT Version
4.2. Both of these standards have newer versions than this, but these are
the versions that are most readily available in the community in all the
We know Version 3 is where HL7 is going, but we weren’t aware of any
adoptions beyond the draft standard for trial uses that are available within
Version 3. So, we thought the more realistic scenario of what is
happening in the states is a Version 2 message to a Scripts version message and
SCRIPT again, we chose Version 4.2 because most of the changes made to SCRIPT
from this point forward have not been that substantial that you couldn’t easily
adapt what you had done from the basis in 4.2.
Additionally, there were so many structural similarities between the way HL7
and their Version 2 series of the standards is similar to SCRIPT and its
messaging and its headers or footers and how it all is structured, that it made
the most sense. Also, we realized to go to Version 3 was going to require
significant training to the NCPDP participants regarding the HL7 reference
information model and how the new standard, is it going to work?
As I stated, we know there have been minor changes to the standards, but
these were the two that were most implemented. So, for the demonstration
— again, our goal is to have two demonstrations of this project in early 2005
to be able to show the successful messaging between a hospital CPOE system and
community pharmacy. The first of that will be at HIMS in February of 2005
in Dallas. We are making that a part of the HL7 demonstration booth and
then follow-up at NCPDP at the annual meeting in March of 2005 in
Phoenix. We want to repeat the demonstration for the pharmacy community.
The participants that would be included in these demonstration projects are
SureScripts, with their electronic prescription transmission and Cleveland
Clinic, who is using EPIC as their CPOE system vendor. The pharmacy
vendor, we are still identifying who that would be, but because of the way a
SureScripts position in this model, we should have a chance to use any variety
of pharmacy vendors to illustrate that.
Additionally, the VA and their CPOE is possibly participating also, but they
are still investigating funding and on the formulary side, RxHub will be
available or will be included as a participant in this transmission and then if
time allows we may include the medication history but we didn’t want to
jeopardize the goal of having the demonstration available in February and in
March, that if the medication history isn’t available or isn’t ready yet, we
didn’t want to compromise the rest of the project.
So, now back to Ross to go through some of the more points on the time line.
DR. MARTIN: Just a note about the demonstration project as well.
We are certainly not limiting the demonstration project to the entities that we
have listed here, but we wanted to make sure that we had a critical mass that
we had at least enough to show how this works. With the demonstration
project, there is historically a fee associated with that. So, for
example, the Veterans Health Administration would not be in a position to
provide the funding to participate in that demonstration. So, it would be
an opportunity for funding, should we be able to get the whole project funded
if you will. There is also a question about how HL7 and NCPDP are going
to try to coordinate sort of a one application for being able to do this
project in both settings, in both NCPDP’s annual meeting and at the prior HIMS
demonstration in the HL7 booth.
But those details, in fact, all of these details that you see here have to
be worked out in terms of agreements and people signing on and actually signing
up to do this. But this is the intent of the group. This is what
people have personally made commitments to and based on history, we think that
their word is good. So, we felt comfortable enough telling you that that
was our intent.
Just one other thing about that. If there is anyone within earshot
that is interested in participating, they certainly can discuss with any one of
the coordinators about participating in that. So, just to give you a
broader overview of the time line of our planned efforts, by August 1, which
has already happened, we have created a workspace using Yahoo groups just so
that we can have a place to put files. Now, this is only available for
people who are actively working on a project. One of the reasons for that
is that we are dealing with fundamentally proprietary information in that the
standards organizations, we are making the standards available to all the
people in the work group. Even if they don’t belong, for example, to HL7,
but only belong to NCPDP, they can still have access to the HL7 standards,
which would normally require a licensing fee and, likewise, the HL7 members can
have access to the relevant NCPDP standards.
Until these different work products are completed, we really want to keep
them within the group, but anybody who has an interest in participating
actively in the process can join — can, again, talk to one of the coordinators
to join. Jim McCain has committed to just right around now to create this
revised project plan. I think it actually has gone through a number of
versions since our initial draft that we went through on the first day of our
kickoff meeting and has done just an outstanding job of mapping out kind of the
details of how we are going to go about this and already work has begun on some
of the mapping specifications that we have gotten a commitment from in
coordination with Bob Beckley, Jim McCain and Chuck Hauris from Cleveland
Clinic to do the mapping specifications and having a draft document ready for
At that point, what we would circulate it among our group, our active task
group, and get comments back from the group, then prepare that for
publication. An implementation guideline would be ready by the end of the
year, but before that time we would already be circulating that draft
implementation guideline to the participants of the demonstration project, so
that we would be ready to present that in the demonstration projects in
February and March of 2005 at HIMS and NCPDP respectively. There are some
other data time line points beyond that, but those were kind of as far as we
could see in terms of really nailing down some dates that we could all commit
But as Karen said earlier, this started a much broader process of looking at
where else we can go with this and clearly there has been evidence of that
already. We have already seen, for example, NCPDP at the NCPDP Workgroup
Meeting last week in San Francisco a number of HL7 members came for the first
time to NCPDP, including Scott Robertson, and not only did they come to the
e-prescribing forum that we held on the 10th, but they also stayed for the
workgroup meetings and actively participated.
This included people like the current chair of HL7, Mark Shaffernon(?) and
also one of the founders, John Quinn and a number of other people who were
there actively participated and contributed and talked about how not only could
they help facilitate some of the things where there are areas where we need to
do things like mapping together, but also how we could start learning from
their methodologies, learning from, for example, the RIM, the reference
information model, and understanding how that might apply to NCPDP standards as
well and maybe take another look at how we within NCPDP develop standards.
So, there is a lot of opportunity for future collaboration, including
mapping, to HL7 Version 3, which again has a prior step of getting the
discussions around and education around how that information model works.
I think that education is not just for the NCPDP members that aren’t familiar
with it, but, frankly, a lot of the HL7 members are not fully familiar with the
RIM yet and the various permutations of that. So, there is a lot of
mutual understanding and education that has to go on.
The people who actually implement this stuff are still very much living in a
2.X HL7 world. There is also an opportunity for us to look at mapping HL7
to the NCPDP telecommunications standard, which is what is primarily used for
pharmacy claims adjudication and more opportunities for bidirectional
medication history information exchange. As we have already discussed, it
could be a part of First One providing that enough resources are attached to
that so it doesn’t slow us down.
In terms of ongoing support, both boards have not only revised — made an
addendum to their current memorandum of understanding specific to this project,
but as I mentioned, they have also actively participated together in ongoing
discussions and we see a real fellowship, if you will, happening there.
At least the interactions that I have observed have been very truly opportunity
driven and, you know, this is in our best interest if not one is looking to
swallow up somebody else or anything like that.
We are also, you know, getting continued in-kind support from stakeholders,
including Pfizer is committed to providing just, you know, functional support
for this ongoing effort, although I do think that responding to an earlier
question of the group about how some of this overall harmonization kind of work
can go on between the different standards organizations, ANSI HISB was
Pfizer does belong to ANSI HISB. I am a member of ANSI HISB. So,
for the last year have been attending those meetings and kind of understand
what that space is about, at least on a surface level. I can see the
possibility of that being a forum, but you have to understand, one of the
initial drivers for even writing the original draft of this was I saw that as
hard as it is to get volunteers who are involved in standards to circle around
their own standard and volunteer their time, get their companies to even
understand why they are there, you know, giving away proprietary information so
information can flow, as hard as that is, it is even harder for two different
organizations to get their volunteers to volunteer to communicate with the
So, I think there is a lot of — you could go a long way to facilitating
that, you being HHS, being AHRQ, being anybody that has the capacity to
facilitate meetings. Since we all tend to be NCVHS groupies these days,
it is like following the Grateful Dead around, except it is always in the same
city. One of the realities was the only time that we could find to meet
was just before the NCVHS meeting. So, maybe that is a way we could make
We are also looking at the possibilities of grant money from — or in-kind
services from like CAFE(?) Rx. You may have seen the announcements
recently that a number of industry stakeholders have come together to try
to facilitate electronic prescribing through an organization called CAFE
Rx. Their broader intent is to do not money grants, but resource grants,
in kind. So, like, if we need a technical writer, we could probably get
one from them and maybe figure out a way to support some of the fees for the
Anybody who wants to get involved, you can join the list serve, which is
RxNCPDP@List.HL7.Org. In order to join that, you do not have to be a
member of either NCPDP or HL7. It is open to all and there are
instructions on the HL7 web site if you look on the right hand side of the home
page at HL7.org, look under “committees,” you will see a place for
list serves and then you scroll down. It is sort of alphabetical, but
this is a subset of pharmacies. So, you will find it under the
“P’s.” Then if you want to join the Yahoo Workgroup, you can
just contact one of us.
Again, I have acknowledged a lot of people in this already, but I just want
to again thank the HL7 and NCPDP board members and their staffs for supporting
all this work. As I said, the HHS staff has been — has really been
fundamental and NCVHS has been — Jeff, especially the conversations that you
and I and also Stan Huff have had about this have really made this go forward
when I don’t think it would have gone quite as quickly. So, your evil
plan to, you know, get us to actually do things is actually
working. I don’t know if I should have told you that.
I want to point out especially Jim McCain’s work from the VHA and the reason
is not just because he has made fundamental contributions to this, but the role
that VHA has been playing in this, he has gotten support from his
leadership to allow him to do this and kind of take a tax dollar and spend it
in a completely different way that is a multiple way instead of just for one
I really think it serves as a model for how some of the federal agencies can
participate in standards that goes beyond, I think, just kind of attending and
reporting back to their agencies about what is happening in standards or that
kind of role. He has really taken a role up your sleeves approach
to this and it has been extremely helpful.
Again, HL7 and just pointing out Scott’s work as an ambassador to NCPDP, I
think, really needs to be highly praised because it has been very significant
On the Pfizer side, I just wanted to point out that my manager, Dr. Steve
Labcough(?), who is a medical informaticist, as well, has basically budgeted,
has, you know, been the one to sign the checks for this because I have no money
and not only that, but he has given me coverage so that I can be out a lot to
take care — to work on these things because he really
fundamentally believes in this stuff. So, I wanted to thank him.
I just wanted to take this one moment to point out something that happened
last week at the NCPDP E-Prescribing Forum. I was moderating that forum,
the educational day, and we had Marco Ray(?) as a speaker, one of our main
speakers, and he was — his plane was late and we had this thing where we had
to rearrange our afternoon in order to accommodate him and we weren’t actually
sure if he was going to make it on time.
Through the magic of my little wireless card there, I was able to look up on
the web and see that he was 300 feet off the ground at about, what was it, 3:37
or something like that, just in time for him to be able to get to our meeting
and be the closing speaker. This is something we can do in medicine today
and for example, the handout that you have, I downloaded from Kinkos FedEx
today for the first time this morning. The program to be able to print
these off the web and send it to Kinkos and I went down the street
and picked it up.
I actually did this in the taxicab this morning on the way here and was able
to send these off for printing and pay for them and just go and pick them up,
walk in the door and pick them up, paid for. We can’t come close to doing
this kind of stuff in health care. What I have been seeing is a lot of
talk about things like if we can send a man to the moon in, you know, less than
ten years — Kennedy said we can send a man to the moon in ten years and we did
it in eight, why can’t we do that in health care?
I want to just kind of get people to throw that analogy away because I think
it is a really bad analogy. I don’t think it is the one that works in
this situation. Transforming health care is not like sending somebody to
the moon because nobody was at the moon. There weren’t, you know,
shuttles that we had to get out of the way of going back and forth. It
had never been done. So, we were doing it for the first time. This
is more like the Big Dig. We have to take a city like Boston, dig it up
— Boston was not shut down during the 35 years that they did the Big
All the attendant problems, this was really hard to do. This was
really hard to accomplish. But I lived in Boston briefly and going back
now and seeing how it is and the promise of — the reality of what it is to
have the Big Dig done or nearly done it is just profound. When they
reclaim all the space, it is just going to be a beautiful thing.
So, I would submit that this is more like a national Big Dig where we are
trying to dig up the entire national highway system and put it underground and,
by the way, not impede traffic while we are doing it. That is why this is
so hard and that is why we need your help so much.
Your role as a facilitator in this is really what needs to be there.
This is one little micro-connection among thousands that have to happen.
It is relatively cheap to do. I mean, for breakfast and lunch, we got 16
people to sit in the room all day and figure out how to get this
little piece done. We have to do that another thousand times. So, I
would just encourage you to keep supporting this kind of work and do more
DR. COHN: Ross, thank you.
I think we have a couple of minutes for questions. I know that I have
one. I don’t know if others do also.
From my own perspective, I consider this to be one of the really critical
things that needs to happen. So, I am sort of scratching my head a little
bit trying to figure out, you know, what it would take to (a) stabilize
this effort as opposed to have it be sort of an interesting experiment, where
we did that. We demonstrated that. That is wonderful. So, I
guess I didn’t really get a sense from you of what it takes to sort of
institutionalize something like this, knowing that at least by my way of
understanding we have got various moving parts and we have got NCPDP continuing
to evolve and change.
We have got HL7 continuing to evolve and change. We have got
terminologies within all of these things, which may or may not be the same,
continuing to evolve and change. I have to disclose, you know, I said I
was on the AMA CPT Editorial Panel. Well, one of my projects is to map
SNOMED to CPT and one of the things we are giving to the NLM, as you can
imagine, because they will try to keep this up and it is not going to be one of
the easier things one could ever imagine.
You need the AMA and you need lots of other people involved. But
having said that, I am not sure that I can — you described it as the Big
Dig. I am sort of not thinking of what you are doing is so much different
than what I am doing and it is great to demonstrate on March — February and
March of 2005 that you can do a Phase 1. I am obviously concerned of what
it is going to look like in October, 2005 and 2006 and 2007 and how if this
becomes an important thing, which it may, because there is some overlap of what
is going on, how are we going to keep this all aligned and assure that
the messages go back and forth with high fidelity as we go forward?
That is my question for you. Karen, you can answer that also.
DR. MARTIN: It is the perfect question. I don’t know that I have
the perfect answer, but it is the question of how to make the stuff really
useful long term. Because just as you said, we can do a project and say
look at what we did. We made this handshake happen. We know it is
possible. We already actually know this is possible because there are
instances where there is some exchange going on like this, but it has all been
based on proprietary mapping. The real question, I think, is going to be
when we can take the implementation guide that we have created from this and it
is enough of a road map for somebody who was not involved in the project to
pick it up, who uses HL7, for example, to pick it up and know how to translate
into an NCPDP SCRIPT message within their system and how to read an NCPDP
SCRIPT message when they get one.
When you asked the question earlier about where should the translation fall,
one of the things that came to mind for me was when we used to do the space
station visits with the Russians in MIR. We would take the shuttle up and
visit with them. They had made an agreement to — the Americans all spoke
Russian and the Russians all spoke English. That is kind of the courtesy
of, you know, you go to your lowest — your own lowest common denominator
if you will. So, I would think that in — the general courtesy would be,
if you will, for the translation to happen on your end so that when you send a
message to somebody, they are hearing what they are used to hearing.
When you send it back to somebody else, you know, you can translate it out
as well. But, again, that can be done through an intermediary.
These are the — once we have shown that somebody can take that implementation
guide, though, and do those translations, that, I think, is really what
the role is of these harmonization efforts. That is my opinion, because
these will always be especially in the not at all locked down world of 2X,
which is out there and will be out there for some time to come with HL7, there
is an awful lot of variability within those — within the implementations that
you can only give guidance. We almost — I can’t remember if we even
decided if we were going to call it an implementation guideline or if it was
going to be called implementation guidance. Because we felt that there
was no way you could concretely say that this is the way to change this message
because everybody’s implementations internally are different.
Now, with Version 3, we have the opportunity to have a mapping standard out
there before it is ever in widespread use. So, you kind of avoid some of
that variability challenge that we are all facing. But, again, for the
reasons that we have already discussed, that Karen already discussed, we chose
to start with this kinder, simpler, gentler task of doing the 2.3 to 4.2
MS. ECKERT: I guess one thing I would add to that is we have heard —
and Scott and Lynne, correct me I am wrong — agreements that this document
would then be maintained and available through the separate standards
organizations. So, I would see any future revisions as people are
implementing it and changes are needed or we add a new — create a new version
of SCRIPT, we would come back and maintain this document as another work
product of that workgroup. So, we are doing the work now, but after we
get done, we are going to hand it back to the standards organizations and it is
something that they would continue to maintain going forward as versions change
and are updated.
DR. MARTIN: I think that is actually the more important work product
of this workgroup was not this little mapping exercise. It is the process
by which from here and, you know, for days to come, HL7 and NCPDP will help —
will do these — will create guidelines for translation and will in an ongoing
fashion support them.
DR. COHN: Jeff, I think you had a question.
MR. BLAIR: Thank you, Karen and Ross.
The question that I am about to ask, it might be in the handout that you
have given to me. So, if it is, just wind up saying it is there.
But in this initial meeting — I think you actually had two meetings, didn’t
you, or at least, you know, one before — what was it, the 27th or 28th of July
and then didn’t you meet again, you know, what was it, last week or something
DR. MARTIN: Well, last week was an NCPDP, normally an NCPDP workgroup,
where a number of HL7 people came, but like, for example, the board members
showed up and they went out to dinner after and had conversations, where they
also received a report about this activity as well, jointly.
MR. BLAIR: When you went through these sessions, are you able to give
us some feeling of the areas where you can do mapping, the data elements where
you can do mapping, those which don’t require mapping because the environment
is simply, you know, different or it is not needed and for those where you can
do the mapping, some feeling for what portion of that might even be ready
for January 1st, 2006, to be included in demonstration projects if you think
that that is needed.
So, I guess that is three questions. Should I repeat them?
DR. MARTIN: No, because I am going to punt it. You will have to
ask the person I am going to punt it to and I think he already knows it because
he is closing his computer and he is starting to get up here. I would
like Scott to address that. I think he has already answered that question
MR. BLAIR: If he sounds the same as you, I won’t know the difference.
MR. ROBERTSON: This is Scott Robertson just so people on the network
In further demonstration, we have generally agreed that there are some
core messages and data elements that must be there to demonstrate the essential
product of moving a prescription order, prescription from one entity to
another. I think from the HL7 side, and I am pretty sure the group agrees
that those core elements exist on both sides. Mapping to them is going to
be relatively easy. Some of it actually is already in existence from some
of the members. They are bringing that forward so that we don’t have to
spend time doing stuff that has already been done.
So, that in the very near term, we have a very focused set of pieces that
need to be mapped and we know we can do them. Going beyond that —
MR. BLAIR: That is kind of what I am looking to hear a little
bit. What are the areas where you feel you can do the mapping?
MR. ROBERTSON: Well, first of all, functionally a new prescription
order and just take these as functions right now, not any detail underneath
them — a new prescription order, a refill authorization request and a refill
authorization. So, that is new order going to the pharmacy, the pharmacy
asking for permission to fill additionally and the prescriber acknowledging and
answering that refill response.
MS. ECKERT: And, Scott, specifically we are talking not just the drug
portion but also the patient identification —
MR. ROBERTSON: Yes, I am going to — within those functions, we need
to identify the person. We need to identify the product. We need to
have directions. We need to have a quantity. The new order has to
say how many fills.
MS. ECKERT: Provider and location.
MR. ROBERTSON: The pharmacy it is supposed to be filled by or that
filled it. So, basically all the essential elements that would have to be
on a piece of paper if a patient received that paper from the doctor and took
it to a pharmacy.
Now, we can go well beyond that and start incorporating things like other
medication history, other known conditions or the indication for use, all these
other things that are supported by SCRIPT are not in this first pass because,
frankly, there are more and more issues. As you go further and further,
you start getting into more and more terminology issues.
Right now, one of the terminology issues that we are just going to sort of
let sit right there is identification of the product is basically going to be a
string name that is passed along, that will have to be interpreted on the other
side and how that interpretation is done. Probably going to end up by
visual inspection just because the terminology doesn’t exist right now.
In the grand scheme of things there is — potentially everything that is in
NCPDP SCRIPT could be mapped into — mapped between NCPDP and HL7. Being
part of the group that is established and maintains the pharmacy messages in
HL7, there are some things in HL7 pharmacy messages that will not be mappable
into SCRIPT and they wouldn’t want them to be mapped into SCRIPT probably, but
that is well beyond the scope that we will have to deal with in
e-prescribing. I hope that is beyond the scope that we will have to deal
with in e-prescribing.
DR. MARTIN: I think this one domain really does speak about the need
for a universal translator like RxNorm, just as we have talked about between
drug knowledge databases when you are talking about system to system
communication, that universal vocabulary nugget can be very useful. It
will be good to see what those specifications are about what is required to
incorporate that into a message. It would be good to put that in.
MS. ECKERT: I guess one thing I would just caution, though, on that
note is if we are still looking at the area of low hanging fruit, including
requiring the RxNorm in the messaging and requiring the RxNorm in people’s
systems to be able to decode what that is, our new requirements that we would
have to take into consideration to the standard and would need to be
reimplemented at all the end user sites and so we would need to go back to some
of that ROI discussion that Lynne had earlier to make sure that there is a
value proposition for all the system vendors for both CPOE and pharmacy to be
able to exchange and use that code set.
That is the caveat to remember.
MR. BLAIR: Let me just add one follow-up question — I am sorry.
Did I interrupt and they were not done?
DR. COHN: No, I think you can go. I was going to try to move us
onto the next issue here pretty soon. You and Steve have one question and
we can wrap up.
MR. BLAIR: This is really kind of expanding a little bit on Simon’s
question, which is what needs to be done to make sure that this momentum keeps
going? I think I heard from you all that you know Pfizer and some others
have stepped in to provide some initial funding. So, is it just funding
and facilitation or does NCVHS or HHS need to do something else to make sure
that the sense of urgency is maintained, the momentum is maintained?
What else besides the funding is needed?
DR. MARTIN: I believe that facilitation is the key. The way I
would define facilitation is project coordination where you have got to
understand, people like — experts like Scott Robertson and Bob Beckley and
Karen and Lynne and all these people, who really know the space that they know
very well, to have them also be responsible for arranging the meetings, writing
up the minutes, writing the technical documents after they have done the work,
but it needs to be cleaned up. Those kinds of things in my view are a
waste of their time. It is not that they are above it. It is that
we need their resources a hundred times over. We want them to be doing
exactly what they are very good at doing and everything else being done by
somebody we can pay to do.
That is where I would say respecting the volunteer’s time is the way that
you can facilitate more than anything. So, grants that provide for
technical writers, for project coordinators, for, you know, people like — I am
using Olivia Wright(?), for example, as a consultant. One of her tasks is
to be the whipping person to keep — that, you know, you told me you
would have this in next week. It is due next week. She does that
for me and she does that for everybody else in the group or if she hasn’t
started doing it yet, she will. That is the kind of thing that we need.
MS. ECKERT: However, I think there still was some discussion about
possibly needing grant money because of some of the training that needs to be
done and the facilitation of getting the trainers in and the logistics of
getting participants from both organizations there for the mutual training,
especially as they go towards Version 3 because that is a little more difficult
Most of what we have done so far has been in kind and things that were
supported in our organization. That is something that is going to take
some more resources and monetary resources that I am not sure how we need to go
MR. STEINDEL: Simon, I basically have two comments and the first
comment is more to the audience and based on something Ross said. How do
you feel like being classified as deadheads?
The second comment also concerns Ross’s analogy of this project, being
related to instead of a flight to the moon, the Big Dig, which has one of the
largest cost overruns anyone has ever seen. I have heard a lot especially
in response to Simon’s question, how are you going to maintain this? In
response to Jeff’s question what has already been mapped and I am still hearing
a lot of things that can lead to these cost overruns.
What I am strongly suggesting is that I still think there needs to be a lot
of ground work laid out as to what exactly this mapping is going to be doing,
how the organizations are going to maintain it and what is the definition of
the result of this mapping?
So, I feel we have made some great strides forward and I, you know,
appreciate that the NCVHS was part of what started this effort, but I think it
is just a start and we need to avoid those cost overruns.
DR. COHN: I think maybe we will finish up the session by just noting
— I think what we are all talking about is moving this from project status to
institutionalizing it. I mean, that is at least the way we think about it
within my organization, where it is not a unique sort of thing, but it is an
I guess I would say if you or others at NCPDP or HL7 have a view of what
sort of resources, you know, would be useful, that would help move it from that
in those statuses, I think that that would be very useful and would be
certainly very timely, not that it might — I mean, who knows where it might
show up, but this would be a very good time to come forward with those sorts of
I do know that HL7 and NCPDP historically works on a volunteer basis and all
of that. I realize as we talked through all of this, we are asking a lot
from all of these organizations. MMA isn’t even us. We are asking
you more specifically because MMA is asking all of us very specifically to
come forward to the challenge, but I think that there is an opportunity
here to be successful in all of this and not just as a single demonstration
DR. MARTIN: I think the volunteers who are giving their time are doing
it for a number of reasons, some of them being that it is in their
organization’s best interest to have them — they see the business case.
But I also see a lot of passion and I don’t think you need to necessarily
reimburse every single person that is involved at all. It is just making
sure that they are not doing any of the work that could be done in a
facilitative way, which does happen in the standards organizations today
through staff and Lynne Gilbertson is a great example of that, where, you know,
she attends these meetings. She is paid to — this is her job and she
comes and facilitates. She makes things happen.
When you don’t have that, you need that person in the middle that is doing
that is doing that both organizations as well.
DR. COHN: Absolutely agree.
Okay. Russ, thank you very much and we will really appreciate your
Our next session is on terminology standards. I thank Randy and Stuart
and Steve for joining us; Randy Levin from the FDA, Steve Brown from the VA and
Stuart Nelson from NLM.
Randy, I believe you are up first. Please.
DR. LEVIN: I am going to provide an update on the FDA medication
identifiers and terminologies for electronic prescribing. I am going to
talk about four general areas. One is medication identifiers from the
FDA, medication terminology, medication information and information exchange
So, with medication identifiers and talking about ingredients and drug
products, terminology, talking about product ingredient names,
manufacturer dosage form, route of administration, appearance, package
types. For medication information, talking about labeling contents and
for information exchange, talking about structured product labeling.
A lot of the work is being done in collaboration with a number of different
groups. One, the National Library of Medicine, which we are working with
people in the National Library of Medicine. Our point person is
Stuart. For the daily med, for exchange of information and the FDA
supporting the RxNorm. Also, collaborators, the people at the Department
of Veterans Affairs and Steve Brown has been our contact person there and
working on medication and information terminology standards.
Then the Agency of Healthcare Quality and Research for Project Support and
Funding and Mike Fitzmaurice has been our contact there. Also the drug
manufacturers, repackers, relabelers and distributors for providing information
on their medication and their products, health information suppliers, who will
provide the user interface for any medication information that we distribute
and Health Level 7 for working on the SPL standard and for work on vocabulary.
First, medication identifiers, we have three types of identifiers for
ingredients; one, the active ingredient, one, the active moiety and one is —
the last is the inactive ingredient. The active ingredient we have
defined as the substance responsible for the effect of the drug product and we
have defined it as either the complete molecular structure or if the precise
marketing structure is not known by an unambiguous definition and we apply for
that, the ingredient, the unique ingredient identifier for the active
ingredient. Here is an example, amoxycillin is an active ingredient and
the uni-code is there for amoxycillin.
The availability of this for electronic prescribing, we have already
actually generated a number of these codes and have shared them with
Steve. I think Steve will talk a little bit about that. We feel we
will have the active ingredients for most of the approved prescription drugs by
early 2005 and then for active ingredients in all of our marketed products,
somewhere in the next three to four years, this is dependent on changes to the
FDA regulations so that is one caveat to that.
Just to show you about — an example here is this is amoxycillin. This
comes from our substance registration system, which we use to generate the
unique ingredient identifiers. The structure is drawn in and then the
name is given and the code is applied.
For active moiety, this is the portion of the active ingredient responsible
for the effect if it is applicable. Sometimes the active ingredient is
there is no active moiety. It is the active ingredient itself.
Sometimes without the counter ion is the effect. So, here in this
situation again the unique ingredient identifier is used and amoxycillin
anhydrous is the active moiety and has a separate uni-code and, again, the
availability is the same as the active ingredient.
Here is the example of the amoxycillin anhydrous. The only difference
here is the trihydrate is eliminated from the active moiety.
The inactive ingredient is all the components of the drug product, other
than the active ingredients and they are given their own unique ingredient
identifiers. An example is titanium dioxide and here is the uni-code for
titanium dioxide. We will have the active ingredients for many of the
approved prescription drug products late in 2005 and for implementation, we
should have most of them available in three to four years, depending on our
Again, here is the uni-4 titanium dioxide from our substance registration
system. We have two identifiers for drug products. One is for the
actual drug product and one is for the package drug product. The drug
product is defined by the active ingredients, the strength, the manufactured
dosage form, the inactive ingredients, the specific appearance of the dosage
form, the proprietary name, the manufacturer and distributor.
A drug product may contain two or more drug products in the same package,
for example, oral contraceptives is a common example of that. We
currently have an identifier for the drug product. It is the NDC labeler
code, plus the NDC product code. That is the identifier for drug
product. An example of a drug product is amoxil(?) 500 milligram tablet
and the code is — here is the code for the amoxil 500 milligram tablet.
Most of the prescription drug products already have these codes. They
are in our national drug code directory and they are also — and a number of
other third party vendors have these codes as well.
Again, implementation, we figure all marketed drug products were aiming for
in three or four years, and dependent on our ability to change the
regulations. The packaged drug product is the drug product that was just
described, plus the packaging and the drug product quantity. This
identifier is the labeler code from the NDC, the product code and the package
An example of that is amoxil 500 milligram tablet in a bottle of 100.
Here is the code for this package drug product. Again, we have a lot of
these just as a drug product. Most prescription drugs are in our national
drug code directory, but all marketed products we are looking for in three to
four years, again, dependent on regulation changes.
DR. COHN: Can I just ask a clarification? I thought the NDC code
was already out there and was accepted as a HIPAA standard. Is this
different than that?
DR. LEVIN: No, this is the same. Just to show you that the
different components of the NDC, because for the drug product, the first two
components of the NDC are distinguishing the drug product. But this is
the NDC. That is currently available.
DR. COHN: It says for implementation three to four years. That
is what I was a little confused about.
DR. LEVIN: What we are talking about here is that there are currently
a number of problems with the national drug code as has been brought up by
previous discussion and what we are looking at is if there are changes in the
regulations, we would improve the national drug code as it exists. That
would take us a number of years to implement that change.
But there are codes available for a pilot if need be.
Now, here is the relationship of these different codes, the active moiety
being a part of the active ingredient. There can be one or more active
ingredients in a drug product. The drug product, we add the strength of
the active ingredients, the manufactured dosage form, the inactive ingredients
and the appearance. That makes up the drug product.
Add in the package type and quantity and you have the package drug
product. In between the active ingredient and the drug product is an
intermediate area and this is where RxNorm will fill in the gaps between those
two areas and that by supplying this kind of information to the National
Library of Medicine, who supports them, be able to generate the correct form.
Stuart will talk about that later. So, this is the identifiers from
FDA with the gaps filled in by RxNorm. Medication terminology, we have
ingredient name, established drug product name, proprietary drug product name,
manufactured dosage form, route of administration, strength, appearance and
The ingredient and product names are available. Most of the
prescription drug products are in the national drug code directory. The
same issues that we talked about just a moment ago with having all the marketed
products would be up to date and comprehensive. That would take those
changes that we talked about.
The manufactured dose form, the way of identifying the drug product in its
physical form, for example, tablet or powder for suspension or injection
solution, those are all different manufactured dose forms. They are
available currently in the FDA’s data standards manual. We are looking to
harmonize this as time goes on within HL7, plus working with the international
regulators to harmonize that across different regions. So, we will be
working on that over time, but there are — we do have manufactured dose form
standards available now.
Route of administration, that is the path by which a drug product is
introduced on or into the body, such as oral or sublingual. These are the
same as the dosage forms that we currently have them available in our data
standards manual and we are looking to harmonize these as well.
Strength is the amount of the concentration of ingredient in a unit of a
drug product, for example, milligrams per tablet or milligrams per
milliliter. This is also available. We have standards in the FDA
data standards manual and, again, we are looking at harmonizing these standards
in with other groups, including the international regulators.
Appearance, there are no current standards for appearance, color, shape size
coding, scoring, logo, imprint codes that are on drug products. This is
with solid manufacturing dosage forms though. We are looking at that
through the HL7 organization. Package type, the container, which the drug
product is enclosed for delivery or display, such as a bottle or a vial, again,
this is available with the FDA’s data standards manual and, again, we are
looking at harmonization.
DR. COHN: I will apologize. We are just sort of asking ourselves
what do you mean by “harmonization”?
DR. LEVIN: Right now we have our own standard for these things.
HL7 has — they are working on looking at the way that different people handle
package type or route of administration doses format. We are looking to
try to bring them together that — we have been working with the Europeans and
the Japanese on how they describe route of administration and try to map
between the two.
MR. STEINDEL: Just a clarification, Randy. Map or the same?
DR. LEVIN: Well, first we are doing a mapping. Then we are going
to try to see if we can bring them to be the same.
DR. COHN: I guess I was just surprised that yours and HL7’s are not
the same given that we are in the same country. I guess I shouldn’t
be. I retract that comment.
DR. LEVIN: Okay. Medication information, this is the labeling
information. The sections, how labeling is divided and the content, we
are trying to standardize on this as well. For example, indications in
usage is a standard part of the labeling or contraindications. We are
taking the labeling and trying to structure it into this structured format and
more into the structured format.
We will have a limited number of approved prescription drug products in this
format late in 2005. We expect to have all the approved prescription
products one year later and then for all marketed products within a three or
four year period, again, depending on the change in FDA regulations, but this
would be that take all labeling content would be in a structured format.
Structured format would be the structured product labeling, which is an HL7
standard that Scott talked about before. It is an XML format with
computer readable tags that devise the labeling into — the content labeling
into sections that are defined by our regulations and includes all the
medication identifiers and terminology standards that we just talked
about. They will all be included in the structure plaque labeling that
are part of the drug description and how supplied section of labeling as it
currently exists. These will all be provided as data elements and passed
on to the National Library of Medicine for distribution in the Daily Med and
for support to create the RxNorm.
In the future, we are looking at providing more information based on new
formatting regulations for structured elements, including things like mechanism
of action, drug interactions food interactions, adverse events, indications and
dosing. Some of these terminologies for these things will be taken from
the work being done at the VA for their terminology.
We have an SPL, structured product labeling, implementation guide for
submission of the FDA content of labeling. This is for pharmaceutical
companies submitting to the FDA and an SPL style sheet, a standard style sheet
for viewing the labeling information.
Here is the XML piece, but one rendered in a style sheet just looks as the
usual package — what you see in a package insert. The FDA data standards
manual is on the FDA web site. I give the web site address here at
FDA.GOV/CDER/DSM. The national drug code is also on the CDER web site at
the NDC/database default.XTDM data set.
Our data standards are also in the consolidated health informatics web site
and structured product labeling information is on the HL7 web site including
the schema, implementation guide and style sheet. The style sheet will
also be on the FDA web site as well.
That is the information I have for you today.
DR. COHN: Randy, thank you.
Steve, you are up next.
DR. BROWN: Thanks for having me. I am going to give you just a
brief update on where we have gone with NDFRT since the last time we talked,
which I guess was last fall or something around those time frames. So,
what we have been doing within VA and beyond VA, as well as the vital
statistics of NDFRT.
For those of you who haven’t heard me this time after time, NDFRT is a
formal terminology. I use it for description logics that we have used of
VA National Drug File as a baseline to build, which is interlinked with
terminologies from a number of other government organizations and at least one
standards development organization with NLM and RxNorm and MESH(?), for
example, and a number of coding systems from the FDA, including uni-codes, as
Randy alluded to.
We are now more closely with the National Cancer Institute and also the HL7
dose forms. In terms of recent development, NDFRT has now published in
the AB metathesaurus, so that is all with where do I get it type of questions
are now reasonably well answered and it is derivable from that. Also, if
you like to browse in the FRT, NCI has it up as part of the browser, NCI terms
.NCI.NIH.GOV. So, it is easy to go and have a look at the contents of it
as well. Those are reasonably updated, reasonably well updated, but not
the most current developmental version or whatever, but still pretty good and
you get a real nice idea of what is out there.
We have been in a couple of areas on additions and the first has to do with
the internal VA needs and a lot of implementing any change in terminology or
anything within an organization like VA, it is really Big Dig. It is a
wonderful analogy. We have NDF is in place. All the medical centers
are up using it. We are doing — it depends on how you count but 75
million prescriptions a year and all that kind of stuff with at least 128
business systems and the like. So, you don’t just change things.
There has to be an organized path and even if there wasn’t just an organized
path, I mean, there is a lot of work to do to alter the software and the like.
One of the first things I can mention is that we spent a lot of time adding
properties that were in the national drug file that we didn’t initially have an
NDFRT that were really of pharmacy interest rather than pharmaceutical
interest. So, the national drug file DSS number. You know, it
matters when you are running however many hundred pharmacies it is that it can
get rolled up to the national level and the DSS is our decision support system
and database in Austin or whether it is on the national formulary
indicator. It is not of particular interest to this audience but it is
still one of those kind of blocking and then tackling, to use a Mark
Tuttlism(?) kind of thing that we just have to have.
So, we have spent a lot of time doing that. We also have been
finishing up with some additions for terms and concepts related to
pharmacogenomics. There was a grant from NIGMS through Mayo that had some
concepts and do some modeling with the FRT to try and make additions that would
help the researchers in the PGRN. So, we have been working on that as
well. So, we have a sample of genes and alleles and enzymes and those
kinds of things. It is a relatively small sample, but it is a whole
different world of modeling.
Within the VA, in addition to having sort of more concepts, we have process
integration in order to make all this happen and there are several things that
we have been working on. Something that we call the new drug transaction,
so a transactional — I will show you pictures of. As we try to — this
is part of an effort to nationalize terminology that has previously been
done largely in a more distributed way and there is a whole lot of things that
you have to do if you are going to — oh, you can’t just put something in the
local file and make your provider happy kind of stuff.
So, again, roll up your sleeves, but real practical things. We are
working on maintenance environment that will help the fellow who is working
with the national drug file now to do a better, faster job using NDA forte(?)
and also we are actively participating as you have probably heard of the VA is
reengineering a lot of its software and we are calling these all different
reengineering projects. There is lab reengineering and pharmacy
reengineering. So, we are actively working with them.
So, what I am trying to tell you, I guess, in short is that we are into the
integration and implementation of this, as opposed to real changes in the
concepts and their interrelationships. New drug transactions, real
quickly, we have in an XML update, we have gotten data from FDA, can parse it,
bring it into the development environment and have it either human reviewed or
where it is a slam dunk, not to try to bring in updates to NDFRT.
This is an example of how that might look. This is modified software
from the development environment where we can actually take in a series of
transactions and pump them in. So, it is not enough just to build it but,
obviously, maintaining it from a variety of authoritative sources and that is
the purpose of that.
New term, rapid turnaround, again, as I mentioned, we have to deal with
field requests for terms and we can’t pretend that we are going to think of
everything first nationally when there is 128 business system implementations
and thousands of providers, who have their needs and their studies and all
these sorts of things. So, we will absolutely not succeed in implementing
if we can’t meet the needs of the folks who are actually providing there in the
So, this is an attempt to accept requests for terms from the field, process
them and move them through and distribute them out to the rest of the
facilities. So, not terribly exciting but that is how we have been
occupying a large portion of our time of late.
In terms of just the sort of general vital statistics, I have come and I
have shown you these things before. Of particular note probably is MOA
and P — these have changed in the year slightly. You see MOA is 256
elements. A year ago it was 232 and P is 1699 and that is 24 higher than
a year ago. So, what I said is there is some change as you might
expect. We have made some improvements and there are some new things, but
for the most part, you know, a year ago it was pretty well set and we have been
focused in other areas and trying to show that it continues to evolve in a
controlled and not terribly surprising fashion.
We have done a fair amount more modeling in a couple of areas. We did
some basic contraindications, you know, don’t give beta blockers in folks with
emphysema kind of stuff, not the very detailed kind of work that knowledge
vendors supply. This is much less than that and I don’t see any way — I
don’t envision this supplanting that in any way, but it was just kind of the
simplest kind of things to do that we have already had modeled in NDF.
The numbers are not usually different from a year ago. Roles,
properties, we have added a few things that have been mentioned, working with
FDA. We have lots of properties that document linkages from the
FDA. We have done updates from NDC databases and as you can see,
these are all the terminologies that Randy mentioned. We do have a draft
of uni-codes. About 1,100 of those are now mapped into the
products. That worked very nicely and we had to change the model around a
little bit and get it uploaded, but it worked fine and, you know, as they come
in, you know, we now have a place to put them and a process for doing that,
which is good.
Obviously, we are working with NOM, lots of RxNorms, drugs and
components. MESH, we are still using. Another thing that you hope,
we hadn’t shown you before is we are working with NCI. We are just
starting to get that going. When we did this immediate snapshot, there
were a few preferred terms in the areas of genes, but that is just a matter of
putting in about 1,500 more of those, based on some review and work that NCI
has done with us.
So, I put both in terms of just to be fair as a media snapshot and real soon
in future work that they have keyed up nicely for us. I included these
just to give you an idea of the kinds of routine, boring stuff that we need to
put into keep — to run the pharmacy systems. What this doesn’t really
include is that we in NDF have a number of supplies and things like diapers and
all that really have no place in NDFRT. These are different things and
this is a problem that we have to solve.
NDF is used for a lot more than drugs just because that is what could
happen. So, I guess that really gives you the brief fly by and where we
have been. I think we have made some relatively small number of
conceptual changes because that was reasonably well mature, even as of a year
ago and are working primarily towards implementation.
With that, I will take questions.
DR. COHN: Steve, thank you. We will let Stuart talk and then we
will do questions and discussion.
DR. NELSON: The good news is I have a very short presentation.
I have submitted written testimony on the spreadsheet and comments about
both Daily Med and RxNorm already and I don’t want to repeat my written
testimony because I think it is just a little too redundant. But I did
notice several times in the spreadsheet references to the Daily Med and I
wanted to be sure that we were all clear about what the Daily Med was.
The Daily Med is expected to be a system hosted at the Library that contains
the structured product label, the RxNorm name, the VA NDFRT information,
possibly links to other sources, such as PubMed and the literature, that it
will — I think the core to it is really going — it is going to begin with
structured product label from the FDA.
So, you heard from Randy just a moment ago that the first ones were going to
come out late 2005. Well, we will be ready and waiting, we hope. It
will be both browsable and addressable from a variety of places at the Library,
as well as downloadable for people who want to download the entire set of
— it is actually Gunwagon(?) in Norway, but — that is my desktop but it is
not what I am showing on my screen. Let me just exit out of this quickly.
MR. STEINDEL: I have essentially the same question but just slightly
different for both Steve and Randy. I will go with Randy first.
You showed a slide where you had a list of terminologies that are available
from the FDA. I want to preface the question to both of you. Right
now, from a certain point of view, your client is CMS and the role is their
role out of e-prescribing under MMA.
A secondary client is the private sector in using an e-prescribing.
Out of those terminologies that you showed, what are the priorities or what is
needed for e-prescribing under MMA and what really we can either put off or
isn’t needed? This is referring to your terminology slide and, Steve,
this is referring to NDFRT in general.
DR. LEVIN: What is needed for the e-prescribing, that I go back to the
people who designed that system. What do you need from those
MR. STEINDEL: So, what we can take from your message is the target
dates that you gave. CMS could go through that laundry list and say,
well, when we define e-prescribing application, FDA will be ready on such and
such a date. Would any of that be ready for piloting before the date that
you indicated in a smaller scale maybe?
DR. LEVIN: A lot of this terminology is already available now so that
they would be available now. That would be things like dosage form and
route administration and strength units and all those are available now.
The product codes and the package product codes are also available now.
So, all those things will be available.
The only things that is not available right now are the unique ingredient
identifier codes. Those will come out early next year, at least for
MR. STEINDEL: Which is probably within the time frame that CMS is —
DR. LEVIN: If you have a certain limited number of products that you
are interested in filing, we could come out with those. You know, it is
possible we could come out with those ingredients earlier.
MR. STEINDEL: Thank you.
Steve, do you have some comments on that? Is it the same type of
answer? If you give me the 1, 2 list, you can pick from it? You
tell me what you want and you can pick.
DR. BROWN: The answer, I guess, is that it is out there in AD and
there are slightly changed different versions that we could get out, but it is
largely there and available. So, it is, you know, asking me how I can
solve your problems, where are your problems?
Just to be fair, I mean, it is there and, you know, I will show you mine and
however that would be useful, great, I would say. There are probably a
lot of things there that I am not under the hood on that I couldn’t comment on
MR. STEINDEL: Okay. Thank you.
Stuart, I think you are ready?
DR. NELSON: Yes, I am back at it.
The core of the Daily Med is going to be the structured product label and
that will not be available until late 2005, but it will —
DR. COHN: Stuart, will you get close to the microphone or are you —
DR. NELSON: I have got to sit where I can see what my slides look
like, too, to remind myself of what I am supposed to say.
I think I have said everything that is on this slide. It will be
centralized around RxNorm names and the structured product label will have the
NDFRT, will have links to the literature and it will be both browsable and
Go back to RxNorm. I am not going to repeat what RxNorm is all
about. I think I have tried to bore you enough with that a number of
different times. I will point out that there is what we hope is
reasonable documentation available on the web, at the Library, which gives
further elucidation if there are any kinds of questions about the model or
I wanted to remind everyone that RxNorm was really built to facilitate
interoperability between vocabularies so that we are trying to address some of
the problems of people having different kinds of vocabularies, proprietary
vocabularies from pharmacy knowledge-based vendors and federal terminologies as
well, so that we are really trying to facilitate that interoperation.
The way I got into the drug vocabulary business was by going to one too many
HL7 meetings and listening to people come and talk about how they wanted this
problem solved and starting to take various kinds of stabs at it. We are
closely linked and will remain to be closely linked with unified medical
language system, which we see as an important vehicle for interoperability of a
number of different vocabularies and of particular importance these days are
the clinical vocabularies that might form the basis of an electronic patient
Where do we stand? Well, I have the documentation available.
Very soon we will see the first updates to RxNorm come out separate from the
UMLS. Our model is that the UMLS will only be released probably on a
quarterly basis but for RxNorm to be worthwhile in a clinical environment, it
has to be available much more frequently than that.
We are going to start off and kind of do the shakedown cruise with monthly
updates and then go to a weekly update model and we are about ready to
implement that system. So the interim releases will be insynchronous
using the same identifiers that are in the extant version of the UMLS, which
leads to an interesting problem, one that happily I am familiar with because of
the medical subjects headings of keeping a vocabulary that is updated
frequently synchronous with one that is updated less frequently.
Where we stand in terms of content right now, it is very hard to measure
completeness, but we have some — I have some interesting measures here.
I think that we have more than 95 percent of all prescription drugs currently
available with an RxNorm ID. We have 110,000 different concepts. I
have shown you some of them, the generic — the ingredients, generic and
branded, the clinical drugs, which we define as an ingredient, a strength and
an orderable dose form, something different than what Randy was talking about
when he talks about manufactured dose form.
Forms, which is something like diazepam oral tablets without specifying the
strength. Recently, Appelon(?) took the version of RxNorm that was
released with the 2004 AA version of the UMLS metathesaurus and mapped it
against the thousand most common prescriptions that were written in the
Department of Defense in 2002.
Out of that, out of the top 1,000, 974 had an RxNorm form, an ID associated
with it. The interesting ones were, of course, the 26 that didn’t.
An analysis of those I will talk about a little bit in the gaps and
limitations, that — there were two major things. One were the kits,
things that up until now have not had RxNorm IDs and the other being the
orderable supplies, whether they be durable or consumable. Things like
glucose test strips show up in both the VA and the DOD as one of the top 1,000
most common prescriptions.
As far as consistency with other standards is concerned, we have focused on
interoperability of vocabulary, but clearly one of the very important things is
consistency with messaging standards and utility in messaging standards.
Based on the discussions that we had with Lynne Gilbertson last week, we
believe that RxNorm could be a representative content of a drug to be
prescribed within the NCPDP messaging standard.
That is a preliminary first inspection and one of the — she suggested and
we heartily agree is to do a more extensive of that with, say, a hundred to two
hundred commonly prescribed drugs that demonstrate just exactly how that would
be done. I would point out that RxNorm doesn’t address some of the
problems with the content of a prescription, including things like the SIG or
the routes of administration, how something is supposed to be given. It
only talks about the form that it was made to be — for which it was made to be
When I start talking about — when I start thinking about the content of
prescriptions, I think about the 20 years I was in practice and the kinds of
things that I used to scribble down on a prescription pad. I think it can
all be fairly characterized into — there are two comments. One is that I
ordered drugs, clinical drugs, and I ordered supplies. In the drugs, I
think that there were three major types of prescriptions that are written
There are those for what I would refer to — what we refer to as clinical
drugs. There are those for things that we would refer to as — that I
refer to as drug delivery devices. Based on a meeting recently where I
was told that drug delivery devices are things like IV pumps, maybe I should
call them drug dispensing devices and those include things that have more than
one clinical drug in them, such as sequential oral contraceptives, packs in
kits and those sorts of things where — and it is interesting because the
nature of the order there is to — the prescription, the quantity to be
dispensed is the number of packs, not the number of actual ingredients or
actual clinical drugs.
The third type, which I never wrote, but some doctors still do, as I learned
when I went to my dermatologist and she gave me a written prescription that
said go to this pharmacy and fill this prescription, are the compounded
medications and I think that those are an interesting thing when you talk about
I understand from Lynne that NCPDP does not accommodate those and one of the
reasons being that state laws are quite variable about compounded medications
require a paper copy of a prescription before they will do a compounding.
By the way, out of a 26 of the thousand most common in the Department of
Defense, most of those were either drug dispensing devices, like the sequential
oral contraceptives or they were the suppliers like the insulin syringes and
glucose test strips and alcohol prep pads that diabetics use frequently.
There are about 2,500 of the drug dispensing devices. Up until this
point, we had not put RxNorm IDs on them, but it is quite clear that we ought
to and we intend to.
In the next versions, you will start seeing the codes on them. We have
not addressed — what we have addressed in a certain way, the level of
equivalence between a brand name and the generics, which contain the same
thing. We have not really tried to represent what is in the orange book
of therapeutic equivalence that is published by the FDA. That is
something that is under consideration about just what we would do and how we
would do it.
Another area that has not been really discussed very well, we don’t really
know quite yet how it should be addressed are these special ingredients, where
you say, you know, this has to be tartrazene(?) free or this is sugar free or
alcohol free or whatever, lactose free.
I suspect that there is a very small list of those particular types of
ingredients of special interest for a variety of reasons and the Library is
willing to undertake — to compile such a list, but I think it is beyond the
sources of information that we have to be able to say that a particular drug
does or does not contain this ingredient, which is otherwise inactive, largely
because we don’t have any good source of information about that.
But Randy is going to change all that for us, aren’t you, Randy?
Similarly, we are willing to do what we can to support efforts for the kinds
of supplies that are ordered on a prescription form for the common types of
supplies that are orderable, that are reimbursable. I don’t think we want
to get into every possible type of collection, you know, jobe(?) stockings and
so forth, but things like glucose test strips and alcohol prep pads and so
forth, I think we could support a vocabulary such as that.
The last thing I wanted to comment on was this whole idea of how do you
coordinate all of this. You know, MESH and MedLine have coexisted and
worked together for over 40 years and they work together pretty well. But
I go around the world actually and tell people that vocabulary is not a
science. It is engineering and that the only way that you can build
something that works together is you have to know what the system is like, what
the parameters that you are functioning in and what are the things that are
permissible moves to make.
So, that is one very important thing is that these — while the first order
approximation you can say that these things can function independently, they
cannot really function independently in the long run. They have to be
tightly meshed together and the only way you get there is you take a number of
things that are incomplete and imperfect and you start working with them and
making them work together and as they work together more and are used more,
then the coverage and the interoperability becomes much more effective.
My example for interdependence here in the messaging standard is SIGs and
the problem of each. Okay? Just how you write a SIG depends
upon what it is that you have written in the order — as the orderable
medication. That is maybe a trivial example compared to some of the other
ones that are going to be a little bit more subtle in all of this.
Okay. Thank you.
DR. COHN: Stuart, thank you. I guess we have time for just a
couple of questions. We are running about a half an hour behind, but we
are sort of used to that by the afternoon.
Jeff and I think Steve then probably has another question.
MR. BLAIR: Stuart, were you here for this morning when we had the
message format developers?
DR. NELSON: Yes.
MR. BLAIR: You probably noticed that a lot of work has been done in
terms of different messages by different SDOs and content developers and
that the subcommittee after awhile began to see, you know, how do you pull this
together so that everybody is marching to the same drummer if we are going to
certainly meet certain objectives and deadlines under MMA?
It just seems to me as if we may need something similar here for the drug
terminologies, to make sure that there is a central point of either
coordination or facilitation or leadership. Is that part of what your
DR. NELSON: I certainly believe so, that there has to be some central
area where it is clear, you know, where everything gets put together and
put together in a workable fashion. The UMLS is a very interesting
artifact and it was really built to facilitate interoperation at lots of
I think that RxNorm becomes a little bit more of a specialized case and
perhaps a little less general than the rest of the UMLS when it comes to
talking about this, but I think it is still much the same thing. We have
to look at it and say, well, yes, our goal is to facilitate interoperation of
vocabularies so that the fact that we are cooperating with pharmacy
knowledge-based vendors for their drug vocabularies, with the FDA for their
drug vocabularies, serving as kind of the core of the communication between the
FDA and the VA and bringing it altogether, I think that is one of the roles
that the Library can play and has some experience in playing within the
DR. COHN: Steve, did you have a comment or question?
MR. STEINDEL: Yes, comment and maybe a question.
What we have heard in these hearings over the last number of months from a
lot of people is that RxNorm or what they term the clinical drug and they
usually then point to RxNorm as being the best example out there, it is
probably the terminology that the clinician would like to use to order the
Then we have also heard that the dispensed form of the medication is the NDC
code. That is what is billed. That is what is dispensed by the
pharmacy. So, the big question that I have had and I think both of you
did a good job of addressing it and I would like to make sure, is how do you
convert the RxNorm code into the NDC code?
One problem that I have here is with generic drugs, where there are a lot of
NDC codes. It sounds like you were going to do a pilot with NCPDP that
might start addressing this problem.
DR. NELSON: I think that our pilot with NCPDP would be to demonstrate
the utility of RxNorm and going from a physician’s order entry system to a
pharmacy, where they might use a proprietary pharmacy knowledge base as their
pharmacy system and interaction checking so forth. So, I think that that
is really what we were talking about with NCPDP.
As far as the NDCs are concerned, we are taking from the FDA, the quarterly
updates of NDC codes that they put — that they make available, which are
essentially the ones that they know about and including them as attributes to
the RxNorm forms so that it is a many to one, so that there may be many NDC
codes associated with a specific RxNorm form. Okay?
I would propose to you that while it may be important for billing to know
about, which the NDC code is, you may even have to have — in a pharmacy, you
might want to have even more specific knowledge, such as lot numbers, about
what was dispensed to a given individual, you know, for epidemiologic reasons
or whatever. Yes, NDC codes are another important level of abstraction
but it is probably not the sole one of interest.
MR. STEINDEL: Randy, do you have anything to add?
DR. LEVIN: Well, I think that once we are moving forward with the
structured product labeling for the products and moving them to NLM, we will be
able to link the RxNorm with the NDCs. The problem is not — the problem
with the NDCs is that they change so much and you can do one mapping but it
will be changing on a daily basis. The sooner we get to a place where we
are doing our listing, you know, electronic format and being able to put it
into the structured product labeling and pass it to the NLM, the better we will
be. So, then you will have an up-to-date linking between the RxNorm and
MR. STEINDEL: Is this a conversion that has to be relatively
extensively tested in the pilots, this process that you are talking about?
DR. LEVIN: I think that that would be something that would — to show
people how that would work.
MR. STEINDEL: That is what I mean. And the structured product
labeling, what I understand in text form, that is essentially available now and
it is going to be starting to be produced in about, what, six months to a year?
DR. LEVIN: What we have is that the standard is now done and we are
going to be starting to get some example labeling in from the pharmaceutical
companies that will show us how that will work. But it not only includes
just the text, the content, it also includes all those listing information like
the codes, the route of administration, all those things are going
to be coded into the structured product labeling.
So, we will start to see some of that and we will be getting those on a test
basis as we implement a system that will be able to manage those changes
because there are so many changes that occur. Then once that system is in
place, which will probably take us a year to develop, then we will have a
process in place to hand them over to the National Library of Medicine.
MR. STEINDEL: And to CMS for a pilot?
DR. LEVIN: It looks like it is going to be — it will be very close
and we will be ready toward the end of 2005, which would be then, at least for
some products, have it available.
DR. COHN: Stan and then Jeff and then we will adjourn, take a break,
take a ten minute break.
MR. HUFF: So, do one or both of you provide or what is the best source
I should say for just the name of a brand substance, like Valium? Is
there a CUIE(?) for that in the UMLS Medi-Thesaurus?
DR. NELSON: There is a CUIE and then if it comes out in between UMLS,
there is an RxNorm ID. Okay? Because the way the CUIEs are assigned
in the UMLS, they are only assigned at the end of editing. It is just
technical detail, but essentially the updates will have CUIEs if the CUIE is
available, but suppose, you know, something comes out on the market tomorrow
that has never been — we have never seen that brand name before, it will have
an RxNorm ID and come out. right, before it gets into the UMLS.
MR. HUFF: I am not talking about Valium 5 milligram tablets. I
am just talking about Valium.
DR. NELSON: In our RxNorm we have IDs for everything from ingredient
down to the level of the clinical drug.
MR. HUFF: So, you also have a CUIE for Valium tablet?
DR. NELSON: Yes.
DR. LEVIN: Just a distinction here, for me, for Valium, is more than
just the active ingredient, diazepam. It is the inactive ingredients
because it is a product. It is an actual drug product. So, for us,
we include inactive ingredients.
MR. HUFF: That would be okay with — I think that is true, too, but
the point is that is the concept without saying that it is a tablet or without
— and actually I guess if you say, yes, because they are Valium injectable,
which would have potentially different inactive ingredients. So, I guess
I am saying something different.
DR. LEVIN: If you are talking about diazepam —
MR. HUFF: I will give you my use case to show you where — I mean,
these things that wouldn’t normally be ordered, though, you can imagine
actually people having an orderable system that would allow you to say
independently in different fields, Valium, 5 milligrams in a separate field,
tablet. In that ordering, if you were able to order that way, then you
need a name for the concept that just means Valium, which is ambiguous as
whether it was a tablet or a liquid or whatever.
So, that is one use case. The second use case is when people say I am
allergic to Valium and you don’t want to specify whether they were allergic to
the tablet or to the — you know, they just say I was allergic to Valium.
You want to capture the information at the same level that they provided it in
DR. LEVIN: For your use case of — are you interested in Valium as the
product, which mean includes the inactive ingredients or are you just
interested in diazepam?
MR. HUFF: I think in my use case, I mean, in the allergy use case, you
are really interested just in the brand name for the active ingredient. I
mean, if it were a combination product, you are worried about all of the
inactive ingredients. So, if it was lower tab or Tylenol No. 3 or
whatever, you are interested — you know, you want a single name that basically
implies those active ingredients without implying a form or a dose or a
strength or route, any of that stuff.
DR. NELSON: We have that as the branded ingredients, as trade name,
which contains these other — the active generic ingredient. That is
DR. COHN: Judy, I think you have the last question and then —
MR. HUFF: That was only my first question. The other may be
You talked about what Daily Med contains, but I didn’t — maybe you said and
I didn’t understand. How is it distributed? Is that a database or
is it set of XML documents? How is Daily Med going to be distributed or
DR. NELSON: It will probably be — it will be made available as some
sort of browsable database at the Library, as well as downloadable, some
downloadable format, presumably XML. We haven’t gotten anywhere close to
that. That is a year from now when we expect to start getting the
structured product labels out.
MR. HUFF: Thank you.
DR. COHN: Any other questions, Stan?
MR. HUFF: No.
DR. COHN: My apologies for breaking in there.
MR. HUFF: No problem.
DR. COHN: Okay. Judy.
MS. WARREN: Mine is for Randy and it is just to double check my
hearing. On your slides several times you had that the standards wouldn’t
be available for three or four years. Can you — due to regulations —
can you explain what that means?
DR. LEVIN: Well, this has come up in the past where as far as the way
the national drug code is handled at this point that there could be the
possibility of national drug codes out there. They are in use that the
FDA does not know about. So, that would be our — our list of national
drug codes may not be complete.
MS. WARREN: And then your other comment that struck me also was a
timing issue, that you said that the NDC codes can change daily.
DR. LEVIN: The NDC code — if you changed your product, then the NDC
code changes. So, if you change — you put a different inactive
ingredient in your product, then you get it — there is a new NDC.
MS. WARREN: But the old one hasn’t gone away. I mean, you are
not reusing codes or anything? It is not unstable.
DR. LEVIN: That is one of the — currently, the rules allow the codes
to be reused, but that is one of the things that we have heard already about
problems with the NDC.
MS. WARREN: Okay. Then I heard it wrong. Thank you.
DR. FITZMAURICE: I just want to follow up on one thing Randy said that
the NDC codes can be reused, but we are trying to change that. Will that
take another regulation in order to change it?
DR. LEVIN: Yes.
DR. FITZMAURICE: Those things don’t happen fast.
DR. COHN: I think that would be what Randy might tell you.
Steve, I think you have the last question.
MR. STEINDEL: Yes, this is just kind of a clarification and, Randy,
you can confirm it.
When you showed the timing, in most of those slides you actually showed two
times. One is for prescription drugs and the other was for everything
else. The prescription drug part is maybe except for a little bit of
tweaks is essentially ready to go now and it is the other part that will take
the longer period of time.
DR. LEVIN: Another slight distinction is the prescription drugs that
are approved and the prescription drugs that are not approved. Those are
under different rules. So, it is mostly for the approved prescription
MR. STEINDEL: Which is what is handled with e-prescribing.
DR. COHN: I want to thank you all for some very interesting
We are going to take a ten minute break. We are running significantly
behind, but that happens and we will come back at 3:30.
DR. COHN: We are going to get started with the next session, which is
on identifiers. Obviously, we are delighted to have Karen Trudel from CMS
joining us, Lee Ann Stember from NCPDP and Phillip, you are also with NCPDP.
MR. SCOTT: Yes, sir.
DR. COHN: Why don’t we get started here. Everyone please be
MS. TRUDEL: Again, I am Karen Trudel. I am the acting director
of the Office of HIPAA Standards at CMS and as such, I have responsibility for
HIPAA transactions and code set standards and also for e-prescribing standards
under MMA. I am going to talk, hopefully very briefly, about enumerating
plans, patients and providers. I will start with plans.
I will preface my remarks by saying that some of the statements I make will
appear to be something narrow in their focus because I want to give a sense of
how some of these issues affect Medicare Part D and it should go without saying
that I am very mindful that that is not the whole universe in terms of
e-prescribing and that there are some issues that Medicare Part D may be able
to work around, that the industry as a whole in attempting to implement
e-prescribing and electronic health records will be forced to address Square
1. So, that is just a bit of context for me.
The first thing I would like to talk about is identification of health
plans. I will start again from the Part D perspective. There are
three types of plans that we will need to enumerate for Medicare Part D.
We will need to enumerate drug programs that are part of a Medicare Advantage
Plan. We will need to identify PDP plans and we will need to identify the
fallback plans that we may need to contract within areas where we need to do
that to provide beneficiaries with adequate choice.
All three of those types of plans would be considered health plans under
HIPAA and, as such, they could be enumerated under the plan ID that is
currently under review in terms of a proposed rule. However, the plan ID
will not be ready probably for three to four years. So, we are forced to
look at what to do in the short term.
We believe that for MMA, we can enumerate those three types of plans,
basically just as an extension of the current process that we have for
identifying Medicare Advantage or M Plus C plans. So, I don’t view this
as a significant operational issue for Part D in the short run. However,
I am also the first to say that Medicare probably is not representative of some
of the complexities that private health plans may have in terms of enumerating
different contracts, different benefit packages and, therefore, there may be
many different kinds of enumerations that may need to go on that would
not be represented in the three that I just described.
I think we do have the opportunity right now to impact what is happening
with the requirements for the plan ID since we are talking about a proposed
rule and I would encourage the subcommittee to provide any additional
requirements back to the Secretary for consideration to be included in the plan
So, I will move on to the patient identifier and, again, for MMA purposes we
do already have a Medicare identifier and we so far don’t have plans for
issuing an additional identifier for Part D. So, the Part A, B and C,
that identifier is basically what we are going to use.
The Medicare identifier is called the Health Insurance Claim Number.
It is basically a two part identifier. It consists of the social security
number of the wage earner on whose account the eligibility is established and a
suffix that shows the beneficiary’s relationship to the wage earner. Very
For instance, the suffix of A means that the beneficiary is the wage
earner. The suffix of B denotes the fact that the beneficiary is the wage
earner’s spouse. Adding another identifier at this point would be
extraordinary expensive and for MMA purposes, I don’t believe that it is
necessary. The Medicare ID would not convey any information about
identity, verification, which was noted as a requirement. It would not
convey any information about beneficiaries personal representatives.
In the broader environment, clearly to interoperate and pass medical history
and medication history back and forth, there would be a need for either a
master patient index of some sort or a return to the concept of the HIPAA
patient identifier, which we all recall with some trepidation, I believe.
I am going to talk about the provider identifier for a moment and before I
start talk about specific requirements, I would like to go back and do a little
mini briefing on the NPI and some of the decisions that were engendered in the
final rule because I think there may not be a clear understanding of what the
NPI does do and what it doesn’t do. So, I would like to do that first.
The compliance date for the NPI is May 23rd, 2007, but the effective date of
the regulation is May 2005. That date was selected as being the date on
which we believed we would begin to issue NPIs to providers. The system
build and the enumeration contract are underway and the time table that I just
received was that they will be pilot testing the system from May to August of
2005 and we will begin to be able to issue identifiers in August of September
So, it appears that if those deadlines are adhered to, NPIs could be
available by 2006. The NPI will enumerate many, many kinds of
providers. What it won’t do is replace some functionality, like the DEA
number or the tax ID number. It won’t enroll a provider in a health
plan. It won’t do rigorous amounts of certification, but any health care
provider is eligible for an NPI, both covered and non-covered providers.
The ones that are not covered entities under HIPAA also are entitled to obtain
We also provide NPIs to subparts of covered organization providers and that
is a major consideration as well. One thing we do not do is to enumerate
each of the service providers, such as carpenters who build ramps for
disability access, taxi drivers, things like that. I think it is highly
unlikely that those people are prescribing in the first place. We can
The subpart concept of the NPI is very critical, especially when you are you
talking about pharmacies. The concept doesn’t apply to individuals, but
it allows the assignment of an NPI to a subpart of an entity when they furnish
health care and they are not legal entities. Essentially, it permits a
provider to determine that subparts are required by federal regulations to have
their own billing number. They may have separate licensure.
They may have a different physical location and all of those are acceptable
reasons for obtaining a subpart ID. Possible subparts could be hospital
components, such as outpatient departments, laboratories, psychiatric units,
could be members of chains, such as pharmacies, DME suppliers, nursing
homes. And, again, they may have a different location.
The NPI, I will just run through it quickly, is the ten position
identifier. It is all numeric. It has one check digit and, again,
does not convey information about the provider or the subpart. We will
know if a provider is unique by collecting additional information about that
provider. We initially proposed collecting a lot of information about
each provider, including for individual providers, where they were trained, et
cetera. At this point, we will collect only information that is
necessary to uniquely identify. There are different data elements
for individuals and organizations.
One of the pieces of data that we will collect is state license
information. For all of the taxonomy codes for physicians, we will
collect both the state license number and the state of licensure. So,
that is a requirement for every physician taxonomy code specialty,
subspecialty, et cetera.
Non-covered providers may apply for NPIs. There is no statutory
requirement under HIPAA that they do so, but we certainly encourage them to
obtain and use NPIs and those providers can include physicians, for instance,
who are employed by a practice do not bill themselves. It can include
nurses or other clinical professionals in a practice.
Now, let me go back to the provider identification requirements and what I
will just mention here is the places that I think that the NPI could work and
the places where there may or may not be gaps. In terms of the dispenser,
we have definitely heard that the NCPDP identifier is the one that are used to
using. However, the NPI is required on the regale pharmacy claim for the
dispenser for the pharmacy. It can, as I said, accommodate locations in
terms of subparts. In terms of the prescriber, again, the NPI is required
for use on the claim. The prescriber, the practice name, I don’t
particularly see gaps there, things that couldn’t be handled.
The requirements around an authorized agent or a supervisor, I am not
tremendously familiar with it. As long as those concepts related to
health care providers, they could receive an NPI, even if they were not a
covered entity. The location of the medical record is where I see a
breakdown and it was a deliberate one on our part. We initially proposed
to enumerate locations, all locations for health care providers. Comments
we received back during the comment period were extremely against that.
They suggested that location changes too often, that having to get an
additional location identifier and provide all this information every time a
practice moved would be nearly impossible for people to keep up with.
Therefore, the location identifier was one of the things that we removed in
the final rule. I see some choices here. One possibility is to
propose or to recommend using HIPAA standards simply because they are HIPAA
standards. They are required in other HIPAA transactions, not in the
prescription, but certainly in the retail pharmacy claim. There is a
drawback to having dual standards. We have told physicians you will only
have an NPI and a DEA number. We could be telling them you will only have
an NPI and a DEA number and another number. We would have to explain why
that was a good idea.
But one option is to recommend the use of the HIPAA standard and to fill any
gaps, for instance, that lack of location identifier, so that you would have a
single identifier that does everything for everyone. Again, we very
particularly did not include a location identifier for some good reasons.
Another possibility is to adopt some other standard that does everything for
everyone and that is another consideration that the subcommittee will need to
think about. The pro there is that obviously you have an identifier that
does everything right out of the box. Again, as I said, the down side is,
you know, we are going to have to explain why we adopted a HIPAA standard that
didn’t do everything that is needed.
The third possibility is to use the HIPAA standard for what it works for and
find other ways to fill the gaps. One of the things that we have heard in
some of the testimony is that some of the manufacturers use, for instance, the
phone number as a proxy for the location where the medical record is. So,
those three are all possibilities.
The pros and cons, I think, of rolling out another standard is that there is
a cost to rolling out a new identifier and we will have to bear that burden,
along with rolling out the NPI, which we are already committed to doing.
There is a cost of all prescribers being required to have and maintain
another ID. I would leave with the thought that in implementing any
standard or even suite of standards that interoperate, mapping is really going
to be critical, mapping the NPIs to the HCIdea, mapping the NPI to other
things, mapping the plan ID to existing identifiers.
I think that those may be also some recommendations that the subcommittee
might want to consider in addition to any recommendations on which identifier
would be appropriate.
DR. COHN: Karen, thank you.
MS. STEMBER: Mr. Chairman and members of this committee, my name is
Lee Ann Stember and I am the president of the National Council for Prescription
Drug Programs. To my right is Phillip Scott, who is our senior vice
president of health care and professional development. One of his roles
is responsibility for HCIdea and its implementation.
NCPDP is an ANSI Accredited Standards Development Organization, whose
mission is to create and promote standards for the transfer of data to and from
the pharmacy services sector of the healthcare industry. The organization
provides a forum and supports its diverse membership to efficiently and
effectively develop and maintain these standards through a consensus
building process. NCPDP also offers its members resources, including
educational opportunities and database services, to better manage their
In 1991, our membership identified a need for a unique prescriber
identifier. Health plans wanted to identify the prescriber for
utilization management purposes. Pharmacies and processors or PBMs
required a unique prescriber identifier across health plans for prescribing,
claims processing and utilization requirements.
A significant portion of the prescription drug industry uses the DEA number
in order to identify prescribers on prescription benefit claims. This is
not the intended use of the DEA number, which was created to act as a barrier
to illegal controlled substance distribution. The DEA and other agencies
would like this number to only be utilized for its intended purpose. The
use of the DEA number for strictly financial transactions compromises the
integrity of the identifier. Not all prescribers have DEA numbers.
Other prescribers have multiple DEA numbers and DEA numbers are also
assigned to non-prescribing entities, such as pharmacies and wholesalers.
As a result of the problematic use and state legislation banning the use of
the DEA number for purposes other than originally intended, NCPDP undertook the
task of creating an alternative to the use of the DEA number as a prescriber
In the early 1990s, NCPDP’s Standard Prescriber Identification Number or
SPIN project was formed to solve the problem and work began in conjunction with
a large healthcare professional association in 1993. Significant progress
was made in the mid 1990s. However, the SPIN project was vacated due to
the announcement of the NPI in HIPAA. In response to ongoing requests by
NCPDP’s membership to create a solution and with the delay in the NPI, NCPDP
announced the development of HCIdea at its 2002 annual conference.
In January of 2002, a contractual agreement was signed with the National
Health Information Network of NHIN, a database development company, to become
NCPDP’s technical partner in developing HCIdea. HCIdea is an NCPDP
prescriber enumeration project where in all prescribers, even those
without the ability to prescribe controlled substances, will be
enumerated. It is different from the NPI in that it contains additional
fields for each prescriber, which facilitate payment processing and lead to
more accurate identification of prescribers.
Ultimately, we envision the HCIdea file to mirror the success of pharmacy
file. NCPDP’s current pharmacy enumerator, the NCPDP Pharmacy Provider
Number, formerly the NABP number, which has been in existence 1977, not only
individually enumerates all pharmacies, but also affiliates individual
pharmacies with their chain headquarters or payees, if applicable. These
additional fields to PBMs, health plans and pharmacy chains alike in processing
claims and providing encounter data to the appropriate entities.
With the eventual implementation of the NPI, NCPDP will create a crosswalk
between the HCIdea identifier and the NCPDP Provider ID Number to the NPI,
facilitating transition to the NPI with the prescription drug industry.
In order to rapidly and accurately populate the HCIdea database and to gain
market acceptance of HCIdea, NCPDP entered into data supplier arrangements with
various healthcare entities to supply prescriber data to our technical partner
for validation and population of the HCIdea database. NHIN receives this
data, which is cleansed and each prescriber is then assigned an HCIdea
enumerator. This model continues to provide to ongoing updates of the
Currently, NCPDP has signed agreements with twelve data suppliers.
Multiple data suppliers have passed all certifications, meaning that criteria
have been met and data can now flow between the two sources. Several
customers are receiving this data, one of which is the Department of Defense,
which has entered into a multi-year license agreement.
The State of Utah Bureau of Medicaid Operations, in conjunction with NC and
a major pharmacy chain, will embark on the first testing for the transmission
of HCIdea data between pharmacies and the Medicaid Bureau in the very near
future, with implementation throughout Utah expected in 2005.
NCPDP wants to emphasize that the HCIdea database is not a substitute for
the NPI, but a large subset of what will be the providers in the NPS. The
current database contains only prescribers, including primary prescribers, such
as MDs and DOs, as well as mid-level practitioners, such as advanced nurse
practitioners, physician assistants and others, such as optometrists.
Our best estimate is that our universe of prescribers currently contain 1.2
million records or in excess of 86 percent. However, we do not yet
consider this to be a mature product. We are in what we term to be a soft
launch phase. But we have made significant strides in a two year time
frame and implemented processes that ensure the integrity of the records that
are contained in the database.
The needs identified in 1991 by our membership remain as timely today as
they were 13 years ago. The HCIdea identifier not only provides for the
claim billing processing initially identified for our industry, but can also
serve as an enumerator for the newly defined requirements of electronic
CMS could certainly use the Medicare ID number to identify the prescriber
for Medicare Prescription Drug Benefits. However, e-prescribing is
gradually being adopted by prescribers, who serve private pay and Medicaid
recipients as well. Using a different number of each payer would add a
level of complexity that will impede adoption by prescribers for all types of
Our NCPDP Pharmacy Identifier has been used since 1977. It is included
on over 4 billion transactions per year and could be used as a resource for
prescribers to identify the community pharmacy of the patient’s choice.
Since the inception of HCIdea, NCPDP has provided information to government
agencies and has offered our assistance in any possible way. I am here to
restate that offer of assistance. Thank you for your time and invitation
to speak to you today.
DR. COHN: Lee Ann, thank you very much. Very interesting
testimony from both of you.
Questions from the subcommittee? Michael.
DR. FITZMAURICE: Since we are in a little bit of a crunch to enumerate
prescribers, is it possible that the development of the NPI could use some
existing data sets like the U-PIN for Medicare and maybe information from NCPDP
to fill in the blanks and then ask prescribers to verify the information as
opposed to going out and collecting it originally. Is that part of a
MS. TRUDEL: Yes. We have always maintained that freeloading with
existing high quality data would be an option. Again, that may be one of
the subcommittee’s recommendations.
DR. FITZMAURICE: But the contractor then could do it? I mean,
there is no clause in the contract that would — I guess we could always modify
the contracts. So, it doesn’t matter.
MS. TRUDEL: We planned for the possibility of mass enumeration.
DR. COHN: Jeff.
MR. BLAIR: You really made it easy for us, didn’t you.
MS. TRUDEL: I try.
MR. BLAIR: Karen, you indicated that the NPI includes is it state
accreditation or certification did you — I forget the phrase. I am
MS. TRUDEL: It includes the license number and the state of licensure
for any practitioner with a taxonomy code in the physician range.
MR. BLAIR: In the physician range —
MS. TRUDEL: Right.
MR. BLAIR: But not necessarily — does that exclude like a nurse
MS. TRUDEL: Yes, I believe it does.
MR. BLAIR: Lee Ann, what method or type or degree of certification
does HCIdea have?
MR. SCOTT: Jeff, this is Phillip Scott. I will answer that for
Our enumeration process takes into consideration all available
licensure. So, in addition to the state license number, if U-PINs and
other identifications are available to us through the value added fields, we
utilize all of those and if we can’t do it in an automated environment, then it
is validated with a call center call directly to the prescriber’s location.
MR. BLAIR: Okay. One other question — and Lee Ann, help me with
this if I am missing something, which I often do. My understanding of the
principle advantage of HCIdea over the NPI was that it can identify
locations. Is that correct?
MS. STEMBER: That is correct. Multiple locations for a single
MR. BLAIR: However, from what Karen was telling us, there is no reason
why an NPI couldn’t be used with separate direct indications, self locations or
as Karen was saying a telephone number to be a proxy for locations. What
limitations or problems would there be if we used that, like a telephone number
or something in addition to NPI? Is there something else that we are
MR. SCOTT: Jeff, I would field that one again. Actually, it
depends on the business need. In the business case that we have developed
HCIdea, the purpose for multiple locations is for on a practical application
when a pharmacy is dealing with a physician with multiple locations,
potentially catching prescriptions from those different environments, you have
to way to verify who that is. Today what happens, if you have got —
because we are driving most of this with the NCPDP number or, excuse me, with
the DEA number, what happens is you may have a physician with as many as one in
Southern California, with 24 different DEA numbers and locations. So, if
you are a practicing — if your pharmacy is in a specific location, it is
important and helpful to know exactly where that physician’s practice is as it
relates to your particular pharmacy.
Also, in order to eliminate the confusion that is generated by having
multiple locations and you can imagine if you have got four or five — I know
how many different ways my name ends up coming into the house as far as mail is
concerned. So, you have got variations on a theme according to whoever
filled out the application. So, it could be Tom Jones. It could be
Tom T. Jones or T. T. Jones, you know, those kind of variations. With one
unique identifier with multiple locations tied to that, it helps to eliminate
the confusion that can happen in that manner.
MR. BLAIR: Let me ask one last question.
If I understand it correctly, Karen indicated that virtually by law or
by regulation, NPIs will be required when they become available for use, both
by prescribers and dispensers. Is that correct, Karen?
MS. TRUDEL: That is correct on HIPAA transactions, which includes the
retail pharmacy drug claim, but does not at this point include the
MR. BLAIR: Okay. And, Lee Ann, you had indicated that NCPDP has
already arranged for mapping —
MS. STEMBER: Yes, that is correct, Jeff.
MR. BLAIR: — between HCIdea and the NPI?
MS. STEMBER: That is correct and the NCPDP pharmacy provider
MR. BLAIR: Okay. Thank you.
DR. COHN: I guess I have a question or two and then Steve and Kathy.
As you know, I am not an expert in HCIdea or the provider ID. So, I
just wanted to make sure I understood a couple of things now. It appears
to me that based on the way you are describing HCIdea that this actually is a
product of NCPDP that is being licensed to people and I presume there is some
sort of a fee involved. Is that a correct statement or am I missing the
MS. STEMBER: Right now there is no fee attached to that.
DR. COHN: Okay. Is that intended to be that way going forward or
MS. STEMBER: That is what we are talking about right now.
DR. COHN: Okay. So, obviously, you are doing this as a public
service then activity to support the membership.
Now, I saw something and I apologize — I have got about 75 pages,
standards, analysis documents all in 9 point type. So, I can’t sometimes
find exactly what you need when you need it, but somewhere in here, I remember
yesterday when I was flying out — and this is one of the values of flying long
distances is that you get a chance to review these things — was some statement
that providers actually need to know what their HCIdea number was. I may
have misunderstood this, but it made me think that perhaps it wasn’t really an
identifier designed for me as a physician to know what that number is, but
something more that was added on by the pharmacy or reflected there, somewhere
in their database to allow them to track where they were or for further
Am I missing something here or am I misrepresenting things?
MR. SCOTT: No, actually you aren’t. For the initial purpose that
we have developed the unique identifier, which is for pharmacy claims
adjudication purposes, to eliminate the use or inappropriate use of the DEA
number, which today is required for a pharmacy claim to be adjudicated.
It is not necessary for a prescriber to know that they even have an HCIdea
As we have continued to grow and inch in our infant sort of way towards
maturity, the interest in knowing this and the interest in using it has really
started to escalate significantly. I think it has driven by the intrigue,
if you would, or the desire for many states to eliminate the use of DEA numbers
in any sort of identification or any sort of financial transaction. As a
result we do have organizations and individuals asking and participating in
getting their members identified, such as the American Optometry Association
recently took a move to include all of their membership in that database
because many places they have no identifiers. Other places, where they
can write prescriptions, what is happening is someone else’s DEA number is
being plugged in to get that claim processed.
So, while we don’t require — it wouldn’t be required for a practitioner to
be aware or to know that they had an HCIdea number, it has becoming apparent to
us that there is a lot of interest in those practitioners being aware of it.
DR. COHN: I think that is sort of my first set of questions.
Steve, you had some?
MR. STEINDEL: Yes, I have a — I hope it will be a simple question.
If you look at our work plan, you will notice in the fall, later in the
fall, we are going to be looking at the question of e-signature and we have had
mixed testimony about the need for e-signature in the e-prescribing world and
are going to pin that down further in the fall. My assumption is that
neither one of these IDs is meant to be used for e-signature purposes. Is
MR. SCOTT: Certainly not, not for HCIdea, no.
MR. STEINDEL: So, there is the possibility that we may have a third
DR. COHN: Was that a simple question or simple comment?
MR. STEINDEL: Neither.
MR. SCOTT: I think it was just simply a threat.
DR. COHN: Kathy.
PARTICIPANT: I just wanted to comment that there may be veterinarians,
people outside of HIPAA or some of the human medicine realm that would be doing
a prescribing as well and very often use human drugs. Just to keep in
mind those other areas or other users of e-prescribing as you move forward.
MR. SCOTT: Just as a point of information, veterinarians are in the
file but they are not included in this number that Lee Ann quoted.
DR. COHN: Harry.
MR. REYNOLDS: It would appear to me that we are on a track of multiple
interim implementations, going towards an end. For example, NPI, I
envision as kind of the end game for I hope what is a single number.
Since there is a national database being developed that assigns those numbers,
it should carry across all environments. Is there an issue with the fact
that as the NPI database is prepared, that it contain these fields for these
other numbers, that if they are used in the interim, obviously, then a person
doesn’t have an NPI and it is not an issue. But as people are assigned
NPI, the whole idea of HCIdea being willing to map and so on allows you some
kind of a continuity because I think the thing we are going to get into is we
keep talking about drug histories and we keep talking about everything
else. If we get too many numbers that don’t map and too many ways to not
— a physician be a physician, a patient be a patient and everybody be who they
are, the day you implement e-prescribing, you have no history.
So, I hope as we look at this, we are able to focus on something like NPI
and wrap all these numbers around it, you know — most systems or most
everything you do have cross reference numbers, always will, always have.
So, I hope that we at least consider that as an option because back to the
whole discussion earlier today, whether it is HCIdea or whether it is NPI or
whether it is anything else. There is going to be a significant
investment by the industry and doing it two or three times and not having some
kind of crosswalk and not having it clear, just seems to be just another way to
flog ourselves as we do this.
MS. TRUDEL: If I could respond to that, that was the point that I made
much less articulately about mapping and the national provider system will
collect and hold other identifiers that are provided by the provider at the
time of enumeration, including the DEA numbers, the U-PIN, et cetera. So,
that is the sort of continuity that we are envisioning and I am sure you all
are doing the same thing.
MR. REYNOLDS: But are you recommending that HCIdea be a part of that
also as one of the other numbers picked up. You mentioned DEA.
MS. TRUDEL: I am not sure which ones we have identified at this point,
but, again, if we were to identify specific files from which we were
enumerating to do a sort of a fast track enumeration, we definitely would be
holding that information in the database.
MR. SCOTT: I just have one follow-up to that, very similar to
Karen’s. I think if you think about it this way at maturity, today
everybody is required to provide a DEA number under the current business
circumstances. It is a common problem that that number is just not
available for whatever the reason is. I don’t expect that the NPI is
going to be readily available in all sets of circumstances, nor would I be
naive enough to think that the HCIdea number would be .
However, if you are working — it is our intent as we work with our client
and membership base that if, in fact, you had the HCIdea number and multiple
locations for a physician, that you could look to that number and it would take
you to the required NPI. However, on the other hand, if you had one of
his license numbers or if you had the NPI number and you needed to identify a
location or know the HCIdea number, you could use that as the reference to get
you to that point. So, while it may appear — and I guess it also goes to
why a prescriber would necessarily have to know he had an HCIdea number.
But if you take that as an idea about how you get to it because we will never
get past the fact that an NPI is required to file certain claims.
So, you are going to have to have that and what better way to do that if you
have got physicians with multiple locations to be sure that you have got the
right NPI with the right physician to be able to use those cross referencing
tools. So, when NPI is introduced as it exists today or as we expect it
to exist, there will only be one identifier. How you get to that and the
other purposes that you use for it, because there are thousands and billions of
claims out that are not necessarily Medicare or Medicaid claims specifically.
DR. COHN: Stan, you were first and then Steve. My apologies.
MR. HUFF: So, I apologize in advance for asking a detailed, technical
question. I haven’t had a chance to read or understand the structure of
the HCIdea database. When I put an HCIdea into a Script prescription to
identify the prescriber, is that a single number and do I have then — does one
individual have multiple numbers, one for each location where he is or is it
somehow concatenated in there with a part that is unique to the individual and
then a suffix that is specific to the location from which he is
prescribing? How is that physically done?
MR. SCOTT: You have one number that is given to the individual and a
dropdown if you would in the existing database, a dropdown if you would that
would give you the multiple locations.
MR. HUFF: So, it is really done in the database. You know
which locations are valid for me as a person.
MR. SCOTT: That is correct.
MR. HUFF: So, it is a database function so that I have — it is truly
a unique identifier of the individual or doc.
MR. SCOTT: Correct.
MR. HUFF: And then the location is done as — and then the location
ends up in a different part of the Script as the originating site or do those
two things get concatenated into that single field for the identifier?
Maybe that is a question for Lynne or somebody.
MR. SCOTT: Actually, I would like for you to ask the question again so
I understand exactly what you are asking.
PARTICIPANT: Do you have two variables or one together?
MR. HUFF: In the Script standard, I mean, there is a prescriber field
and there is a prescriber ID field and then there is name and, you know, other
demographic information associated with the prescriber. In that field, in
the Script standard, what number goes in there? Is it just the identifier
for the individual or is it some concatenation of the individual and their
location or what is —
MR. SCOTT: It is for the individual. And the Script — HCIdea
was developed to fall within all of the NCPDP standards so that it could be
used as the identifier, which currently is functioning as the DEA number for
the most part. So, as you are substituting that, we would place the
HCIdea number in that identifier. There is also — if I remember
correctly, there is also a secondary field for an identifier if it is — if, in
fact, an NPI or some other form would be required.
There is another location that could be used for that. But it is only
one individual. It does not identify the location. The location
actually I think is more of a physical action around the filling of this
prescription itself just to be sure you have got the right location.
DR. COHN: Can I just follow up on that one to make sure I really
I guess I had thought that this HCIdea was an identifier that included
location and the unique identifier for the individual and that there was a
different ID for each location for that individual.
MR. SCOTT: No. One individual, one number and then cross
reference to all of his DEA or her DEA numbers and multiple locations.
DR. COHN: Thank you. I am sorry, Stan. Please go forward.
DR. FITZMAURICE: Is it true then location is a separate variable
on the Script standard? You have the unique ID, the HCIdea, the notes to
the prescriber and you have a separate variable that shows location and maybe a
payment schedule is tied to that location.
MR. SCOTT: Yes, that is correct.
MR. STEINDEL: Do you handle location in HCIdea differently than NPI is
handling location with its supplemental ID?
MR. SCOTT: Well, we have multiple locations and the NPI is a primary
MS. TRUDEL: The NPI collects a primary location and a mailing address.
MR. STEINDEL: I thought there was the supplemental location code that
could be added.
MS. STEMBER: That was in the proposed rule, not in the final.
MR. STEINDEL: Okay. Thank you for the clarification.
MR. HUFF: My second question actually is then a follow up to his — I
understood in your testimony, though, that basically if NCVH and the community
decided that, in fact, we needed that correlation to location, you would be
amenable to possibly adding that as additional information in the NPI
database? Or did I misunderstand?
MS. TRUDEL: What I was saying was that there were several different
ways to go about that. One was to go back, revisit that original decision
and repropose an NPI with a location code and the other was to try to arrive at
a completely separate piece of data to act as a proxy for that location, which
was my point about the phone number because we had just heard that some people
were using it that way. So, I think those are the options that we
have. The question is do we — these are requirements that are specific
to e-prescribing and do we want to impose these requirements for an all HIPAA
transaction, like claims?
That is the balance and I think that is what the subcommittee needs to
MR. HUFF: Well, again, just to make sure I understand, right now as
part of the NPI database, you have the ability to track state license numbers.
MS. TRUDEL: That is correct.
MR. HUFF: So, you are basically doing a combination of the name of the
state and the license number within that state and that will provide the unique
map to the NPI.
MS. TRUDEL: No, that is just one means of identification to make sure
that we don’t have duplicates. We found in doing some basic data washing
with the U-PIN file that we could identify duplicate based on state license
number and date of licensure that we could not identify as duplicates using
name or date of birth or a number of other things. So, the state
licensure does not correlate to any kind of a location. It simply —
MR. HUFF: No, I understand that.
MS. TRUDEL: It is simply one of the pieces of information that we will
MR. HUFF: My point wasn’t — yes, my point was that it was a two part
key. I guess I am confused because I don’t — even if I made location
specific ideas, I am not sure why that would — in other words, if you did
exactly what HCIdea is doing and made it possible to know for a given
individual, the locations from which they can practice or which they are — I
mean, in an individual situation, I mean, why would that force you to put the
combination of those IDs in HIPAA transactions, for instance? I mean, you
could use it where it was appropriate and not use it where it wasn’t
appropriate. So, there is no reason just because you had the location
specific IDs that you would be forced to use it if it wasn’t appropriate.
So, you could use it in prescriptions and not use it in claims if that were
appropriate. Right? Just the fact that you had it in your database
would not force you to use it in all transactions. Somebody is
maintaining it already.
The United States has a value for it. Why doesn’t the government pay
for stuff that everybody has value for?
DR. COHN: Steve and then Jeff and then maybe I will —
MR. STEINDEL: Karen, I would like a follow-up clarification to where I
see your response to Harry going. You know, he asked the question as,
paraphrasing it, you know, can you include the HCIdea in the NPS and you seemed
to indicate, yes, you could. You know, if you are enumerating from
HCIdea, you know, you were talking about talking about the U-PIN number in and
other numbers. The NPS then is not fixed in structure by
regulation. You can modify it on an as needed basis. Is that
MS. TRUDEL: I would say the specific data elements that we are
collecting from individuals is specified. In other words, we couldn’t —
I don’t think we could go back at this point and say we changed our
minds. We now want to start collecting, you know, the medical school that
you went to, Mr. —
MR. STEINDEL: No, I —
MS. TRUDEL: But in terms of we already have a requirement to provide
identifiers that are already assigned to the person. Now, if the
physician doesn’t happen to know what his HCIdea is, he wouldn’t be able to
provide it at the time that he has enumerated, but if we were enumerating from
a file, we certainly could capture the information.
MR. STEINDEL: Then that clarifies my question. I wasn’t asking
that you go back and ask him for the information that wasn’t required. I
was just asking from a structural point of view. Can we modify the
database? It sounds like the answer is yes, as the need develops.
MS. TRUDEL: I am answering off the top of my head, but I don’t see any
regulatory reason why not.
MR. STEINDEL: Okay. Good. Thank you.
MR. REYNOLDS: That then starts making the NPI database a repository
for identifying a person and it allows people as we transition to use whatever
they have in place. I mean, as we look at going from here to a standard
that we really all want to get to, the transition is always the most
difficult. Being able to do that allows a transition so people can
continue doing what they are doing as we move forward.
MS. TRUDEL: At this point I would say that — I may be wrong about
this — what most people are using is their DEA number.
MR. REYNOLDS: Yes, but, again, some are using DEA. Some would
use HCIdea. Some would be moving to NPI, but the point is you have a
repository. Right now, the problem is this stuff is all over the place.
DR. COHN: Jeff.
MR. BLAIR: I think that Harry has really covered — well, maybe if I
built on that. What problems are there that we are not aware of if we
were to take the suggestion that apparently Steve and Harry are positing, which
is that the HCIdea number were included as appropriate for prescribers with
their NPI number, what problems does that create, if any?
DR. COHN: Jeff, maybe you can ask that again?
MR. BLAIR: What problems will it create if we wind up including the
HCIdea number as appropriate with —
MS. STEMBER: None that I would anticipate on behalf of NCPDP.
MR. BLAIR: You know, Karen, is there a problem with that that we need
MS. TRUDEL: Not that is springing immediately to mind.
DR. COHN: I guess I am a little confused. But, Margaret, I will
let you start and then I will try to ask another question.
MS. AMATAYAKUL: Since the NPI database today cannot collect data from
a provider that is not specified in the regulation, does that mean that to get
the additional information that is in the HCIdea database, the HCIdea database
must continuously be maintained and you continuously would preload over time to
get that new information, which basically then means the industry as a whole is
having the cost of maintaining two databases?
MS. TRUDEL: I am not sure I understand the basic premise here. I
don’t think we were talking about loading additional data from the HCIdea
database. I thought we were talking about simply making sure that the
identifier was crosswalked and contained in the NPS database.
The other thing, I see two differences here and we have talked about this
some time ago. I see the HCIdea database being almost serving the purpose
of a master prescriber index, in which case it would continue to have a life
and a value of its own, even after NPIs are assigned to everyone.
DR. COHN: You know, Karen, thank you. That was actually the
point that I was going to try to make because I come away feeling that the
HCIdea database is a tremendously valuable product for the industry. If
you are a provider that has multiple locations, this is the way to figure out
likely where this prescription might have come from, what address or whatever
it is they put down on there.
So, I sort of saw this as a value added and I don’t necessarily believe that
the NPS or the NPI needs to solve every problem for everyone.
MS. TRUDEL: And you are absolutely right. That is the
overwhelming flavor of the comments that we received in the final rule that
caused us to cut back on some of the content, things like location code and
other data because we were trying to do more than just enumerate. People
came back and frankly said that was a mistake.
MR. BLAIR: Okay. It seems as if we are converging on a best of
both worlds compromise solution here and so then the next thing is the
mechanics because does this mean that NCPDP needs to get to CMS by a certain
date, real soon, like within this next year, the HCIdea numbers — they say
they are already doing the mappings. Is there either a deadline or a
mechanism or something else to make sure those mappings are in place, certainly
to support the demonstration projects, January 1 of 2006?
DR. COHN: Maybe I will start and then maybe — I guess I am missing
this one. I actually didn’t — I mean, it seems to me that the NCPDP may
choose to help CMS with their enumeration process, just because they seem to be
far ahead in terms of identifying all the providers. I actually wasn’t at
all clear that I thought that the NPS should have the HCIdea number in it or
that actually the HCIdea number was even that important except to help us link
all of this together as a value added product for NCPDP. But maybe I was
Karen, help me with this one. Am I thinking upwards, backwards?
MS. TRUDEL: I don’t want to speak to what the subcommittee’s
recommendation should be, but I would suggest that whatever the recommendation
is, it simply tells the Secretary, you know, we recommend that you whatever and
then we will figure out how to make it happen.
If the recommendation is that, you know, make sure that you map to as many
existing identifiers as you possibly can, that might be one recommendation.
MR. BLAIR: Maybe the essence of my question is a little
different. Lee Ann, you indicated that HCIdea is already — I thought you
said it was already doing the mapping, but how could you map to NPIs that
aren’t out there yet?
MS. STEMBER: We have the availability to begin the mapping.
MR. BLAIR: Okay. Now, the thing is that if you do the mapping,
how does that mapping and the linkages get to CMS so that when they start to
issue the NPIs, that that linkage is there? What needs to be done to make
MR. SCOTT: You probably need to call some folks in Fort Worth, Texas
and ask that question. That is probably the first thing that needs to be
done, but I think, Jeff, the reality is is that if there is a collaboration
between the NPI and HCIdea, it is not necessarily about the number, HCIdea
number, it is about the fact we have identified and cleaned up and standardized
a database of multiply identified single individuals that gives Karen and her
team essentially a data source, as opposed to a data supplier, which means that
it doesn’t require nearly as much verification or validation as it would if it
were coming from a dirty source, a source that hasn’t been cleaned up and
DR. COHN: I don’t think we have resolved this issue but thank you for
clarifying this one. Let me just ask one question. I am almost
afraid to bring this one up, only because we will be here to 7:00 p.m. by the
way we are going. We obviously have been focusing all of our time talking
about HCIdea and we haven’t talked about the dispenser number. I am
praying that that is the right term. It used to be the old NADP —
MS. STEMBER: The NCPDP provider identification number?
DR. COHN: Thank you.
Now, we obviously — what I was hearing about that, just let me ask the
question. Is that a different case or is this sort of the same case
around all this, given that pharmacies are going to have to be using the NPI in
their transactions? Let me just ask because I don’t know whether the same
issue we have just been discussing is part and parcel of this other issue.
MR. SCOTT: I think it is the same case.
MS. STEMBER: My opinion is it would be the same.
MR. SCOTT: We are making the same sort of strides across mapping
across reference purposes with the NPI to the provider from a map and have
reached out to CMS to see if there is not a way that we can help them in some
bulk enumerations. So, I think they are very similar, just a different
type of provider.
DR. COHN: Very useful.
Any other comments? I don’t know that we are done with this
conversation, but it is a — I think we will have to think about it tonight and
we may ask you some more questions as we go along here.
Other comments? Okay. Well, thank you, Lee Ann, Phillip,
Karen. We really appreciate your participation and enlightening us on
Our next discussion is really around standards for formularies, messages,
structures and prior — even in terms of prior authorization into this.
We have Teri Byrne from RxHub and Brian Bamberger from MediMedia. This
is Brian’s first time actually testifying. He has been an active
participant from the audience in our last set of discussions.
MR. BLAIR: Wonderful written testimony that has been very helpful,
DR. COHN: Teri, I think we are going to let you start out the
MS. BYRNE: I am Teri Byrne from RxHub and I would like to thank the
committee for inviting us back once again to testify. I want to let you
know this is probably going to be very brief. We don’t plan to go through
the R Club model again. I think we have testified to that enough. I
am not going to explain what a formulary is versus a benefit, et cetera.
The purpose of our testimony today is just to let the committee know what we
plan to do as far as taking our proprietary standards through an SDO to get
them to be accredited standards. I want to give you a little bit of
background just on what I have done and what I do at RxHub to help you
understand where we have been with these standards. I was brought into
RxHub almost from the beginning. My job basically has been to develop —
to understand the requirements from a technology vendor perspective, from an
EMR perspective and from a health plan and PBM perspective on what data and
information needs to be transferred back and forth in the e-prescribing
That has to do with formulating benefit information, medication history, as
well as prescriptions and refills. So, what we did was we developed a set
of standards early on and we put them out for public comment and we have had a
series of workgroups over the past several years to modify the standards.
We have had some pilots to prove the standards.
We actually are now using our standards in production applications.
So, the participants that we are working with today include point of care
technology vendors, electronic medical record vendors, EMR vendors, pharmacy
benefit managers or PBMs, health plans, hospitals and pharmacies or pharmacy
We have twelve participants in production today that are actually using our
standards or presenting information to physicians. This is one thing that
I want to stress because I think, you know, whenever I present or talk
sometimes at NCPDP or in other areas, people don’t realize that we are actually
doing this in the industry. I have seen a statistic that says 3 percent
of the physicians are using electronics on prescribing applications. I
believe it is higher, just based on our experience, but I don’t have any
statistics to disprove or prove that statistic just based on our experience in
our roll out.
We have ten additional participants, who are certified on our transactions,
ready to move into production as soon as they are ready to roll out to their
physicians and we have eight participants currently certifying on our
transactions. In addition, we are also working with other participants,
who are developing flow transactions. They have signed NDAs. They
have received our implementation guide. We just don’t have them in the
implementation process yet.
So, there are a lot of people who are already using these standards out
there. I am going to briefly go over the standards that we have talked
about and have been testified to and talked about from other testimony just to
help you understand which things we are going to address and which ones we
aren’t going to address.
The first thing we have talked about in our testimony is our master patient
index or our MPI. This is proprietary software that we developed in order
to locate a patient or a patient’s health plan, to uniquely identify that
patient and the payers. We have a proprietary file transfer to load that
data from the health plans and the PBMs.
The next transaction is the eligibility transaction. We utilize the
X-12 270, 271 inquiry response and I believe we use it in a correct way and we
do answer the questions that are asked, not just the “yes/no.”
We have testified to that. Formulary and benefit information file
transfer, that is the RxHub developed file transfer that we preload the
formulary and benefit information to the point of care application.
Information is provided by the health plans and the PBMs.
Then we have the formulary and benefit real time transaction to the
PBMs. That is the transaction that we developed that was supposed to be
utilized in a real time mode for a physician application to ask the question of
a PBM or health plan, is this drug covered.
We don’t have a lot of utilization of that transaction because the
technology vendors want to preload the information ahead of time. Patient
medication history, we actually use multiple formats and the formats that we
develop using NCPDP SCRIPT standard, the first format that we use, it basically
looks like — it is another script type of transaction, if you will.
It provides information from the PBM for health plans of medications that
have been paid for by that health plan. We also utilize a couple of
different HL7 formats. One is the ORU transaction or messaging format and
that is basically just a report format. I am not an HL7 expert. So,
Scott, feel free to correct me if I misspeak on any of this stuff.
But that is basically just a report where we take in — we receive a request
and then we send an NCPDP-like transaction to the PBM to get the information
back from them, but then we translate it back into a report format of
medication history for the physician. This is used primarily in the
We also have a message format that we developed called the RDS message
format or that we utilize where we actually translate between the NCPDP-like
translation data format into an HL7-like data format. That is for those
vendors that want to utilize the discrete data versus the report format.
That is where we are participating in the effort to map between NCPDP SCRIPT
and HL7 because we have already done that.
The provider list is a list of participating physicians and pharmacies and
we receive information from those participants and share it back out to the
pharmacies and physicians. Then NCPDP SCRIPT we use for prescription
So, the standards that we are talking about that are right now were
developed by RxHub but we want to bring through — NCPDP is the organization
that we have chosen because it makes the most sense — are the ones that I have
highlighted and those are formulary and benefit information file transfer,
patient medication history, the NCPDP-like format because the HL7 formats
already exist and they are accredited, and the provider list format.
I want to talk about formulary and benefit file transfer and I am just going
to talk about the data at a high level and what it consists of and then the
process that we are planning to follow with NCPDP. This is our batch file
transfer. Again, it is information that we have received directly from
the PBM or health plans. Today, that is what our format provides.
We transfer the information to the point of care applications and the RxHub
process basically breaks out based on contracts the information that can be
transferred or based on preference by the technology vendors. The
technology vendor says, you know, I don’t need the entire list of formularies
from a particular health plan because I am just working in a certain region
right now. We can break that out for them. But the standard doesn’t
need to support that. That is a function that RxHub does.
The formulary data that we are talking about is formulary status and
alternatives. There are some non-listed drug indicators just to help the
technology vendor understand, you know, if there isn’t a designation, a
formulary designation in the file, what should I do with generic drugs, what
should I do with branded drugs.
And then some text messages. The benefit data is drug requirements for
coverage exclusions, prior authorization, step therapy, age and gender limits
and quantity limits. There is also the ability to provide some text
messages with that information and it also provides co-pay information and that
co-pay information can be specified as tiers or percentages or flat dollar
amounts. It really depends on the benefit design and each health plan and
PBM can have multiple ways of defining their co-pay information.
It is not the specific amount that a patient would pay for that drug in that
instance. It is generalized based on general criteria, such as brand and
generic indicator or formulary status or whatever is in other indications the
health plan uses to determine the co-pay.
I want to clarify prior authorization a little bit here because what we
provide as far as prior authorization is concerned is information that a drug
requires a prior authorization. It is not part of the formulary. It
is part of the benefit information and it just tells a physician that a drug
will require a prior authorization.
The format does not allow the physician to send an electronic request for
prior authorization. I also want to clarify that for co-pay we actually
just added this to our format probably about six or seven months ago and we are
actually in the process of completing a pilot for utilization of that
As far as submitting this to NCPDP for approval, this format, there is one
file format that kind of contains a bunch of different file structures that
supply this information. As Lynne testified at the August 11th workgroup
meeting, a task group was proposed and formed and I did volunteer to lead that
task group. I think they took me up on it. Just so you understand
the process, the task group process, which differs from the DERF process, which
I am going to talk about for a drug history, and when there isn’t already a
standard that can be modified, a task group is formed and a task group is open
to anyone who wants to participate. You don’t have to be an NCPDP member
and I suspect that we will have a lot of participation in this task group, at
least I hope we will.
What we plan to do is just basically take our format forward and, you know,
get feedback, have a series of workgroups and such to include other formats, as
well. The time frame for approval, we testified — we are estimating at
two years. I think that is optimistic. It really just depends on
how well the RxHub format supports the rest of the industry’s needs that have
not participated in our workgroups.
I can tell you that we had a series of workgroups for formulary that didn’t
take very long, but they were intense, if you will, because you need to be able
to support the flexibility of PBMs and health plans to have different
information but you also need to have it structured enough so that the
technology vendors can understand the information and present it in a common
So, I am hoping that it goes well, but I want to say that that is optimistic
because the process does take awhile. You know, the first year is
probably just the task group working through getting consensus on the format
and then the process after that is basically you take it forward to the actual
workgroup, which I am assuming will be Workgroup 11, which is the pharmacy
workgroup, and you submit it for approval by the entire workgroup.
I suspect that this will take some time to get through the workgroup as well
then. The process — and this is where I will need Lynne to make sure I
am speaking correctly on this — is that if you propose it as a new standard
and it doesn’t get through in three workgroup meetings or three tries, it dies
and needs to be resubmitted again.
We are hoping that won’t happen, but — that is where I am saying we are
being optimistic. Once it is approved by the workgroup, it then goes out
for a vote or adjudication by the entire membership and if the entire
membership — I think that is a majority vote, but the membership can vote, you
know, yes, with comments or, no, with comments and then it comes back to the
workgroup. You review the comments. You may end up modifying it and
then it goes out for adjudication again.
The workgroups are held quarterly. So, that is why the process, you
know, can take longer because you only have those quarters. So, we are
hoping to get consensus from the task group and then, hopefully, when we get to
the work group, it will go quickly.
Okay. As far as patient medication history, again, currently the
format is supplies medication history from the PBM or health plan and it
includes medications paid for by the PBM or health plan. The data that is
included is patient information, the medication dispensed and that is at the
NDC level, quantity/days supply and other information about that medication
that was dispensed, the dispensing pharmacy and the prescriber.
We did use the NCPDP SCRIPT format to develop this transaction. So, it
basically uses all of the same information that would be on a new
prescription. The process for bringing this through NCPDP is to submit a
data element request form or DERF on which basically is a request to change an
existing standard. So, that is what we are planning to do and we are
planning to submit our DERF at the November 2004 workgroup. Time frame
for approval was estimated at one year. So, assuming that it takes us
those three tries to get it through and it goes out for the vote, and maybe it
goes out for a vote twice and so estimated completion would be January of 2006
Again, that is the optimistic view. What I am planning to do with this
one instead of just presenting the DERF, I am planning to bring some of
industry participants together ahead of time to review our format so that they
are not seeing it for the first time at the November workgroup. We have
done that with some other DERF changes that we have requested to Script and it
has worked really well. So, hopefully, we will get some consensus ahead
of time on that.
Same process, so I don’t need to go over that again. If it is
approved, great. If it is not in three, then we have to resubmit
it. As far as the provider list, again, this format provides a list of
participating physicians and their demographics from the point of care
application vendors and a list of participating pharmacies and their
demographics from the pharmacies or pharmacy networks. We distribute that
information — you know, we store it and we distribute it back out to all
participants. The data is basically the physician name, the
IDs and their location, demographic information about the physician.
Pharmacy is the same thing; name. We use the NCPDP provider ID, the
location and demographic information.
The format that we use, again, was a modification to NCPDP SCRIPT and it
actually was a format that had been proposed at one point at NCPDP called the
provider broadcast. We actually modified that somewhat in our utilization
and NCPDP has requested that ProxyMed resubmit that change request or start
that process again. I think they were in the middle of it and they
withdrew their request at one point.
So, we aren’t actually going to champion that one through but we have
requested to work with ProxyMed and I am assuming SURE Script will probably
work with us as well, as the three of us have all had different versions of
that transaction. Same process. It will be a — I am assuming it
will be a DERF or a change request to NCPDP SCRIPT. We are estimating —
I would estimate one year from the time of submission, but since I can’t speak
for ProxyMed, I don’t know when they are planning to resubmit that.
In conclusion, our recommendations to NCVHS is we would recommend
utilization of the RxHub standards that are in use today for demonstration
projects in January of 2006. The reason that, you know, we are requesting
that is that these transactions again are being utilized today and it will take
some time to get them through the approval process at NCPDP and it probably
will be helpful for us to be, you know, using them in the demonstration
projects as we are taking them through the process.
We, of course, would like you to actively involve RxHub in these
demonstration projects and we certainly volunteer to be part of them. We
recommend that HHS adopt the RxHub standards under MMA and we also would like
to recommend that the NCVHS committee stay involved in NCPDP’s consideration of
the RxHub standards. You know, we would like to provide you with our
plans, with the DERF, et cetera, just to keep you abreast of what is going on.
DR. COHN: Teri, thank you. I am sure we will have a couple of
questions for you.
MR. BAMBERGER: On behalf of MediMedia, it is my pleasure to testify
before the subcommittee again and make myself available for questions. As
I testified in July, we distribute the InfoScan Formulary Database, which
contains 3,400 health plan, PBM, PPO and self-insured employer formulary.
In addition, most of the plans associated — in addition to most of the plans
that are associated with RxHub and CAQH, we represent many of the smaller plans
and PBMs, who thus far have not chosen to affiliate with RxHub and CAQH.
As I testified as well, we have been providing the formulary database to
electronic health records and e-prescribing software companies since 1997,
predating many of the standards that we are discussing today. Our clients
include many of the e-prescribing and EHRs that have testified to this
committee, plus many of the larger EMR vendors, like Web M.D.s Medical Manager,
E-Medical Systems Medicalogics, Serner(?), X-Gennisis(?) and others, a
veritable whose who of the mature health care information technology providers
that are out in the marketplace.
My goal is to represent the interests of all of these stakeholders in
providing you with an alternate but relative viewpoint in the areas of
formulary structure and prior auth. In terms of the submission formats
and formulary information, as we testified in July, our health plan and PBM
clients provide us with a variety of formats, including flat files, Excel
files, word files, e-mail calls, faxes, whatever it takes to get the
In the absence of standards for collecting this data, we have become experts
at doing so and validating the information by aggregating it into single
standardized data sets. A submission standard for us would be desirable,
but not necessary. We have been there and we have worked through that
Frequent formulary decisions that we receive, frequently they are before
they have reached the claims processing system, working with the P&T
committee and we are working at a drug level, rather than an NDC level.
So, we are working at a much higher decision level when they are making
decisions on the P&T.
That said, we have been cautioned that if not constructed intelligently,
standards for submission could limit plan creativity or put a ceiling on types
and the quantity of information. Therefore, we believe that submission
format should allow for maximum flexibility and creativity on the part of the
plan. In May, the CAQH testified that they had developed a standard
submission format and we would be open to the committee’s recommending CAQH
submission formats as a starting point.
However, it is possible that they would have to be expanded as some of our
clients submit a great deal more information than I understand is submitted to
CAQH. Such an expansion would allow for the previous mentioned need for
flexibility in the future. Furthermore, I would argue that the submission
format should also be developed and maintained by an ANSI accredited SDO, like
NCPDP or HL7.
We collect, validate and aggregate and maintain out rich set of data by
providing it to clients on a weekly or monthly schedule, depending upon what
best meets the needs of the software company and their physician
prescribers. This has evolved over time and, in fact, at this point, we
are testing some more real time in some very specific situations.
We believe that our distribution electronically in a batch format is
what is required today and we are looking for the future to build that, what is
needed. As I testified in July, I think that an ongoing data set could be
standardized to minimize custom coding requirements and we would be happy to
provide our formats as a starting point or work with any formulary aggregators
on these formats.
Factors to consider when thinking about this recommendation, you know, the
need for a wider data set than maybe is seen today in some of the alternate
standards, that our database again predates many of the standards. So, we
have had to — there is going to be a need or a potential to force vendors to
reconfigure their existing software over time to fit this. So, we need to
build in the adequate time frames to allow for any of those transitions as the
data may change or we certainly don’t want to penalize the existing vendors who
have been out there in the marketplace.
In addition to drug status and tiers, we also have step therapy rules,
relative price ranking, drug policies and some of the criteria for prior or
preauthorization, as well as therapeutic guidelines and advisories and
restrictive notes as part of the database. The content and the breadth of
a formulary, as well as the definition of terms varies by plans. This
creates a challenge for EHRs and e-prescribing companies, just as it has for
the processors, who have had to try to adopt these new definitions into their
legacy claim systems.
While there is a benefit in terms of standardization, again, we don’t want
to dampen the creativity and competition between plans as we move
forward. Nevertheless, this will be a tricky process that will require
input from all stakeholders to keep costs in check while improving outcomes and
it is important that the plans and PBMs have that flexibility that is needed
for their respective customer base, employer base, et cetera.
Developing formulary standards and messaging standards should be led by
plans with input from stakeholders while it isn’t an ANSI accredited SDO, an
organization such as AMCP, the managed care pharmacy organization, may also be
an entity to include in the process as they look to the future of where
formularies are headed and where they see the business is headed.
We agreed that the testimony at the July hearing was one of the biggest
challenges in e-prescribing is obtaining the patient’s demographic information
from the practice management system. However, getting this information
into the EHR or the e-prescribing application only overcomes one of the
barriers. The other is linking the formulary with the health plan card
information obtained by the practice.
You see, often an employer will self-insure for a prescription benefit,
contracting directly with a PBM. In this case, the patient ends up with
two cards, one for the pharmacy benefit and the other for the health
benefit. The office staff collect only the information require to submit
the medical claim today and the information on the health plan card, which may
not be that information on the health plan card.
That may not be necessary or even the link to the formulary for that
particular patient. Used as an example, United Health Care, where they
have got tens of millions of patients and members in their system, but as many
as 25 percent of those patients may have another benefit, other than United
Health Care for the pharmacy benefit.
RxHub has worked around this by creating the master patient index.
However, not every PBM or plan feels it is in its best interest to join RxHub
and I question whether or not forcing these affiliations is wise or
necessary. In the standards, there is a way of identifying health plans
and identifying and displaying this number on the plan cards. This would
facilitate a linking of a specific plan to a formulary. It could be used
as a way to create this linkage at an individual card level for an individual
Changing the topic to prior authorization or preauthorization, it is, of
course, a process of obtaining a certification or authorization from a health
plan for a specified medication or specified quantities of medication. It
also involves appropriateness review against preestablished criteria.
Those criteria can vary by plan and within a plan by drug. The number of
medicines that require prior authorization vary greatly between plans.
The process of approving or receiving, obtaining an approval for a prior
authorization is an onerous one by design. Its purpose is to encourage
appropriate use of medications, most likely to have certain risk factors and
the approval criteria is generally developed and endorsed by the plans P&T
committee based on information from the FDA, manufacturers, medical literature,
active practicing consultants, physicians and appropriate external
Failure to obtain a preauthorization often results in a financial penalty
for the patient. So, physicians are highly reluctant to prescribe these
drugs thus labeled. In fact, almost any physician’s office that has even
a moderate number of managed care patients will tell you that the prior
authorization is tops on the list of the pain scale. For this reason, the
e-prescribing system that can reduce that pain of prior authorizations will be
making a substantial positive impact on the practice.
I also believe that if e-prescribing becomes more commonplace, the rate of
unformulary prescribing will increase making prior authorizations a more
attractive cost containment tactic. Automating the process will allow a
clinically appropriate prescribing. In today’s paper world, the
prescriber doesn’t know if the drug is on prior authorization or not.
While he or she quickly learns that it is likely that growth hormones or any
fungal agents have been designated requiring prior auth, with a trip team or
the therapeutic categories that are less consistent across plans. For
example, COX(?) 2s or proton pump inhibitors where availability or lower cost
options have created prior authorization restriction on many medications in
Should the office want to continue with the prior authorization, the staff
would obtain the form from the plan web site. The form has a series of
questions on it to help the clinician determine if the description is medically
necessary. While it is more complex than “yes” or
“no,” according to our analysis, the example presented by MedCo in
the July testimony was on the complex end versus something more simple that has
more “yes,” “no” answers in questions.
The fact is computers were designed to automate paper processes like
this. At a minimum when the prescriber is using the software solution
that leverages a formulary database, these drugs will be flagged as requiring a
prior auth. The physician can access the notes and restrictions on the
drug and make a decision as to whether or not to put in practice the pain of
requesting the formulary exception. But that is only a first step.
An algorithm can run either on the software system or interactively that
allows the physician to enter diagnosis codes, answer questions and document
his or her clinical judgment. Some plans for some drugs might issue an
approval code right at this moment from the physician side. In other
cases a form would be created either electronically or paper and transmitted to
the plans’ clinicians for approval. When the approval is obtained, the
code can be transmitted with the prescription to the pharmacy where it can be
included in the system and transmitted to the prescription or payer.
Now, agreed, telemarketing guy — and I can make that sound really simple
and I am sure that there is a lot of process that goes into actually working
and connecting all the “i’s” and the “t’s” to make sure
that the dots and crosses are all in the right place. But we would offer
to work on a volunteer basis to create a demonstration project maybe with one
of our software vendors and a health plan to start to work through that process
much the same way we started working through formulary information back in
DR. COHN: Thank you both very much for some very interesting
Teri, maybe I will start out with a question. I am actually looking to
see if Lynne — I am not sure if the question is for your or for Lynne, but let
me just ask it. Obviously, I am looking at a time line for working on
standards within the NCPDP world. I am very understanding. I have
lived a lot of my life in standards work and I know that it takes awhile to
move things along.
I also know that at times the government has an opportunity to help
accelerate sometimes through financial or other arrangements activity and I was
thinking, for example, the HL7 function, you know, EHR functional standards
work that occurred recently where appropriate funding seemed to accelerate
development of at least a draft standard for trial use.
Now, I am actually wondering — I don’t know if we asked you this before,
Lynne, whether that would be something that might be helpful here to move
forward. I guess what I am looking at is this recommendation that
standards that occur that are being demonstrated and tested in January 2006 are
RxHub standards as opposed to NCPDP standards, draft standards, for trial
use. I don’t know if — Teri, you are the one to answer that question or
whether it is Lynne. Maybe you can advise.
MS. BYRNE: Well, as far as whether they are draft standards for trial
use, I am not sure that NCPDP — does NCPDP have that concept?
MS. GILBERTSON: It is ANSI.
MS. BYRNE: My only issue is I am not sure how far — I mean, if we
take wherever we are at at that point, it might be difficult because the
technology vendors who are already using these transactions would have to
redevelop to essentially new standards that still may not get approved.
So, that would be my concern.
DR. COHN: Lynne, is that your comment, too?
MS. GILBERTSON: It is kind of a mixed bag. I mean, for the
demonstration project if you want to show that there are opportunities today
that are working, if the standards turned out to be slightly different by the
time we got to the demonstration time frame, you would be expecting the
demonstration folks to have moved to that. I don’t necessarily think that
that would be a fair return on investment because they would want to wait until
things are locked down maybe before they actually moved.
As far as monetary assistance, I think the ones that have — the standards
that have been proposed to come forward are either based on Script, which
shouldn’t be that hard of a fit. There is always the possibility that
consensus could not be reached. So, that may be the time factor versus do
we have something to work forward on. I think as Teri said, the pre-work
with the different organizations, who are involved in this phase, for example,
that will go a long way to showing everybody all the gaps and things like that
between themselves and the next guy and all that.
This formulary — and we have two organizations sitting right there and you
also mentioned ANCP, which we are aware of as well, who are already in this
space. So, once again, it becomes the collaborative effort that
recognizing that what we come out with has got to be consensus between users of
the data. So, yes, it is kind of painful to, you know, decide where you
are going to give up pieces and add your pieces and your interests and things
like that, but I think we know the charges. We have to come up with a
standard for the industry.
So, that is the goal and we won’t back down until we have that goal.
You know, at some point proprietary just isn’t going to work in this
environment. I am not sure if I answered your question. I am not
sure where we could use the funding. I think it is the same pool of
volunteers to get the job done and the incentive is we have to be ready for the
2007ish time frame.
DR. COHN: Well, maybe my understanding is a little different than
yours. I don’t think of the 2006 pilots as demonstrating what we already
know to work. I mean, those are the — the Secretary has an opportunity
to decide if something is industry standard works and just to prove
it as a standard without pilot. The point of the 2006 pilot is to
identify standards that would likely become finalized in 2009 and test those
That is why I was sort of struggling with what I was hearing in terms of the
MS. GILBERTSON: Well, part of that, though, is what we were talking
about earlier, that for one thing we have a new payer in this realm and that is
Medicare and the prescription drug benefit environment. So, we can’t just
assume that because, for example, all the PBMs and the health plans that are
participating in the existing environment, that whatever the Medicare rules
might be will fit. We hope so because that would be a great place to
start from without having to add a whole bunch of new requirements.
The other thing is the demonstration projects may not be necessarily to
demonstrate that the standards in existence work, but what return on
investment, what educational opportunities, what studies can be done of those
as part of the demonstration that would help us going forward as part of, okay,
we are going to name X as a standard or whatever and here are the benefits that
these studies showed or here are, you know, the findings. So, I think
that might be also part of the demonstration as well.
DR. COHN: Steve, do you have a comment?
MR. STEINDEL: I have a question, but I have a comment on this
particular discussion. I mean, obviously, since we were just handed the
RxHub specifications this morning, you know, I can’t say I had any chance to
really read them or — but I did glance through them very quickly and the first
thing that had struck me was this was basically an EDI standard already and
then I glanced through it some more and I realized it was based on NCPDP,
including NCP envelope, et cetera.
My observation on this is essentially this is already an NCPDP
standard or set to be an NCPDP standard that is instituted just strictly by
trading partner agreement and moving to full standard may not be that
difficult, as Lynne indicated because it is very close.
MS. BYRNE: Our experience in the development of this is I agree with
Lynne in that it is probably more getting consensus on, you know, the data
formats, et cetera. You know, our experience, when we implemented the
CAQH health plans, CAQH had their own standard. We had our own standard
that we had developed and we did modify our standard to support some additional
information that CAQH has. So, I fully expect that as we bring this
standard through the task group and through the NCPDP, it will change.
So, you know, to address your point, Simon, maybe what we do is take what we
have got — I mean, hopefully, we have something that we are proposing in
January of 2006 and maybe for those applications that are currently using the
standard as it is and want to modify to the new standard that is being
proposed, if that is the standard you want to test. Maybe the
incentive is to help, you know, technology vendors pay for the development to
that new standard or anybody else who needs to do that.
MR. STEINDEL: This question is addressed to Brian and I couldn’t
really tell whether or not you were going to (a) participate in the development
of the formulary exchange standards, whether you would support the formulary
exchange standards that RxHub is talking about. I did get an indication
that with your customers you probably would not use that because you already
have proprietary systems in place.
Is that a true observation and what would your participation be with the
development through NCPDP?
MR. BAMBERGER: In the July meeting, I believe, I said that we would be
willing to participate and would look at it. We don’t understand enough
of the RxHub information, haven’t seen the document that you were reviewing
earlier today, to see how well it suits the needs of what we see the industry
needing today and in the future.
So, I think that we would be — we have said we are willing to participate,
that we are very careful not to take steps backwards with the
existing vendors that we have to support them for as long as they need the
support because that is the bread and butter that is out there today, the
20,000 physicians we have using our file right now as the key for the future of
the file and the work that we have done with the plans as well.
MS. BYRNE: I would like to make one comment on that as well.
When we implemented CAQH again with RxHub, the technology vendors that were
using CAQH were in a lot of instances also using the RxHub formulary. So,
they were very pleased to find out that they could move to one format versus
having to support two. So, I think that, you know, as we move forward, we
need to recommend one format, again. I mean, that is the idea of a
standard so that they don’t have to develop to multiple proprietary standards.
We are expecting to work with, you know, like I said the task group is open
to anybody, members or not members of NCPDP. So, we are hoping that many
media will participate on the task group as well as others in this room.
MR. ROBINICK: Bill Robinick with ExpressScripts.
I just wanted to add on that note — you said since 1997, which, you know,
when electronic prescribing first got started, that is what everyone
used. When we joined with MedCo and advanced to create RxHub, it was
partly because we didn’t want to develop to 20 different point of care vendors,
but it was also because we didn’t find the level of granularity that was
available at the time to be sufficient.
So, if you are talking about plan level formulary, it is one piece of
information. If you get to group level, it is better. If you get to
member level, it is even better. So, we don’t see the two as necessarily
mutually exclusive, but not every plan will join RxHub or even if they don’t
have to join RxHub to use the standard, they may not have the size or scale to
do any development work. So, there is a place for basic plan level
formulary, but getting to a more detailed level is better information.
So, it shouldn’t be seen as either or. There is really a place for both.
MR. REYNOLDS: Teri, I need to understand how this works a little
more. I know your formulary and benefit information file transfer, you
talk about a batch file transfer to the point of care application. Are
you talking about during the day if they know that X people are coming in
tomorrow, do you do it based on them or do you download an entire formulary for
the United States?
MS. BYRNE: Actually, it is the entire formulary and it is based on,
again, on the technology vendors need. So, a PBM or health plan can
distribute all of their formularies to a particular technology vendor or some
of them based on, you know, if that vendor is regionalized versus across the
U.S. However, the standard basically just provides a way to send
formulary information. So, you can send two files or you can send 200
files, 200 formularies.
MR. REYNOLDS: Then further, when you talk about patient medication
history, that, though, is an individual situation. It is not batch.
Would it be real time?
MS. BYRNE: Right. It is real time but it is not during the
prescribing process itself. When the PBMs testified in July, they talked
about the RxHub model and that was for the plans to download the formulary
information ahead of time, do the eligibility request ahead of time due to
eligibility request in real time made, but prior to the physician seeing the
patient, so they have the information in what formula file to use.
The same with medication history. They can get that information, you
know, when the patient walks in the door, when they register, when they call
and make the appointment, whenever. But it is typically not while the
physician is prescribing because they want to use that information during the
MR. REYNOLDS: So, this standard would have to be tightly coupled with
the point of care application vendors.
MS. BYRNE: Yes. And it works just like the other NCPDP
transactions. It is a real time request and a real time response.
DR. COHN: Other questions? Jeff.
MR. BLAIR: I think it is appropriate for a thank you to RxHub, that
they have made the decision to work with NCPDP to bring these messages to
become ANSI accredited standards to the consensus process. I just want to
indicate that I think that is a very positive step.
MS. BYRNE: Thank you.
DR. COHN: I think I will second that comment. The subcommittee
members are sort of nodding their heads. As I commented before the
morning break, our work is a lot easier when everyone is working
together. There are enough issues as it is, even when we are working
together. So, it is great to see that sort of synergy developing.
Other questions, comments? I want to thank you. I think what we
will do is to move into sort of open microphone for a couple of minutes, offer
the opportunity for anyone in the audience to comment, knowing that we have
only kept you here for about ten hours today.
Yes, would you like to come up and just introduce yourself and whatever
MR. WHITTEMORE: Thank you, Mr. Chairman. Ken Whittemore with
I am a little reluctant to call this a follow-up after what I heard
earlier. Just two points. With respect to what Mr. Reynolds
mentioned earlier, I think it was before lunch, with respect to deal breakers,
I just wanted to share a little of our experience dealing with the physician
As you folks will recall from our testimony, we have to contract with both
physician application vendors and also pharmacy vendors. To date we now
have contracts with 26 different physician side vendors and while we did run
into issues of deal breakers with respect to economics or business models, the
fact that these entities would have to code to the NCPDP standard wasn’t a deal
breaker in our experience per se. So, I just thought I would share that
The other item was one that Dr. Steindel brought up earlier with respect to
ROI. We have talked about ROI a couple of times and it has also been our
experience that there is a dearth of good, unbiased, sound information with
respect to ROI for either the physicians or the pharmacies. There are a
few studies out there, I would almost classify them as anecdotal and in as much
as there is a lack of ROI information that you can point physicians to or
pharmacies to, there is a little bit of reluctance to adopt the technology.
So, to the extent which financial incentives might be offered to those
providers that would allow them to kind of get over that hurdle of not really
knowing whether there would be an ROI, I think that is a way that CMS could
help the situation.
Thank you very much.
DR. COHN: And certainly DMMA, NPRM does talk about financial
incentives that could be offered by prescription drug plans to providers to
help them with implementation.
MS. BICKFORD: Carol Bickford, American Nurses Association.
This is a question for CMS and also IHIdea or whatever the term is.
There is a discussion about inclusion of the mailing address for the
individual practitioner. My question surrounds the fact that we are doing
electronic business. Are we using electronic mail as an identifier for
those individuals as well?
MS. TRUDEL: I would have to check those facts, but I know that for
anyone who is going to obtain identifier on line, we would get their e-mail
MS. ECKERT: Karen Eckert with Medi-Span.
Just a clarification to build upon something Stan had said earlier regarding
RxNorm for people to be aware of as you make your evaluations and
recommendations tomorrow. There were several different term types within
RxNorm, talking about the ingredient name, the brand name, the semantic
clinical drugs, the semantic brand name drugs. So, I would just ask that
as you make recommendations tomorrow that you don’t use the umbrella of
RxNorm. You may use the specifics if you mean specifics or if you mean
all of them, then say RxNorm. But it would help clarify it in the end
when people are going to try and use this.
DR. COHN: Thank you.
MR. BROWN: Jeff Brown, outside counsel to our sub.
Two questions or comments. The first is talking about prior auth, if
we do talk about having a standard implemented for that, I am concerned if it
is not on the non-retail pharmacy side, 278 or on the retail pharmacy side,
NCPDP telecommunications, because their HIPAA standards, the way I read the
HIPAA regs, they are mandated. So, we couldn’t use another standard until
we went through the HIPAA modification process. I don’t know if there is
another viewpoint on that, but it is a worry of mine, especially since I don’t
think that 278 is ready for prime time, at least in the standard that is
mandated for HIPAA. That is my first question and point.
My second is getting to a comment that was made about the demonstration
project starting in 2006. I read the language and I think the key words
from my point of view is that those projects are to, quote — to show, quote,
the efficient implementation of the requirements. It is not so much that
we are testing to make sure that the EDI standards work from a technical point
of view. It is making sure that we can implement those in e-prescribing
generally, efficiently and getting back to some of the comments that we heard
about HIPAA and some of the experiences that I have had with people, not
understanding what was going on with HIPAA until the deadline was a few months
away and especially the actual providers not understanding because I think the
health plans understood early on and they had lots of people and money working
on it for roll out, but it was when the hospitals called out their lawyers and
said what does this mean to me, do you mean I have to change all of this.
I don’t want us to have that happen to the doctors on Medicare.
I think if we do demonstration projects that are well publicized, that that
can help the implementation even if doesn’t help change the standards so much.
DR. COHN: Thank you.
MR. ROBINSON: George Robinson, the First DataBank.
Just a point of clarification from what I saw earlier with Stuart Nelson’s
presentation and of Randy. It is my understanding of RxNorm itself and
the links to the NDC and the links that will be needed for the structured
product label. My understanding today is the only concept that is linked
to the NDC is the semantic clinical drug. That is the generic
representation of the drug and not the brand. So, the additional work
that I am seeing for the drug knowledge base vendors, such as ourselves, is
that we need to take the branded representation, for example, Valium 5
milligram tablets oral, as we will need to link that ourselves directly to the
NDC or directly to our name base concept so that it dispenser’s written order
can be properly communicated and so that identifier if it is selected to
be used within the Script standard itself can be there.
So, all I am saying is there is quite a bit of cross referencing work that
is ahead yet for the drug knowledge base vendors to make all of this work.
DR. COHN: Thank you.
MR. SHEATH: My name is Tony Sheath(?). I testified a couple of
times last week or last session. It seems like last week.
I wanted to bring up a point. I am not sure. It may be a little
out of your scope, but I think it is a point that I would recommend that you
keep in the back of your minds. Last session there was considerable
discussion among pharmaceutical manufacturers talking about electronic
prescribing companies that have come to them and offered the opportunity for
commercial messaging within the electronic prescribing process. I think
there was some reaction from some other folks in the audience where they said,
well, whoa, those that have contracted with us, you know, that is not allowed
and I don’t understand how that could be happening.
One of the individuals testified that we need to sort of work on the
business models. So, the point I would like to bring to the attention of
the committee is that the electronic prescribing business models call for each
individual that receives some sort of value from electronic prescribing to pay
something corresponding to that amount, that value that they receive.
So, instead of, you know, going to one source for revenue, the electronic
prescribing and electronic medical record companies, who will move into more
robust electronic prescribing, they go for revenue from multiple different
sources. Some of those sources have pretty strong negotiating positions
and, therefore, don’t reimburse them perhaps fairly. You know,
“fair” is a relative term in terms of whose perspective that you are
talking about. But what you have are, you know, particularly some of the
little e-prescribing companies are really struggling financially from the
standpoint of them not being reimbursed financially as perhaps in corresponding
to the value that they receive.
Some of the large electronic health record companies, they allow for a
prescription to be sort of documented in the process. They haven’t gone
to the robustness of being able to transmit a prescription electronically or
having some of the robust other clinical abilities or even formulary
abilities. Part of the reason perhaps might be that it is not that
attractive for them to move into that world.
I am posing this as sort of a problem without a solution. I hate to do
that. Usually when I testify in these kinds of things what I like to do
is pose some sort of a solution. One thing that was mentioned earlier
today that I agree with is it is in some way some of these companies could be
reimbursed, helped out sort of financially for, you know, maybe changing, you
know, if they are going to adopt a formulary standard that they don’t have
today, if they could somehow be reimbursed for that.
But I think it is something and I would recommend that it is something that
the committee should keep in the back of their minds is — and to quote
somebody else that said this last time, that the business models need to be
looked at closely. Perhaps out of the purview of this group, but
nonetheless something I think you should keep in the back of your minds.
DR. COHN: Thank you. An important associated issue.
Other comments, questions?
I want to thank everyone. We have had a very interesting and useful
day and I really want to thank our experts and testifiers who are surrounding
us and behind me here for what has been some really excellent testimony.
Now, tomorrow we start out at 8:30 in the morning and we go until 10:00 at
night. No. Only being partially facetious. We will be going
obviously throughout the day working through sort of issues that you have
brought up, working through items that we have identified, working through a
lot of responses that we have all received from various questionnaires.
Now, obviously, we invite you to join us. We are going to have to sort
of figure out exactly how it is going to work in terms of comment. I
mean, the subcommittee needs to get work done, but I also don’t want to unduly
prevent public comment as we go forward. So, we will be working on that
to make sure that if it isn’t continuous, at least it is there very frequently
where there is open microphone time for conversation because we do understand
that we are working on areas that are — you all are concerned about, that you
are also experts about. So, we do want to be respectful of that.
Obviously, we have had, I think, an ongoing collaborative conversation so
far. Once again, it is just going to be sort of a balance and I ask for
all of your forbearances as we go through with that tomorrow and probably on
With that, we will adjourn and we start at 8:30 tomorrow morning.
[Whereupon, at 5:50 p.m. the meeting was recessed, to reconvene at 8:30
a.m., the following morning, Wednesday, August 18, 2004.]