Transcript of the May 26, 2004 NCVHS Subcommittee on Standards and Security Hearings

[This Transcript is Unedited]

NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS

SUBCOMMITTEE ON STANDARDS AND SECURITY

May 26, 2004

Hubert H. Humphrey Building
Room 505A
200 Independence Avenue, SW
Washington, D.C.

Proceedings by:
CASET Associates, Ltd.
10201 Lee Highway
Fairfax, Virginia 22030

---

TABLE OF CONTENTS

• Call to Order, Welcome, and Introductions – Dr. Cohn
• E-Prescribing: The Software Vendors’ Perspective:
  • Tom Doerr, Wellinx
  • John Faughnan, McKesson Provider Technologies
  • Peter Kaufman, DrFirst
  • Jill Helm and Larry Trusky, AllScripts
  • Rohit Nayak, MedPlus, Inc.
  • Daniel S. Nutkis, Zix Corp.
  • Discussion
• E-Prescribing: The Drug Knowledge Base Vendors’ Perspective:
  • George Robinson, Tom Bizzaro, First DataBank, Inc.
  • Mark Millikan, MediSpan
  • Gerry Hobson, Multum
  • Open Microphone
• Follow-up on HIPAA Implementation Issues – Ed Jones, Jim Daley, Jim Schuping, and Steve Lazarus
• Follow-up on DSMO Request – Donna Pickett
• Development of Potential Letter

---

P R O C E E D I N G S (8:45 a.m.)

Agenda Item: Call to Order and Welcome and Introductions – Simon Cohn, MD

DR. COHN:  Good morning, everyone.  I’m going to call this
meeting to order.  This is the second day of three days of hearings 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 US Department of Health and Human Services on national health information
policy.

I am Simon Cohn.  I’m chairman of the subcommittee, and the National
Director for Health Information Policy for at Kaiser Permanente, and a
practicing physician.

I want to welcome the subcommittee members, HHS staff, and others here in
person.  I believe this morning we are also starting out on the Internet,
just to let everyone know.  And since are on the Internet, I want to
remind everyone presenting, as well as asking questions, both from the
subcommittee and from the audience to speak clearly and into the
microphone.  As you notice, I am very close to the microphone, because
these microphones, unless you get very close to them, they don’t project very
well.

This morning we continue our work regarding prescribing standards.  We
start the day with a session and discussions of software vendors, which I think
will take most of the morning.  After lunch we continue with presentations
and discussion to look at drug knowledge base vendors.

After the afternoon break we will have an open microphone session, and
following that we move into some HIPAA issues, a number of which were sort of
left over from our last meeting, one related to some WEDI implementation issues
and recommendations, and another relating to a specific DSMO issue that was
once again, an issue that was initially brought up at our last session, and we
are hearing more information coming forward about that.

I want to emphasize this is an open session.  Those in attendance are
welcome to make brief remarks if you have information pertinent to the subject
being discussed.  We obviously have time for an open microphone later in
the day, so obviously, there will time for open testimony from those in the
audience.

Finally, for those on the Internet, we do welcome e-mails and other
communications from you on any of the issues coming before the subcommittee.

Now, with that, let’s have introductions around the table, and then around
the room.  For those on the National Committee, I would ask if there are
any conflicts of interest relating to any issues coming before the subcommittee
today, would you so please publicly announce during your introduction.

Jeff.

MR. BLAIR:  Thank you.  Jeff Blair, Vice President of the Medical
Records Institute, and Vice Chair of the Subcommittee on Standards and
Security.  There are no conflicts of interest that I’m aware of related to
our topic that e-prescribing.

DR. STEINDEL:  Steve Steindel, Centers for Disease Control and
Prevention, staff to the subcommittee, and liaison to the full committee.

DR. HUFF:  Stan Huff, Intermountain Health Care and the University of
Utah in Salt Lake City.  I have potential conflicts, because of my
involvement as the vocabulary co-chair in HL7, and also as a co-chair for the
LOINC committee, if those subjects come up.

MS. AMATAYAKUL:  Margaret Amatayakul.  I’m an independent
consultant assisting the committee.

DR. HELM:  Jill Helms, All scripts.

DR. KAUFMAN:  Peter Kaufman, DrFirst.

DR. FAUGHNAN:  John Faughnan, McKesson.

DR. DOERR:  My name is Tom Doerr.  I’m a physician with an
electronic prescribing company called Wellinx.  And my only conflict is
that I like to sell my products.

DR. COHN:  Those members of the audience do not need to mention
conflicts, just the subcommittee members.  You’re off the hook.

MR. TRUSKY:  Good morning, Larry Trusky with All scripts.

MR. NAYAK:  Rohit Nayak with MedPlus, a division of Quest Diagnostics.

MR. NUTKIS:  Dan Nutkis with Zix Corp.

MR. MANTOOTH:  Mark Mantooth, Health and Human Services, also on
general counsel.

MS. PICKETT:  Donna Pickett, National Center for Health Statistics,
CDC, staff to the subcommittee.

MS. AULD:  Vivian Auld, National Library of Medicine, staff to the
subcommittee.

MR. REYNOLDS:  Harry Reynolds, Blue Cross-Blue Shield North Carolina,
a member of the subcommittee and no conflicts.

DR. WARREN:  Judy Warren, University of Kansas, member of the
subcommittee.  And I’m not aware of any conflicts today.

MS. FRIEDMAN:  Maria Friedman, Centers for Medicaid and Medicaid
Services, and lead staff to the subcommittee.

[Additional introductions were made.]

DR. COHN:  So, welcome everyone.

Jeff, do you have any introductory comments for the morning?

MR. BLAIR:  I just wanted to welcome the testifiers.  I wanted to
just briefly indicate for those folks that might be here today that these
hearings are in response to the directive to the NCVHS from Congress via the
Medicare Drug Prescription Improvement and Modernization Act.  The NCVHS
is charged with the responsibility of evaluating, selecting, and recommending
standards for e-prescribing.

While it didn’t give the committee a date, the secretary has a deadline of
September 2005, to set forth initial e-prescribing standards that will support
demonstration projects in 2006.  And this is the second of an array of
hearings.  If you would like to have the perspective on that, you could go
to the NCVHS Website.  There is a work plan that we are at least working
with.  It’s not cast in concrete, but at least you could get some idea of
the sense of the hearings that are scheduled for July and August and September.

And with that, I just want to thank everyone for being here.  Thank
you, several of you that filled in the written testimony.  And for those
testifiers that have not yet filled in the written testimony, can I please
encourage you to fill that in, and get that to us as well.

DR. COHN:  With that, why don’t we start our first testifier?
I’ll go from left to right.  Tom, why don’t you start out?

Agenda Item:  E-Prescribing:  The Software
Vendors’ Perspective – Tom Doerr, Wellinx

DR. DOERR:  Thank you.

My name is Tom Doerr.  I’m a physician with a part-time internal
medicine practice that is limited to Medicare beneficiaries.  I am also
one of the two physicians who founded an electronic prescribing company called
Wellinx.

I read the March 31st testimony, and also the April 14th e-Health
initiative reports with great interest.  So, my goal in my oral testimony
is to complement those sources.  I will briefly mention a few issues that
those sources addressed, but I would really like to discuss two issues that
those particular sources have underweighted or under recognized.

Near the end of the March 31st meeting there was discussion about the
Medicare Modernization Act’s directive to provide the patient’s history
along with the prescriptions.  Most of my prepared oral testimony will
address the benefits of writing prescriptions that are associated with
diagnosis codes.  And the other topic I would like to address is the
barriers to creating real time interfaces between practice management systems
and electronic prescribing software.

The bottom line for my presentation is it’s my recommendation that at a
minimum, prescriptions should be accompanied by a list of the patient’s
diagnosis codes, assuming the patient has signed a HIPAA release.  But I
think it’s optimal that along with the electronic problem list, prescriptions
should be accompanied by a diagnosis code that indicates the condition that is
being treated with that particular prescription.

The value that can be created when physicians write electronic
prescriptions and associate them with diagnosis codes is enormous.  This
value deserves some emphasis and some explanation.  Basically, with
diagnosis-driven work flow, the safety and the cost effectiveness of
prescribing can be substantially improved, and that can happen in four ways.

First, we can automate the checking for drug-disease interactions, and also
deliver messages about omitted drugs that would typically be appropriate for
those diagnoses.  And that can happen at several points in the
prescription process.  Ideally, this will happen at the point that
prescribing decisions are made, and then that way the prescribing decisions can
be reconsidered and changed most efficiently.

Secondly, we can improve the level of discussion in communication between
PBMs, pharmacists, and payers, because they will all be able to better
understand the patient’s health condition that is being treated with a
particular prescription.  So, this can facilitate understanding, and
improve the appropriateness of communication among the stakeholders.  I
think unnecessary phone calls can be prevented.

And the third point I would like to make is that we have an opportunity to
dramatically lower the costs of treatment if we can deliver intelligent
messaging about high quality, cost effective pharmacotherapy, and if that kind
of messaging is delivered at the point of prescribing.

The Medicare Modernization Act clearly calls for electronic prescribing
standards that accommodate messages about lower cost therapeutic
alternatives.  I think it’s going to be more effective if these messages
are delivered to doctors that are specific to the diagnosis that the doctor is
trying to treat.

It is optimal to give physicians objective comparative information about
the efficacy of treatment alternatives, and about the trade offs between the
costs and the efficacy and the adverse effects for that particular diagnosis,
rather than having it focused on the drug.

We physicians need something analogous to Consumer Reports for doctors
integrated into our work flow.  This kind of content must be transparently
evidence-based in order for it to be credible.  And the most logical way
to do this is to promote a diagnosis-driven work flow.  And the most
effective way to promote a diagnosis-driven work flow is to set a requirement
that the standard is that the prescriptions are associated with diagnosis
codes.

My company implemented diagnosis-driven electronic prescribing in a 70
doctor group in Missouri several years. We measured a significant improvement
in the prescribing behavior.  So, we set out to study the impact of this
system in a more rigorous, controlled trial.  This was a trial of about 38
physicians who were practicing at an integrated delivery network in Wisconsin.

So, in that study, compared to the control group, the cost for new
prescriptions and their refills of those new prescriptions decreased by 8.3
percent during the first six month period.  And during the next six month
period the cost to the payer decreased by about 16 percent.  In this study
the payer had a 40 percent market share, so the savings to the payer were $584
per doctor per month, averaged over each of the first six months, and then
$1,101 per doctor per month, averaged over months 7-12.

Then we found there were almost identical improvements after the system was
implemented in the control group in that integrated delivery network when we
completed that study.  Overall, in this particular study the payer
measured their return on investment to be about 6 to 1.

These results were then subsequently confirmed independently in a study of
100 doctors that was sponsored by a payer in Maine.  Now, both of these
studies are limited by the small size, they are 100 doctors or less, and they
were focused on commercial population, because the sponsor and payer in these
studies was not CMS.

Nonetheless, there is a remarkable consistency of the results here.
And when we consider all of the physicians who are using our system, looking at
all the prescriptions that are written, this is a system with a
diagnosis-driven work flow that has something analogous to Consumer Reports for
doctors integrated in the work flow, 65 percent of the prescriptions that are
written on this system are for generic medications.  That could be
referenced to a national benchmark that is probably somewhere around 50
percent.

And in this particular slide, if we follow the blue curve, that is the very
first 70 doctor group in Missouri.  They started with a generic
prescribing rate of 57 percent, and they settled in about 10 points higher, the
upper 60s.  The green curve is the group in Wisconsin in the integrated
delivery network that started at53 percent, and they are up around 64
percent.  And then the red curve is a study of a high performing group in
Maine.

Now, during this time period several important brand name drugs became
available in generic form such as Prilosec and Prozac, so that it is important
to have a reference group or a control, and that is why we did the control
trial in the IDN in Wisconsin.  And that’s why we also did a control study
in Maine.

I think that requiring diagnosis codes will not impose an undue burden on
physicians.  We doctors are routinely associating diagnoses codes with our
orders for blood tests, with our orders for radiology tests, and with our bills
for our professional services.  It’s not that difficult to make that work
flow pretty smooth.

And many of the electronic health record companies already have diagnosis
codes associated in their work flow.  If you looked at just electronic
prescribing companies, three of the companies on the market already are
generating prescriptions that are associated with ICD-9 diagnosis codes.

I think most of us are aware that that less than two months after the
Medicare Modernization Act was signed into law, the OMB estimates in the cost
was about a third higher than the costs that were estimated by CBO.  So,
we have to figure out some way to reel in the costs of pharmacotherapy without
compromising the quality.  And I think that nationwide adoption of
diagnosis-associated prescriptions will enable the promotion of fiscally
responsible, high quality care.

The next topic I would like to discuss is this issue about the interfaces
that are needed between electronic prescribing systems and practice management
systems.  And I think this is probably going to be the greatest hurdle to
widespread adoption of electronic prescribing.

As far as I can tell, this is a matter that has not been discussed very
much before this committee, and it was only briefly mentioned in the April 14th
e-health initiative report, which is why I bring this up then.  If you
don’t have a real time interface between the doctor’s scheduling and billing
software and the electronic prescribing system, the office staff has to do
double entry of the patient data before the doctor can write
prescriptions.  And this is usually too time consuming to be practical.

My written testimony describes some of the alternatives to real time
interfaces that are sometimes used, and the problems that are associated with
each of those.  The current standards for interfaces between practice
management systems and electronic prescribing systems are inadequate, even when
the practice management system and the electronic prescribing software are
compatible with HL7 2.4, which is what is commonly in use.  It takes at
least two to three person days of labor, and sometimes a week for each of the
two vendors to complete the transactions.

So, Version 3, when it’s available, will reduce this labor
requirement.  And most importantly, Version 3 of HL7 will require
performance testing.  But it doesn’t address the other aspect of the
problem, which are the exorbitant expenses that are often charged to the
medical groups who need these interfaces.

The total cost of these real time interfaces typically run between
$5,000-10,000, but in some cases they are as much as $30,000.  And when
you get to higher figures of in between $10,000-30,000, they are usually
arising from anti-competitive business tactics, or price gouging, or other
reasons that are discussed in my written testimony.

But even if you are at the level of a $5,000-10,000 charge for the
interface, that’s just a huge barrier for adoption of electronic prescribing
for the 1 to 2 doctor offices that make up most of the medical practices in
this country.  So, we have an issue here with business practices, and we
also need to hoe to hasten the completion of the work on the HL7 version for
interface.

I have a few other brief comments about Dr. Tice’s (?) testimony on
March 31st and the April 14th e-health initiative.  I just want to touch
on three of those issues.

The first issue involved the code sets for drug names.  The government
has encouraged the existence of several competing code sets for drugs at higher
level of abstraction than the NDC codes.  And we saw this when the
government forced the divesture of MediSpan from First DataBank.  And this
encourages competition, but it also causes some barriers to interoperability,
and thus to rapid adoption of electronic prescribing.

My understanding is the NCPDP protocol doesn’t resolve this
interoperability issue.  This is a significant barrier to smooth
communication among stakeholders, specifically in between the prescribing
applications and the pharmacy network, and the pharmacy management systems and
the PBMs and practice management systems.

Ideally, there would be one, well defined, hierarchy of numeric identifiers
and associated text descriptions that is supported by all the payers and the
commercial drug database vendors.  I think RxNorm promises to be the
solution for this, but it’s very important that we resolve this issue.

We have been blocked with some attempts to do some implementations because
of these incompatibilities, and had to just turn away certain customers.
It’s really a barrier, the way the system is currently in place.

A second concern I would like to emphasize is the welter of state
requirements for the detailed printed facts and electronic prescriptions.
Each state pharmacy board has its own approach to really the same essential
needs.  The result is a bewildering set of sporadically changing
requirements that act as barriers to rapid implementations and adoption of
electronic prescribing.  These really need to be standardized at the
national level.

And finally, the need for standardized tables for managing formulary status
information is becoming increasingly apparent.  Insurance companies, PBMs,
and application vendors, and formulary data vendors do not share any widely
accepted standards.  I’m specifically referring to the definition and
sharing of the formulary lists, and the special policies that go with them to
define a formulary.  Setting standards for describing formularies will
significantly reduce the cost of managing this irregular information.

So, in conclusion for my oral testimony, the challenges here really are
formidable, but pharmacotherapy is really well positioned for
improvement.  And the potential to create value here is very
exciting.  Evidence-based medicine is often admired in health care, but it
is seldom practiced.  And in order to get evidence-based medicine out of
academia, and get it into patients’ bodies, we need to give doctors
informational tools and incentives to embrace the fiscally responsible care.

We are at a crossroads here with electronic prescribing standards.  We
have an opportunity to create standards that would include a framework for
seamlessly weaving context-specific decision support information into the work
flow of prescribers, and that is a huge part of the battle.  I have
additional comments about incentives in my written testimony.

Thank you for considering my opinions, and I would be happy to take any
questions as time allows.

DR. COHN:  Tom, thank you very much.  We’ll just have discussion
after everyone has had a chance to testify.  Thank you for your testimony.

John Faughnan from McKesson.

Agenda Item:  E-Prescribing:  The
Software Vendors’ Perspective – John Faughnan, MD, MS, McKesson Provider
Technologies

DR. FAUGHNAN:  Thank you.  Thanks again for inviting me to speak
today.  I’m John Faughnan.  I’m a physician and informaticist,
product architect at McKesson Provider Technologies.

McKesson is a big company.  Provider Technologies is one part of
it.  We are by some measures, the world’s largest health care services and
health care IT company.  And we span the whole spectrum of e-prescribing
from our robotic automations, through our medication and supply management
services, through software use by nurses and physicians, and other health care
workers at the point of care.  So, we have a lot of perspectives.

For the testimony today, in addition to my written response, I have written
up my oral testimony, and I will provide that in an electronic form.  It
is based on discussions within the company, with our business partners, and in
some cases on my personal experience too.

I chose to follow the outline we were given, so I’ll proceed with the
questions, with emphasis on the latter part.

The first set of questions was to ask what were the best standards for code
sets.  And I combined the ones that we use.  In brief, to provide an
orientation, we have a longstanding relationship with First DataBank, and we
have made extensive use of NDDF Plus, their medication terminology in one form
or another really over the past eight years.  And we have seen that evolve
quite a bit over that time.

We are using now, SNOMED-CT to represent patient conditions that may impact
e-prescribing.  We expect to use SNOMED-CT and LOINC to represent
measurement and test results that also affect medication usage.  We
recognize that in the real world, you frequently have to deal with ICD-9,
however, our experience with ICD-9 as a way to represent patient conditions to
support indication and contradiction checking was quite negative.  We
didn’t find it satisfactory for those purposes.

We use NCPDP Script, and we expect to continue using it as it grows and
evolves.  I will talk about that in a minute.  And we expect to use
emerging standards for medication history, drug utilization review at the point
of care, and queries and responses and structures really, around formulary and
tiered formulary inquiries.  And we have a history of using those in the
past, but expect to be moving more standard methods for those.

We do support RxNorm as an external representation of abstract and
dispensable medications when we must communicate with the outside world.
And I look forward to seeing that develop with more support for the SIG and
aspects of clinical prescribing beyond the dispensable, including things like
the clinical route.

We are big supporters, and perhaps me particularly of the National Library
of Medicine’s effort to facilitate interoperability and coordination between
First DataBank’s commercial versions, NCPDP, SNOMED-CT, and RxNorm.

Dealing very briefly with some selected strengths and weaknesses, and I go
into more detail in the written terminology.  We know that there are some
process and structural issues with regard to NDC, and the particular versions
of NDC.  However, we also recognize that NDC is deeply embedded in all of
our applications and all the different aspects of medication
transactions.  So, we would ask that as those issues are addressed, they
are done in an evolutionary way to preserve NDC.

We know that it’s possible to use NDC codes as a substitute for a
dispensable identifier.  I mean you can use it.  You can take any old
NDC that works for the dispensable, send it, and recover the dispensable at the
other end of the transaction.  But we do look forward to a true
dispensable identifier.

And we have used NDDF Plus for quite a while, and we are very pleased with
how that is evolving.  And we applaud the collaboration that may emerge
with the National Library of Medicine to enhance RxNorm’s management of
packaged medications and medication ingredients.

We also likewise, talking about collaboration, encourage a collaboration
between SNOMED-CT, the National Library of Medicine, and vendors like First
DataBank to facilitate communication of patient conditions.  Prescribing
vendors represent patients typically in an internal terminology that has been
optimized for medication use.  SNOMED-CT has a very basic terminology
that’s somewhat different, and we need to bring the two into alignment.

We see mapping errors when you transmit patient conditions between the
medical record and a decision support system are a big issue.  We really
can’t have mapping problems.  So, we need to get this terminology
integrated so that there aren’t mapping issues.

Then moving along, is nationwide adoption necessary?  Yes.  I
won’t go into detail there.

What are the code gaps?  Well, with Script we know that it
works.  It has evolved to fit people’s needs.  But primarily it came
from the communication between the pharmacies and the payer, although it has
quite a bit of support for communication between the prescriber and the
dispenser.  And we expect that it will evolve to include a number of
important additional data elements, including acknowledgement of receipt, so
that the prescriber knows that the dispension was received in the transaction.

The clinical SIG, in other words doses, tags, frequencies, clinical routes
may be different from dispensable routes.  An IV medicine may be dropped
in the eye instead of injected in a vein.  Text messages, and in
particular the encoding of prescriber review of a warning note and formulary
notes.  What those prescribers warned about, what they overrode, and why,
so that when the pharmacists get the same kind of warning messages, they can
understand what the dispensing physician had in mind.

Other gaps we are looking forward to seeing completed is standard
transactions for formularies and tiered formularies, medication history
messaging, a politically acceptable method for identifying the patient,
electronic signature and authentication that is accepted nationally,
standardized encoding for the reason the claim was rejected, and identifiers
for durable medical equipment and supplies that would be analogous to NDC code
and dispensables, ideally consistent with RFID standards.

What are the barriers to the development and adoption standards, and how
can they be addressed?  There are many, but I have chosen to emphasize
four, a little different perspective perhaps.

Number one, customer readiness.  Are customers ready for these
transaction sets and these standard terminologies?  Even if we vendors
are, even we are personally, we would like the customers to be eagerly
demanding these solutions, rather than passively accepting them.  We hate
being too far ahead of our customers.  We like to be a little bit ahead.

In some cases, although perhaps less for e-prescribing than in some other
messaging structures, the message really does come to dictate the structure of
the application.  What has to go in the message, the way the knowledge has
to be presented, the terminologies and feedback can alter the deep structure of
the application.  And so to implement fully some of the more powerful
standards requires new applications, often difficult to reverse engineer to
existing applications.  So, that has some issues of how long it takes us
to bring on those board.

Coordination and collaboration between messaging standards and terminology
standards.  People who have worked in this world know that messaging
people and terminology people come from different points in the world.
Folks on the committee understand this.  And they needed this.  The
borders are sharp and demarcated.  There is a lot of give and take that
has gone on back and forth, a lot of painful things that we encourage support
and mutual acceptance of.

I do worry, perhaps myself more than others, about how we are going to put
together all of the diverse terminologies and transaction standards that have
been partitioned by domain.  So, we have had things that are partitioned
by procedures, partitioned by some kinds of tests, partitioned by
conditions.  And I think in practice, how that is all going to be
maintained, how that is going to fit with different messaging models.
How, since patients have all those things together, and since they overlap
domains in medicine, I think we will discover some issues there.

How can we address these things when we are concerned that licensing costs
and business models have to work for the standards bodies?  We, of course,
like the costs to be predictable for the long-term and sufficient to keep the
standards bodies healthy, but not to add a lot of extra expenses, which after
all, gets passed onto our customers.  And we would also ask for
consideration of management of international licensing costs as well.

The incentives would be graduated and include partial credit.  It’s
going to take a while to get to the full solution.  We would like to have
some good milestones with partial development.

We would like to support the delivery of knowledge of all types to
prescriber, dispenser, and patient.  We know that there has been some
concern about the volume of data that might be going to the prescribing
physician.  And perhaps some of that will be overwhelming, particularly on
the formulary side of things.

We have done this in the past at McKesson.  We have had products that
dealt with that flow of data, and we do think it’s manageable, and want it to
be provided to patients, providers, and dispensers.

In the near-term we recognize that detailed payment information with actual
co-pays costs for the patient may initially be first provided to the dispenser,
as a first step.  We also recognize that some of that information
commercially is very sensitive to payers, and sympathize with the need to
recognize their business needs.

The last thing I would like to mention is the incent to
collaboration.  Again, we think the National Library of Medicine is
playing a major role with RxNorm, NCPDP Script data, and SNOMED-CT medication
data, and we would like to see a trusted agency or organization monitor the
health of all of these terminologies and messaging organizations over time,
including relationships to the administrative code sets of ICD-9.

Thank you very much for the opportunity to speak today.

DR. COHN:  John, thank you very much.

Our next presenter is Peter Kaufman.

Agenda Item:  E-Prescribing:  The
Software Vendors’ Perspective – Peter Kaufman, MD, DrFirst

DR. KAUFMAN:  Thank you for allowing DrFirst to present the
vendors’ perspective on e-prescribing and the need for standards.  We
greatly appreciate the work you are doing, and believe this is a critical step
in improving the deployment and adoption of electronic health care.
DrFirst is a four and a half year old company, focusing on connectivity and
electronic prescribing.

There are three points I intend to stress today, and hope you agree with
their importance.  They are that connectivity is critical to efficient and
safe e-prescribing, that RxNorm should be supported for standard naming systems
for electronic prescribing, and that a standard SIG, or how a patient is to
take a medication will improve interoperability of electronic prescribing.

I am giving the vendor’s perspective.  You may hear like it
sounds like I am giving the physician’s perspective.  It’s not
because I am a physician.  It’s because we realize that in order for
this to be successful, we need to be physician-centric.  That the company
has to think about what the physician wants, what the physician will use,
because if the physicians don’t use the applications, then all is lost.

I don’t think I need to dwell on the need to improve safety and
efficiency in medicine.  With over half a million preventable events,
40,000-100,000 preventable deaths, there is room for improvement.  As a
member of the EHI Zion Implementation Working Group, I had the privilege to
help create the report that was discussed a few weeks ago by Dr. Tice.  I
support that report fully.  Today, I appear as a vendor, and emphasize
those issues most important to the vendor community.

From a high level, it is obvious that the impact of prescriptions which are
directly touched only by the physician, the patient, and the pharmacist is much
wider.  For the vast majority of providers, even most of those encountered
in surveys as using electronic prescribing, this is how the process works
currently.

Paper is involved somewhere in the chain, and communication is limited
internally to the applications being used, and perhaps the pharmacy.  I
just want to point out quickly that even if the prescription is using
electronic prescriber that is often faxed to the pharmacy and sometimes the
patient will fax their handwritten prescription, but there is often paper
involved there, and all that it entails.

Handwritten prescriptions are especially prone to errors.  Almost 25
percent are never even filled.  Having the prescription checked for
insurance coverage to allow the lowest co-pay, and sending it directly to the
pharmacy so the patient does not need to wait may improve the compliance.
Finally, electronic prescribing is far more efficient for the pharmacy and
medical office staff.

How many times have you seen this pyramid so far?  I thought the two
previous speakers were going to show it also.  So, let me just go into
what the levels that we discuss in the pyramid are.  Here is a breakdown
of the different levels of electronic prescribing as depicted on the previous
slide.

I want to point out that level six, in order to be superior to level five
in terms of prescribing efficiency and probably safety, must also contain the
connectivity to pharmacies and payers described in level five.

MR. BLAIR:  Peter, for my benefit since I can’t see it, could you
elaborate slightly more than the levels, and tell me what each of the levels
are?  Then I’ll be able to follow you.

DR. KAUFMAN:  Absolutely, I apologize.

MR. BLAIR:  Not a problem.

DR. KAUFMAN:  Level one is reference only.  Basic electronic
information only.  Drug information, dosing calculators and formulary
information available, but not automatically assumed nor prescribed.

Level two is stand alone, searching by drug name, and creating a
prescription, but there is no long-term data involved.  Level three is
stand alone plus, where you do provide basic patient information for
demographics, drug list, allergy list, and a non-specific formulary.

Level four goes to medication management, where the prior medications are
available not just for renewal, but also for checking interactions against
those prior medications.  Level five is connected e-prescribing,
integrating with a practice management system for patient demographics and
schedules, the pharmacies for new scripts and renewals, for pharmacy benefit
managers and payers for eligibility, patient specific formulary, and drug
history.

Level six is EMR, electronic medical records, integration to prescribing
into a full electronic medical record.  Our current systems don’t often
have the full connectivity of level five.  In fact, I’m not aware of any
that have the full connectivity, but they are getting pretty close.  But
some that say that they are e-prescribing really are just creating a printed
prescription and handing it to the patient, or faxing it to the pharmacy.

We believe that the next level will also involve connectivity across the
continuum of care between inpatient and outpatient, between different providers
and health systems.  And this is where connectivity really is critical.

This is not easy.  You can help.  Appropriate standards for data
exchange and electronic prescribing will ease this exchange of massive amounts
of data through markedly disparate systems.  We need a unifying naming
system like RxNorm.  We need standardized classifications, like RDFN T,
which I will mention a little later, and we need a mature descriptive language
like Scripts with a fixed standard.

I would like to mention continuity of care record at this point.  As
you can see from this slide, which shows a million different — prescription
manufacturers, pharmacies, physicians, distributors, hospitals, and everything
with lines in between them all, the e-health connectivity is a daunting
prospect.

The CCR is a small, but exceedingly important early step.  Not only is
the data in the CCR critical for health care, but the fact that so many
organizations have signed on to the concept is equally recognizable.  This
demonstrates the importance of connectivity.

Some connectivity is happening already.  You heard yesterday from
SureScripts and RxHub.  Their backbones were in place, although neither
had complete and universal functionality yet.  These both work well, but
connecting to them was not straightforward, because there was no standardized
connection.  Still, their value to the electronic prescribing process
cannot be discounted.  This is what we are talking about.

People think e-prescribing is simple.  Pick a patient, pick a
pharmacy, pick a drug.  Truly, e-prescribing is much more than that.
To illustrate the importance of connectivity, let me walk you through what
happens during the generation of a typical prescription.

Demographic information, as we heard earlier, about the patient comes to
the e-prescribing application from the patient’s practice management software,
or directly from the hospital information system, or through an intermediary
application like Patient Keeper.  Of course, it can also be entered
directly.

Eligibility, patient-specific formulary, and medication history are
available through RxHub directly from the patient’s pharmacy benefit manager.

The medication database provides a list of medications from which the
provider chooses, and he or she can access further information about the drug
from an integrated drug reference application such as Hippocrates.

As the medication and dose are chosen, the system will utilize information
from the medication database and the insurance formulary, as well as the
patient allergy list, and generate alerts regarding interactions, allergies,
and formulary coverage.  The provider can follow these alerts, or can
choose to ignore them.

The prescription is then sent through SureScripts or NDC Health to the
patient’s pharmacy of choice, if local or to RxHub to a PBM mail order firm if
desired.  Currently, pharmacies that are not accessible electronically are
accessible by fax.

Connectivity is key to the e-prescribing process.  It improves the
efficiency of the process, and improves value for the end users — physicians,
patients, and pharmacists.  The number of connections is not trivial, but
the number of connections is less important with appropriate standards making
those connections easier to build.

Now, I’m ready to discuss specific standards.  Except for RxNorm,
these are the standards which we use.  HL7 is a loose standard, because
many vendors use the optional fields for important information, which must then
be specifically coded by other vendors’ applications.

The Scripts standard from NCPDP is great, but it needs a way to describe
how the medication is to be taken.  We use NDC codes for naming
medications, but they are not really appropriate for the prescribing
process.  If the physician is trying to prescribe Amoxicillin 250
milligrams, there are multiple NDC codes to describe it.  There are
separate NDC codes from each manufacturer of Amoxicillin 250 milligrams, and
even separate NDC codes for different sized stockpiles.

DrFirst transmits an example NDC code to our communication partners, and
that example code is then exploded to encompass the other way of describing the
medication which the physician chose to prescribe.  This is not an
efficient way to pass this information.  That is why RxNorm should be
useful for e-prescribing.

This is one of the points I really want to stress today.  RxNorm is a
standard for naming each medication.  It is designed to specify the
medication as prescribed by the physician, and should allow various entities to
share information regarding this information.  Providers do not care about
the number of pills in a pharmacy stock bottle, or even the provider of a
generic version of the medication.

There is an area in my written testimony where I mention that RxNorm
doesn’t yet include specific packaging, such as for tampering doses of oral
steroids for oral contraceptives, when actually it does.  According to Dr.
Stewart Nelson at the National Library of Medicine, these packages are
considered drug delivery devices rather than medications, but are linked to the
drugs that are in them.  For example, Ortho Novum 777, contains 7 pills
each of three different combination of formulas.

RxNorm does not include radiopharmaceutics or radiology contrast materials,
and may not contain the materials used for compounding of drugs.  They
also are still adding over-the-counter medications.

We believe, however, that RxNorm is the strong solution for a drug naming
standard that should be supported and adopted.  NDFRT, or the National
Drug Trial Reference Terminology, specifies the contents of medications and
other features of the medications that is classification.  This
potentially will simplify the interaction and formulary management used in
e-prescribing applications.  Standardization of medication classes would
also allow more accurate interpretation of drug-drug and drug-allergy
interactions, and insurance formulary lists.

Current insurance formulary lists vary from payer to payer.  While not
the highest priority, we would like to see a uniform format specification, even
if the suggestion or example that would include a standardized tier rating
scheme, as well as an exclusion list, that is, if this drug is not in the
formulary, what drug is on the formulary to replace it?

The thick(?) standard is another important point I hope to make
today.  Different vendors describe patient directions differently.
Some use abbreviations, some use data elements, describing full language text,
and some just provide a text string allowing anything the provider wishes to
write without structure.

We agree that e-prescribing systems should be flexible to allow
prescriptions as the provider wishes them to be.  The point of having a
free text field is for outliers.  But most things should be described as
data elements and shared as XML.  I have included an example in my written
testimony.

As more and more physicians utilize electronic health care applications,
there will be more need to share this information.  By utilizing the data
elements, SIGs written in various applications can be converted to appear
native in any application the provider is using.  This may further reduce
errors by adding homogeneity.

An XML-based SIG standard will also increase the value of the CCR by
allowing native display and the importation of prescribing information between
e-health applications.

There are three other issues I wish to address.  I continue to see
pricing for electronic prescribing in the press listing a very high cost for
the first year.  While it is possible to spend that much, especially if
you want to buy a car, many e-prescribing applications, even with hardware, are
less than $1,000 per provider per year.

Physicians need to understand the benefits of e-prescribing.  Our
experience is that once they understand these benefits, they sign up for it,
and they are willing to pay for it.  It is a misconception that doctors
are technophobic.  Doctors are not unwilling to pay for and use
technology, but they require the technology to be simple, powerful, and fast.

Finally, there are concerns, especially among physicians, that the speed of
e-prescribing applications is not fast enough.  While nothing is faster
than using a roll-based system like ours to have someone else write the
prescription for you, let me show you how fast it could be to write
prescriptions electronically.

This next slide is actually going to show you a very quick video of writing
prescriptions on a PDA, and I will describe it as we go through it.  This
is me prescribing, but we have many users who are faster than me, because they
write 250  scripts a week, not 10.

First, I’m going to select the patient.  We can select the
patient from the schedule, or by hitting prescribe.  By typing in a couple
of letters, I can sort the list to the appropriate patients.  I can choose
a patient’s medication list, and see their current medications, and request for
a one touch refill.

I can use of favorite list, and pull up the medications on the favorite
list.  The bottom shows more information that is prescribed, and
that’s the summary screen.  I can put in a new medication by typing a
couple of letters of the name, and the list of the sort of that
medication.  I can choose the medication by double clicking, show the
drugs it is available in.

I can pre-fill the top of the formulary for common drugs, the quantity and
the number of refills.  And I have just written three prescriptions.
I can review the prescriptions and edit them after they have been written, and
I can send them all at once to their various pharmacies.

In summary, connectivity is critical to efficient and safe electronic
prescribing.  Appropriate standards will improve the process and maximize
value.  RxNorm and NDFRT should be supported and considered for naming and
classification in e-prescribing.  Finally, a SIG standard will improve
interoperability in e-prescribing.

Thank you very much for allowing me to testify.

DR. COHN:  Peter, thank you very much as well.  And also, thank
you for the demonstration.

Our next presenter is Jill Helm and Larry Trusky.  You’re doing a
joint presentation.

Agenda Item:  E-Prescribing:  The Software
Vendors’ Perspective – Jill Helm and Larry Trusky, AllScripts

DR. HELM:  Well, thank you very much.  Larry and I thank you for
the opportunity to testify before you on the AllScripts experience.  I’ll
be presenting our oral testimony this morning.

To give you a little background, AllScripts is a provider of clinical
software and information solutions for physicians at the point of care.
AllScripts has significant numbers of physicians using one or more of our
solutions.  We have offered e-prescribing software and medication
management solutions for physicians since the early 1990s.  And we are
currently creating and routing millions of prescriptions a year.

We represent a very broad physician experience.  We have single
physician and small physician practices, all the way to very large,
multi-specialty groups utilizing one or more of our software solutions.
We develop and deploy stand alone e-prescribing systems, as well as fully
integrated electronic medical records.

We work with a number of health plans that provide funding for adoption of
e-prescribing.  We have seen funding for systems and utilization-based
incentives effective in getting physicians started.  Continued use depends
on recognition of benefits by the prescriber.  We have discovered that
even free isn’t good enough.

There is much evidence to suggest electronic prescribing can improve
quality and reduce health care costs.  We have seen successful physician
groups utilize electronic prescribing to prescribe more cost effectively by
increasing generic utilization, and driving formulary compliance, reducing
costs to both consumers and health plans.

Automation of the prescribing process and the refill process has reduced
the number of calls received by clinics from pharmacies by up to 50
percent.  We have experienced tremendous growth in the number of
prescriptions written electronically using our application.  But we can do
better.

There are a number of components we would like to cover today, including
transmission and messaging, drug history, eligibility, prior authorization,
formulary, and identifiers.

Eligibility — identification and association of patients with a health
plan and drug benefit is key to driving information delivery.  Formulary,
co-pay information, and pharmacy benefit information are all driven off of
eligibility.  Today, there are member IDs, plans, and group codes for
prescription claims submission.  But these identifiers are challenging for
physicians to collect and enter at the point of care.  A standard health
plan identification number would be helpful in this process.

Prescription plan identifiers are not always highly visible on insurance
cards.  We should not have to depend on clinic staff to be experts in
deciphering prescription drug benefits.  We support companies like RxHub
that offer solutions to this via X12, 270, 271 exchange, and it’s important
that we move forward in identifying formulary and benefit information at a
patient level, as well as a plan level, because the plan level information
alone does not meet the need.

Moving on to formulary and prior authorization.  Standard formats for
formulary information need to be adopted to facilitate the consistent
presentation, and to provide as much formulary information as is practical to
the prescriber at the point of care.

Today, we take formulary information from a number of sources, including
RxHub, direct feeds from PBM and health plans, as well as independent third
parties.  We believe that health plans and PBMs will continue to offer a
variety of formulary designs and co-payment options to meet the needs of the
marketplace, but there needs to be a core set of formulary attributes that can
be used to describe preferences, or to designate those products with an
unfavorable formulary status.

Prior authorization should be communicated to the prescriber electronically
prior to the creation of the prescription.  We are supportive of
developing standards for electronic prior authorization, and for exchange of
that information between the interested parties.

Medication history — this is another transaction that we support.
It’s a real benefit to jump starting the physician experience.  It can
serve as a proxy for patient compliance.  A challenge with this is the
reconciliation of multiple sources of information to a single medication list
for a patient.

Given the diversity of sources for medication, for example, patient
reported medications, ambulatory medications, including those prescriptions
created within e-prescribing systems by the physicians, as well as inpatient
medication histories, we anticipate that we will have to support multiple
standards for these transactions.  But that just highlights the need and
the importance of having standard data definitions for the data elements within
those transactions.

CBI transmission of prescriptions and identifiers.  We feel that NCPDP
Script standard is a solid and established format for communication between
physicians and pharmacies.  It’s expandable, and it is maintained by a
well represented body.

Some of the challenges we face today are really a result of lack of
appropriate national identifiers for lapses in operational processes.
Success comes when we have a targeted campaign of physician and pharmacy
education, and a coordinated roll out of script standards.

Continued acceleration of e-prescribing will be contingent on the adoption
of national identifiers for drug terminology.  Today, the NDC code is
useful for pharmacies to communicate dispensed medication information to a PBM
or payer for claims processing and reimbursement.  It specifies the
manufacturer, brand, form, and even package size of the medication.

Current drug classifications are based on mechanism of action or chemical
structure, a classification that is meaningful for pharmacists, but not all
clinicians. A more meaningful classification for physicians would be based on
therapeutic use or indications.  Drug databases need to expand their
listings to include a more comprehensive list of prescribed treatments to
include medical devices as well as DME.

A couple of data sets which hold promise for the future.  RxNorm is a
compilation of clinically orderable drugs described by its project team as
capturing what the doctor ordered.  It contains a high level of things
that someone might find on a pharmacist’s shelf.

The NDFRT or National Drug Trials Reference Terminology from the Department
of Veterans Affairs also establishes detailed facts about active drug
ingredients, such as mechanism of action, physiological effects, therapeutic
indications, drug class, and other ingredients. This of course is substantial
information to understand the effects, group allergies, interactions, and
contraindications.  RxNorm listings are built from NDFRT components, and
together should provide substantial amounts of clinically relevant information
for electronic prescribing.

Pharmacy identifiers.  Today, identification of pharmacies is via the
NCPDP script number.  That has worked well for us, and we don’t see this
as a standards gap for prescribers.

As you have heard from other testimony, the DEA is the most prevalent
identifier used for physicians with all of its accompanying limitations.
The national identifier, like the proposed NPI would go a long way towards
enabling the entire prescribing base to transmit electronic prescriptions.

And last but not least, I would like to address some of the regulatory
challenges that we face.  Today, we support a variety of different
prescription formats to meet the state and federal requirements.  Even
within states, there are differing requirements for printed prescriptions,
faxed prescriptions, and electronically transmitted prescriptions.  We
support all of these today, but it’s a complex process.

At the federal level, the DEA does not recognize electronic signatures for
controlled substances.  States have drafted regulations without fully
understanding electronic prescribing, or the technology involved.  States
also impose requirements such as approval of e-prescribing systems prior to
their use, which can slow implementation efforts.  Accelerated adoption of
electronic prescribing depends on simplifying these requirements.

There is an evolutionary aspect to messaging standards.  Not
everything can be expected to be there day one.  Also, electronic
prescriptions need to be seen as comparable to other methods of prescription
communication in order for adoption to occur.  State and federal
regulations need to address and recognize electronic prescriptions, but not to
make the requirements so insurmountable that they are too cumbersome for anyone
to implement.

Also, all parties involved should not hold electronic prescriptions to a
significantly different standard than other methods of prescription creation
and transmission.

In conclusion, we have processes today that are creating and sending
millions of prescriptions a year electronically, but there is room for
improvement.  Standards must create value through meaningful expansion of
available information to drive measurable outcomes, and promote sufficient
connectivity to a diverse group of health care providers, while maintaining
flexibility and allow for a national condition path is guided by provider
incentives for adoption.

I thank you.

DR. COHN:  Thank you very much.

Larry, did you have any comments, or were you just here for questions and
discussion?  Okay.

Our next presenter is Rohit Nayak.

Agenda Item:  E-Prescribing:  The Software
Vendors’ Perspective – Rohit Nayak, MedPlus, Inc.

MR. NAYAK:  Good morning.  My name is Rohit Nayak.  I’m the
Vice President for Clinical Information Solutions for MedPlus, which is a
wholly owned subsidiary of Quest Diagnostics.

Unfortunately, as you might have detected, I don’t have any written
statements or haven’t submitted any of that.  I seem to have fallen out of
the loop in terms of the submission cycle over here.  So, I will make an
attempt to catch up with this process after this meeting.  I apologize.

My background, or at least the pertinent part of my background is that I
would be a co-founder and CEO of an e-prescribing company that I started in
2000 called Rx Nuts and Bolts.  And I followed that adventurous journey
through to the summer of last year when the company was acquired by Quest and
MedPlus.  So, a lot of what I’m sharing with you today is my experiences
as a part of that journey, and what I see as a lot of momentum, driving factors
that occur in the industry today.

Clearly, the fact that this committee exists and that we are all here today
has gone a long way in demonstrating that we all believe that e-prescribing
provides significant value to all of us in the community, from a quality
standpoint, as well as from a safety standpoint, so I won’t really go into that
in any detail here.

When we started deploying our software many years ago, there were several
companies out there at the time.  E-prescribing was already a buzz
word.  And as a lot of you can recall, there were a lot of pictures out
there with a physician in a white coat, carrying a PDA, and that was the
presentation of e-prescribing, suggesting that the physician and the PDA can
solve all problems associated with e-prescribing.

As a lot of the other speakers over here have already recounted,
e-prescribing is more of what I like to call prescription management, the
management of prescribing that occurs inside of a physician’s office.  And
so, we spent a lot of time trying to understand the work flow issues that occur
inside of a physician’s office.

I will admit that I am not a standards geek.  I don’t understand them,
and I haven’t spent a lot of time appreciating the various aspects of various
standards.  But what I have spent a lot of time in looking at is various
adoption issues that occur inside of a physician’s office.

As a result of being number five on this list over here, one of the
drawbacks is that most of the good points have been covered.  And so, what
I will do is reiterate some of those points, and maybe expand on that a little
bit with my perspective.

One of the points that we believe bears some attention is one that was just
raised a few minutes ago, and that is associated with the physician
identifier.  Clearly, we have the DEA number that uniquely identifies a
physician, but it doesn’t necessarily uniquely identify a physician in a
particular care setting, which is key when we are trying to have a two-way
communication between a physician practice and other entities such as
pharmacies, for example.

Pharmacy board issues have been mentioned a couple of times today, and as a
small company that started out regionally, we were lucky that we operated in
Ohio.  We found that the state board of pharmacy in Ohio was extremely
forward thinking.  They had an approval process.  The good news was
once I went through that, it became a lot easier for me in other parts of the
country.

Nevertheless, there are significant legal and compliance issues now that
I’m part of a larger organization, that I have to look at very, very
careful.  And it certainly is a detriment to a widespread
deployment.  So, we would certainly encourage any effort in that
particular direction.

The whole concept of controlled substances relating to drugs as well is a
sticky issue, because we can do a lot with the electronic prescribing
today.  Organizations such as RxHealth we believe, and SureScripts lately
are doing a lot to make this real for us in a lot of different ways.  But
if we have a manual process that is inserted in a large electronic process,
efficiencies are limited in a natural perspective.

Some of the things that we have learned in communications with pharmacies
is that while we are talking about a lot of different information and
technology issues that occur within a physician’s office, surely the pharmacy
would love to get more information on their patient as well, because right now,
what we have is multiple entities maintaining similar sets of information on a
particular patient, but sometimes they are not complete on either side of the
spectrum.

So, for example, I believe it was the first speaker who talked about
diagnosis information.  And certainly, health the ultimate dispensing
entity with that additional piece of information.

A second point that I believe was also mentioned here in the last several
minutes is some information based on the physician’s actions that were taken
while they were prescribing a particular medication.  For example, if they
were allergic with a drug contraindication, and the physician chose to override
that, that piece of information would be extremely useful to the pharmacy,
because that will alleviate a lot of the touch points that occur back and
forth.

The NCPDP Scripts standard does have I guess the ability to incorporate
allergy information, but I know for a fact that it is seldom used.  So,
that piece of information, or the receipt of that information at the pharmacy
will certainly help a lot from a safety standpoint.

A couple of people have talked about NDC codes, and I’ve got to tell you,
it was extremely puzzling to me when I got into this business, that you have
numerous NDC codes that represent a particular drug.  And for all the good
stuff that we do in the physician’s office in terms of selecting an NDC code
and representing that as part of the script, ultimately when it is dispensed,
we know that the chances of it being the same NDC code are extremely slim.

So, I guess that’s just to reinforce some of the concepts associated with
NDC.  While I have made my comment about the NDC, I would also like see
modifications to that be gradual, because we have invested a heck of a lot in
utilizing those codes as part of our technology black box.  So, a little
bit of selfishness there.

There was a lot of activity over the years with respect to the various
stakeholders, especially the PBMs and the payers on one side that are trying to
promote formulary compliance at the physician level.  And we have gone
through a lot of evolution on that, and I believe RxHealth as an entity, has
made that real to a great extent.

One of the key pieces that I recall that was missing from earlier
deployments relatively simple to incorporate formulary lists, there was no
specific mechanism to identify a patient with that formulary.  And that
went all over the board.  RxHealth has, by virtue of its MPI, has made
that, by using standards that are out there a lot more real, so that we have a
real time ability to check for medication insurance, which as most of you know
is very, very different than health insurance information.

On that point, we certainly encourage, and I think everybody here will join
me, other plans and payers to join up with RxHealth, because I believe that
clearly will help us from an adoption standpoint.

We talked about the medication history here as well, and the aspect of
getting medication history, and maintaining medication history.  Clearly,
there are at least a couple of mechanisms to do that today.  So, some
attention on that particular note will certainly help us as well.

On the pharmacy side, SureScripts has, by virtue of its technology
platform, and its associations with various retail pharmacies, has certainly
allowed for the two way communication to occur between pharmacies and
prescribers, specifically with respect to the renewal of a particular script.

One piece of information that would be neat that I believe doesn’t exist
today is a notification back to the prescriber when a particular drug is
dispensed.  We do get receipt of when a particular prescription is
received at a particular pharmacy, but the dispense information will certainly
help make medication history the true history.  And I believe there are
enough statistics out there about the number of prescriptions that are really
not picked up.

With respect to adoption issues, and I made a reference to that when I
started off my conversation here today, price is something comes up with
respect to whenever anybody talks about adoption.  I firmly believe that
e-prescribing or integrated e-prescribing in conjunction with the EMR systems
that category five or six alludes to provide a distinct value if done right, to
the physician and the physician practice.

And as long as a value is there, I believe that physicians are going to pay
for it.  So, I see a lot of the different initiatives around the industry
today that are around offering free — and free is not good enough sometimes,
right? — free hardware or software to promote e-prescribing.  Money
talks, and that is certainly a good thing.

But I do believe the implementation and the deployment issues associated
with software are equally important, because we do know that there are a fair
amount of PDAs out there that are absolutely not used, and it has not a
significant impact, if you will, on adoption for the country.

Adapting to the work flow that occurs within a physician’s office.
I’m not sure how standards fit into that particular domain.  I’m sure
there is a way.  But that is clearly an important part of making it all
happen for us. There are a lot of touch points that occur inside of a
physician’s office with staff and medical assistants, in addition to the
physician.  And clearly, we have got to be cognizant of that.  It’s
not just the physician and the PDA that makes this thing work.

These applications truly need to be integrated, and I think the EHR report
and its gradations make a good statement of what integration means.
Though I would like to caution us technocrats, and that’s clearly my
background, do not get too excited about providing a lot of information to the
physician, because that’s not what they are used to today.

I think a lot of what we have to look at is where is the industry
today.  We certainly know where it need to be at some point in the future,
but we believe that there are incremental steps that take place in the middle
before we get to that particular point.

So, for example, providing a lot of what I would believe to be onerous
pop-ups to the physician, informing them of very assets about the medication,
or what the treatment that they are doing might now help adoption, as much
sense as it makes to all of us from an information, safety, and quality
standpoint.  So, a little bit of a precautionary note over there.
That’s my experience in observing physicians practice medicine.

The other aspect about workload is that no two physicians do the same thing
in the same way.  We have heard that many, many times, and I can assure as
a part of the sales and the development role, I found it extremely
frustrating.  The good news is it makes us as software vendors, provide
better technology that is adaptable and flexible.  And that’s clearly
something that we need to observe.

As we develop and deploy standards, we should always think about what the
physician needs.  We have a lot of stakeholders in this room today, and
they are the right stakeholders, there is no question about that.  But we
have to focus on is this going to be a hindrance to the physician to use a
particular piece of information, or is it going to take more time?

And when I say time, I’m not talking about the aspect of writing a
prescription, as was aptly demonstrated, but just the general scheme over the
history of a prescription, through its life cycle, through its renewal cycle,
and everything from there.

I think I have covered the various points that I wanted to cover in my
presentation.  Thank you for the opportunity.

DR. COHN:  Thank you.

Our last presenter is Daniel Nutkis.  Thank you for joining us.

Agenda Item:  E-Prescribing:  The
Software Vendors’ Perspective – Daniel S. Nutkis, Zix Corp.

MR. NUTKIS:  Good morning, and thank you for letting me participate in
these proceedings.

My name is Dan Nutkis, and I am Chief Strategy Officer.  It’s
interesting, we didn’t send our CTO or CMO or head of regulatory affairs.
We sent the individual that drives our business strategy.  And anything
that either enhances or impedes that strategy becomes my problem.

I really don’t have responsibility for implementing anything, but I do have
responsibility for insuring that anything that either can accelerate or detract
from our ability to meet our goals is considered either as a positive
assistance, or as a potential threat to our success.

We have provided — I did spend quite a bit of time last night throwing
away the testimony.  I apologize to the three people that are here from
Zix or the four that wasted some time on my old oral testimony.  We have
provided I think 22 or 23 pages of written testimony, which goes into I think
substantial detail of our position on the various standards, where we think
there are deficiencies, where enhancements are, where we believe that in fact
it can be enhanced.

Let me give you a little bit of sense of the experience, and therefore our
perspective, because I do believe, since I don’t have a medical degree,
although I do have an informatics degree, and a minimal amount of time in that
area — so, I’ll take Zix first, since most people have no idea what Zix is,
except for the ones that remember it as your video store called Blockbuster
Entertainment, and some of you may remember it as Toll Tag Company, EZ
Pass.  And Toll Tag is also one of our products as well.

So, we have now made the transformation into health care, and we’re in
various lines of business, of secure messaging.  About a third of the Blue
plans or health plans in this country use our secure message portals, the
technology behind the technology.  Actually, BASF uses our technology as
well.

We provide a suite of other protection services, transaction
services.  And then we have a division that provides care delivery
services.  And in that group, there are a suite of services supported by
e-prescribing, or a whole suite of services we refer to as e-prescription
management, online doctor visits, patient education and compliance and
adherence programs, laboratory ordering and results reporting, and a whole
bunch of secure messaging infrastructure.

Our e-prescribing services are currently in use in about 20 states in this
country, and we have specific focus in 8 states, a large focus in
Massachusetts, Maryland, Connecticut, and New York.  And now with our
recent partnership with Microsoft and Wellpoint, we have expanded to Missouri,
California, Georgia, and Wisconsin.

We, probably like many of the others, work closely with the plans, PBM, ICA
Health Systems in whatever structure there may be to get to a physician or to
find a model to promoting e-prescribing.  In certain cases we are working
— I think this was discussed yesterday — as part of something that is
referred to here as a collaborative, where plans are banding together as in
Massachusetts, where Blue Cross and Zix have come together to try to coordinate
an effort, and I think you will see additional organizations become part of
that, to promote e-prescribing within the state of Massachusetts.

We also have activities as well in Massachusetts with the Mass Share
Program, specifically their Medsinfo, which is an initiative to provide drug
dispense history to patients admitted into the emergency room of local
hospitals, and that is to be with all the plans throughout the state of
Massachusetts.  Initially, I think that is starting with four plans, and I
think three emergency rooms.

Today, we connect to various sources to get the information we need from
the PBMs.  So, we use RxHub.  I believe we are still the only
organization that transacts will three big PBMs.  We also go direct to
other PBMs in a number of cases.  In a few cases the data is actually
stored in one of our data centers, where they provide us with their
eligibility, drug dispense history, and formulary information, and we process
it.

So, from our perspective, the that we want people to understand is that we
have found a way to work around what would be considered a lack of
standards.  And have found that that has not been a substantial impediment
to being able to deliver our services, although as you will see in the oral and
our written testimony, we believe that there is certainly a substantial amount
of progress that can be made.

We, as I think was discussed, like many others deliver our services either
in Web-based, PDA.  We use various wire line and wireless services, either
cell-based or WiFi.  We are relatively platform agnostic.  We do
today, and I think was also discussed, have the ability to deliver contract
level formulary information, patient co-pay at the point of prescribing.
We can do it in a real time manner.  And as was also discussed, we have
the ability to provide real time laboratory results through our laboratory
results reporting systems at the point of prescribing, and we link that with
the diagnosis code.

The other thing that I would also point out is that the functionality is
evolving at a relatively rapid pace.  And we see that as the formularies
continue to evolve, and we see new models that are starting to exist in the
marketplace, we will see the technology and the functionality need keep pace.

So, as we see new formularies, as we see new trend management programs, and
we see more drugs that are becoming step therapy drugs, we are seeing that the
requirements from the plans and PBMs to be able to keep up with that is at a
relatively feverish pace.  And therefore, again, as we consider standards,
and as we consider implementing guidelines, that they take into account
honestly, things that probably I am not qualified, and I’m not sure who is at
this point, to decipher and determine what they will be.

Just to give you some background, I am the chief strategy officer.
That does make me extremely difficult to do this, as my background is making
sure, and my responsibility that these impediments don’t take place.

Just to give you a sense where I draw my experiences from, I did start out
years ago relatively active in the health care standards drafting organization,
a part of HISDT, which was the old health care informatics standards panel, one
of the original members of WEDI, and the Computer-based Patient Record
Institute.

So, I do have at least some experience in the standards process and
understanding the complexity in being able to draft and fine commonality around
standards.  I also did spend some time as the national director of health
care emerging health care technology for Ernst and Young.  So, I was also
very good at taking anything that was simple, and making it complicated.

But I also have some experience in understanding what the lack of standards
would do in emerging markets, especially in health care, where we were trying
to work with at the time, computer-based patient records, and a lot of evolving
technologies with a lack of standards.

Also, I think to discuss what the subject matter was yesterday with the
panel on the merits and virtues of e-prescribing, I’m not sure that anyone will
question the fact that e-prescribing has enormous potential with health study.
And there are obviously a number of other studies that have been released on
the merits.

The thing that I thought was interesting yesterday was, and it something
that comes out from time to time, is that the standards were not, or the lack
of the implementation was not related to standards.  I remember hearing
that the pharmacy didn’t know what to do, and the physicians wanted to be more
incented.  And there were a whole bunch of obstacles that were preventing
widespread adoption.

And although there was some question about nomenclature or terminology and
vocabulary in some standards, I think we would support the fact that our
limitations are not related to standards.  That’s not in anyway to
say that the standards process is not extremely important; not that we don’t
support it 100 percent.  But the pain that we see today, and the pain in
the marketplace was around being able to put technology, emerging or not, there
is a lot of question around the fact that it is emerging technology.

People like me would say the tipping point has just occurred, and we’re now
on the down swing.  Some people, of course people that were here long
before me in 1998 and 1999 developing this technology, would clearly say that
it started some time ago, and this has been a very long road.

So, just from a standards perspective, and as I said, I have gone into a
fair amount of detail in the written testimony, NCPDP Script standard, HL7, and
the AST X12, there are three standards by which we operate today, and are able
to perform the majority of our functions either through the various PBM
partners.  Obviously, as was discussed, RxHub has their own set of
standards, and we also work with those as de facto standards per se.

Specifically, I think everyone has gone into the details on the NCPDP
Script, so I’m not sure it makes sense for me to spend a lot of time on going
through the NCPDP, the X12.  We use the 270 and 271.  And with
regards to the HL7, a transaction set, and some of the other standards, we
currently are faced with about 120 either hostile information practice
management or EMR systems through standard interfaces or proprietary
interfaces.

It is heavy lifting, it was what we need to do.  And interestingly
enough, we don’t charge for our interface.  We provide the HL7
interfaces.  We provide our interfaces all free of charge.  Our
experience has been, and certainly others here have greater experience, has
been that it comes out to be a sophistication and cost issue at the practice
guide, not always the availability of an interface.

Irrespective of a standard or not, theoretically, the standard was so
simplistic, and the interface was so easy, that the costs would not be
prohibitive, and it would be much easier.  We have actually found that
people have said there is no incentive, there are no dollars, there is no
sophistication.

I think we recently did a poll of the facilities that we work with, and the
majority of the smaller practices weren’t even under maintenance, so they
couldn’t call the vendors to even find out if they had an interface, and in
fact, to get any resources to try to work with the standard, to work the
interface.

So, we looked at some of the standards.  We also looked at levels of
sophistication, and the ability for organizations or institutions to be able to
use the technology or tools that are provided to them.  And that is
certainly one that we have found irrespective of the differences, such as the
HL7 standards.

I’m going to spend a lot of time going through the strengths and
weaknesses.  It’s in my written testimony.  And again, we find that
the standards have not been a significant limitation.  There is no
question that we would like to see nationwide adoption.  We run into it
all the time, especially in some of the cross-border states where we’ve got a
lot of people in New York, New Jersey, Connecticut, Massachusetts, Rhode
Island, New Hampshire, and so on and so forth.  We have the Washington, DC
area as well.

Some other areas, I’m not going to spend a ton of time here, although there
is one that hasn’t come up, so I will bring it up, and I am surprised, given
the magnitude of the vendors that are here.  We have a significant problem
with authentication.  Because of organizations wanting more integration,
and wanting to have seamless access, either on the privilege side or the access
control side, they are asking for e-prescribing and a lot of the other
functionalities be more indebtable in what they do.

So, the doctor that is currently doing lab ordering, and also wants to be
able to also write the prescription.  And they have already signed now for
a third time on their PDA, and they don’t want to have to sign on a fourth, or
on their browser.

And there is — obviously, there is X7 IT standard, or digital
certificate.  And as someone mentioned yesterday, they were using the AMA
Verisign digital certificate  or digital identity program.  But there
is really no standard in that area for vetting and authentication, and we can
determine what the control space should be, and then have a single vetting and
a single authentication.

Formulary, it’s been mentioned and I’ll mention it again that clearly for
us, formulary is a significant issue. The bigger issue is misunderstanding
where formularies are going.

Co-pay, obviously in drug benefits this is another area where
standardization would be important for us as we looked at being able to assure
that we can display the information and understand the potentials.

Drug utilization review.  This is one that we have spent a fair amount
of time trying to I think internally determine what role standards groups
should play in the whole area of DUR.  I think we would like to put it out
there that DUR is an area where there should be standards, at least for the
messaging, for the protocols, and at least for being able to identify how best
to report this.  As everyone probably knows, some of the standards on the
adverse events reporting system probably would also have potential to be
carried forward for some of this.

The area that also, and I’m not sure where it goes, but it’s kind of a
consistent issue for us in the market space is we operate a number of certified
data centers.  We operate them within pretty stringent guidelines.
And it’s just a consistent issue on standards for what is appropriate for
transaction sets.

This is more an area of SLAs, but the problem is as we develop these more
complicated and sophisticated standards, 36 seconds for a DUR transaction to
come from point to point, negates the whole transaction, because people put the
device down.  So, we looked at some of these to try to understand what in
fact people’s expectations are for being able to allow the transactions to take
place.  For simpler transactions it’s not an issue; more complicated,
timing becomes an issue.

Incentives — we are very much in favor of incentives.  As was
mentioned yesterday, there is a number of PCPIP programs or PIP programs,
physician incentive, or primary care incentive payment programs that exist
today.  Blue Cross talked about one.  We are now involved in a number
of them where physicians are being compensated on a pay for performance basis
for using e-prescribing, and we see that those models probably will continue to
evolve, as there does appear to be significant interest.

The question will be the limitations that are put on those programs.
We have some organizations that want to incent for writing scripts
electronically, some that want to incent for writing scripts in accordance with
formulary, and then you see some ratcheting up of some wanting to do incentives
based on enrolling in adherence or compliance educational programs where the
numbers can be substantial, where the physician can be compensated.  And
the question is, where do those programs fit into the model?

We also believe that the adoption by the Medicare program and the federal
government will become de facto standards, so the acceleration in adoption by
Medicare will in effect, take care of that.

And I think that’s really the extent of my testimony, and I would obviously
be willing to answer any questions.

DR. COHN:  Okay, well thank you.  I think I’m going to allow
everybody to take a break, and then we will come back for discussions.  It
is now 10:30 am.  Why don’t we give everyone a 15 minute break?

Thank you all for great testimony.  We will continue with questions
and discussion after the break.

[Brief recess.]

Agenda Item:  E-Prescribing:  The
Software Vendors’ Perspective – Discussion

DR. COHN:  Jeff, you have one comment, and then we are going to get
into questions.

MR. BLAIR:  Yes, since the topic of the continuity of care records
came up, I need to recuse myself from any comments or discussions related to
the CCR.  It is an apparent conflict of interest that I may have on that.

DR. COHN:  Okay.  Does that mean you will not be able to ask any
questions for the remainder of the session? No, I’m sorry.

Any questions from the subcommittee members?  Jeff do you have any
questions or comments?

MR. BLAIR:  Let me pull my thoughts together a little bit
further.  Let me give someone else the time.

DR. COHN:  Judy.

DR. WARREN:  This is for Jill.  You had mentioned that you wanted
to include medical devices.  Could you just talk more about that?

DR. HELM:  Sure.  We would like to expand the available data
standards to accommodate those items that are traditionally orders on a
prescription pad today.  So, those would be devices that would be
prescribed in an ambulatory care setting as part of a treatment plan, or
devices that would be used to administer medications, and include things like
oxygen, DMEs, home health.

DR. COHN:  Great.  I think Jeff has a question, and then I’ll
follow-up afterwards.

MR. BLAIR:  I think Jill may have mentioned the issue of prescriber
identifiers.  This really is open to all of you who could help me
understand this.  Yesterday in our testimony there were three standards
mentioned that get at the area of prescriber identifiers — the DEA, the
emerging, upcoming NPI, and something developed by NCPDP which is called
HCIDEA.

The comment yesterday was that none of the three is ideal.  They all
have certain strengths or weaknesses.  I would love to hear from all of
you regarding your perceptions with the strengths and weaknesses of those three
prescriber identifiers, and any guidance or recommendations you have for the
subcommittee in terms of our recommendations on that topic.

MR. TRUSKY:  I guess I’ll take a shot.  Larry Trusky.

Certainly, I think yesterday some of the challenges were mentioned about
the DEA number, and I think those are common challenges.  Not every
prescribing entity will have a DEA number, so it doesn’t cover the entire base.

The NPI, that is intended that every prescribing entity could have a unique
identifier.  The challenge I think that Richard mentioned yesterday from
ProxyMed was that doesn’t really represent a clinic.  So, does a patient
belong to the physician, or do they belong to the clinic?  When the
physician leaves, how does the pharmacy system target back to the clinic,
versus targeting back to a physician?

And I’m not going to say that there is a great answer for that yet.  I
think that requires more study and more understanding of how the pharmacy
systems operate.  Today, I believe they do target by DEA number, so I
think that is the challenge either way with the DEA or NPI.  And HCIDEA, I
believe is similar to the NPI.  It’s another representation of the unique
provider ID.

So, I don’t think any of them answer the question that Richard was asking
yesterday, which is if a physician leaves, how does that targeted message from
the pharmacist back to the site, stay essentially with the patient chart — the
patient chart is going to stay in the office.  How do we keep targeting,
or how do we message back to where that chart is going to be?

MR. BLAIR:  Let me prompt you a little bit.  My understanding is
that HCIDEA is able to identify the prescriber may order from many different
sites.  And that the facility and the site is important, as well as the
prescriber ID.  However, they say that there are some shortcomings, and
that is what I’m not familiar with.  And of course, the NPI won’t be
available for a couple of years in terms of usage.  And it doesn’t
identify site.  It would just be the individual.

MR. TRUSKY:  My guess on that, not knowing enough is that even the DEA
number today, you can have multiple DEA numbers based on your location.
You can, state by state, get a DEA number.  I don’t think that’s how
people operate.  I think even though the standard may allow for it, there
may be challenges in implementing that, because as I understand HCIDEA now,
they are just taking data feeds from multiple sources, and trying to match the
patient and provider.  I don’t think that accounts for multiple
registrations.  Again, the standard may, but I don’t think the data
support that today.

DR. KAUFMAN:  This is one of those instances where the message
structure is probably going to determine what identifiers are needed to fill
it.  So, certainly we need an identifier that is going to uniquely
identify a prescriber.  When as a message model evolves to support the use
cases, we’ll identify the need for additional identifiers such as the location
or ownership of the patient’s persistent medical record, site at which the care
is provided, and other things like that.

But I think it will be more coming out of the use cases and the messaging
models that will define the additional identifiers that may be needed beyond
the obvious one of the prescribing agent.

MR. BLAIR:  Peter, you are with DrFirst, and you gave us this
demonstration of how quickly you could wind up going through the flow to order
a prescription, which has tremendous value, and is really important.

On the other hand, we also saw the pyramid where we could go up levels of
that pyramid and improve patient safety in doing so, and quality of care, and
previous costs.  But as we do so, we are giving more information to the
prescriber to make decisions with.

Could you and the other folks please give me your thoughts about as we go
up the levels of the pyramid, and increase value from that standpoint, are
there concerns that we have to consider when we are selecting standards which
would make the process of prescribing more time consuming?  And is there a
threshold where the physician will feel uncomfortable that it’s taking too
long?

DR. KAUFMAN:  Those prescriptions that I wrote, at the time I was
writing them, were checked for allergy, drug interactions, and insurance
formulary.  None of them had any alerts, or the alerts would have come
up.  The level six in the pyramid, which is including the full EMR, would
allow more data to be placed into the system without the physician or the
prescriber, whoever it is, having to specifically enter it, notably the
patient’s weight, the patient’s age, although our system does contain the
patient’s age, and the diagnoses.

Some systems get around this by adding value using the diagnosis to help to
choose medications, where they are able to capture the diagnosis without the
doctor minding that they are actually having to add a little bit of extra data.

But if the system is not doing that, the doctors really are concerned very
much for safety, but even more so with their time, and being able to change
their practice patterns to include electronic prescribing, without having to
see fewer patients, without having to — but I do think that it’s very
valuable.

However, I think we can add more value to the system by getting doctors to
accept and adopt electronic prescribing in a connected form level, than
requiring doctors to jump through the hoops it takes to do level six, which
most doctors are not ready for.  But I do agree that we should always at
least mention the use of comparing the patient’s weight and the patient’s age
and the patient’s diagnoses with the medication and the dose prescribed, and
not just allergies and drug interactions and insurance formulary./P>

MR. BLAIR:  And Dan Nutkis, on this issue you referred to 36 seconds
doctors would have to wait for something.  What were you referring to?

MR. NUTKIS:  Well, that was theoretical.  I think it was
mentioned yesterday that we held some focus groups and we brought the
physicians in recently to go through some of this.  And we have been given
a whole bunch of anecdotal information.  And the information was the
physician didn’t want co-pay.  And all of these alternative messages that
may change their behavior would be negatively perceived by the physician.

Now, again, in this commercial society that we have, that means the
physician puts our device down and they pick one of the other devices that
doesn’t in fact require them to do all these other things.  By the way,
our devices don’t require any of those things.  But having it mandated by
a third party appear to be somewhat difficult for us to accept.  Our focus
groups have said about 20 seconds.

But what we have found now is there are a number of things that will change
that factor.  It’s called value composition.  So, again, it goes back
to the analogy of cell phones.  They were big, and they had bad battery
life, and they had terrible coverage, and they cost a lot.  And there was
a dipping point, or there was this curve when the early majority and the late
majority and then your adopters came in.

And there were some things that have been added now, so we know it is 22
seconds.  We know they want full blown DUR.  We know they want
prescription history.  But the biggest function that we have now found
that they are willing to wait for is real time fire up.  And this is the
one now that we have found that they will wait a significant amount of time for
real time fire up.  So, we have been developing it.  We’re testing it
now for doing step therapy.

So, we’ll be able to tell them real time in fact what the step therapy
is.  Now, for a physician, they just don’t know, right?  They go and
prescribe Celebrex.  You know that 10 percent needed to be a prerequisite,
and they get the call.  And the pre-authorization call is the most
difficult to deal with, because of the charge, the formulary.

So, we have now found that even though 22 seconds — and by the way, 5
seconds is what we strive for.  Five seconds, bar none, is what we
absolutely try to get everybody in and out for in 5 seconds; 22 seconds is
about enough time the physician will do before they sit down the pad, and walk
away.

MR. BLAIR:  Is the prior authorization work that you are doing, is
that being done in conjunction with X12 or HL7, where it could be a
standardized message?

MR. NUTKIS:  I think the answer is yes, X12.

DR. COHN:  We have a couple of other comments.  Are you done,
Dan?

MR. NUTKIS:  Yes.

DR. COHN:  Tom, do you want to go first?

DR. DOERR:  Yes, I’d like to suggest that the market is going to find
several sweet spots in the trade off between the complexity of the work flow,
and the value that is created when physicians use the particular combination of
features and functionality.  And I think that one sweet spot may well be
what Peter shows us with DrFirst, where you are highly connected.  You
have this very fast work flow.

And in my oral testimony I have been lobbying for arguing there is another
sweet spot that says let’s make the work flow diagnosis-driven, because then we
can hook decision support information to combinations of diagnoses and drugs,
and create a lot of value.

And the third sweet spot would be the full blown integration with the EMR,
the level six.  So, level five is what DrFirst offers.  I’m
suggesting there is a level five plus that says that in the big picture,
physicians are directing about 80 percent of spending in a $1.7 trillion
market.  I saw three estimates that were in that range of around 80
percent published.

And the information that we doctors are bringing to our decisions in terms
of I’ve got to make a decision here to treat a condition.  What are my
alternatives?  How effective is each of these options?  How expensive
is it?  And what are the short-term side effects, long-term side
effects?  Because I don’t have information.

So, that my vision in founding my company was to bring this into doctors’
work flow.  And what we demonstrated in one controlled trial that I sent
in earlier, it’s in press with The Annals of Family Medicine, was that the
payer was saving about $1,000 per doctor per month when the payer had a 40
percent market share.

And so that the proposition that I make payers is that it’s not enough to
have the payer pay for the cost of the system.  The payer needs to be
sharing some of the value that is created when physicians are conscientious and
practice evidence-based medicine, and invest in these tools.  So, that
there is a sweet spot there where the payer will get some of those savings from
physicians above and beyond the cost of the system.  And I think that the
market has been finding sweet spots, and I would say there are trade offs.

DR. KAUFMAN:  I simply would affirm that we want to provide all of the
information.  Let vendors figure out how to make that very efficient and
effective for providers.  We have done that in the past.  I think
it’s doable.

Two really detailed things that actually might pertain to things you could
look at doing, two of the things that we found out — well, the act of
prescribing — are spurious or unwanted alerts.  And that is the decision
support system is warning prescribers of problems, which they don’t want to
hear about.  And frequently, we know that it’s pretty well demonstrated
that once people start discarding and missing these alerts, they stop paying
attention to them altogether, and they stop being effective.

There are two things that tend to produce those that you guys might be able
to impact.  One is when we used to use ICD-9 to try to — it was kind of
the interlingua between our applications and knowledge-based vendors’
applications.  It produced a wide array of false alerts for indications
and contraindications, because of the big problem with ICD-9, the fact of the
classification, the fact that you have codes which about 2,000 out of — they
don’t know what they mean.  So, that is a delusion problem.  We had
to turn off contraindication checking.

So, going into a clinical terminology like SNOMED-CT, integrating that with
the knowledge-based vendor representation of patient conditions can really
decrease the amount of false alerts, and reduce the time span of dealing with
them during prescribing.

My understanding of this, I ran into this in acute care applications we
had, the Food and Drug Administration over years of work, has provided these
black box warnings that go on package inserts.  These are warnings of drug
interactions or contraindications in which you are never supposed to give it.

Well, what I was told by people who work with this, and clinicians who work
in this domain is that many of those probably are not correct any more.
Some of them are historical data of studies that were done in the fifties or
sixties with 10 people.  They have identified them, but they are still
there.

And neither we, nor our customers are comfortable turning those off as long
as they are on that black box on the package insert.  And so, they are
constantly producing unwanted alerts, and nonsensical warnings that people turn
off.  And they are category one, and they interrupt work flow.

So, the only way I can see that getting cleaned up is going back to the FDA
and saying, we’ve got a problem with these black box alerts.  Can they be
re-evaluated for ongoing clinical reasonableness?

DR. DOERR:  I’d like to briefly add I very much support that idea of
going back to the FDA, because they are driving the standard of care.
There is no physician who is going to ignore a black box warning, because of
the threat of malpractice liability.

And I would like to amend my earlier comments to add that the Achilles’
heel electronic prescribing systems has always been utilization.  There
was a whole wave of these companies that came out in the late 1990s.  And
they came and they went, because they weren’t creating enough value.  The
value proposition to the physician was inadequate.

And there are work flow challenges that are important in making the system
usable, but there is this whole other concept of how much value is being
created when a physician uses the system in any one of these sweet spots?
And then what happens to that value?  How much of it gets back to the
doctor?

With the system that my company made, which is ICD-9 diagnosis-driven, we
average 15 new prescriptions per doctor per day across the entire user
base.  And in two of the studies we did that involved payers, they told us
that those represented capture of more than 98 percent of all the prescriptions
that were associated with those doctors.  So, that we think it’s very
close to the whole utilization.

And there were some work flow efficiencies that made that happen, but the
other thing is this whole concept of the doctor can create several thousand
dollars worth of value per doctor per month using a system such as ours, or
others could develop something similar to that.  Give some of that back to
the doctor, and then we’ll get their attention, because my own belief is that
it takes about $10,000 a year in new revenue to get primary care physicians’
attention, and really get them to want to change their work flow.

Really, it comes down to the concept that I believe primary care physicians
are economically stressed enough that we are economically driven.  And it
is just the brutal truth.

DR. COHN:  Tom, thank you very much.  Obviously, most people are
driven economically.

Larry, I think you were going to comment?

MR. TRUSKY:  The only other comment, Jeff alluded to it, the other
transactions that happen during the process.  There was pressure early on
for some of these formulary drug histories that some of this is just too
hard.  You can’t do that.  You need to hit up against our databases
while the prescribing is happening.

And of course, we worked through that, but I worry about the danger of
saying that some of these transactions, that these standards are too hard to
evolve, and not allowing for that data to come down to the prescribing system,
and get through these as fast as you can, in three seconds, five seconds,
whatever the right time is.  Not be 20 second transactions for a drug
history, or a 20 second transaction for an eligibility, a 20 second transaction
for formulary.  That’s all got to happen pretty quick.

DR. COHN:  Final comment on this one before we move to another
question?

MR. NAYAK:  Actually, the previous question I guess, and that was
related to work flow and adoption, and the fear that standards might impact
that.  I just wanted to provide by way of example of the instructions, the
directions that are the meat of the prescription that we have been referring
to.

There clearly are many, many ways to do that, and you will find that the
prescribers use different — everybody does it slightly differently.  So,
the ability to provide recommendations of things, this is how it needs to be
done, but also providing sort of the release valve to say, yes, you can also do
it your own way, in some way, shape, or form might be the right balance that
you are looking for in standards and so forth.  Just a commentary.

DR. COHN:  Harry?

MR. REYNOLDS:  Can you put your presentation back up again, and put up
that level — if you’re disconnected, fine, don’t worry about it.

MR. NUTKIS:  The pyramid?

MR. REYNOLDS:  No, the other one that had the levels.  While
we’re getting it set up, I’ll up what Jeff had to say and some of the other
comments.  And then yesterday, for some of you who weren’t here, Stan Huff
got up towards the end of the session and talked about this is a journey.
And the pyramid is a long journey.  And segments of the pyramid are not a
long journey necessarily, because some of you are pretty much already there.

Also, since you happen to be the segment of the industry — if you look at
this, this is really in the end, all about whether the doctors will use
it.  The rest of the industry has pretty much said we’re going to circle
the wagons, and everybody is kind of involved in one way or another
already.  So, it’s really about who touches the doctors.  It’s a
similar issue to what we ran into with HIPAA and other things.  It’s
really where it meets the road in the care setting.

So, if can even in just short period of time, and playing again off Stan’s
comments yesterday, are there things on that pyramid, going from level one to
level six, that you think are things that could be set down, could be put in
place that would allow you to develop the rest of the things on the pyramid?

I’ll just use an example.  You can pick NDC for example, just using a
layman’s example.  That could be put in place so that we are not
developing standards right now all the way to six, which are going to take a
long way to do, which means that your process is longer.  Again, just to
try to play off a little bit of our discussion yesterday.  And then what
time frame roughly would you have on some of these?

DR. COHN:  Harry, just to make sure I understand it, are you talking
about low hanging fruit?

MR. REYNOLDS:  That would be a real succinct way of asking it.

MR. BLAIR:  Just one clarification with respect to low hanging fruit,
because this is a very good question.  But the pyramid is really saying
levels of functionality.  It is not necessarily saying that the top of the
pyramid is a lot way off compared to other things.  There are folks that
have EHR systems today that are adding e-prescribing, and they have some of the
aspects of five and six, but they may not have number one, which is the
separate referencing information.

So, I’m not sure that it’s a linear path line from the bottom to the top,
but it is a very good, useful way of representing functionality.  Did I
make a misstatement here, or do our testifiers agree with what I have said?

DR. KAUFMAN:  You made an excellent point.  Level one is kind of
separate from all the other levels.  I believe that from 2-5, they are
inclusive.  You can’t really have level five without level four.  You
can’t have four without three.  The level one is something that some
e-prescribing applications, even level five applications do not have.  And
as I mentioned during my testimony, level six EMR applications have that.

MR. BLAIR:  Thank you.

DR. HELM:  I would submit to you that some of the gradations on the
pyramid actually are independent of standards and availability of
standards.  Level one, if you are going to have any type of stand alone
formulary information, any application that is functioning in isolation,
standards are much less important than when you have an integrated set of
applications, and you need to exchange data between multiple applications and
multiple partners.

And that’s where I would argue that standards become critically
important.  It’s really so that we can play at those higher, more
sophisticated levels of functionality, and to get there more quickly.
Now, does that mean that I think we need to have universal adoption of a
standard across the entire health care delivery system?  I really don’t
know if I feel that way.

I think there needs to be interoperability.  I think we need to be
able to talk to one another.  But I don’t think we need to always speak
the same language.  And I would argue that there are certain attributes of
different standards that may play well with certain health care delivery
systems or for certain applications, where I don’t know that you would want to
make that trade off between that level of sophistication between that data
richness for the sake of uniformity.

DR. KAUFMAN:  Can you give an example?

DR. HELM:  I think that ICD-9 or SNOMED, I think vocabulary for
diagnosis is the perfect example.  You may want to use nomenclature that
is more symptom-driven if you are a clinician practicing at a health care
delivery setting.  You may not — when a person walks in with a runny
nose, a physician may not want to say, oh, this is allergic rhinitis.
They may want to say this patient has a runny nose.

Now, that eventually needs to lead to an ICD-9 for purposes of billing, for
purposes of transactions.  But it doesn’t necessarily have to be the only
identifier that is utilized.

DR. COHN:  Harry, I’d like to get back to your original question,
because I think we are sort of going off.  And I apologize, but really, I
think what Harry was asking was really a very reasonable question having to do
with low hanging fruit.  And I would like to redirect this there, rather a
more general discussion about the value of standards.

MR. REYNOLDS:  And again, as was pointed out yesterday is sometimes
low hanging fruit is taken as a quick way to an answer.  I have heard
everybody on this panel talk about things like NDC is embedded in your
system.  Well, if that doesn’t become a standard, it’s obviously going to
be a burden on you, and change what you do or don’t do.  And it’s going to
cost a lot more, and it’s going to take a lot more time.

So, I’m just trying to understand you as a segment, what do you have that
you like?  How does it relate to what we are talking about here, which was
kind of the end game.  That’s why you are all walking around with the same
pyramid.  It’s kind of the end game in all facets.  That’s kind of
what I was after.  So, it’s not just a quick answer and we all go
home.  It’s which are the pieces.  And I think that’s what Stan
touched on yesterday.

Which are the pieces?  If we had passed earlier standards and said
those are in place.  You don’t have to worry about those coming out from
under you, you could move forward.  That may not be an easy answer.
It may not be a clear one, but it’s at least a discussion we are having.

It came out yesterday, and I would at least like to see it explored by the
people who are actually dealing directly with the care setting, because you are
going to be the ones that have to deliver to them, and they are going to have
to pretty much agree with it.

DR. FAUGHNAN:  There are a lot of sophisticated systems that have been
in place throughout the nineties and continuing on to this day.  So, we
can’t have left off what’s out there.  If you look at the very minimal
things that most systems that are in place today have used that you would want
to put in your standard, NDC11, certainly.

And there are some issues with NDC, issues with how the codes are assigned
and packaged throughout, process issues with who handles that.  Is it the
manufacturer?  Is it the FDA?  There are ways this can be done to
work around those process issues.  There are improvements that may perhaps
be needed in some domains, when they might be running out of digits they could
use for packages.

But by and large, work can be extended and grow.  Script works, and
CPDP Script, and they’ve got a lot of pieces in it.  It hasn’t been fully
tested or fully used.  They have a good starter and a good process to
direct it.

If you want to send diagnostic information, most people — and you are not
worried about contraindication checking, you can use ICD-9, and that is
traditionally what people have done, or just a text string, at least to give
something to the pharmacy.  But probably you wouldn’t want to do automated
contraindication checking on it.

And then you’ve got all the X12, and the well defined transactions on
demographic, and probably the biggest one — they are not so much
essential.  We can and have worked without them, but they reduce costs for
implementation.  And speeding implementation are the standards around
formulary transactions, drug utilization review.  Getting the medical
history back to age you are prescribing.  They’re nice.  They’re good
to have.

MR. NAYAK:  If I can just focus on one, maybe people know it’s
important, but just to sort of try listing these out.  The listing of
formulary items for one would be extremely beneficial, just because it may be
some done formats, but it would certainly be great to get people to correlate
and manage to a single file.  It would certainly help the cause.

MR. REYNOLDS:  Are there dominant ones out there?

MR. NAYAK:  I know RxHub has done a great job of trying to normalize
this between the three largest manufacturers that they have, between
MedColi(?), SI, and Care Mark.  So, that’s the reference point.

DR. KAUFMAN:  I think if you are talking about low hanging fruit, we
really need to deal with a couple of issues.  One is patient safety.
What is the low hanging fruit in terms of patient safety?  We are talking
about allergies, drug interactions, diagnosis perhaps, but I don’t view the
drugs for disease complications as low hanging fruit, because it’s hard to get
the data in there.

Whereas, if you have medication history and an allergy history, which is
very little data, and its persistence, that is the low hanging fruit.  And
having a good classification of medications, that would be a standard across
different systems, and I know you will hear this afternoon from the drug
database vendors, is extremely useful.

I heard this morning during our coffee break about a company that is using
one of the vendors.  And they want to supply the e-prescribing to a
practice that has EMR, that uses another vendor.  And it’s very difficult
to compare those classifications of drugs.  So, that’s for safety, I think
some of it would be useful, and would help with low hanging fruit.

For sharing prescriptions, and I know that over on my right I want to talk
to a couple of privacy people.  And one of the things that I think this
committee is not, I believe responsible for doing, but it’s a really big issue,
how do we get this one patient’s medication list to be shared between different
providers in a way that is safe, and still values the patient’s privacy.

The Scripts would say delete the patient, and carry it around with
themselves, as long as they didn’t get it the day they end up in the emergency
room.  But having some way for a good medication history, RxHub is doing
this now, but it’s limited to the PBMs who are their partners.  And RxNorm
I believe also to have a way of calling the drugs the same thing between
applications, so that prescription medications would be shared and be
analogous.

And the last thing is formularies, in terms of low hanging fruit, providing
value.  E-prescribing doesn’t help if the prescribers are not using
it.  We want to provide them with the greatest value, with the least
cost.  And that’s not just monetary.  Sometimes they are willing to
pay for something that works better for them, but time as well.

And the comment made earlier by Dan about the prior authorization and
automated step therapy, all of this is extremely important to provide value so
that the physicians will use the system.  And then the system can have the
medication history shared between the different doctors, however we’re going to
do that, and provide the safety that is inherent in electronic prescribing.

MR. BLAIR:  Could you elaborate slightly on that last one?  With
prior authorization, exactly what are the messages that need to be
standardized?  And who do they go from, and who do they go to?

DR. KAUFMAN:  I believe the TH Links are going to be testifying
tomorrow, is that correct?

DR. COHN:  That was yesterday.

DR. KAUFMAN:  Yesterday, too late.  They are trying to put
together a standardized prior authorization system.  And they have had
some difficulties with it.  Many of the medications, the prior
authorizations sustained between many of their — 24 I believe it is now
members — some of the members have it a different.  And then there are
other medications where every member has it a little different, and they
require prior authorization for different drugs.

At least if there could be a standard format, the format might require a
patient’s diagnosis.  The company could require or not require it.
It might require prior medications if the patient has been on medication or
medications.

It’s very easy for the vendors up here to write a format like that, which
would appear differently for different vendors, different medications based on
which of the specifications were stipulated by the payer.  But it’s
complicated to do it when each of them has a different specification, just like
each state is not a state, but many of the states have different requirements
for a faxed prescription, and it has to be in a specific form, for a controlled
drug it has to be on a specific form.

It makes it needlessly extra work, and it’s just because different people
are writing it.  While the whole purpose of having a great standards
committee is that they can make at least suggestions, if not standards, to say
why don’t we all do it this way.  And then we vendors can all yell at the
companies who don’t follow along.

MR. NUTKIS:  Let me give a slightly different perspective.  I
guess in my little experiment how we have defined pyramid is really an
evolution of functionality.  It’s really how people look at this, and how
they evolved.  And I think that was really the perspective that we had,
and again, e-prescribing is a good example.  What we consider e-script
writing, what we consider e-prescribing, and it’s vernacular,
terminology.  And the pyramid gives us a commonality around the
terminology.

So, now I think the question about I guess I look at it as the
prioritization of obstacles, or increased value propositions, or anything to
increase the value, which is what I think we put in our testimony, which is
from our perspective, it’s all about adoption.

Adoption is about value.  Obviously, someone has got to find value,
either to pay for it or to use it.  And what are the obstacles for me to
be able to find that value?  I think as we went through those obstacles,
we again would highlight the same issues.

DUR is one that continues to come up in patient safety.  But it is one
that there is an inconsistent formulary.  And again, formulary us, we look
around.  We would like to see it standardized, but by the way there are
some there are work around and some there aren’t.

And that goes to I guess really my main point, is I guess being somewhat
jaded, having been there with Richard Dick when the computerized patient
records, those that were there — Paul Tang, Peter Whately(?), I think you were
there, right, Jeff?  We had this grandiose plan.  And I want to say
that was 1991-92, but again, I don’t remember anymore.

And we had this great plan of a computerized patient record.  And I
think it’s all quite a bit now.  We haven’t gotten that criteria.
That’s fine.  And there were some spin offs, I think it was with the MCC
program, the Microelectronics Computer Technology Consortium.  We were
going to get all the technology vendors.

And I guess it comes down to is complexity, comprehensiveness, versus
speed.  And where is the happy medium here?  Are we better off
defining less — and I think that was the point — that the standards be
slightly less effective, but get out there sooner, so that they can meet the
specific component of our requirements?  Or do we take the time and really
develop comprehensive standards, and take the extra year to do them right?

And again, the cost of re-implementation and upgrading.  And those are
some bigger questions that I think someone has to give a lot of consideration
to.  But for us, we get by with heavy lifting.  So, formulary, and we
talked about prior authorization.  And by hook or by crook, we make them
work, because there is a business justification to do so.

DR. COHN:  Steve, I think you had a question, and I have one after
you.

DR. STEINDEL:  Actually, I have developed several questions during the
discussion, and I think they are somewhat related in the sense that I have
heard multiple opinions about standards expressed as we go through the
panel.  We hear NDC.  Everybody uses NDC.  But then we hear that
RxNorm is a better way to order.  Let’s get RxNorm into place.

I asked the group yesterday, and I’m asking this group the same, last year
both the CHI process and the NCVHS process coordinated on a recommendation for
medical terminology that involved FDA terminologies, which include NDC, but the
improved form of NDC, RxNorm fully implemented, and the FDA also included
structured product labels in its expanded form.

And I haven’t heard any of that mentioned in any of the discussions.
I would like to know how that fits in with your strategies, especially when I
hear at one end of the table that we can work around most standards.  Will
this help standardize the medication issue?

I have also heard with regard to the diagnosis code, various forms of this
expressed.  One is that ICD works.  That’s what we use for a
diagnosis code, and that’s what we should continue to use.  And then I
also heard that we need to go to the granularity of SNOMED to talk about
medical history.

But I have heard that more in the context of running physician support
systems than in the context of actually providing that information for making
this decision whether to dispense or not.  And I’m wondering if what I
just made is a true statement or not?

I’m also concerned about the use of the diagnosis code, which gets to the
discussion that came up earlier about the temporal nature of FDA
recommendations, and how they change in time.  And this is with regard to
all places that use the drug.  If we are requiring diagnosis codes, and
then we’re looking at an off-label use of the drug, which also gets in line
with the drug-to-drug interactions that have been proven over time to be false.

Where should the medical literature come into this system, and where should
the regulatory system come into this system?  And if we are going to bring
in the medical literature, what’s actually going on in practice, we have no
standards to do that.  How do we do it?  How do we make decisions?

So, I would just like your comments on what I see as some of the confusing
issues that are coming from the panel on how we should resolve them.

DR. FAUGHNAN:  What do standards let you do?  We already do a
lot.  Standards let you do it much less expensively.  It reduces the
cost of implementation.  And it reduces the cost of
interoperability.  So, typically if you have enough money, and you have
enough time, you can get around most of the things.

Now, some translations between terminologies are just impossible.  You
can get around a lot of things, you can work around a lot of stuff, but it
reduces cost, and let’s us start to do more advanced things somewhere further
down the road.  That’s what they are about.

In terms of the diagnosis codes, it depends on what you are trying to
do.  If you are trying to take those diagnostic conditions and throw up a
list of medications that might be indicated, we still do that with using
ICD-9.  Every so often you get some really whacko recommendations that
way, but prescribing, for someone who knows about medicine, and so they can
sort of guide those recommendations.

If you try to do contraindication checking, and that does not work.  I
can definitely tell you, that does not work.  You can’t use ICD and do
contraindications.  In terms of RxNorm, I think it’s really about
interoperability.  Within our world and our applications, we work with
First DataBank; if we need to work with Multum.

If we need to send drug allergy information to a pharmacy, we can always
use NDC codes to fudge, because you can pick and NDC code, any one of 100, and
back reference it, and you’ll find that it’s dispensable in a local
representation.  But I’ve always been a little nervous about that.  I
think it would be maybe a little safer to have a standard way to represent the
medicine at the dispensable level at least.

In terms of regulatory, the issue that I raised about the black box labels
and drug interactions.  I don’t have a process recommendation for that
other than I think that you guys are the experts in knowing how the government
works — or doesn’t work.  It’s seems to me it’s something the FDA has to
be tapped to address.  I don’t know how to do that.

MR. TRUSKY:  I guess I would add I hope you heard today that NDC is an
available identifier.  It’s not a good one for our group.  And there
are plenty of organizations that refuse to use them.  One of the script
messages are renewal requests from physicians.  Imagine if you can get
that drug identifier, and you are not using free text.  You can save that
physician a whole lot of time reconciling things to his database.

And going back to this pyramid, I think you can get to level four through a
lot of heavy lifting.  When you try and get to level five, that’s where I
think the most that it’s going to be connectivity.  The message is crucial
here, is interoperability.  The minute we start to go there, we have to
rely on national identifiers for drug or for allergies or for diagnoses.

That’s the biggest thing in the bucket we can come out of here with is
getting those identifiers in place such that we can allow the interoperability
to happen.

MR. NUTKIS:  Just as a clarification on the point, I think that my
response was related to prioritization.  And the question is prioritized
based on the pain today, where things are inefficient, and we can do some heavy
lifting, or do we do areas that actually do require new standards, although
they may not be the greatest pain, but there is no work around other than
standardization.

That was really where I was going with that, is which one do you attack
first?  The rest of the domains, including standardization is very
beneficial, but we it today, but there are some things that we just
can’t.  And as I was looking at the list of the presenters from the data
sources that is coming up later, I realized I think we actually contracted
every one of them.

And that’s really become the model for us, because we do have everyone
accepts different models, everyone wants reference guides.  And we have a
hard time with assistance in that area.

DR. HELM:  Just one more thing to add about NDCs versus RxNorm.
Today we force electronic prescribing into an NDC model.  So, we ask
physicians to make a level of drug designation that may or may not be
clinically meaningful to them, and that they may not necessarily want to make.

So, we are competing with the paper prescription pad, and we want to be
able to emulate the functionality that a physician has with a piece of paper,
that if he wants to write a prescription for Tylenol, and doesn’t particularly
want to specify whether it’s a tablet or a capsule, that he is free to do that,
and an electronic environment can communicate that to the pharmacy of the
patient’s choice.

So, that’s really — as we heard and as we speak about different standards,
what we have today is very good, but it’s just kind of what it’s going to take
for us to get to the next level, and with that, increase adoption.

DR. KAUFMAN:  Just a quick point.  One is yes, we can work around
using the NDC codes.  I say we shouldn’t.  RxNorm is basically just
about done, and we should certainly take that.  But the more important
point is that if we step back for a second and look at the whole process, we
have to do e-prescribing right.  This is just the very beginning.

When we talk about the national health information infrastructure, and
sharing the data with an electronic medical record and prescription data, we
are talking about a much bigger deal.  You have to get it right with
electronic prescribing to eliminate all the heavy lifting we can, because the
real heavy lifting is yet to come.

DR. DOERR:  I’d like to comment, I’m supposed to be answering
questions, not asking them.  But what seems to me what has emerged from
all this is this question of how much juice do you get from this squeeze of
these various standards that are being tossed around here?

And I think it might be fruitful, to keep on with this metaphor — no, I
think that in the case of RxNorm, the advantage that is enabled is you have
some electronic prescribing companies who are very well connected.  You
have some electronic health record companies who have gotten diagnosis
codes.  They have physician support, but they are not connected.

And the problem is if the electronic prescribing is using First DataBank,
at least for my company, and here an EHR company that wants to replace their
electronic prescribing module with ours, we can’t, because they are using
MediSpan.

And when those systems for naming drugs, they change at different times, so
it’s just not interoperable.  So, that I think that in the case of naming
drugs, my opinion is there is a pretty compelling justification to go forward
with RxNorm.  It’s pretty well worked out.

I think there is overwhelming justification to have a nationwide standard
for presenting prescriptions.  This idea that every state has their
different way of doing it, that’s just incredibly wasteful now.  And sure,
there are work arounds, but there is not a good reason to keep it the way it
is.

In the case of formularies, there is benefit in standardizing the
expression of formularies.  It really comes from the payers who want to
have formulary status information that is granular down to the level of the
employer group.  And there are services that license formularies that
information like MediSpan.

So, in the case of formularies, there is a work around.  It gets the
job done reasonably well.  So, there is an alternative to forcing the
standards there.  In the case of physician identifiers, it’s really
problematic, when we can’t, and I think there is a need for that.  And I
don’t have the expertise that John has to comment on all the standards, but
these are my thoughts about these few.  I don’t know if you want to add
any comments on any other ones?

DR. FAUGHNAN:  Script we have mentioned a few times, and everybody is
okay with that.  Along with that, signatures, electronic signatures and
authentication.  And then the way to uniquely identify the patient,
however that is done.  And with X12(?) we have algorithms to try a unique
way to identify patients that seems to work well, and we are not in a position
to suggest an approach to take.  There is a need for a way to do that.

MR. NAYAK:  I’ve heard the questions and listened to the
responses.  My general question back is are we responding to the
questions?  The question is what are the low hanging fruits that we can
tackle?  And then there was an expansive question on specific standards
that you asked.

But the general response that I think all of us tend to give is that we
have working around this for a long time.  We can do it, and don’t develop
something that will hurt us all in a particular sense.

Are we really responding to the questions?  I think in the earlier
testimony that we all gave, we sort of identified five or six points that I
think you can boil down that there are areas that need attention I guess.

DR. COHN:  Well, I don’t know if we are going to pursue this any
further.  I think Steve was asking about some additional areas, and I
think we got some fair response on that.

I have a question I want to ask.  This follows along from Dan, your
comment about this issue of authentication and all of that.  And
obviously, it causes me to think about sort of an issue we haven’t been talking
about.  It really gets beyond our agenda for later on, but it’s this issue
of controlled substances.

And I’m curious about what happens in your experience when a physician has
to write a controlled substance.  You said potentially there are systems
where you couldn’t write it, but are they going down to paper?  Are there
ways that they use it?  Are there things that we need to be thinking about
as of now or a little bit down the road to enable this one, or what?

Peter, do you want to start out?

DR. KAUFMAN:  Well, our system is written mostly by Brendan
Berloski(?), who wrote PubMed.  He at the National Library of
Medicine.  Some of you may know him.  And we put a lot of rules into
the system that will take care of these issues.  They take care of it, but
again, work around is horrible.

If someone writes a prescription for a new drug, and then sends it to the
pharmacy, they will get an alert that says it can’t be sent to the pharmacy,
please print it out.  And you have to print it out, sign it, and hand it
to the patient.  Of course that’s a problem in New York, which is
triplicate form.  But it can be done, it can be worked around.

We are scheduled to do 325, those can’t be sent electronically
either.  The SureScripts, those can get faxed using whatever standard form
the state has, as long as it has been in triplicate form for the 325, and New
York has I think a number of drugs that are required to go on the triplicate
form.  But again, that can be faxed on the state standard form.

Again, I understand that end of it.  I understand the basis of
it.  It makes no sense to me, given how much states — the digital
signature, and we are requiring a separate PIN code, user name and password,
and it’s a two password system.  Biometrics is a possibility.  Our
system is triple scripted for strong SSLs.

This is so much more paper than paper prescriptions.  I have had
patients in the past take paper prescriptions and write their products on them,
or change the refills and stuff like that.  This system is preventing
that.

Yes, it’s not perfect, but it’s much, much better than what we had, and I
would strongly urge for some — I don’t the FDA is supposed to be coming out
with regulations.  I imagine everybody at this end of the table and our
company are very good on security, but we’re a little worried about what they
are going to come out with, and how strong those requirements are going to be
for authentication, biometrics, TKI, things that really are not ready.
The infrastructure is not in place yet, and the machinery is expensive.

I know people don’t want to hear it’s good enough, it’s better than what we
have.  I’m going to say it’s certainly much better than what we have, and
I would hope that we could figure out some way to accept firm authentication
for controlled substances.

MR. NUTKIS:  I always find this kind of interesting.  This
actually came up in a focus group last week or the week before.  We were
talking about our security now.  We are rooted at a security
company.  We have full time cryptographers with various certifications
from the Department of Treasury.  The messaging goes to our data centers
as does a number of other large financial institutions as well.  We take
security very seriously.

It was very interesting when we were in the focus groups, and I think in
this case it was the physicians and maybe some of their office staff as well,
who were not really sure that the technology is really secure.  And we’re
all looking at each other behind the glass going, you mean more secure than
picking the phone up, calling the pharmacy?

And we weren’t even talking about controlled substances here.  We are
just talking about regular prescriptions that they, for some reason, thought
that there would be the possibility of abuse.  And what was interesting to
us is that there are a number of pharmacies now that will take things like
Vicotin(?) and others, and will take them as call ins if they happen to know
the physician or the nurse.

So, it was interesting, the perception, and I think people from the
security side will know that security-wise we can, there is no question with a
standardized authentication and authorization process, this isn’t an issue.

We have implemented TKI and strong encryption, Defense Department systems
already that have extremely — a greater degree of requirements.  And a
lot it comes as what is the authentication process?  What are the security
mechanisms?  How do we vet the individual to know that the individual
actually writing that prescription is in fact the person that they say they
are, and is in control of that device at that time?

And there are standards that exist for various other related topics.
So, there is the Federal Bridge Project.  There are a whole bunch of other
projects that are already at the federal level.  We level it’s something
that can be overcome pretty quickly with education.

I think there also is a VHA — isn’t the VHA doing a pilot right now?

MR. BLUSY(?):  Luca Blusy from VHA, the Pharmacy Benefits Management
Group.  Right now, they are writing a pilot.  They are just about
ready I think to test it, with the hopes that by the end of the year they will
have requests to electronic prescription.  It may be tried first at the
VA.  That’s some of the information that we have received, and work out
the wrinkles.

But they have been working on it for about four years now.  The main
concern that they have is that they don’t want to put up a window of
opportunity.  They know that the prescription process as it is now, is not
perfect.  But they don’t want to come up with something that can really
open up it for people to do fraud(?).

DR. COHN:  Thank you.  We’ll just sort of note that as an ongoing
issue that we’ll be talking about I think later on this year.  I guess I’m
not hearing from any of you that it’s a major barrier to the success of your
systems, which is good.

Other questions, comments?  Lynn, did you have a question?  Come
and introduce yourself.

MS. GILBERTSON:  Hi, I’m Lynn Gilbertson from the National Council for
Prescription Drug Programs.  Jeff, you had some questions about DCP
IDEA.  I don’t know that I can answer them, but I will be glad to provide
to both you and Maria, the contact people, so that you can get any questions
you have answered about the DCP IDEA.

MR. BLAIR:  Thank you.

MS. GILBERTSON:  Just a couple of just very brief comments.  The
Scripts standard does have the ability to do a compliance transaction.
It’s a fill, no fill, partial fill, and refill information.  For example,
if a new prescription is sent from the prescriber to the pharmacy, and the
patient never picks it up, a no fill transaction can go back to the
prescriber.  Likewise with a refill that the patient never picks up,
that’s refill one, that can go back as well.  The standards of course is
just when the industry is ready to play together on that.

On prior authorization, one of the things that complicates it is well, who
is doing the prior authorization.  Some authorizations are required at the
pharmacy to perform it, and there are standards for doing that, and they are
normal telecommunications standards.  There is also the prior
authorization by the prescriber itself.  So, even prior authorization,
like formulary, is not an exact science either, and there are different flavors
to it, depending on what the health plan has set up.

And the last item is I have a plea.  And that is over the last couple
of days I have heard lots of comments about Scripts should be enhanced, to
Scripts should support this.  Please don’t be afraid to submit changes to
the standards, because for example, I just want to make sure it is clear for
example, NCPDP, the organization, does not create the standards for the
industry.  The industry creates the standards.

So, if there are things that need to be added to it, anyone, members,
non-members, anybody breathing can submit a change request, and get it
discussed in the consensus body for what you might need.  So, don’t be
afraid that you have to wait for somebody else to do it, because if it’s not
you, who?

DR. HUFF:  Now, I do have a question.  Why would they want the
pharmacist to be responsible for pre-authorization?  They are scared of
the physicians?

MS. GILBERTSON:  I’m going to rapidly get over my experience in
this.  There are a couple people in the audience who I think can probably
answer a lot better.  There are requirements at the prescription level, at
the drug level that is related to the number of prescriptions allowed in that
particular time frame.  That might be one example of something that needs
to be prior authorized.

There are some situations where the pharmacy can submit the prior
authorization.  They may go back to the prescriber and ask for further
information, and then provide that, submit the prior authorization transaction,
and then bill the client.

DR. HUFF:  Half of it doesn’t make sense.

MS. GILBERTSON:  I don’t claim to understand why things do or don’t
make sense.

MS. BURN(?):  This is Carrie Burn at RCloud(?).  I’m actually
speaking from my PBM experience.  I’m not sure that the pharmacist can
actually ever prior authorization something.  I think typically they will
get a prior authorization by running the claim, or by calling the payer, and
then put the number on, or tell the doctor at Scripts that they need one.

But I think typically the prior authorization is most likely questioned
between the doctor and the payer.  And criteria needs to be met.  The
pharmacy sometimes intervenes and makes the phone call for help.  But I’m
not sure that the pharmacist can actually ever give the prior
authorization.  Is that new?  Not in any system that I know of.

MS. GILBERTSON:  The pharmacist requests the prior
authorization.  That’s the function.

MS. BURN:  Right, that’s the process, because they are the ones
running the claim, and the claim comes back and says prior authorization is
needed.  The prescriber should understand that prior authorization is
needed, and then prior authorization should be requested at that point.
The pharmacy would not intervene in that case.

DR. COHN:  Good.

DR. HUFF:  This may be a level of detail that maybe we shouldn’t try
to do here in this meeting.  It’s hard to design something in a meeting
this big.  But it seems to me that there are things in the work flow that
are work arounds, because you assume that somebody isn’t available, or can’t do
this.

It would seem like that if you could get all the information to the
prescriber, the pre-authorization information, the formulary information, the
drug utilization review information, all of that at the point where the
prescription is being done, and then the right prescription is done, and you
know that that prescription is now correct according to what the physician
wants to do, what the co-pay is going to be, what the benefits are allowed in
this program, all that sort of stuff.

Then it wouldn’t have to be redundant at the pharmacy, it wouldn’t have to
be redone and rechecked.  Now, it might be that we have to put a
pre-authorization number then with the prescription that flows all the way
down, so that they know that that met requirements, and there might be other
things like that, that need to be happening.

But I guess I would appreciate your perspective on that.  Is that
really true?  Or are the things that really need to be redone in the
pharmacy and at the payer in order to make it work right?

DR. COHN:  Can I maybe try to deflect this one just a little bit in
the sense that I’m looking at the issue as really pertinent for what is more
academic in the sense that for the next many years, you are going to have to
enable the pharmacist to do a prior authorization if a physician isn’t using
e-prescribing.  Right now, the pharmacist is wired and the physician
isn’t, so that we are always going to have the capability in case the physician
is not wired for e-prescribing capability, for the pharmacist to be able to
handle it.

So, am I off on this one?

MS. GILBERTSON:  I think that would be an excellent question to pose,
especially to the pharmacy and perhaps the prescriber testimony as to what
functions do they perform that they feel are pushed down to them, or pushed up
to them, or what could be performed early or later in the process.

I think there is a lot of de facto stuff that — it would be interesting,
because you are going to hear positive, and you’re also going to hear negatives
about somebody coming in between something that is an established
practice.  So, I think that would be very interesting to ask them for
responses back.

DR. HUFF:  You’re right, we need to allow the transaction to happen
the way it’s done today.

DR. COHN:  I agree with your redesign.  It’s just the flexibility
needs to be in these standards.

DR. KAUFMAN:  I just had a quick comment.  You are talking that
if the people who write the software for the backbones, and the people that
write the software for the physicians to use.

And I don’t know whether you have any power over the people who write the
software that the pharmacists use, but perhaps the pharmacy software could be
required or at least asked to have the capability in their system to score a
serial number and company name for each prescription.  And then the
ability to match that when a prior authorization comes in directly from the
insurance company.

So, the 26 seconds or so that Dan talked about before, could happen
actually offline, and the prior authorization could come directly from the
insurance company into the pharmacy computer system, and the pharmacist could
see that the prior authorization was received.

MR. NUTKIS:  And by the way, that is how our system works.
That’s how we have done it again, based on the limitations that existed
today.  We built around that.

DR. HELM:  And I think the way we see prior authorization working is
that there is as much information as practically available to the physician at
the point of care, so we can do that prior authorization.  But then
communicate that information to the pharmacy, because they are going to be the
ones that are going to be seeking reimbursement for that particular medication
from a health plan.

So, not to have them redo the work, but to know that the work is there,
with sufficient confidence that they are going to get reimbursement, which is
kind of the whole purpose for prior authorization anyway.

DR. FAUGHNAN:  And speaking from the pharmacy side, one of the things
I have heard from our group who works on that side is that — and this is I
think more in a nice to have, not in the essential stages — is encoded reasons
for refusal of prior authorization.

Pharmacists just that are handlers of a lot of transactions, that are doing
authorization requests could then interpret the coded response and resubmit in
some cases without human intervention.  And so, that is desirable for some
of our high volume pharmacies who automate response to an initial denial.
That’s a potential standard.

MR. RODMAN:  Bill Rodman with Express Groups.  I just wanted to
clarify — I don’t know how seriously you are considering standardizing prior
authorization as part of the electronic prescribing, but there are probably a
lot of issues that haven’t been raised here that need to be put on the table if
we are even thinking of going down that road.

Sometimes it’s as simple as it has been described, where it could be
automated.  But very often it isn’t.  There are different reasons for
prior authorization, and there are different payers with different philosophies
about prior authorization, that you can’t fit everything into a neat little
bucket.

So, if it’s a simple step therapy and all you have to do is confirm that
somebody did the first step, then you might be able to automate it, assuming
all the payers that are doing the prior authorization, have done the work to
automate it.  Sometimes there are quantity level limits.  There are
different kinds of things that might fall into a bucket of prior authorization
that you might not be able to standardize.  And there are certain payers
who use prior authorization really to drive different utilization decisions, as
opposed to just making sure the people get the right drug at the right time.

So, some payers would welcome the opportunity to do automated prior
authorization.  If you answer the questions right, you get the prior
authorization.  And others would rather have a barrier in place, because
what they really want is the physician to prescribe something else.

And so, I don’t think you can talk about automating prior authorization in
a vacuum, because what you are talking about is standardizing benefits, and you
can’t really go there, I don’t think.

DR. COHN:  Well, I think there is a difference between standardizing
the benefit, and enabling the process.

MR. RODMAN:  If you say prior authorization has to be done in a
certain way, be careful not say the answer has to be preordained, because
people need the flexibility to be able to use prior authorization the way they
want to use it.  So, it isn’t necessarily always something that putting
those rules in place, would —

MS. GILBERTSON:  And whether or not it would be offered in the
Medicare program at all.  If it’s not part of that, you may put that aside
for a while.

DR. COHN:  Jeff, did you have a final comment?

MR. BLAIR:  I just think that one of the observations that you made is
something that I think there are many portions of this e-prescribing community
has anxiety about.  And you mentioned that there might be intentional
barriers related to this issue.  And so, I think it would be good for us
to discuss this later at another time, as to whether those barriers were
appropriate barriers, or whether standards need to be developed to address
them.

DR. COHN:  Okay, with that note, we want to thank our panelists for
really fascinating presentations and discussion from a lot of people, and a lot
of information.  People are taking notes all over, so we’ll be putting
this all together.  So, once again, thank you.

The meeting will adjourn for lunch.  We will come back at 1:15
pm.  Thank you.

[Whereupon, the meeting was recessed for lunch at 12:15 pm, to reconvene at
1:15 pm.]

A F T E R N O O
N  S E S S I O
N    (1:15 pm)

DR. COHN:  Well, I want to welcome everyone to the afternoon session,
post-lunch.  I want to thank you all for coming, and I would like to
welcome our presenters for this afternoon.  Obviously, the topic here
continues to be e-prescribing, and really the focus here is the drug knowledge
base, the vendors and your thoughts or perspectives on all of this, and we want
to thank you very much for coming.

Now, could we just have you introduce yourselves?

MR. BIZZARO:  Tom Bizzaro, pharmacist, and I’m Vice President of Sales
and Marketing for First DataBank.  I’m also very active with NCPDP, as a
representative of First DataBank.

MR. ROBINSON:  I’m George Robinson, Vice President of Knowledge Base
Development for First DataBank.  I’m also a pharmacist.

DR. MILLIKAN:  I’m Mark Millikan.  I am the Editorial Director of
Product Development for Medi-Span.

MR. HOBSON:  I’m Gerry Hobson.  I’m the Research Manager for
Cerner Multum.  I’m also a pharmacist.

DR. COHN:  Okay, well, thank you for joining us.  George, you are
leading off the discussion?  Please.

Agenda Item:  E-Prescribing:  The Drug
Knowledge Base Vendors’ Perspective – George Robinson and Tom Bizzaro, First
DataBank, Inc.

MR. ROBINSON:  Certainly, it’s a pleasure to be here.  And I
think both Tom and I have really seen a very constructive attitude taken by
participants and committee members in the last couple of days.  So, it’s a
very encouraging development.

I think the drug knowledge base vendors have been really working very hard
over really the last 10 years trying to prepare database structures to support
full electronic prescriptions, and not only ordering, but order
fulfillment.  So, we are very pleased with what happened here.

We have had several vendors talk in detail about work flow issues, and I
think I’m going to take a little different approach here.  I’m going to
talk about the current role of knowledge base vendors in the industry, and then
talk about what I think our continued role will be once standard terminologies
become present, once we start seeing increased standardization of messaging,
and see if that makes sense.

And then Tom will jump in and save me when I make some egregious mistakes,
and also he has some comments to share.

Fundamentally, what we do is we collection information, we codify it.
We link it to concepts that are useful at point of care applications,
distribute it to the industry, and that is used within point of care
applications.  So, in a nutshell, that’s what we do.

From the drug formation collection, what we rely on is we really come
through the product labeling information that is distributed by the
manufacturers.  We also look at information related to over-the-counter
medications, alternative medications, and a limited subset of medical supplies.

Above and beyond that, we also look at primary literature.  We look at
FDA publications such as MedWatch.  We look at manufacturing letters and
publications, the dear doctor letters, as they come out.  We’ll also
compare our information with the official compendia, as more of a measure of
are we in the right ball park with that we are doing.

As far as drug information codification, our roots go back really
supporting electronic claims and inventory management for retail
pharmacy.  So, we have a well developed database of NDCs that also extend
to the universal product codes and health-related identifiers.  So, in the
NDC field of our products, we also have numbers of abandoned DCs, to make that
clear.

In addition to that packaging information, we also link other knowledge
based content such as images, imprints, pricing information for drug form
information, which defines it as a HLM or gram, which is a central component of
electronic plan installments.

What we do with the clinical contents that we gather, we aggregate to a
formulation level.  It would be analogous to a clinical drug level of
RxNorm.  That particular concept includes a distinct set of active
ingredients, route of administration, dosage form, therapeutic classes, and
also related to clinical knowledge.  So, you have a National Drug Code or
package product that is linked to a formulation-type of identifier that has
further relationship between drugs(?).

Also, over the last few years we have started using of use of content in
either nursing applications, prescriber applications, order entry
applications.  We observed that the world used drugs differently than the
pharmacist.  In other words, drugs are used as a name base type of
identifier, not as a number for an NDC.  So, many of the drug knowledge
base vendors through the years have introduced name-based concepts that are
very useful, and are used today in several prescribing applications.

We also have introduced several versions of therapeutic classes that have
been used through the years.  Many of our therapeutic classes are actually
imbedded either in state Medicaid plans or in Medicare.  I think somebody
in Part B was using the enhanced therapeutic classification of First DataBank.

The way we use the therapeutic classification system, it’s more a process
of looking at how drugs are most commonly prescribed by class.  So, it can
be all over the place.  It can be chemical.  It can anatomic.
It can pharmacologic.  So, it’s a hodge-podge of grouping things.

Most knowledge base vendors also have a drug indications database.
And embedded with the drug indications database are the actual nomenclature
describing condition.  It may be a higher presentation of conditions, so
they roll up.  You can look at cardiovascular drugs, and then just take it
down to hypertension or to whatever type of specific indication you want.

I just want to get that clear that there is drug indications database, and
then there is the therapy classification.  And they really serve separate
needs.

We provide allergens information.  It really covers three or four
different domains.  There is an allergy class, which is the most general
way of capturing an allergen; individual ingredients are put in the context of
the patient allergy; or either a name-based drug identifier, or actually an
NDC.  An NDC can actually be put in the context of the patient allergies.

As of this date, there is no national standard representing allergen
information.  We have worked cooperatively at least with one customer
looking potentially at SNOMED-CT as an interoperability layer for that.
But as of right now, there is no standard.

In order to support point of care screening applications and navigations
related content, we really break I guess our vocabulary content into four
different areas.  There is the drug contents that we talked about,
allergens, which we have discussed, and there is the domain of disease, and the
domain of lab.

As far as the disease space content, in order to respond rapidly to our
need to present concepts that are condition-based which described what
indications, precautions, drug-disease contraindications, and side effects, we
maintain our own proprietary disease vocabulary.  It’s a fairly small,
controlled subset.  Our strategy with that is to be able to express the
sort of concepts with a lot of flexibility, and then link that disease database
back to standard terminology.

So, through the years ICD-9-CM is all we had.  We are now looking at
cross-referencing to SNOMED-CT.  So, that will take us a few months, but
that is something we are looking at.

The important thing I guess about the disease is where the knowledge is
aggregated is you will take a drug concept, and you will relate it to for
example hypertension in the context of a drug to treat.  For another drug,
hypertension could be presented in the context of a side effect.  In
other, that context can be as a contraindication.  So, they are the same
term, but they are linked to the drug concept, depending on the relationship
that you state, and they have a different purpose.

Also, behind the scenes the hierarchical relationship that you maintain
with the disease vocabulary may lead to differences based on the purpose you
are trying to resolve.  You may make one approach, and when you are
rolling up the condition for the contraindication screening, and another
functioning when you are rolling a condition to do drug indications.

I think a lot of the issues that perhaps we had raised over how difficult
it was to do drug contraindication screening with ICD-9 is based on perhaps the
roll up, and the lack of granularity in the ICD-9.

As far as lab based content, it’s similar to the disease database.
Our attempt is just to provide the lab concepts that we need to describe the
context of labs that in combination with the drug, may give you a false reading
with the drug-lab interference.  It may be there in the context of if you
are on this drug, and you have an elevated lab result, perhaps we should do
something with your drug therapy.  Should you continue the drug, hold the
drug?  It’s more of a drug-lab warning.

The other type of link that you do between drugs and labs would be drug
therapy assessment.  For example, if I’m on a particular statin, then I
should have a liver function that is done every three months or whatever that
is.

As far as dosage information and screening, again, we use our proprietary
drug concepts.  It may be linked to our disease concepts, because based on
diagnosis you may have a dramatically different dose, especially with some of
the antibiotics, with end use priority methods of prescribing a dose string,
and also representing that dose in discrete parsed elements.  So, that’s
the state of affairs as far as the knowledge.

I think commercial of the knowledge base providers —

DR. HUFF:  You mentioned you use lab codes, that you have lab
associations.  Do you use an internal coding system?

MR. ROBINSON:  We use a proprietary system and provide links to
moiety(?) strategy.  And it’s a fairly small database.  One thing I
did want to mention, because one of the requirements that were discussed was
drug-lab interference.  Pure drug-lab interference, we are talking about
analytical effects.  You need to go all the way through methods used for
the lab.

So, in other words, if you are looking at serum fat, and there is a drug
honestly that may give you a false reading, it’s not on that that you measure
the serum fat, and have that problem.  So, if you are actually trying to
introduce this in the screening environment, what you actually need to do is
associate with your lab information system, the actual methods used to measure
the lab through the particular vocabulary codes that are used in the system to
do the true screening.

So, I guess what I’m just trying to throw out there is some of the
complexity of doing screening at the point of care environment are not
necessarily as easy as it looks from the outside.  There are some factors
that have to be considered.

Common to most knowledge based vendors is they may provide API codes,
application-program interfaces that encapsulate the data, and they may use
object technology that are based on an input parameter that will pass into the
program, and will retrieve the text and discrete messages of screening
results.  So, that’s a convenience option that we provide to our
customers.  What we cannot do is provide the actual user interface.
That is up to either our self-programming customers or our vendors.

Our product delivery, it’s pretty common I think with the vendors that
pricing may be delivered as frequently as daily for the more high volume type
of enterprises.  They may get our information on a weekly basis, or you
may get a monthly or quarterly delivery schedule.

Once the vendor or self-programmer receives our content, what they need to
do then is load our content into their database.  Some large vendors
actually maintain their own centralized terminology layer, so they may be doing
some interfaces there.  And then once they have integrated the data into
their internal database structure, then they push it to their customers.
It may be CD.  It may be an Internet push.  There are all different
ways to do that.

So, that’s the state of current affairs.  Is there anything you want
to add, Tom, before I go into how we see the world changing a little bit?

MR. BIZZARO:  I think what we wanted to do was to show the complexity
of what we as database providers give to our customers, and the breadth of the
screening functions that we have.  I think that this committee’s — one of
the committee’s roles is going to be to see how much of that is part being
prescribed, or is that a separate function, and not an e-prescribing function.

Because by improving patient care by providing screening at the point of
care in the physician’s office for all these different things, you also enter a
level of complexity and a time issue.  So, what is e-prescribing, and what
is a screen that is separate from e-prescribing?

MR. ROBINSON:  Okay, moving forward in the future enlightened
environment that we hope comes our way soon, is we welcome of course the
structured product labeling that is coming from the manufacturers based on the
new labeling standards that are being I guess approved as we speak, this week.

What this is, is perhaps the means to get some internal efficiencies in
looking at the product labeling itself.  We still will have a process of
needing to review that content and import it essentially to our database
structures.  And along with that also, continue the compilation of
complementary content that comes again from primary literature.

So, for example, we would not get into the structured product labeling with
the optimal uses.  So, that type of information, again, we would take an
evidence-based approach, and broaden the scope of the information we delivered
to be more than just what limited the manufacturer information and the
labeling.

We do see the opportunity that we could perhaps build additional links to
another proprietary database in the structure product label.  So, that may
give the opportunity for an end user to make a direct call to perhaps a black
box warning section of the labeling, or the side effects, more from a reference
perspective.  So, I think that would be in the real of possibilities.

As far as drug information notification, we are not really seeing anything
changing with the product ID or the NDC other than we may be in a position to
redistribute links between the NDC and the RxNorm clinical drugs.  That
would be in the realm of possibilities.

As far as formulation base or clinical drug identifiers, I think we would
see the opportunity to public links between our FEBC(?) code, our DECFC(?)
number, or clinical simulation ID, and the generics of some clinical drug that
will be coming from RxNorm.

And at the ingredient level we would see the opportunity to provide a
cross-walk between the FDA UNI created as an identifier, and our ingredients.

So, the trend we are seeing here is in order to minimize the impact to our
current knowledge base, is that we would like to have the opportunity to
continue to use our contents really without any apparent change to the end
user, but have links to the interoperability standards set, so that when they
are working with other trading partners, the information can be translated and
communicated in that fashion.

Name-based identifiers, this is where we would have the opportunity to link
with additional branded semantic clinical drugs’ IDs that may be coming from
RxNorm.

As far as therapeutic classification, this is still kind of an
unknown.  I understanding in reviewing the act that the USC has been
challenged to rate some therapeutic classifications.  I think we’ll have
to see how that turns out.  One therapeutic class compared to another can
be really be an apples and oranges situation, so I don’t know if that’s a
situation where we’ll actually be able to offer a cross-reference or not.
I guess we’ll just have to see how that pans out.

Allergen information, again, that’s an open issue I think from a standards
perspective.  All I can say is First DataBank would be willing to
cooperate with whatever standard is selected there, or work with our standards
customers to on a case-by-case basis cross-walk.

Disease-based content, we certainly have plans to link to SNOMED-CT.
One issue with SNOMED-CT will continue to be the issue of how willing the
College of American Pathologists is willing to modify their editorial policy to
introduce pre-coordinated terms.  If you look at a lot of the drug
indications, they actually can be very specific and sometimes creative in how
they are published in the labeling.

There may be a reference to a co-morbidity.  So, it may have a
particular indication for an infection in combination with immune-suppressed
populations or with HIV.  So, those are concepts that many times you may
not find in exact SNOMED-CT concepts.  So, there is that dilemma of do you
do post-coordination of qualifiers, or do you join multiple SNOMED concepts,
active mapping.

The more post-coordination we do in the mapping, the greater the complexity
is going to be.  Again, someone making a call using a SNOMED-CT concept
into our knowledge, needs to get additional information.

As far as laboratory, we are already committed to coding for the LOINC
database.  As far as dosage information, we would be happy to cooperate
with any third party that would want to step up to the plate and come up with a
standardized grid terminology.

API tools, I think about the only change I would see there is what we may
need to do is code to handle standard terminology as the input property.
In lieu of coming in with proprietary drug code or proprietary disease
concepts, we would need a handle to recall using SNOMED-CT as I have this
problem, what are drugs to treat?  So, then we would make crosswalk for
indications database and provide the information.

Product delivery.  We would be open to potentially delivering some of
the standard terminology along with our proprietary content.  I think we
have to get down to obviously the copyright issues.  But what I am seeing
is that it may be somewhat of a dilemma for customers to coordinate delivery of
standard terminology when they are getting it from virtually five or six
different courses.  So, there may be a convenience opportunity for a drug
knowledge base to redistribute some of those in a terminology.

Finally, at the customer level, again, what we would hope is that we could
do this in a way that is somewhat transparent.  One of the big mandates
that’s on the actors, can this be done with minimal impact existing
systems?  Obviously, they will have to code in order to send a standard
code set, and to use it as an inference parameter.  What we would hope is
that we could minimize the impact at least to the user interface.

The goal would be for an end user to write a prescription as they had
previously, but behind the scenes the miracle happens.  The drug is
converted to an RxNorm concept.  They make the call to formulary if they
need to, or it gets sent to the pharmacy.  And all those things that can
happen behind the scenes, and not mess up the physician’s day would certainly
serve us all.

So, based on that, I’ll turn it over to Mr. Bizzaro.

MR. BIZZARO:  What I have found in the last day and a half, and I want
to tell the committee that I found its testimony both very interesting and
educational.  I’ve been taking some notes in some areas I see as
problematic, and I want to just comment on them from both my experience with
First DataBank, and also with my experience with customers, and from the
standards organizations that I have dealt with.

RxNorm I think is a name-based identifier, and is an identifier that
hopefully represents the way doctors think.  It makes an awful lot of
sense, and George has talked about our willingness to cross-reference to
it.  I also need to keep in mind the NDC does not go away.  We still
need to have links to the NDC, to the RxNorm, because in the pharmacy
eventually you have to bill for it, and to bill for any packaging spotting
information at the NDC.  And I think we have heard that reiterated a
number of times here.

The issue of who sets the standard I think is going to be a very thorny
problem.  NCPDP, over the last six or seven years, has tried to get
consensus between the payers, providers, and other organizations that attend
that meeting for standards today.  And we could not come up with consensus
for establishment of a standard set.

The main issue was that in other systems, and I guess any system now that
is writing orders, there are numerous code sets used within those systems right
now.  And they are all individual proprietary code sets.  And the
problem that pharmacy had, even if we could come up with standard code sets for
SIG, they thought there was a difficulty in cross-referencing what they
currently had in the pharmacy, and which of the standards that the code set
would be.  And they were worried about the liability of making an error in
that cross-referencing.

So, the standard SIG I think is a great idea, and I think it has great
value.  And I also think it’s very difficult to get.

I also believe the pharmacies — and I wish there was more representation
here from retail pharmacies — it is my belief that retail pharmacy is going to
require a contextual representation of the prescription even if it is
codified.  I think they are always going to want the name of the
drug.  They are going to want the SIG in the contextual format, because
they feel that that gives them a fall check, and it helps them avoid the
liability if you would have an error in the code, the identifier.  I think
we really need to keep that in mind.

I think we are going to have to also deal with the duplication of effort
with clinical screening.  We do clinical screening in the physician’s
office, and then we do clinical screening once it gets to the pharmacy.
And we do clinical screening once it goes out to the PBM, whatever order that
happens.  I think we are going to have duplication of effort there that we
are going to need to learn to deal with.

I think there is an opportunity for messaging that goes back and forth
between those entities that says I’ve already forgotten this alert.  I
have reviewed it, and I have decided to still go forward with it.  So,
that messaging complicates issues, but I think that’s the only way you will get
these entities to give up what they would see as their responsibility to
clinical screening.  And also, again, the liability issue if they would
not do that clinical screening themselves.

Yesterday, Harry Reynolds mentioned that he thought that there seemed to be
in the industry, an acceptance to this e-prescribing initiative.  And I
would absolutely agree with that.  I think that we all can see the benefit
of e-prescribing, both as a patient safety issue.  Also as the
efficiencies that we gain by doing e-prescribing.

I also think that in the industry there is a concern that we may make the
requirements for e-prescribing so complex that it becomes very difficult to
implement, and also very difficult to get acceptance in the marketplace.
So, I think those are some of the things we need to look at as what can we do
to make e-prescribing an advantage in the business place to those who want to
use it, because we keep talking about the economic impact, and I think that is
certainly true.  And there are certainly benefits to e-prescribing that we
can make the physician aware of.

And in closing, I would just like to again thank the subcommittee for the
opportunity to speak here today.

MR. ROBINSON:  I have one additional closing though.  I think I
read yesterday from the complexity that we are seeing with so many components
of the standard terminology coming from different providers.  So, speaking
for First DataBank, we certainly would be open to a champion for aggregating
this information.  If that’s UMLS or the National Library of Medicine,
that would be great.

And perhaps bring together a work group or a forum where we could have a
roll up our sleeves meeting between all the participants to really figure out
how we are going to make this work, because there are several
dependencies.  You have to have an agreement for example, before you can
link the ingredients in a packaged product.  We need that therapeutic
class in place.  If it’s a new, innovative drug, the therapeutic class has
got to be figured out before we can associate drug to the therapeutic class.

If it’s a new and unique indication, there has to be a SNOMED code to
represent that indication on the labeling.  So that we have all these
interdependencies that if we are trying to build systems around it, you’ve got
to be able to get that information out quickly, and also figure out how that is
going to passed then to the entities that are actually distributing standard
terminology, and getting it to these under systems, and eventually sort of the
eyeball for the end user.

So, my humble suggestion would be probably the National Library of Medicine
is probably in the best position to do that in a non-proprietary, competitive
form.  So, that would be useful.

DR. COHN:  Thank you very much.

Mark, I believe you’re next.

Agenda Item:  E-Prescribing:  The Drug
Knowledge Base Vendors’ Perspective – Mark Millikan, PharmD, MBA, MediSpan

DR. MILLIKAN:  Thank you.  I am Mark Millikan, the Editorial
Director for Product Development of MediSpan, a part of Wolters Kluwer
Health.  With me today is my colleague Karen Ebhart, which many of you may
know from previous NCVHS meetings, and she is a project manager at MediSpan,
and my technical back-up for these discussions.

MediSpan has been in the business of providing drug product information
since the late 1970s.  My initial comments will focus on NCPB Script
standards, which MediSpan recommends for the transmission of electronic
prescriptions. Script’s wide use and acceptance is our strength, that to
physicians, this standard for the enhanced e-prescribing as specified by the
Medicare Prescription Drug Improvement and Modernization Act of 2003.

The Script standard developed a foundation to build upon, and contains the
data necessary for basic communications between organizations involved in
e-prescribing.  As noted in previous testimony by NCPDP in March 2004,
enhancements to the current Scripts standards are needed.  MediSpan
recommends the following enhancements, with the proviso of Lynn’s comments
earlier today, that all of this is subject to the normal processes at NCPDP,
including the balloting process.

First of all, MediSpan recommends that nationally adopted drug products and
prescription instructions, terminology code sets be added to the Script
standards, which already include proprietary drug terminology code sets and
text descriptions for the purpose of identifying drug products.  The
inclusion of nationally adopted code sets will facilitate interoperability
between e-prescribing companies that use different proprietary drug and
technology code sets.

The current Script standards allows drug products to be identified by
proprietary drug terminology code sets such as those MediSpan.  In
addition, drug products and prescription directions are also identified by text
description of the drug product.  MediSpan supports the addition of a
nationally adopted code set such as RxNorm to the Script standard for the
purpose of identifying drug products.

MediSpan also supports the addition of a nationally adopted code set that
would identify prescription instructions for the SIG.  At this point, this
is a limitation of the Scripts standard, the ability to transmit descriptions
and instructions by any means other than a text field, and is an area for
enhanced support.

In addition, a code set that tries to take elements of the prescription
instructions has yet to be adopted on a national basis, as we have discussed
over the last couple of days.

We believe that continued inclusion of text fields and proprietary code
sets will facilitate quicker adoption of e-prescribing systems.  Limiting
the transmission of electronic prescriptions to nationally adopted codes sets
code slow down the adoption.  While nationally adopted code sets are
evolving, the proprietary code sets and text descriptions are available for use
today.

In addition, this approach also provides the opportunity for a human to
read the transmitted text field against the transmitted code sets, therefore,
providing the health care professional the opportunity for a final check
against the transmitting codes.

MediSpan recommends that drug utilization review alerts, summaries, and
communications between organizations involved in e-prescribing, as noted in the
act, be incorporated into the Script standards.  The Medicare Act
describes the change of information on drug-drug interactions, warnings and
cautions, and dosage alerts.

We recommend that this information be supplied as a summary of the DUR
screening alert generated in the prescribing, dispensing, or reporting
processes, along with the actions taken.  For example, if a prescriber is
alerted of an interaction between a new prescription and a drug already on the
patient’s profile, but decides that the use of the prescription is appropriate,
this action should be transmitted to the dispensing, so that the dispenser, the
pharmacist is aware of the prescriber’s position.

As a part of NCPDP’s health communication standard, it is routine to
transmit DUR results between pharmacies and insurers or payers.
MediSpan’s recommendation is to incorporate this same approach into the Script
standards by allowing for the transmission of actions resulting from the DUR
alert from the prescriber to the dispenser.

We do not, however, recommend transmitting complex monographs of DUR
information.  This information should probably reside locally in the
e-prescribing system, but is not necessary for the electronic transmission of a
prescription.

MediSpan recommends that the version of the Script chosen for e-prescribing
include a longer drug product description, or that the dosage form, dose
strength field available in Script standards are made mandatory.
Currently, the drug main field of Scripts is mandatory, while the dosage form,
dose strength, and dose strength information fields are treated as conditions.

There are, therefore, implementations of Scripts that incorporate the
dosage form, dose strength, and dose strength unit measure into the 35
character drug name or item description.  Thirty-five characters for the
drug description is not sufficient.  Recent JCAHO and IECC guidelines and
recommendation are to remove dangerous medical abbreviations from drug
orders.  These guidelines cannot be accommodated within the constraints of
a 35 character field length.

We recommend either the addition of a drug prescription extension for drug
descriptions longer than 35 characters, or changing the dosage form, dose
strength, and dose strength unit of measure fields to be mandatory in the
Scripts standards.

And finally, with regards to the Scripts standards, NCPDP’s portfolio for
standards includes a standard for transmitting an entire formulary.  Based
on information that we learned yesterday about RxHub and CAQH, we could see
that incorporating what NCPDP is already begun to develop in the various
formularies, as well as what RxHub and CAQH are developing should be eventually
rolled into Scripts.

I would now like to answer questions that we were asked to address during
our testimony today.  First of all, best standards for codes sets meet the
requirement of the law.  As stated in my earlier comments, MediSpan
recommends the use of the Scripts standard, along with the necessary code sets
to meet the e-prescribing requirements established by the act.

A nationally adopted drug terminology code set isn’t necessary for the
interoperability between proprietary drug terminology code sets.  We
support the use of RxNorm for this purpose, and are exploring ways that we will
be able to make use of that, and be able to create mapping between RxNorm and
the MediSpan code set that we provide to our customer base.

Which of the standard code sets do we use?  MediSpan has a long
history of supplying drug-related code sets for use within standards
established by NCPDP, including the Scripts standard.  And this spans
current and future use of standards, along with the further development of
proprietary code sets and mapping to nationally adopted code sets are
customer-driven.

What are the strengths and weaknesses of the standards and code sets?
As we have mentioned, the strength for NCPDP is it’s widespread use and
acceptance.  And I have also touched on some opportunities for
enhancement, including the 35 character drug product description when it is
implemented as the full description of a drug product.

In later versions, separate fields were included for strength, strength
units, and dosage forms, as well as repeating fields for longer product
names.  It appears, however, at this point in time that these later
versions of the Script have not been widely used.  So, it’s therefore, our
recommendation that these later versions and future versions be adopted for the
purpose of e-prescribing.

Another opportunity, as we have also discussed a couple of times today,
would be enhancing Scripts standards for the addition of codified or partially
codified field of the prescription instructions or SIG.  Partial
codification of the prescription directions would frequency per day, unit per
day, and duration of therapy information to be included or expressed in
separate numeric fields.  This would then allow checking to dosage alerts
and that sort of thing.  We could do the appropriate match.

It is our understanding that HL7 is starting to create structured
prescription instructions with the cooperation of NCPDP.  MediSpan would
support this effort, and recommend that this be considered for inclusion in
Scripts some time in the future.

What are the e-prescribing standard and code set gaps?  Again, several
things have already been mentioned for the past couple of days.  One would
be what we have talked about with the SIG in terms of being able to map for the
purpose of checking, and some kind of standard way to represent SIG information
or prescription instructions.

And second would be a code set for drug allergies, which again is required
for interoperability between the different proprietary drug allergy screening
tools.  It is MediSpan’s understanding that such a code set is available
through SNOMED.  It is not available under the federal government’s
licensure of SNOMED-CT for terminology.

And then just to close, I would like to echo George Robinson’s comments in
that MediSpan is open to mapping to interoperability code sets, and cooperating
with organizations that are developing those, because we are very interested in
advancing the cause of e-prescribing.

Thank you very much for your attention, and the opportunity to present
today.

DR. COHN:  Mark, thank you very much.

Gerry, our last presenter of this session.

Agenda Item:  E-Prescribing:  The Drug
Knowledge Base Vendors’ Perspective – Gerry Hobson, Multum

MR. HOBSON:  My name is Gerry Hobson.  I’m the Research Manager
for Cerner Multum.  Cerner Multum is a subsidiary of the Cerner
Corporation.  I’m a pharmacist by training.  My responsibilities at
Cerner Multum encompass the management of our clinical content creation
process.  I appreciate the opportunity to testify before the NCVHS on
electronic prescribing standards today.

The Cerner Corporation is the leading supplier of clinical and management
information knowledge systems to health care organizations in the United States
and abroad.  Services include more than 2,000 clients in the US and around
the world.  These clients include:  health delivery care systems,
community hospitals, inventory clinics, physician practices, health management
organizations, blood banks, laboratories, and some health agencies.

As a subsidiary of the Cerner Corporation, Cerner Multum specializes in the
creation, maintenance, and distribution of clinical drug information.
Cerner Multum’s clients include the Cerner Corporation, other health care
software vendors, as well as several private and public organizations.
These would include:  electronic medical records systems, physician
offices, hospitals, insurance companies, PBMs, and government
organizations.  These entities license our drug and production
information.

Cerner Multum’s drug terminology is a reference terminology in the National
Council of Prescription Drug Program, NCPDP, and its map requires terminology
with UMLS Medical Thesaurus.

We even supply our drug information to electronic medical records solutions
and buyers, giving some experience and some insight into clinical decision
support opportunities.  Some of these opportunities have been previously
addressed.

For some of our clients, patient medical history already exists
today.  Patient medical history is the basis for providing
patient-specific information.  This information would be valuable in
decision support.  The level of content can be as basic as providing
common prescriptions or physicians, which improves not only the efficiency of
writing the prescription, but also the accuracy of the prescription.

It expands to provider the prescription writer, and by taking active
choices in the prescription, they can use logical choices to come to
appropriate drug therapies.  For example, for given drug, it can be
available with over many routes, many dosage forms, many frequencies may be
appropriate.

If a route is selected, and those dosages that are appropriate for that
drug are narrowed down to a specific number of dosages.  The specific
dosage is selected.  The frequency for that order narrows down to a
limited number of frequencies.  If the frequency is selected, the ratios
may be narrowed down as well.  Having this information available will
improve the work flow of a prescriber in implementing electronic order entry.

The next point to consider centers around the choice of the drug.  A
listing of the appropriate drugs to treat a condition is the starting
point.  After the selection of the drug, the appropriate dose for the
patient based on condition being treated, the patient’s renal and liver
function, their age, their weight, and a variety of other factors can be used
to come up with a recommended dose for the patient to provide safety and
quality in the ordering of the product.

These recommendations can be developed based on these parameters.
Based on clinical evidence, these parameters support this data.  In a
model where this information exists in the a patient record, use applications
can be developed, and are developed today.

Moving onto dose range checking.  Dose range checking provides an
invaluable service for patient safety.  Dose range checking can be
granulized based on patient-specific information which makes it a more valuable
service.  Particularly when you are dealing with pediatric patients, often
times information on dose range checking is not available on the package insert
for appropriate dosages.

This information available in the patient’s electronic record can be
utilized to narrow the range of the dose for pediatric patients.  For
patients that have a renal dysfunction or liver dysfunction, a dose range for
the medication can be narrowed down to a more appropriate range, thereby
eliminating and streamlining the number of alerts that are displayed to the
clinicians with the order entry.

This is a very important point.  As we move on to more advanced
clinical information, the time that a clinician gets to process this needs to
be streamlined as much as possible.

For drug interactions, there was an interesting statistic yesterday that
the CHUA reported that all but 4 percent of drug interactions were
overridden.  At Cerner Multum, this is not surprising to us.  We have
various severity levels of drug interactions, between 5-10 percent of those
fall under the major interactions category.

We can easily see how 4 percent of the interactions, all but 4, could be
overridden.  Now, that doesn’t mean that interactions at a majority
severity level aren’t important interactions in a subset of the
population.  For example, the geriatric patients, when multiple drugs are
ordered that cause sedation, that’s a very important interaction to display.

If the patient in fact receives an ACE inhibitor and has a potassium level
that is elevated, that’s a very important interaction to display.  A drug
interaction, subsetting them to those that have the most majority severity is
an important thing to do, but also coordinating them with patient-specific data
to make a much more robust system for determining and improving the patient
safety.

At Cerner Multum, the terminologies that we use for the NDC codes mentioned
before.  The NDC code has several limitations.  Number one, they are
not readily available from any one source.  The initiative that the FDA is
seeking to provide these NDC codes will go a long way to streamlining our
processes in gathering these NDC codes.

The structured product labeling will help dramatically in listing inactive
ingredients for products, and active ingredients are valuable in certain alerts
such as allergy checking, where an inactive ingredient may be available in one
NDC code, but not in another.  Having that information come directly from
the FDA would be invaluable.

Other areas for structured product labeling that would help range from FDA
approved indications, black box warnings, side effects, drug interactions, and
as I mentioned inactivity ingredients.

Some areas for consideration of the FDA in creating the structured product
labeling is inconsistencies that the manufacturers provide and their product
labeling.  One area that comes to mind is drug interactions.
Psychoactive drugs may be listed as a drug interaction, but what drugs are
constituted in psychoactive drugs.

Another area would 2t-interval polymers(?) which received a general
statement which might include a couple of examples of what are those drugs that
are 2t-interval polymers.  The last example might be, which is prominent
in drug interaction is the actual liver enzymes that are used to metabolize
drugs.  If a drug is considered a cytochrome p450 inhibitor, which drugs
are included in that list?  So, if the FDA goes on to create structured
product labeling, we would like them to clear up those implications.

In conclusion, I would like to support use of RxNorm in connecting
different databases’ content with the drug products and the common drug
terminologies.  We support the NCPDP’s Script standards and SNOMED
terminology for conditions.

Thank you very much.

DR. COHN:  Well, thank you all for some very interesting testimony.

Steve, I think you had your hand up.

DR. STEINDEL:  Yes, before we get into any more substantive questions,
I just have a clarification question for Mr. Millikan.  You mentioned that
the SNOMED allergen tables were not available under the federal
licensing.  And yet, I’m looking at SNOMED right now, and under the
disease axis, there are drug-related disorders, adverse reactions to drugs,
allergic drug reactions, and then an extensive table under that.  And
that’s available under the license.  So, I would like some more
information about where you got that.

DR. MILLIKAN:  I spoke to a representative of SNOMED last week about
the allergy information, and clarified that this is not available nationally.

MR. FONOW(?):  There are three parts to it.  There is the —

DR. COHN:  Who are you first?

MR. FONOW:  I’m John Fonow, speaking about SNOMED.  There are
actually two parts to it.  The medications of course could be part of the
thing that you are allergic to.  My understanding is, is it not part of
the license.  The allergen subset, which is defining like substance
identifiers, that subset — so, that’s what tells you which of all the SNOMED
classes are considered part of the allergen subset.  That is not covered
by the license, although it’s very cheap.

So, the concepts themselves are covered by the license.  So, you get
SNOMED-CT, you download in the UMLS, and you get it from SNOMED, and you would
get all the concepts.  But what you wouldn’t get is the definition of
which particular portion of all of those different things are considered to be
part of the allergen subset for the use in a physician support system.

MR. ROBINSON:  Can I jump in here?  I’ve been looking at SNOMED
as well.  What is available, and there are allergen descriptions all over
the place in the content and different domains.  What we found is in the
best work with is in the substance domain.  There is a relationship of
allergen class in the substance domain that you do get with the core set.
So, that’s where you may see betalactins, penicillin, cephalosporin.  That
is there.  What you have to pay an additional fee to get is a default tick
list that includes ingredients, classes, many different things pulled together.

Now, in the model that we have been doing, the most effective
interoperability standard would be just simply at the  class level
itself.  What you are trying to do is if you express the class, that’s
kind of an abstract concept, penicillins for example.  If you are defining
an allergen using a particular ingredient, probably the best path for that
would to be use the ingredient that is related to the proprietary system that
you are using, and then ultimately tap the Unicode from the FDA.

Similarly, if a clinical drug is in the context of an allergen, again, the
best thing to pass would be an RxNorm concept for that.  So, when I looked
at SNOMED, I believe there were about 400 concepts right now that are in the
substance hierarchy.

What the caution you live is kind of stay away from the disorder domain,
because first of all the relationship of the ingredients in a substance
hierarchy is not consistent when you start looking at causative agents for
example, which would be in the disorder domain, versus I think what you see in
the substance hierarchy if they have ingredients in relationship.

That’s probably more than you want to know, but the point —

DR. STEINDEL:  Thank you.  As you point out, those subsets and
specific for the US drug domain were excluded in the license.

DR. COHN:  Other questions?  Jeff?

MR. BLAIR:  Could you help me understand what work has been done, if
any, on the possibility of standardizing the messages that go from each of
these drug knowledge bases for DUR?  What is the status of those?

MR. BIZZARO:  Let me speak a little bit, because I’m going to talk
about it from the point of view of NCPDP and the way messaging happens
there.  You make the telecommunications standards, and I believe this is
true in Script also, there are qualifiers that identify what the source of the
information is.

And so, those qualifiers identify — First DataBank has a source of the
drug interaction information.  We then have values of let’s say severe,
moderate, minor, and I can’t remember if those are the three actual values, but
there are three different values that give you three different levels of
interactions.  But those interaction levels are specific to that
proprietary organization giving you the alert.

As far as I know, Jeff, the only —

MR. BLAIR:  The interaction with the proprietary message is a
standardized NCPDP message?

MR. BIZZARO:  That’s correct.

MR. BLAIR:  Thank you.

MS. EBHART:  Can I add onto that?  Karen Ebhart.

I believe also for the message — I’m looking at Lynn as I say this —
there is a codified field telling you what type of DUR was found.  What’s
the drug interaction.  Was it too high of a dose, too low of a dose?
As well as a pretest field that currently allows you to identify in the case of
a drug interaction, what drug it was interacting with on a profile.  You
know the drug coming in is one half of the interaction.  You need to have
somewhere be identified what drug it is interacting with.  And so, the
structure allows you to identify that portion of all this.

MR. BIZZARO:  This is part of where my concern is as we looking at
messaging, and what we’re going to do in e-prescribing, and what type of
electronic messaging is going to go back and forth between the physician and
the pharmacy.  It can become very complicated and time consuming as you
send that messaging back and forth.  And how do we manage that and still
meet the requirements of a timely message for presenting an electronic
prescription to the pharmacy?

I’m going to give a personal opinion here now.  I think that the FDA’s
initiative regarding the linear bar code on packaging got through the
rulemaking office on a very short period of time.  I think that happened,
because it was really a simple description of what they wanted.

I think the benefit from having that linear bar code on packaging for drugs
is going to provide both economic and patient safety value in the future.
And as it changes and becomes more sophisticated, that will create — but its
simplicity is its greatest value.  And I hope that we do the same thing
with the e-prescribing.  We’ve got to get people doing
e-prescribing.  We’ve got to get them doing.

Being a classical pharmacist for 15 years, I can’t tell you how many phone
calls I made to a physician simply because I could not read the prescription,
or I couldn’t read the strength, or he forgot to put the refills on.  If
we look at what we gain by just doing something that simple.

And then let’s add some of these things we’ve talked about, some of these
added-value we can do here.  Let’s do that also.  But let’s keep it
simple enough that we get physicians writing scripts electronically.

DR. HUFF:  I’m going to ask the question I thought Jeff would
ask.  Were there any standards for distributing the knowledge or the
relationships, or however you want to describe it, of a drug interaction
database or an indications database?  So, the idea being that at some
point it might reduce the work everywhere of having one group or one consortium
do that, and then you guys can distribute those value-added parts.  But
there would be a standard way of representing.

And I wonder if any work has been done in that area, or people have talked
about making standards for distribution of the knowledge itself?

MR. ROBINSON:  The only standards discussion that has taken place is
—

MR. BIZZARO:  AMCP in association with a number of other organizations
a few years ago tried to come up with a methodology to come up with a common
list of drug interactions that we could all, on a national level, agree upon as
to their severity and what those drug interactions are.

My understanding on that — this is kind of an untimely question, because I
have a meeting with them Friday when I get back to the office, to talk about
what they decided to do going forward.  My understanding now is that what
they have decided to do going forward is to see if they can come up with a
methodology that the different compendium would agree to as they develop their
own proprietary drug interaction databases, but use that common methodology,
and see if that can kind of get partway to where you want to go, but not all
the way.  And I’m not sure where they have gone.

The issue with not doing that on a national basis was the timeliness of the
information, and having a national group coming together every six
months.  And we all know that in the drug world, our business depends on
the timeliness of the data on a daily and weekly basis.  And trying to do
that I guess using a non-proprietary national standards body where a blue
ribbon, it was just too difficult to do.

MR. ROBINSON:  I think we need to get on the table what we would be
willing to work with, and what we would consider an infringement on our right
to being a viable business.  And I think we are open to common messaging
standards.  I think we would be open to if a third party wants to say here
is the standards criteria you should use.  I’m saying something is a
severe drug interaction, or something falls into a particular category.
So, perhaps that actionable knowledge can represented in a common way.

I think each of us want to reserve the right independently to research and
codify drug knowledge and medical knowledge as an intellectual property asset,
and distribute it in a competitive environment.  So, we kind of again,
broke into the standards and the communication, but we really need to protect
our business for intellectual property.

MR. HOBSON:  I would agree that standards would be helpful in this
area.  I agree that intellectual property is helpful.  It’s important
as well.  As I mentioned before, interactions vary based on the patient
taking the two drugs.  And use of that information, though it may not be a
severe interaction, it may be very important.

MR. ROBINSON:  You mentioned don’t go too far.  So, I’m going to
take the opportunity and say from an alert standpoint, what does far enough
mean?

MR. BIZZARO:  For me, I’m very pragmatic and practical.  As I see
— and some of the things again, are going to be personal opinions — but I
don’t know how you make that decision, Gerry, I really don’t, because different
practitioners have different levels of experience and knowledge.

For example, if you are in a hospital that has an oncology unit, your
specialty is those oncology patients, the doses that you are going to use for
certain products is going to be way out of the range of what you would normally
see in the retail pharmacy.

That’s probably the most difficult question is how do we take care of this
message overload of our physicians and pharmacists are just hammered with, but
still protect the patient, so that we don’t give them too little
information?  I wish I had an answer for that.  I don’t know,
although we never want to — in our business, we don’t ever see ourselves as
the experts in the practice of health care.  Our customers are those
experts.  They assume the liability and the responsibility for the care of
the patient.  We see ourselves as providing the information that helps
them do their job.

But I think that’s the million dollar question, who do we cut down on the
useless information that physicians and pharmacists see day after day, without
getting into a liability issue?

MR. ROBINSON:  Do you see a demarcation between e-scripting and
e-prescribing?  In other words, e-scripting to me came across as the
process of electronically submitting prescriptions, electronically passing the
information around.  Whereas, e-prescribing to me goes one step further,
which is helping understand the medical side of the situation.

MR. BIZZARO:  I actually see it as four different, distinct things
that you do as a physician.  Not being a physician, I may be speaking a
little from just my impression.  But when we are talking about
e-prescribing, I see four different things that can happen that benefit the
physician.

One is eligibility of that patient for a particular health plan.
Second is access to preliminary(?) for that patient.  Knowing what drugs
are covered for that patient, and what it’s going to cost them in
general.  I don’t think we get the exact price for that patient, but we
tell them about what it’s going to cost if we know which plan and group
information.

The next thing is DUR.  And that level of DUR happens in physicians’
offices up to that vendor who is supplying that information to that physician,
and what the physician wants.  And I think the fourth thing is what you
are talking about as e-scripting, I think that’s the simple part of it.

Now, what we want the physician to do before they get to the point of
sending that old time prescription to a pharmacy.  The infrastructure was
not available for years.  And that was the biggest problem I saw with it,
and that was because RxHub and those organizations didn’t not exist.

DR. COHN:  Jeff?

MR. BLAIR:  It’s my understanding that the Anilin(?) with the UMLS
will be mapping RxNorm and the drug classifications to each of your drug
knowledge bases.  But there are two things that I wanted a little bit of
clarification on if I may, because one of you — I don’t remember which one —
I think said it as if it was going the opposite way around.  That you were
mapping to RxNorm, and so I wanted to understand that a little better, if I
heard that correctly.

But the other piece is a second question besides that, is could you give me
some feeling for the time frames when you feel like you will be able to use
those mappings?  And what’s behind the second question is if we are to be
ready for the demonstration projects beginning in 2006, January 2006 for
e-prescribing, does it look like we are on track to have those mappings in
place for the demonstration projects?  So, those are the two questions.

MR. ROBINSON:  This is George Robinson.  First of all, I think we
need a representative from the National Library of Medicine to address what
progress has been made as far as RxNorm, as far as when it’s going to be ready
for prime time.

My understanding of its development is it’s focused first on the generic
and semantic clinical drug level, and that there may be some relationships to
brands I think as a synonym.  And then coming down the road would be the
branded clinical drugs.

The importance of that, if you are trying to write a prescription as a
dispense as written, and you’re stuck with a particular brand, that function
has got to be there.  So, from a cross-mapping would be one approach if
you are linking to the generic clinical drug, and then you are also having to
link to the synonym brand name versus an entire concept.

As far as the process of mapping right now, this is the part of the roll up
our sleeves meeting that I’m saying that we need.  Initially, my
understanding is the National Library of Medicine — speaking for First
DataBank, and making our generic formulation identifiers, and we’ll be
assimilating them so to speak.  And what we would get back from them would
be the clinical drug with a relationship back to a First DataBank concept.

That being said, we still, I think, will need to work with it, and offer
for example mapping back to some of our other concepts.  I don’t have a
good handle yet on the ingredients side, because I haven’t seen the Unicodes
yet.  So, I’m not quite sure what the time table is for that.

But, Jeff, you are exactly right, you work backwards on a timeline
according to project management methodology, to the present.  There are
several tendencies before this could work.  Now, we have had several
people say repeatedly they found a way to make e-prescribing working in the
absence of standards.  But if the goal is to create a model program that
uses all standard terminology, we have a short timeline all the way across,
especially when you’re not turning up recommendations to the secretary until
September 1, 2005.  This is what I heard.

MR. BLAIR:  Well, the interim will be in November of this year.

MR. ROBINSON:  I guess there will have to be scope setting on what the
initial pilot program will actually be achieving.

MR. BLAIR:  Right.  Other perceptions on these two questions?

DR. MILLIKAN:  Yes, this is Mark from MediSpan.  First of all, we
are interested in doing the mapping to RxNorm.  The mapping is available
currently through the UMLS to one of our drug concepts.  And some of our
customers have already begun to ask if we have had an opportunity to validate
that.  So, we are going to want to be able to do it ourselves.

Initially, this seemed to be a fairly simple situation, because it was
mapping to the clinical drug.  Now, RxNorm is becoming a more robust drug
terminology code set.  It can link into — or I should say map to various
drug concepts that MediSpan has.  So, it’s not just a matter of linking to
one particular drug concept.

In terms of our timeline in terms of how we are looking at RxNorm, we are
in the evaluation stage for 2004, and trying to start work on it in 2005.
But again, I would agree with the comments of George, that we are not quite
sure yet what the scope of all of this is.  And there are also other
things to be considered besides RxNorm.

MR. HOBSON:  This is Gerry from Cerner Multum.

I believe that our identifiers have been mapped by RxNorm for the generic
name, as well as the products.  And we are in the same boat as the other
providers in terms of how we use that linkature of content.  So, I do
believe there is very close mapping work on content concepts through the
product level.  And we’ll be looking at the branded levels, and what level
of that information that applies to the drug content.

MR. BIZZARO:  One of the vendors this morning talked about not having
just drug products in that database, but also those things that are commonly
prescribed medical devices.  He even mentioned DME.  And I think that
presents another level of complexity, because syringes are very often
prescribed by physicians.  We don’t have syringes in RxNorm.  Is
there going to be a different methodology to write a prescription for a syringe
than there is for a drug?

So, that was a good point that I hadn’t thought of before.  In our
database, we are not a DME database.  We are not medical devices.  We
are a drug database.  But in it there are thousands of non-drug items that
are commonly prescribed and commonly paid for by third party payers.

DR. COHN:  Well, Tom, just to remind you, those aren’t in NDC either.

MR. BIZZARO:  And that’s another conversation.

DR. COHN:  Steve?

DR. STEINDEL:  Yes, Simon, this is not Steve.  This is our
distinguished colleague, Mike Fitzmaurice, who I’m afraid couldn’t make it to
this meeting.  He had to attend a meeting in Jackson Hole, Wyoming, and I
couldn’t convince him to trade places with me.

But I would just like to point out in somewhat response to Jeff’s question,
and Randy should chime in a little bit, because he is on the receiving end,
AHRQ this year provided a bit over I think $6 million through the National
Library of Medicine and the FDA for support of developing the drug
terminologies that were recommended by CHI and NCVHS.  So, we expect to
see an acceleration of that project in the next year.  So, AHRQ took the
recommendation of NCVHS and is moving forward with it.

DR. COHN:  Thank you.  And I think we will be hearing from them
in September, if not before, with more information on how they are doing.

MS. AULD:  I think we are scheduled for September to testify.
And I’m going to go back and talk with Dusty and see if we can’t get something
in writing that gives kind of status report of where we are, what is the
schedule for this year.

DR. COHN:  Yes, I think based on this conversation, that would be
helpful.

MR. BLAIR:  Could I ask for one additional thing?  Maybe in
preparing the comments on the status of how far along RxNorm is, and the
representation of actions and physiological effect of drugs from NDFRT for
mapping, maybe if a conversation could also take place with the drug knowledge
base vendors, where there is some feeling for how much time they need after
RxNorm and the other actors from NDFRT are ready.  Is that going to be
enough time for them to have all those mappings in place by January 2006?

DR. COHN:  Other questions or comments from the panel?

DR. STEINDEL:  Simon, I don’t have to ask my question.  They did
a very good of addressing it in their presentations.

DR. COHN:  Okay.  Just following up on a comment that Mike
Millikan made, where you were talking about the way drugs or dosages are
handled within the transactions.  You were describing Script.  I’m
just curious, you were the only one who brought up this issue about how the
information is bundled together appears to be an optional feature of Script,
which I presume has some implications for other knowledge base vendors when you
are trying to alert some reminders.

Does everyone have that issue?  Is that just a MediSpan issue, or is
that others?

MR. BIZZARO:  Actually, it’s more an issue for those vendors who are
using our content within their applications, and then sending information into
Script.  Mark mentioned about the drug name.  We talk about it not
being long enough with the 35 characters currently, and having repeatable
possibly to take care of that.

Let’s look at it from the pharmacy’s point of view.  Well, the
pharmacy doesn’t care if you send them 35 characters, because most of their
systems can only effectively use a maximum of 30 or 28 characters.  But
even though we may be building a system on the standards side for Script,
provide more information to the pharmacy, and not to be able to automatically
remove that directly into their dispensing system, because their dispensing
system is going to have limits on the application itself.

I mention it only to give a sense of the complexity of what we are talking
about here in all the different applications that are affected, and what they
may already have in place as a standard for field lengths and repeatable fields
and messaging that we are talking about.

I must have said something wrong.

DR. COHN:  Lynn.

MS. GILBERTSON:  Lynn Gilbertson from NCPDP.

The standards development process really works.  The things you asked
for are in the most recent, almost approved standards, so they are already
there.

MS. EBHART:  Aren’t they conditional, and it’s still not safe that
everybody’s program can hit the correct version, and a lot of people are on an
earlier version.

MS. GILBERTSON:  Right, that is correct.  It’s one of the newest
ones available, but in response to the industry, there is room for label names,
as well as the longer names.  And previous versions had the support of all
the DUR codes, as you mentioned earlier.

MS. EBHART:  I think that was part of a recommendation that as a
version of Script was identified that we want people to map to such as it was
in HIPAA, there were versions and standards, as the PDP Version 3.1 or 5 or
beyond.  It’s our recommendation that the version of Script that we
recommend being implemented would be this version or one later that has these
additional identifiers, and not use an earlier format where that’s the only
field identifier.

DR. COHN:  Yes.

DR. HUFF:  The fundamental issue is, or the fundamental complexity
maybe is that when you order a drug, you need to allow lots of different levels
of things.  And so, you can put a pre-coordinated code into that drug code
that is like a clinical drug, or you can use a branded clinical drug if you
want to specify that it be dispensed as ordered if you are trying to fix the
brand names in the prescription.

Or you can do it in pieces.  So, instead of saying penicillin 500
milligram tablets, you can say penicillin, and down in additional fields you
can say 500 milligrams oral tablets.  And so, you want that flexibility in
the ability of how people might prescribe things.  But at the end, you
want all of those to be computable.

And in some cases, what that probably means is you put in the drug, but
then you redundantly put in the pieces that are parsed out, so that you don’t
have to go to a knowledge base to find out that this — ultimately, there is a
knowledge base behind there.

DR. COHN:  Yes, and my issue was the computable part.  If you put
it in as one long stream, including the dose, the question was how you deal
with that.  Of course NDC has a solution, but beyond that, how do you deal
with that issue if you are looking for dosage issues and all of that.

DR. HUFF:  It always gets back to what do you put in what slot.
If you put that code there, what does it mean for the other slots?

DR. COHN:  Thank you.

MR. BIZZARO:  Stan has a good point.  I think we look at the
NDC.  The NDC in my mind, was meant to be for the package modeling.
And you had this discussion in the past in this organization at the full
committee.  But we do need to make sure that as we kind of look forward,
we are doing this.

And I think we are all doing this with the idea that we do want to have
those tiny little boxes that are very distinct, and they mean exactly what they
mean, and they don’t have multiple uses.  And you try to design your
system so they are computable with their standards.  But that’s looking to
the future, and we’ve got all these issues that we need to deal with.

DR. COHN:  Any final questions or comments?  Now, we are running
about 15 minutes early.  Why don’t we take a 15 minute break, and when we
get back we’ll have open microphone, and then we’ll move on to our HIPAA
topic?  We’ll be back at 3:00 pm.

[Brief recess.]

Agenda Item:  Open Microphone

DR. COHN:  I guess I should ask, does anybody want to testify or give
some comments about the e-prescribing testimony, at least until now?
Well, okay.  I think we’re going to get more information between now and
tomorrow morning.  We’ll be having more e-prescribing discussions.
We will have another open microphone tomorrow.

With that, what we are going to do this afternoon with the remainder of our
time is to follow-up on some HIPAA implementation issues, and then we were
going to talk about an NCPDP, DSMO issue that we had talked about at the last
session.

So, Jim, would you like to introduce yourselves?

Agenda Item:  Follow-up on HIPAA Implementation
Issues – Ed Jones, Jim Daley, Jim Schuping, and Steve Lazarus, WEDI

MR. SCHUPING:  Jim Schuping, Executive Vice President for WEDI.

MR. JONES:  Ed Jones, I’m Executive Vice President of Vizicure(?), and
the chair of WEDI.

MR. DALEY:  Jim Daley, vice chair of WEDI Planning and
Operations.  I spoke here at the WEDI HIPAA hearings we held earlier this
year on campus.

DR. COHN:  Please lead on.  We actually have a copy of a document
from you it looks like.

MR. DALEY:  This is the letter.  When we here before, as you
recall, you asked at that time to take the primary recommendations and try to
make an attempt at prioritizing some of the ones that we thought were most
important.  So, what you have in front of you here is a draft
document.  We are presenting a summary of that.

I think Ed is going to lead off with some background for the folks here,
and then we’ll get into it.

DR. COHN:  Okay, fine.  I wasn’t sure if we should use this as a
guide, or just to start the conversation.

Thank you.

MR. JONES:  We’re pleased to be back here before the subcommittee
today to follow-up on our earlier meetings.  Just as a little background
for everybody, WEDI held a one day hearing in Tampa in late January to evaluate
the challenges and successes of the HIPAA administrative simplification
implementation of the transaction and code set standards.

Following that hearing — we had testimony from I believe 56 persons and
corporations, both in oral and in written format — we sent a letter to the
secretary, and brought it here in late March, indicating a variety of areas
that mainly were challenges and some successes with regard to the
implementation of the transaction code sets.

We were asked at I believe it was the March 29th meeting to prioritize the
recommendations that we were bringing forth, and to return today to go over
those, which Jim Daley and I will.

We have broken these down into three areas:  things that we believe
need immediate attention in one way or another; ongoing issues; and then some
strategic longer term issues.  There are four issues that I would
highlight and turn it over to Jim, that we believe are current implementation
issues.  And these are in some sense of order of priority.

The first is to continue the focus on full compliance while maintaining the
contingency plans, but with some set of timelines to move away from that.
We also believe, especially based on the WEDI meeting we had in LaJolla last
week and talking to people, that progress is being made.

One of the areas that does need attention here is there is a need for a
definitive, authoritative source for frequently asked questions.  And if
that is CMS, we need to get the word out more on that.

The second priority — this came up quite clear in the meeting in Tampa —
is that everyone in terms of focusing on implementing the transactions, is
really focusing on implementing the claims transactions, and not any
others.  So, we think that a renewed effort is needed to identify and
promote the net benefits of these other transactions, and to have movement on
those.  And this relates to contingency plans, because if those aren’t
being addressed, obviously the contingency period is going to be somewhat
longer.

With regards to this area as well, and this was a great concern last week
in discussion at the WEDI meeting is we need to really bring some
standardization to the data content, particularly with what are the situational
variables.  And there is a lot of variability with it.

The third is to focus on — and this really did not come up in the
implementation specifications — we really need a requirement for the use of a
standard acknowledge, such as the 997, and we need some more definition on area
reporting, because a number of transactions are being kicked out as part of
this transitional phase of implementation.

And then the fourth area is to conduct a comprehensive review of
coordination of benefits issues.  And probably the most immediate concern
here is getting the health plan identifier out, because without that particular
identifier, you really cannot move efficiently in electronic formats from plan
to plan.

I’m going to turn it over to Jim to give you a little more context on
these, and on the ongoing strategic issues.

MR. DALEY:  Thanks, Ed.

Ed has pretty much described the highlights of this first item.
Again, the rest of the document, the final four or five pages have a little bit
more context for those particular items.  I would rather just go over
these first couple of summary pages, and then open it up for questions.

In regards to COB, there were a couple more recommendations that we did
have, including clear instructions to Medicare contractors on how to populate
crossover items in a standard manner.  That’s been a real problem in
industry, getting information from the various contractors through those
payers, and having those items accepted.

And another major COB issue is associated with serving the various claim
form types.  If a claim comes in, in a professional format, crosses over
to another payer who requires that in an institutional format, it does cause
some problems for the provider.  Now, many other things in there in terms
of code value, whether they do COB at line levels, claim levels.  There
are a whole bunch of items in there that are causing problem when you do try to
cross from one payer to another.

In terms of ongoing implementation issues, I do want to make one other
statement before I go into this.  Even though we have separated these into
various categories, some of the strategic issues do apply to some of the
current issues as well.  You can’t do one without the other.

For example, looking at getting the right input, as you might want to make
a change to the standards.  Making sure that there are no unintended
consequences at either a pilot, or some measure such as that.  You really
do have to look at all of these together.

Developing the standardized test data is an ongoing issue.  It would
be great for the industry to have a way to say if you can accept this
information into your system, we can send it out with these standards.
Everybody would be able to use it.  We have it certain to be that way with
some of these testing facilities out there.

But even that is not going sufficient to really guarantee acceptance from
one entity to another.  It would be good to develop some kind of a test
bed, set of test data that would accomplish that.  I guess I could liken
that to your light bulb.  You plug it into your standard socket, and know
it’s going to work.

Expediting change standards.  Our recommendation is that CMS, WEDI,
and the DSMO stress the standards development process to determine whether an
expedient change process should be developed.  There is a need at times to
fix what’s wrong with the standards.  But again, we shouldn’t disregard
the regulatory process as we are doing so.

We do need the industry input, and concurrence that it’s the right thing to
do, because even minor changes can result in significant efforts.  Even a
new value in a field means maybe some different edits, and changes the value of
the entry screen, possible business changes.  There are a lot of
implications of even minor changes that need to be thought out before we just
make a quick fix.

We need to address in a timely manner, some of the data content issues that
exist.  One of the big things to come up recently is the fact that the 835
remittance does not have enough information in those code sets to really tell
providers what went wrong.

Some of this complicated by what Ed just mentioned in terms of the
acknowledgement and the error reporting.  Some may get accepted, some may
get rejected.  Some lines may be accepted.  Some may not be
paid.  And in the long run, remittance by the system with a limited set of
codes may determine how much money you are getting.

Tracking what really happens during that whole set of processes has been
very difficult, because a lot of the problems in getting everyone transitioned
to the new HIPAA standards.

Code set maintenance would be restructured to a new schedule of only once a
year where appropriate.  We look to the implementation guides for some
internal code sets, which are maintained by the DSMO.  They issue the
guide, and you’ve got to issue the guide to change those.  That wouldn’t
be a problem.

But there are many external code sets, for example CPT codes, zip codes,
you name it, there are many in those guides, each with potentially different
schedules.  And since they are incorporated as part of the HIPAA standard,
conceivably we can change the HIPAA standards every month.  And if we
could bring some order to that process, that would be helpful to the industry.

Code values on instructions should be specified.  One of the examples
I’ve heard is that yes, we are specifying too much with the CPT code for your
procedure, for the professional services.  But there are apparently
specific instructions as to when and where you use those procedure codes.

And I recall for anesthesia you can either code that as a surgical charge,
or you can use a separate CPT code to say this is the anesthesia, so there were
some questions about that.  And one of the suggestions was that there is a
standard way to say this is the way you do those codes, and that’s how it
should be followed.  And we’re doing things somewhat differently.

Then there were strategic implementation issues.  The first three
listed down here can be somewhat grouped together — in way in which the
providers are included in standards development; balanced representation on the
DSMO; and encourage decision-making and participation in standards development.

In other words, we really need a good, balanced industry representation to
come up with standards that are workable for everybody.  We don’t just
need testimony from people.  We don’t just need those who can afford to
attend the meetings for both the dollars involved and the time involved.
We don’t just need people such as the vendors or payers or the associations.

We need the people right out there, the real doctors for example, to say
this is how it affects me.  How can we get all of those people in here so
we really get a standard that is workable for everybody involved?

We just pilot standards prior to issuing the final rule.  In other
words, a standard is out there on the table.  Let’s see if it really works
before we tell the entire country to adopt that standard.  Through that
pilot, we will be able to see if there are changes needed before the standard
is adopted, and we can validate the costs and benefits of that proposed change.

One final recommendation, that we really do look at enhancing the
implementation process.  Specifically, we need to look at the realistic
time frame for all aspects needed to complete the implementation.  For
example, if a standard must be implemented, the vendor products need to be
designed and developed and rolled out and published in reported sites.
Those can be provider sites, clearinghouse sites, payer sites, whoever needs
the crossover in order to perform their function.

That takes some time.  Once that is done, everybody believes they are
compliant, we still need to migrate all the different trade department
connections that are out there.  There could be thousands of connections
for a payer, a clearinghouse.  Even providers have dozens or hundreds of
connections they have to deal with, and that takes time as well.  All that
needs to be rolled into the process, so that when we say make a change by a
certain point in time, we have a real expectation that we’ll meet that
particular role.

Accept high level economy(?) of what the recommendations are.

And I would able to give Jim one last chance to add a couple of bullets, if
they have any, otherwise we’ll answer questions.

MR. JONES:  I thought it was interesting listening to the discussion
this morning and earlier this afternoon, and in particular I think it was
Harry’s question on the difference between e-scripting and e-prescribing.
And it seems very similar in context to what we have been dealing, the
transaction, the code sets from a format standpoint if you look at e-scripting,
and a content standard if you look at e-prescribing.

And I think we have learned a lot of valuable lessons that may be
applicable to this.  And I remember on May 6 when Sec. Thompson had his
summit at the Willard, he said he wanted to expedite the time frame for dealing
with the e-prescribing.  That we have been working on these transactions
for eight years, which is I think kind of like the time frame we are dealing
with in the MMA.

But we really need to focus on what is the standard, and getting that
common content together for the format to make it work.

DR. COHN:  Well, just a reminder.  As you remember, HIPAA was
supposed to be implemented by 1999 completely.  Sometimes complete
implementation takes a little longer than initially anticipated.

I do want to start out with rather than discussion, since we didn’t have
any microphone for e-prescribing, but I did just want to make a couple of
comments.  Obviously, I have not recused myself from any of the
conversations so far today.

At the last meeting, I, at that point, recused myself from being the chair,
only because at that point I was at least nominally on the board.  I do
want to disclose to the committee that while I was elected last year, because
of being unable to ever get a satisfactorily read from HHS about the impact
that that might have on my station and involvement with any recommendations, I
really never served.  And I have since vacated that seat.

My understanding from Marjorie Greenberg, our executive secretary, is that
since I didn’t serve, I really am able now to take testimony in all of
this.  So, if you any of you have any questions about that, I’m happy to
chair the hearing for the moment.

The other thing is obviously with the code guidelines and instructions,
which as you know, I sit on the CPT editorial panel, and a lot of those
actually relate to CPT, although I do have to say that I consistently see this
one, so I should understand what this is, but my memory of the reason why there
were no guidelines specified for CPT at the time that came out as a final rule
is that there were separate guidelines for CPT.  And to my understanding,
there still aren’t.

So, I guess at some point I will give the chair to someone else so, maybe
we can the AMA come and explain to us what this is all about.

Anyway, having said all that, I guess to me the issue that I’m most
concerned about as I listen to all of this, and I know Karen may talk a little
bit about contingency plans hopefully when she returns, is really this issue of
moving to full implementation of all the HIPAA regulations.

I think I actually spoke at the meeting last week, and at that meeting
actually asked — in two weeks, full implementation in two weeks — but I
actually asked people to raise their hands about how many were working on
things beyond the claims transaction.  And I would say it was not a
majority in the room that were raising their hands at that point.

Clearly, it was a situation where people are striving to have compliance of
claims transactions.  And I really think being somewhat forward looking,
as well as responsible at the NCVHS, the subcommittee for monitoring how we are
doing with administration simplification, really the focus we all need to have
to now is both how to monitor, as well as promote the implementation of all the
regulations.

And I guess as I look at that one, that’s really the area I think we really
need to be focused on.  And I am curious about what thoughts WEDI had, in
terms of our subcommittee thoughts.  I do know as I’m listening to all of
you, it seems to me that maybe one of the things we need to do as a
subcommittee, given that we are responsible for monitoring this for Congress is
to launch some sort of an evaluation, or I would say not really a model(?) of
the industry, but getting a very clear view of where we are in terms of
implementation, so we can report on that.

But I’m open for ideas, because that really, of all the things I have seen
here, that’s the one really galvanizes me.

Steve?

DR. STEINDEL:  Yes, Simon, this question came up at the NUCC
discussion a couple of weeks ago as well.  And as people are aware, CMS
publishes I think approximately a weekly list of the compliance with primarily
the 837, and how it is improving on a monthly basis.

But it also has several other transactions, one of which is coordination of
benefits.  And the NUCC meeting, we pointed that it would be a good idea
for them to start doing the same type of effort that they are doing to put the
837 into widespread use by tracking and trying to improve the figures on
coordination of benefits.

In part because as we understand it, some of the other claims also involve
coordination of benefits, and will probably go into effect at the same
time.  So, that was our suggestion to try to improve it, to have CMS start
tracking more vigorously.

MR. SCHUPING:  Simon, I think you will probably remember there was one
presentation early in the program last week where the comment was need made
that some of the feedback they were getting was that some of plans already were
hitting like 90 percent or better on a lot of the compliance conditions.
And that they were thinking of discontinuing their contingency plans.

But my guess is that the eyes are going to be on CMS, since they are the
big player in the game.  And when they put the stake in the sand, I think
that’s what is going to create the momentum for the full implementation.
But I think we are going to see some of the plans begin to discontinue their
contingency plans if they are hitting those kind of percentages.  I don’t
know when that will be.

DR. COHN:  Yes.  And that obviously needs to be the 837.
Did they say anything about all the other transactions?

MR. SCHUPING:  They were just referring to that.

DR. COHN:  Okay.

MR. SCHUPING:  That’s the dilemma that we’re in right now.  We
need to get moving on these other transactions.

MR. DALEY:  I think from a perception standpoint, one of the positive
developments that goes back to the dialogue we just had on how long this has
taken, since August 1996.  And my feeling on that is that the
Administrative Procedures Act is really not a useful guide when you are using
technology, particularly when technology is moving very rapidly.

The market is outpacing the ability to promulgate regulations.  So, I
think one of the important things beginning last year, particularly on the
clinical side was the federal government taking the initiative through a
variety of vehicles, the EHR, the SNOMED-CT, the CHI initiatives, et cetera, to
use its market power if you will in the health care field to show an impetus, a
need to move forward.

And I think one of the interesting things, going back to that May 6 meeting
with 15 CHI initiatives were adopted, one them happened to be the HIPAA
standards, which puts that in the context of the overall health care
environment.  And I think particularly on the provider side, where they
kind of look at it, and see these electronic transactions, but they really
don’t have any context on the administrative with what they are doing.

All of the sudden they think you have put the whole thing together in a
whole.  And I think if you look at the provider initiatives that have been
taken, and the data that flow through those clinical processes in a practice or
in a hospital, those data that really populate the administrative transactions,
and I think over time that’s where we are going to see the benefits.

But I think we really need, from an educational and outreach standpoint, to
really hammer that message home to push these HIPAA transactions.
Otherwise, I think other than the 837, these things are just going to kind of
move at a relatively slow pace.

MS. FRIEDMAN:  I’m interested in the second goal under full compliance
on page 2 about the FAQs.  Because CMS and OCR both have a number of FAQs
on their Websites.  And I know ours are updated frequently in response to
inquiries we get, both ones that come on e-mails or “Ask HIPAA,” and
also from the hotline.  OCR does as well.  I’m not so sure what you
are referring to, and how would this work?

MR. SCHUPING:  I think what we need to do is designate one particular
source.  There are a number of sources that people look to for guidance
for this area.  And not everyone is looking to CMS or the FAQs that CMS
has out there.  It does not over the issue, and you are dealing with a
vendor, and you want to get an answer in a couple of days, this type of thing.

So, we have just found in the testimony that was given in January, and in
the discussions last week, people want more information.  So, we’re
positing this as an idea that would have some immediate benefit.  And it
probably is a greater outreach effort.

MS. FRIEDMAN:  Well, one of the challenges, having managed the FAQs is
that a lot of people want us to solve their business problems for them.
And there is just no way that we can do it as the government.  But more
importantly, I think people want a very down and dirty call on every business
decision they make, and I don’t see that happening.

MR. SCHUPING:  Well, I can give you one example, then Jim can
comment.  This came up again last week.  It’s been coming up for
quite a while, is what is the status of a bank as a covered entity?  Is it
a covered entity?  What are its responsibilities?  And this has been
hanging out there, and has been an issue on the payment side.

That’s just one example, but that’s a fairly significant example, because
all of the sudden what I like to call the medical banking sector is really
heating up in terms of wanting to be innovative in terms of transactions with
the 835 in place, et cetera.

MR. JONES:  One of the things I’ve heard is that the FAQs are a great
reference source.  However, it does take some time to develop an
appropriate answer in many situations.  Sometimes the questions really
aren’t a policy question, a business question, or a regulation question, but
the fact that were told to use the IG.  And I’m not quite sure what the IG
says.

And someone is going to have a different interpretation from may have
meant.  And with the hundreds of thousands of people reading the IG and
trying to comply with HIPAA, there are a lot of interpretations floating
around.  And often, they just want to say by reading this language, we
read it that way.

And I seen some of the IGs go to lawyers to say legally, what should we or
shouldn’t we do in putting the IG.  It doesn’t sound like the right way to
implement HIPAA.  If we had a source that we could go this is a policy
decision, a regulatory decision.  We really need CMS, the government to
write an FAQ, then fine.

But for the other ones, as we said, this is strictly just what did they
mean when they wrote this line in the IG?  Perhaps whoever wrote the line
could say that’s what we meant, and clarify that maybe more rapidly.  To
coordinate the resources of WEDI, DSMO, and CMS to come up with a more
expedient way to answer the various types of questions that exist.

MS. FRIEDMAN:  And that’s an example of what I just mentioned before
about the business issue.  I want to find out if I can do X.  And the
implementation guide may or may not help me.  This is kind of a gray area,
a new area.  So, CMS, tell me what I can do, so I can go make money.
And we see that a lot actually.

MR. SCHUPING:  Obviously, some of the questions that you are going to
get, we experience that on the WEDI’s tip list.  So, we get just thousands
and thousands of tips every week, and people looking for answers and
solutions.  And some of them you just flat can’t answer without doing a
fair amount of research.

But I think the message that was coming through here again and again in the
hearing and what we got is that they would like to be able, to the extent that
it’s feasible, if industry and government could coordinate to have a central
point where they could go to so they have a reasonable change of getting some
of their questions answered in a timely fashion, instead of kind of having to
go to a dozen different Websites.  Well, I go over here to find this, and
I go here to find that.

If there was some way that collectively we could work to bring a reasonable
amount of data together, so that we have a chance to answer the questions.

DR. COHN:  I want to change back to I think our previous frame, if
that’s okay.  Karen, what we were talking about, WEDI was proposing
obviously what they saw as current issues.  And obviously, on the one hand
there is the issue of complete compliance of the 837.  And I think you
might have some recent statistics on how the Medicare and CMS is doing in terms
of all of that.

But then we were also returning to this issue of how are we going to be
mobilizing the full compliance of all the recommendations?  And I want to
open that up for brainstorming for everyone.  I think everyone has a role
in all of this.  Certainly, the role of the National Committee is try to
get that published and monitor what is going on for all of that.

But I think you might have some ideas both in reference to how we’re doing
with the 837, what your sense is of how all this is going, and all of that.

MS. TRUDEL:  The statistics for last week in Medicare compliance in
terms of the inbound claims were published and were at about 85 percent, which
is excellent.  And we continue to inch up week after week, so that’s good.

The Medicare interim steps that we have put into place, where we will pay
electronic non-compliant claims effective July 1 as though they were paper, in
other words, we’ll pay them more slowly, I think is beginning to have an
effect.  And that was intended to be an interim measure that was not as
dire as cutting off the contingency altogether.

I think what we are trying to do right now, what all payers are trying to
do right now is to get over the hump with the claims, and to work with
clearinghouses and vendors, and providers have all of that.  That has to
be taken care of.  Get everyone into production, and let’s begin to turn
our attention to other transactions.

I know we are tracking percentages on the remittance device and the
outbound claims, COB claims, as well as the inbound claims, because we have
already identified a number of COB issues that are causing our trading partners
to not be able to process our files, and have therefore put in some system
changes which will also go into effect in July, which we think will really
improve that percentage.

And I think that we and everybody else in the industry is going to keep
having to keep chipping away at these other transactions, at the claims status,
at the eligibility.  And to make sure that providers understand the
benefits of using those transactions, and to constantly think about doing
outreach, or even putting incentives into place possibly.  Some plans may
think that they can put incentives into place to encourage people to use these
transactions.

MR. REYNOLDS:  A couple of comments.  A couple of on the ground
comments first, and then maybe questions for the gentlemen from WEDI.

The other transactions are slow in our environment, however, people are
waking up, and we are seeing much more action.  The thing that I think is
still out there that is real, we just held eight seminars across the state,
1,150 people attended from provider organizations.  The most heavily asked
question was exactly how do my work?

Please don’t underestimate the difference in what we heard in
e-prescribing, where they are saying we have an industry.  We’ve got a lot
of things in place.  Let’s add to it.  HIPAA was the rebirth of
everything related to what we wanted.  And that rebirth is coming
slowly.  And it’s coming really slowly in the individual environments.

So, remember, that’s 1,100 people that came of their own choosing, and it’s
1,100 people who asked of us — this was not a prompt question, this was not
what’s your biggest issue, this was open forum.  And those asked the
question, we have all rebuilt so many new relations.

I can tell you for example, our second largest submitters, two major
clearinghouses, one that was here yesterday, have yet to submit an 837, not
because of their strategy, but because they can’t get around the country, all
the way around to get everybody going, because everything is a little bit
different.

So, as you look at what we’ve got here, the exciting thing to look at this
is this is a short letter.  I mean that.  That is very important to
keep in mind.

The other thing that’s exciting, if you listened this morning, we are
talking about e-prescribing, and we heard 270, 271.  You didn’t go
mainstream.  So, we can’t lose sight.

So, I think as you look at what they have here, and you look at the
contingencies, and I know I have testified, Karen says 85.  I’m at
45.  Three of my top five haven’t submitted any claims yet.  I can go
pretty quickly, but the 1,100 people that attended probably make that 15
percent, and don’t even know where their claims are going, because a lot of
times their IT people have made some contract agreement with somebody, and the
actual person working on the ground doesn’t have a clue.  So, their claims
are flying around.

But as I look at the successes on the ground, it was sobering.  It was
very sobering to get that list of the top questions.

MR. BLAIR:  But their number one question was what is the claims
process?

MR. REYNOLDS:  Yes, what is the claims process for my plan.  How
do I get them to you?  How do they get there?  Because what you have
is there are many vendors, when the claim goes to the vendor, hospitals and
doctors don’t necessarily know that the vendor sends it to a clearinghouse,
because you’ve got all these arrangements.

So, understand we have literally from them, out of all the EDI we used to
have, and it is being built back.  We are consolidating them in
industry.  And it continues to be consolidated.  So, what happens is
the H5 is good.  We’ve got some questions, and I think a lot of those will
be answered by the COB, because that will take Medicare from they are.
That level may then become philosophically a little more of an industry thing.

This is really starting to go, but panic is not a key.  But I do have
some questions on what you actually said here.  So, that was just kind of
an on the ground, day-to-day, all day, every day looking at it.

The contingency plans that you are going to see, I think you guys are
recommending that once Medicare does what it does, everybody else needs to make
sure that they are far enough along to make their own decisions.  That’s
what you’re here for.  I think most people aren’t going to blow their
business up.

MR. DALEY:  That recommendation is based on the fact that the average
physician, about 43 percent of the claims going to their Medicare-related —
unlike at the turn of the 20th century in Philadelphia, where if you lived in
the north and the south, you had two phones, because they didn’t
interconnect.  In the physician’s office, they are going to be using the
same system.  So, I think we can piggy-back on that benefit, but that’s
only one of them.

MR. REYNOLDS:  Yes, but I think it’s going to take another key though
is they have clearly stated, these vendors have clearly stated they are doing
Medicare and Medicaid, but then the commercial.  For example, we had
Medicaid actually attend the conference.  We actually invited them from
North Carolina, both the Blues and Medicaid.

Second, I wondered on your second, your non-claims transaction, are those
in priority one?

MR. DALEY:  No, there is no particular order.  Because from the
ROI benefit to the providers, we can get the right information to them from the
payer.  The eligibility is going to have a huge pay off.

MR. REYNOLDS:  Because I think another thing that is occurring in
reality is that now that HIPAA is real, free enterprise is going to get on it I
think in a lot of ways.  So, you are seeing a lot of people expanding
these.  I know for example the Blues are.  We have
requirements.  We have I think I have read six requirements over the next
two years from the standards transactions.

CAQH is in the process of answering the CAQH questionnaire that is out
right now on eligibility, where the industry groups have grouped up and said
here is what should be in the eligibility terms.  So, again, the standards
are there.  None of them said yes or no.  But the industry is going
not really.

Now, we are kind of there, and that’s not the right answer from
industry.  We need to move.  So, there are no frequently asked
questions that you’ve got to move.  There is no IG change that you’ve to
move.  But there is a groundswell that is moving.  Because yes or no
is a problem.

Now, the next question I have is you mentioned that the health plan
identifier for COB, because that would help COB, but right now the COB process
that has always gone on, most of the Medicare intermediaries have taken a
number from the payer, and have been able to do that.

So, I would say that probably 80 of the COB transactions are probably
Medicare-related.  Is that fair?  So, I think that even without this
as a new identifier, it’s good to get the new identifier, but I think once some
of these COB things are settled, we’re probably going to make an 80 percent
move as an industry without having to wait for the national provider ID — or
excuse me, the health plan identifier to be there.

Because Blue to commercial or commercial to Blue or commercial to
commercial isn’t going to flow nicely, because there is not enough volume that
people usually pass their individual number.

MR. DALEY:  I think the self-funded it is not quite that
high.  At least from what I’m getting from the TPAs and most of the large
companies are self-funded and they are multiple plans and so forth.  That
would not be that high.

MR. REYNOLDS:  One of the things that I’m aware of in a number of
states, the prompt pay laws in states says — you look at number three on your
paper where you talk about a timely manner of data content coding
assessment.  One of you mentioned the 835.  The prompt pay laws in
the states require certain messaging going back on remittance devices to the
provider.

So, I would be anxious to see WEDI work with the people who are
establishing the code sets’ meanings for the 835.  So, it’s not a code,
it’s an identifier, because I know in at least two or three prompt pay states,
I know we will be required to send a paper notification of payment, as well as
an 831.

That is not where we are trying to head, because what the state does with
their messaging are not included in the federal codes that are under the
835.  So, what happens is — now, we are working with our department of
insurance and others are too, but those kind of costs, where you get inside the
transaction and you have like we talk about the DSMO and everything, where you
have the messages that aren’t — they are just message codes.  Anyway that
WEDI and others could help with those things.

I think also it starts to take these rough edges, and starts to take these
frequently asked questions, and these things start going away.

MS. FRIEDMAN:  It is the same situation in our state, where you have
to do the papers on electronic billing.

MR. REYNOLDS:  Yes, I don’t remember the exact ones.  In other
words, there are probably 30 states — 20 or 30 states that have prompt
pay.  I can guarantee you this prompt pay law went in ahead of HIPAA.

So, the size of the letter is exciting.  The fact that you did roll
down versus what you had is exciting.  And I think that we have these
recommendations, and I guess I’m going to be interested in what is WEDI asking
us to do with these recommendations?

MR. JONES:  May I add to comment to what you said, Harry?  Many
people would agree with you.  You really hit on something I said earlier,
that some of these recommendations are interrelated.  The real problem I
see, and many of the things that I have heard from the people I have talked to
is getting the money side working.

Just getting the claim in, getting something back that is a
remittance.  And then also knowing am I ever going to get paid through the
eligibility impacts?  So, between those three, to know upfront that the
patient is covered, you are able to submit a claim that can be processed.
You can track through the process to know that it actually got to the
payer.  And then that’s how you get remittance that says here is how much
you are going to paid.

That would solve a lot of the problems and make people move forward.
Some of them require the use of these proper codes or recodes on the
remittance, so they get more information, and they understand it perhaps a
little bit better.  They are all interrelated.  Given that we can
direct the money in place is the key to making people move forward.

DR. COHN:  Karen, did you have a comment?

MS. TRUDEL:  I would just like to make two points.  One to build
on Harry’s, and I think that having the standard transactions, standard code
sets, standard data content has served to cause people to be able to realize
that there are perhaps slight nuances and differences in business practices
from plan to plan that aren’t necessarily relevant to their actual operation.

And it presents an opportunity for people to begin to look at clarifying
and making the eligibility business process more consistent.  I think that
the transaction standards are serving as the best kind of harmonization
processes where that makes commercial sense.

The other thing was, although it wasn’t necessarily directly related, there
have been some questions about the Medicaid contingency plan, and when at some
point in time we announced that we are going to lift that, how that leaves
physicians other plans.  And it has come up often enough that we have
drafted, and in are in the process of clearing a frequently asked question that
will address that.

MR. DALEY:  I think that would be useful.  There is a lot of
uncertainty about that, particularly with regard to the various percentages of
compliance on these other standards.  And it is obviously going to take
time to deal with that.  I think that would be very helpful.

The other point, based on what you were talking about, what Harry was
talking about, and it really relates to some of the discussion earlier
today.  In the health care industry there is a lot of confusion about what
the standards are.  There is a lot of confusion about what the standards
are.  There is a lot of confusion.  People believe that a lot of the
electronic processes, and data systems, software, et cetera, somehow lend a
competitive advantage.

There are lots of counter examples, such as in a very recent in the way
kiosks are used at the airport for getting your ticket, and those more or less
standardized.  The way ATM networks now are seamless around the world, the
way cell phone technology is evolving, although there are three different world
standards.

And I think that’s what we have to convey to get people to move ahead in
the health care industry is we are not taking away competitive
advantages.  What we are really trying to do is create a tool or a utility
that is going to convey benefits to the investment in this, and that there are
other business-related factors that are going to have an impact on
competitiveness.

This is a very confused notion in the health care market.  And I think
as we do outreach and education, we really have to get that message across, and
use examples that people see in everyday life.

DR. COHN:  Harry did ask an interesting question, which is I think you
asked what should we do with the recommendations.  What would WEDI like us
to do with these recommendations?  And obviously, there are various views
on all of this stuff.

Harry is happy with only one page.  I’m not sure I’m happy because
this is a two page summary, but anyway, it isn’t a 100 page book.

But what are your thoughts?  We actually heard from you.  There
are things that I have already commented that I think there is a need for us to
be monitoring, though hopefully not in extending hearings.  We getting
sort of hearing-ed out.  But we need to come out with a better methodology
to do more in-depth monitoring.

MR. DALEY:  WEDI is prepared to put together task groups.  And in
fact what I called kind of a herd of true confession of that May 6 meeting when
we all went and volunteered to do something under Sec. Thompson’s 120 day
schedule, probably in a fit of madness I went up to the microphone and said
that we would be delighted to marshal our troops to work on incentives on both
the provider and payer side, to look at how we could foster faster movement on
compliance.

And so, we, at our board meeting last week, put together a group that will
be meeting.  I believe we have a conversation with David Brailler(?)
coming up to focus on this issue, and to kind of get his thoughts on it.
I think the incentives, and that was brought up earlier, is going to be very
important for moving the industry faster in this.  So, that would be one
thing that we would bring back to the table.

I think a second issue here, and we talked about it just a minute ago is
diminishing the uncertainty with regard to these contingency plans.
People, I think, if there were for example sequencing, if we are going to go —
as Harry said, are these in priority order — if we are going to go from
eligibility into however we end up going down the certain time frames or ranges
of time frames, I think that would focus industry attention, and move out of
kind of this uncertain zone.

MR. SCHUPING:  I would just add that I think we would be very
interested in getting any feedback that the committee would care to offer as to
how useful these are these recommendations and priorities.  We have found
that we’ve had a lot of success by just moving with task groups, where you kind
of focus on a single project like this, give them a timeline, and give them the
resources to get it done and make it happen.

A lot of these things I think we need to have it sooner, rather than
later.  And in some instances, as you look at the different
recommendations and see that they are not going to be solo flights, they are
going to require collaboration with more than one organization or entity, and
many industry and government stakeholders.

So, if you feel that these are on target, and they make sense to you, then
with that blessing, we would be happy to take a lead role in making any one or
several of these happen, because we think they are critical.  This is the
feedback that we heard from our constituency.  These are the things that
were important to them, and we would like to move it forward.

MR. JONES:  Simon, just a couple of comments as well.  As Jim
indicated, we would certainly like NCVHS to endorse these
recommendations.  One of the other things that we consistently heard in
the testimony and other places is that with the outreach that CMS has been
doing in various ways, anything that the committee here could do to help
enhance outreach and continue the support of outreach programs, that would be
very beneficial.

Was the answer acceptable, Harry?

MR. REYNOLDS:  Obviously, the devil is in all details.  There is
no question that each of these subjects, I feel is very valid.  But they
are issues that are clearly out there.

I think that if WEDI works with CMS, because I think we’re all kidding
ourselves, the whole initiative is kidding itself, CMS, whether they want it or
not, has the ball, since they are upfront, working with them.

So, if WEDI works with them, because one of the things I think is
interesting about one of the recommendations, the recommendation that CMS do
the frequently asked questions, in many, is that CMS site of care, or is that
the CMS?  Because remember, some people out there don’t know the
difference.

MS. TRUDEL:  It could be both.

MR. REYNOLDS:  That’s right.  But I think the key thing is the
recommendation that you team up with DSMOs and WEDI, and then team up with CMS
so it evolves with the frequently asked questions, even if there is a new
frequently asked question coming out from CMS, it is vetted, at least so that
you could comment to your constituency about what you think it means.

One voice — the one voice doesn’t have to be agreement.  One voice
has to be all understand what we’re saying and we all do it.  So, that
recommendation I think moves people along.  But the rest of them are
valid.  It’s just my opinion, I don’t think there is anything in here that
either CMS or WEDI has the authority to do without any recommendations.

DR. COHN:  Say that again Harry.

MR. REYNOLDS:  I’m not sure if there is anything in these
recommendations they don’t have the authority to continue to do.

MS. TRUDEL:  That’s absolutely correct.

MR. REYNOLDS:  In other words, that was my last question, because none
of these recommendations, these are all good recommendations about good
business process.  And there are already ways to do it now.  Speeding
up additional ones, things are clearly identified.  And then staying close
with CMS, so that their message when it comes out, comes out to an entire
industry and a group.

Like you mentioned a contingency.  On the contingency matter, the CMS
changes its own, well, that isn’t clear which CMS has.  That one is a
resounding message.

MR. DALEY:  See, most of the private health care community,
irrespective of what CMS says, is going to view that as the federal government
and the imprimatur.  And on the same token, if some sort of schedule is
laid out that people could anticipate.  Right now they are in an uncertain
zone.  And they don’t know how to proceed.

MR. REYNOLDS:  That was my point.  And you guys, the three of you
worked directly with CMS to help as they are about to come with something — I
would hope between the two of you, before it comes out.  You are
representing the industry, and they are representing the government aspect.

But there needs to be some reviewing so that the message that comes out —
I think CMS will say obviously whatever CMS wants to say.  But you guys
being side by side with them, explaining what it really means, because I would
a little bit disagree.  I don’t think most of the private sector is going
to do a dramatic link when Medicare pulls off there compliance if their client
level is not sufficient to keep their business going in a reasonable way.

But the messaging of that is still key.  I’m not playing
point-counterpoint.  And so, those are the things, that you guys working
together with CMS especially, and then industry on these other things, I should
keep going, and keep going more vigorously.  I think you should set your
committees up, and I think they should go, because all these things still have
to get fixed.

But I don’t think that writing anything, putting together a letter and
writing to the secretary, or WEDI writing to the secretary, I don’t think
that’s changing.  I think that’s changing.  I think now that we’re in
the business process, making sure that the right people are grouping up.

DR. COHN:  Actually, I think WEDI has already written to the secretary
about this, as I understand.

MR. DALEY:  We haven’t.  We do have an excellent collaborative
relationship with the Office of HIPAA Standards.

DR. COHN:  I think Karen also has a point.  My own view is that
you are doing absolutely the right sort of thing.  You are an advisor to
the entire process as identified in the legislation.  This sort of advice,
and work in the industry, which is I would say WEDI at its finest, but one of
its finer moments.

And as I look at all of this, I’m not sure I think it’s necessary just only
working with CMS.  But really your strength is working also with the
industry.  A lot of the things are things that we need to be working
directly with DSMOs about.  This issue of external code sets, which we
have already gone on record sort of saying, gee, there needs to be improvements
to the process, improvements to the updating, and all of that.  I mean you
are dealing with very specific examples at ground zero on that.

And it’s something where I would think by bringing this to them, and
working with them around all of this, that one should be able to get
resolution.  And if not then CMS, and NCVHS might need to get
involved.  But again, there are really a lot of things here.

As I said, I don’t think this is something we should approve or
disapprove.  There are too many pages here, too many nuances, including a
number of things I would have to recuse myself on anyway.  But having said
that, I think the direction is correct.

MR. REYNOLDS:  I totally agree.  I wasn’t excluding any of those
others.  Since they represent the industry, I would agree.

DR. COHN:  Karen, any comments?

MS. TRUDEL:  I just wanted to reassure some committee members that HHS
is already looking at a whole 20 page report or whatever, and analyzing the
recommendations.  So, that activity is ongoing irrespective of whether the
committee chooses to or not to make any of a recommendation.

MR. DALEY:  Using Harry’s model, that shows how much progress we’ve
made over the last two months, fewer pages.

MR. JONES:  On some of these items there has also been progress.
For example, the announcement on error reporting, WEDI has done a white paper
on that, and several other things are still being worked on.

DR. COHN:  Is there something anybody else wants to comment on?
Is there agreement with everything that has been said so far?  Jeff do you
have any comments?

MR. BLAIR:  No, I feel very comfortable with your observations.
I think WEDI has just done an outstanding job in a very tough situation in
trying to help and shepherd the adoption of HIPAA standards.  So, no, just
complete support.

DR. COHN:  Yes.  Stan?  No.

Lynn, I saw you had your hand up.  Do you have a comment or a
question?  Come to the microphone.

MS. GILBERTSON:  Lynn Gilbertson from the National Council for the
Prescription Drug Program.

Just one or two things I wanted to make brief responses on to the summary
document.  And that is when you talk about the frequently asked questions
or the industry place to go to, NCPDP members saw the need immediately when
HIPAA became real.  And as part of the standards development process,
since we could not make modifications to the main standards, we created a
document lovingly called the Version 5 Editorial.

That is available on the public Website.  It’s available to anyone who
purchases standards.  And as it is updated, which is usually on at least a
quarterly basis, industry is notified that this document has been
updated.  It contains editorial corrections, with some typos.  It
contains clarifications.  It contains new examples.  It contains
anything that comes forth.

We have an industry task group that spends a lot of time going through
questions that come in from the public and say, how do I do this, or what did
you mean by, or whatever.  So, we have already built that model so that
the industry has some place to go to.  Because basically, to be honest,
NCPDP and its members are the experts of this standard.  So, we didn’t
feel a separate body needed to formed.

One of the things to make you aware is that the DSMOs have spent an
enormous amount of time with the Office of HIPAA standards discussing
information that could be put into the next NPRM having to do with a
modification and maintenance process, and trying to build as much into moving
forward with the industry as quickly as possible within the realm of what we
are allowed to under the regulations.

We have spent an inordinate amount of time doing that.  So, that will
be available when HHS is ready to publish in an NPRM, hopefully later this
year.  So, it’s not like there are complaints, and no one is
listening.  We are spending a lot of time trying to make things,
recognizing that when it comes to naming new versions of existing standards,
for example, the rulemaking process has got to take place.

So, as long as we’re under those guides, we have to work in that time
frame.  But as far as maintenance type things, there might some
suggestions we can offer in industry comment.

The other is as I look at the certification, and when you
post-certification maintenance should be restructured for a version release
scheduled only once a year.  But where appropriate, it’s real important
that NDC better not be that way, or we’ll shut down the industry.

The other thing, just as an FYI, there is a comment in the COB section, and
we are still trying to track it down.  Someone made a comment about the
NCPDP standard.  And we have not yet been able to find who submitted that
comment so we can take action on it.  So, it’s out there.  It’s in
the document, but until we can track down who submitted the comment, we can’t
take a lot of action on that.

And then I guess the other thing, in the summary, just as a question, item
number two says balanced representation in the DSMOs.  And when I went
back to the original testimony, it asked that the maintenance committees should
insure balanced representation.  And I’m wondering if new information came
in that something will be sent to the DSMO about this item?  Or was it
just a typographical change from under the code set, maintain a committee that
acts in the biz(?) mode?

MR. DALEY:  That’s a typo.

MR. REYNOLDS:  At the last meeting we had, Jack Embry(?) testified,
and I’ll paraphrase what he said.  I can’t quote him.  His comment
was that it’s hard to figure out how to get providers involved in WEDI or any
of these other things, because who really represents them.  And his
comment was we’re so diversified, who represents us?

But I continue to ask as we move forward, especially with WEDI, whether
it’s CMS, whether it’s the DSMO, that we somehow don’t forget the people that
are actually on the ground.  The same thing we talked about with leaving
scrap(?) where we can do all the dancing we do, but if we don’t make that work,
we don’t do that electronic prescriptions, we don’t have a deal.

And so, I think as we go forward and make these changes anyway, any of the
three main groups that can reach out and make sure that they at least make
extra effort to touch some segment of that provider group, I think is going to
be a plus.  Because otherwise we will continue to hear testimony about we
really weren’t involved.  We didn’t get involved.  We didn’t know how
to get involved.

So, I continue to recommend to WEDI and others that they would push this to
make sure that they work with the main associations, that the people either get
involved, or they had a chance to get involved in this.  Because I think
that absolutely uninvolvement is not going to be a positive for us as we
continue implementing, especially with the questions I mentioned that we heard
in the seminars.  And especially as we try to go forward and make changes
to something that is not even done yet.

MR. DALEY:  Can I comment on that?  I learned early on in the
early nineties when I started a company, I had two young doctors who wanted to
make an investment.  And I asked them why, and they said, well, when you
go home at night, your meter is still running.  When we leave the office,
our ability to generate revenue ceases.

So, in terms of outreaching to physicians, you’ve got to get them on the
weekend.  You’ve got to focus on their office managers.  It’s a
totally different outreach and marketing.  We did a seminar for OCR in
March 2003 in Chicago on a Sunday that was snowing.  We had 1,202 mainly
providers show up from a five state area.  We certainly wouldn’t have had
that response on a Wednesday or other week day.

So, we have to kind of bait the hook to suit the fish.  And down at
HIMS, when there were a number of very good presentations on the electronic
health record, Linda Schady(?) of VA was talking about they made an extra
effort to go out on these weekends and so forth, to draw the providers on
getting their input on the EHR.  So, I think we need to do more of the
same to get them here.  And we are doing a lot of that in collaboration
with CMS.

DR. COHN:  Thank you.  I have been called many things in my time,
but never a fish.

I just want to make a comment.  First of all, I want to express I
think for the full subcommittee and the full committee that we really very much
appreciated obviously your advice and guidance as we move through all of
this.  Obviously, this continues to be a very important relationship.

We don’t always vote on all your recommendations, but I think we can
certainly sense that you have made an impact with these recommendations.
They are being significantly scrutinized by the CMS Office of HIPAA Standards,
as well as by the secretary.  And I think that that’s an important piece
here.

Obviously, there is a lot you can do without anyone else, without us saying
anything.  But I think the important thing to keep looking at what’s going
on, and figure out ways to move forward.  Now, certainly to my view, as I
look at all of these things, my sense of the subcommittee, obviously, our main
goal in all of this for the moment needs to be to figure out some robust way to
do monitoring of the industry.

And if you have ideas that can be of assistance in that, that would be very
helpful.  You know, one of our main responsibilities really is that
monitoring and advising Congress and the secretary about how things are
going.  I know for a number of the years involved, implementation was just
sort of in all of our minds’ eye.  That was really more of a pro forma
activity.

But given the issues that I think we are all talking about, I think it’s
going to be very important for the next while for us to be doing this in a more
active sort of fashion.  And as I say that, we also realize for the next
three or four months, you know what the subcommittee is going to be
doing.  We are going to be meeting nights, weekends, and every day of the
week on e-prescribing.

So, we will need to figure out some way to do this a little bit offline,
and it will be discussions that we have with our staff and all of that.
But we obviously have to find an efficient to really understand how the
industry is moving forward on all of this.

Also, certainly if you have some ideas about consensus.  Once again, I
don’t know about that in terms of NHII or e-prescribing or whatever, but it may
make sense, knowing that I think we repeatedly commented that a lot of the
benefits of these transactions actually doesn’t go to those who are actually
having to implement them, at least in the providers’ offices.  And so, if
indeed there is significant value to be derived from the industry as a whole,
it’s something that we all ought to be looking at, and going from there.

Anyway, I want to thank you.  I wonder if any subcommittee members
have any final comments to make?  We will obviously continue the
conversation, and we’re just going to have some very good way of keeping our
eye on how the implementation is going.

MR. DALEY:  We appreciate the opportunity to be before you.

Agenda Item:  Follow-up on DSMO Request –
Donna Pickett, NCHS, CDC

DR. COHN:  Thank you so much for joining us.

Okay, our next item is the DSMO issue from before.  Donna, should I
remind the subcommittee of where we are at this point?  If you will
remember, it was an issue that came before us as a request from the DSMOs back
in 2001, related to the issue of use the NCPDP transactions in building and
supplies.  That was an issue in the year 2001, that the subcommittee, for
a variety of issues, decided to defer a recommendation on.  And it was
brought back to our attention late last year or early this year.

There have been sort of two issues that we have been mulling about in
relationship to what to do with that recommendation.  One relates to the
actual use of the standard for the building and supplies.  And then the
other one had to do with the issues related to use of the code sets in
relationship to those supplies.

I think last time we talked about the use of the standards, but we were
sort of left waiting for HHS further evaluation and assistance and guidance
with the issues of the code sets, and what sort of recommendation we should
have around those.  I think at this point Donna Pickett has done some work
for us, and has come forward with some conclusions.

MS. PICKETT:  Thanks to Lynn Gilbertson and Karen Ebhart, Tom Bizzaro,
and also John Roberts at the Uniform Code Council for being so patient with my
questions and my phone calls.  We had a wonderful time trying to connect
with all the meeting we were attending.

Basically, what I have tried to do is pull together a broad overview of the
Universal Procedure Codes, UPC codes, and the HRI codes, again, kind of like a
high level overview.  Fortunately, Lynn, Karen, and Tom are still here, so
if anybody has questions that are down in the weeds, I guess turn to them and
say help.  So, thanks for your help in advance.

I’m going to start with a description of the Universal Product Codes.
The UPC codes are assigned to over-the-counter products such as vitamins and
St. John’s Wort.  To obtain a UPC code, the company needs to become a
member of the Uniform Code Council.  And the Uniform Code Council is the
not-for-profit organization that administers the UPC.

In order to become a member of the UCC, the company needs to provide some
basic information.  That information includes current or projected sales
revenues, the number of products that will be identified with a UPC symbol, and
the number of locations.  And the fee schedule is based on that
information.

And as mentioned earlier, both the UPC and the HRI codes are used within
the NCPDP standards.  They are external code sets to the NCPDP standards.

There are guidelines for the use of the UPC codes.  The manual is on
the UCC Website.  The UPC consists of a labeler portion.

Well, first of all, it’s an 11 digit number.  The first six digits are
assigned by the UCC, and the last five digits are assigned by the manufacturer
and represents the item reference number.  The number identifies the
individual products for the company’s prefix.  Each product has its own
item reference number in order to be uniquely identified within the UPC bar
code.

The HRI, Health Related Item code sets is a system for identification and
numbering to market device packages.  That is a category of other
numbering systems such as the NDC or UPC.  It is a 10 digit code and can
appear in two formats, a 4-6 format or a 5-5 format, and is assigned for such
things as syringes and band-aids.

Manufacturers who desire to use an HRI number for a unique product
identification must first apply to the FDA for the labeler code, which is the
first portion of the code.  Participating manufacturers and distributors
then complete the code by identifying their devices with a sequential number.

The manufacturer or distributor assumes responsibility for maintaining the
number.  There are specific directions governing the use of the product
package sections of the NDC HRI codes that are prescribed by the FDA in
voluntary compliance regulations.

The other thing that I have written in reviewing these two code sets was to
look at the codes as it relates to the 10 guiding principles that we had under
HIPAA.  And for those of you who may not recall of them, I will just read
through them quickly.  There are ten.

The first one is to improve the efficiency and effectiveness of the health
care system by leading to cost reductions for or improvements and benefits from
electronic health care transactions.

Number two is to meet the needs of the health care data standards user
community, particularly health care providers, health plans, and health care
clearinghouses.

Three, be consistent and uniform with other standards required under this
part, data elements, names, definitions, and the codes, and the privacy and
security requirements, and with other private and public sector health data
standards to be the extent possible.

Four, have low additional development and implementation costs relative to
the benefits of using the standards.

Five, be supported by an ANSI accredited standards setting organization, or
other private or public organizations that will insure continuity and efficient
updating of the standard over time.

Six, have timely development, testing, implementation, and updating
procedures to achieve administrative simplification benefits faster.

Seven, be technologically independent of the computer platforms and
transmission protocols used in electronic health transactions, except where
they are exclusively part of the standards.

Eight, be precise and ambiguous, but as simple as possible.

Nine, keep data collection and paper work burden on users as low as is
feasible.

And ten, incorporate flexibility to adapt more easily to changes in the
health care infrastructure, such as new services, organizations, provider
types, and information technology.

The UPC and the HRI codes do not meet two of the ten principles — not
being precise or unambiguous, and lastly, though it’s not specifically
identified within the 10 guiding principles, inherent through the updating
process was the fact that the updating process should be an open process.
And obviously, since these code sets are product-driven, having an open process
obviously does not lend itself to how proposed sets of these supply codes would
be updated.

The codes have been used successfully for a number of years.  And I
think as Lynn Gilbertson’s presentation at the last meeting, she had referenced
issues related to what would happen in the industry if they were not able to
continue to use the UPC and HRI codes.

That is kind of the high level overview of each of the two code sets.
And hopefully, I have not done anything inappropriately.

Agenda Item:  Development of Potential
Letter – Simon Cohn, MD

DR. COHN:  You know the issue of supply codes, I think we have all
realized this one is a difficult area.  It seems to be particularly
problematic in health care.  Having said that of course, NDC is really a
supply code.  And the bar code has recently was approved by use in the
hospitals, is once again another application of supply level codes.

Certainly, this issue has brought to me, I think the way we need to take a
hard look at the whole issue of supply codes.  We have already had some
conversations about how late in this year or early next year, or whatever, we
probably need to have a session where we take a hard look at the whole issue of
supply codes.

HCPCS is part of it, but there is really a whole lot more there that
relates to the world of supply codes as it relates to health care.  This
is not in any way to say that there is a solution that I have to offer.
But more just to sort of set the stage for a letter that we tried to draft.

Once again, I wasn’t quite as on target as I think we should — once again,
I think we should mull over where it represents some of the potential
subcommittee views on what is it that we might want to say.  As you all
know, we have waited from coming forward and showing the letter on
recommendations until we had at least a little more information on the various
code sets.

Now, I will say just sort of by introduction to the letter, and I’ll put it
up here so you can read it, this letter is intended to give the department some
latitude in terms of moving forward, and really it’s designed to do that,
recognizing that I think we all thought that there needed to be some further
evaluation aspects of all this as the department moves forward and considers
any possible rulemaking.

Does anybody have a comment about this before we go into the letter?
Tom?

MR. BIZZARO:  I think part of the confusion that was caused because of
us identifying NDC as the code that we would use within the telecommunications
standard has been brought about because when we talk about the files for
proprietary drug files, we talk about those files as NDC files.

Those files across the board I believe for all the competitors in that
arena, have always continued NDCs, UPCs, and HRIs, because even though we are
drug databases, the industry demanded certain medical devices and items that
had UPC codes and HRIs as identifiers.

Donna mentioned that there are not unambiguous codes.  And that’s a
true statement, because there can be duplicates of those codes, and they are
not strictly regulated in the way that they can be reused by the member
organization.

I guess I would argue in a more practical sense that they are unambiguous,
because in the databases we are required — we use those as unique, either UPC
codes, HRI codes, and NDC codes.  We do not have duplicates in the codes
at a single point in time.

So, I think that the industry, and retail pharmacy especially, when they
talk about how they use the telecommunications standards, and we talk about NDC
codes, we inadvertently did not also say well we use NDC codes, we use HRIs,
and we use UPCs.  It’s been a common practice for as long as I know there
has been electronic telecommunication and adjudication of claims.

DR. COHN:  Okay, any questions?  And I think that’s exactly the
issue that I think we are confronted with.

Anyway, should we read through the letter, and just sort of go from
paragraph to paragraph, and we can sort of see if indeed this captures the
sense of where the subcommittee wants to go with the letter of recommendation
on this.

The first paragraph hopefully can be relatively non-controversial.  As
part of its responsibilities under HIPAA, the NCVHS is responsible for
studying, selecting, and recommending standards under HIPAA.  This letter
addresses a request made by the DSMOs with NCPDP 5.1, the adopted HIPAA
standard for electronic link for supplies via retail pharmacy.

Is everyone okay so far?

MR. PACE(?):  Scott Pace with the National Community Pharmacists
Association.

I would it just to be consistent with the language of the second paragraph,
and Ms. Gilbertson’s testimony back in March.  It probably ought to be
reflected as certain supplies, rather than saying — it should say certain
supplies.

DR. COHN:  Okay, thank you.

Okay, next paragraph.  The NCVHS did not make a recommendation to the
secretary on this issue when it was first raised in 2001.  In response to
a second request by the DSMO steering committee in February 2004, the issue was
revisited in a hearing on March 31, 2004.  Testimony indicated that a
business case by retail pharmacies to have a real time standard under HIPAA for
building and for supplies, and the current widespread use of the NCPDP
telecommunications standard by the industry for this purpose.

NCVHS recommends the continued use of NCPDP standards for building and
supplies that are consumed or used during or a consequence of the
administration of a drug therapy or commonly dispensed via a retail pharmacy
channel.

We further recommend that the department investigate this issue further,
especially as it relates to possible overlaps and redundancies in the use of
NCPDP and X12 standards for the building and supplies by pharmacies.

So, we changed therapy to pharmacy, correct?

MS. FRIEDMAN:  Do you want to say just certain supplies?

MR. BLAIR:  Yes, please.

DR. COHN:  And it probably isn’t redundancies, it’s probably just
overlaps.  And I want to get rid of redundancies.  I think it’s
really more overlaps, on the second line from the bottom in the paragraph.

Okay, Jeff, do you have a question?

MR. BLAIR:  Yes.  When we indicated that we didn’t make a
recommendation, but that we revisited things, I get the impression that we/and
or NCPDP Script, and/or both of us overlooked — somehow it was overlooked that
these codes were being used, and therefore that they were not included.

And I think that if we include that in the letter, it would strengthen the
fact that really they should have been in there from the beginning.  Is
what’s said correct or accurate and/or helpful?

DR. COHN:  Jeff, this is not an issue related to the beginning.
The beginning was many years before 2001.  This has to do with issues that
came before the subcommittee as a result of the annual DSMO review
process.  That was a fast track process, or it’s part of the DSMO
review.  It was an issue that came before us.

So, this is not an issue of well, there was something neglected in the
initial regulation.  This is more a request by us for specific of use
here.  So, I would speak against trying to go back as far as you are
describing in terms of records.  Is that okay?  I think this is fine
letter without trying to improve the strength of it.

MR. BLAIR:  Okay.

DR. COHN:  Are there any other comments, questions on this one?
If we need to strengthen it, we’ll go back and do other things.

The next paragraph is the DSMO steering committee also requested that when
supplies are billed by the dispensing pharmacy, the NDC be designated as the
standard code set that must be used for supplies with the NCPDP standard when
the NDC number is available.  If there is no NDC, they recommend that the
parties may use any codes in the NCPDP standard that they mutually agree to
such as the Uniform Product Code or the Health Related Item Code or the HCPCS
Code, where non-specific codes will not lead to potential problems in patient
care from their use.

MR. LEVIN:  This is Randy from the FDA.

There shouldn’t any NDCs for supplies.

DR. COHN:  Tom, do you want to come to a microphone and clarify that?

MR. BIZZARO:  Yes, there are some items that have NDCs, that I believe
were assigned many years ago.  And the thing that comes to my mind, and I
am not in that part of the business any more, but syringes have NDC codes on
them.  In fact, syringes may have an NDC code, an HRI code, and UPC code.

MR. LEVIN:  A number of years ago there was series of lay bar codes
that was given for devices to be coded.  And this is the not an NDC per
se, but codes with the other code that you had described.  It’s not a NDC
specifically.  So, whether you want the same type of codes for supplies,
that’s another issue, but as far as it currently is right now, those are not
officially NDC codes.

MS. FRIEDMAN:  Does somebody have the original incoming from them, so
we can double check the language?

DR. COHN:  Well, Lynn is here.  Let’s ask her for —

MS. PICKETT:  Lynn, do you have yours?  I’m not sure that that is
the language that was in the testimony from March.

MR. LEVIN:  They may say that they are NDCs, but they are not.
The health related items are the series of supplies that were given a set of
labeler codes so they could provide codes for them.

MR. PACE:  I believe Becton-Dickinson(?) actually uses the NDC as a
break in the number on their syringes.  In fact, they do on their Website.

MR. BIZZARO:  In fact, their labeling still says NDC, and then gives
the code.

MS. TRUDEL:  I need to clarify this with Randy, because it sounds as
though an HRI code is about as close a relative to an NDC code as you can
get.  It has still got a manufacturer labeler code in the first position
that is issued by the FDA in just the same way as the NDC.  And they are
not the same structure, but the same size codes.

MR. LEVIN:  The same size.  So, by looking at it, it looks like
it could be an NDC code as a labeler code.  But they are there —
officially, NDCs are reserved for a drug, and the HRIs were for supplies.
And that was many, many years ago when these were first given out, and I
don’t even know if the Center for Devices even continues that activity
now.

MS. TRUDEL:  Are they included in the same database?

MR. LEVIN:  At the FDA?

MS. TRUDEL:  Yes.

MR. LEVIN:  No.

MS. TRUDEL:  Okay.

DR. COHN:  Steve.

DR. STEINDEL:  Randy’s discussion brings up a very perplexing
question.  Since these aren’t considered NDC codes by the FDA, are
they a HIPAA code set?

DR. COHN:  I’m sorry, what does that have to do with this?

DR. STEINDEL:  Think of the user then, external code sets.

MS. TRUDEL:  No, this is where we are going to have to get some expert
legal advice, but at least on a very amateur version of the reading of the
regulations, we adopted HCPCS as kind of a catch all for devices, supplies, and
other whatever you use in health care, regardless of the setting.  So,
it’s sort of devolved.  If you are talking about something that is a
medical item, and there is no other code set in place for it, it is possible to
argue that it is HCPCS, simply because of the way we wrote it.

Where my question comes is in is although the FDA says that these
aren’t NDCs, it appears to some of the people who are using them, that
they are NDCs.

MR. LEVIN:  There is one caveat.  Certain drugs have devices that
stay with them.  So, there may be some mix there.  But in general,
there is a whole group of supplies that have a code that looks like an NDC, and
could be an NDC, but it’s not.

DR. COHN:  Well, Randy, I think this is sort of an issue, fact that
the individual users don’t know the answers.  I think for the moment
I would just leave the wording the way it is, but if you can present –
once again, the committee will simply do what it chooses to do with this
letter.  Obviously, it will to go the full committee to pass in mid-June.

And there is certainly ample time that if you can bring forward
information, we can certainly modify it.  But this is what I think the
industry thinks is going on.  So, we probably ought to keep it that way.

MR. LEVIN:  The statement is okay actually.  If there were no
NDCs, you would just go on – I think that NDCs are solving an issue for
supply.

DR. COHN:  Okay, so the next statement?  Currently, the NDC is
the HIPAA standard for reporting drugs and biologics by regional pharmacies,
while HCPCS codes were adopted as the HIPAA standard for supplies, equipment,
and other items without regard to setting.  Testimony indicated that the
UPC and HRI codes currently are used by the industry to identify items on
retail pharmacy claims.  These codes are used within the NCPDP standard,
but are external codes that are separately being included and updated out of
the NCPDP standard setting process.  Their continued use is necessary to
provide the industry with the level of detail needed to handle claims
processing, ordering, supply management.  Testifiers noted that HCPCS
codes do not provide the specificity needed for the day-to-day operations for
retail pharmacies.  NCVHS anticipates holding hearings on the use of
supply codes in the near future.

Are we okay so far?  The next paragraph is NCVHS urges the Department
to investigate these issues, and determine if they need to be addressed in
future rulemaking.

MR. REYNOLDS:  We state a position, and then we talk about holding
hearings.  And then we ask the department to investigate.  Are you a
little confused?  In other words, we’re asking them to investigate,
I’m not sure we wouldn’t have hearings until after that.  But
once we ask them to investigate it, I don’t think we need to have
hearings, because once they look at it, they decide something.

DR. COHN:  Yes, I think what we are trying to do, and this has not
quite come together yet, but I think what we are trying to say is we want the
department to further investigate all of this, and come to some future
rulemaking on this, knowing about our feelings and general views on this.
And the last part, the NCVHS anticipates holding hearings on these supply codes
in the future.

MR. BLAIR:  Investigates these issues to determine whether they would
require future rulemaking.  That, I think, clarifies the distinction
between holding NVCHS hearings and what we are asking them to investigate.

DR. COHN:  Okay, so we’re saying that we’re going to move
that emphasis on NCVHS after that.  Now, I think that maybe the sentence,
maybe it’s a separate paragraph.  What do people think?
I’m sort of thinking about the first sentence in the second paragraph
needs to be moved up to the last paragraph, but I’m not sure about that as
I’m looking at it.

Margaret?

MS. AMATAYAKU:  What are NCVHS hearings going to do?  Are they
going to make recommendations then to the department in terms of their thinking
about whether they want to –-

DR. COHN:  No, maybe it’s hold hearings on the larger issues of
supply codes, not the specific issues.  NCVHS holds hearing on the larger
issues.

MR. BLAIR:  Larger issues surrounding.

DR. COHN:  Surrounding is it use of supply codes or just supply codes?

Margaret?

MS. AMATAYAKU:  Can I ask again, first of all, you might want to
separate those two into two separate paragraphs.

DR. COHN:  Absolutely.

MS. AMATAYAKU:  So, what are these hearings on the larger issue going
to do with respect to this specific issue?

DR. STEINDEL:  Simon, by filling in this letter, will you ask the
department to investigate the issues and determine if it needs to be addressed
in future rulemaking, and not say anything about us holding hearings.  Or
we should say we are going to hold hearings about this issue, and not ask the
department to do it.  One or the other.

DR. COHN:  Should we get rid of all comment about our upcoming
hearings?  Okay.

MR. LEVIN:  The question is about having codes for the supplies, which
codes to use.  Is that what the question is?  Right now, we can
supply codes for the license, and that’s a supply.  It’s
different than saying we need those codes.  Are you looking at what the
need is for the community as far as having codes for medical supplies?

DR. COHN:  Are you asking about our hearings?

DR. LEVIN:  Yes, what are you asking the department to do?  If
you are talking about supplying codes for the medical devices are supplies, we
already can do that.  Now, whether you want those codes, do you want
people to supply those codes?

MS. PICKETT:  I think what Randy is getting at is there is already a
process for creating the codes, using the codes.  But I think what we are
talking about might be the use of the code sets in HIPAA transactions, which we
haven’t really quite said well in the sentence.  And I think that is
what Margaret is talking about.  What is it we want to do with it?

And we have NDC which is in the docket as the HIPAA code, and it is the
supply code somebody mentioned.  But then we have these two, plus there
was another one that we referenced in the final rule on transactions and code
sets that we said bears watching, but because it wasn’t being used in the
claims environment, we would wait until after the pilot was over.  So,
we’ve got a couple of supply codes that have all been sitting out here
together.

DR. COHN:  Yes, that was my sentence about when we talked about this
later on.

MS. AMATAYAKUL:  Would it help to say NCVHS urges the department to
investigate whether the specification of the UPC and HRI code sets need
separate or additional future rulemaking to be used?  Is that the crux of
the issue?  Because they weren’t included in the rule
specifically?  You only mentioned the NDC and not these others, is that
the issue?

DR. COHN:  I think it’s because of how they use it.  How
should this go forward?  I think what we are saying is that the idea of
the NCPDP uses those supplies is I think a reasonable step forward.  We
are trying to give the department fairly wide latitude in terms of what it
wants to do to further investigate and figure out this issue with the code
sets.  That is really sort of the intent here.  And obviously, we are
not seeing it very well, because we are sort of falling over ourselves.

MR. BIZZARO:  Remember, we are kind of biased about how we see this
being used.  But we know that industry in the past has used
telecommunications standards, first 5.1 and in earlier versions using all codes
and CPCs, as well as NDCs.  So, we are not asking for the creation of any
new code sets.  What we are asking for is simply permission to use the
telecomm version 5.1 UPCs, NDCs, HRIs for those appropriate products.

MR. BLAIR:  So, you’re saying you’re asking for HHS to
recognize these code sets as the code sets?

MR. BIZZARO:  Actually, I think what we’re asking is that within
telecommunications, within the NCPDP telecommunications standards, we are
asking permission to use these three code sets to identify drug products or
supplies.

MR. BLAIR:  How is asking for permission different than recognizing
them as HIPAA code sets?

MR. BIZZARO:  I guess in my mind maybe it’s the terminology,
because I’m not familiar with the rulemaking process.  In my mind,
I’m not sure the committee wants to make UPC codes entirely in the NDC
code set as a HIPAA main standard, because UPC codes are on bananas and just
whatever you can think of.  So, I don’t know if you need to name that
code set.  We are really talking about health-related items in some
manner.

Now, I don’t want to limit that, because we are talking about
diapers.  We are talking about certain things that a Medicaid agency
identifies as a necessary supply, that they will pay for it, but it’s
purchased through the pharmacy for the pharmacy benefit.

DR. COHN:  Let me try this one, because this sort of makes
sense.  NCVHS recommends that the department further investigate the use
of these specific code sets in the NCPDP transactions, and if appropriate,
initiate the rulemaking process.

Margaret, you are laughing.

MS. GILBERTSON:  Here is a comment.  Just to come full circle,
the pharmacy industry that is needing to use the HRIs and the UPC codes are now
under contingency plans, because that was the guidance given, continue doing
business.  So, without a rulemaking process or something, we’ll have
contingencies forever, heaven help us.  We need that step of the
rulemaking process or something to get it moving, because we have to get out of
contingency at some point.

So, I think you have backed into the fact that it’s got to somehow go
through whatever process OHS has to go through, because we can’t stay on
contingency forever.

DR. COHN:  Right, so you are agreeing about the rulemaking process.

MR. PACE:  Scott Pace with NCPA again.  My only comment to Lynn
was that I think she said it pretty well back in March when she related it to
the Kleenex.  Everybody picks up a Kleenex, whether it’s a Puffs or a
Kleenex, and calls it a Kleenex.  And I think that’s how the industry
is with these three code sets.  We just call it an NDC.

And it seems to me that maybe what happened in this case, it was called
NDC, and it is referring to everything, but in literal translation we are only
taking it as the National Drug Code as it applies to prescription products
approved by the FDA.

It seems to me that if there was just a clear definition of NDC as
interpreted by the industry, that would alleviate this.

DR. COHN:  I don’t think that’s going to be our solution
here.  Interesting, but I don’t think it’s going to quite work.

MS. EBHART:  Karen Ebhart with MediSpan.  I again agree with what
Lynn was saying.  The intent of this request here was to allow us to
continue doing business today.  This is the de facto standard that we are
using.  Just as we acknowledge that there are problems with the NDC, and
there is an NPRM in the future coming to help clarify that, we also know that
there are problems with these code sets that we would like to hopefully
identify and also resolve in the future, but that’s the future.

This is for business today.  So, for me, part of the investigation is
where do we need to go in the future to make it better, but also, we need to
acknowledge what we need to do to continue business today.  So, again,
just trying to clarify what you mean by investigate the code sets.
Investigate to make them better in the future?  Or just to include in the
NPRM that this is a valid way to build certain supplies into the pharmacy sets?

DR. COHN:  I think I need the assistance of the subcommittee here in
terms of how we want to go.  I think our intent here initially was to try
to create something that said the standard needs to go forward.  We
obviously think that there is a business case here, but want the department to
further investigate the use of these code sets, and moving forward with
rulemaking.

Yes, Margaret A.?

MS. AMATAYAKUL:  I have another suggestion, just because this is a
clear brain.  I haven’t heard this discussion before.  I
don’t know if this is right, but NCVHS recommends that the department
grant permission to continue use of the UPC and HRI codes, or whatever codes as
part of whatever full compliance with the NCPCP telecommunications standard
version 5.1.

MR. BIZZARO:  That’s exactly what we are asking.  Now,
I’m not sure I understand the rulemaking process, but that’s exactly
–

DR. STEINDEL:  And investigate the use of rulemaking if necessary.

DR. COHN:  Why don’t we take a look at it here?  You know
what I’m going to suggest for the moment?  It’s now 5:06
pm.  I’m going to suggest that we take a break for 10 minutes.
And then we’ll get back to this when we get back at 5:15 pm.
Let’s take a 10 minute break.

[Brief recess.]

DR. COHN:  Okay, everyone, I think we are ready to come back for our
final session.

I think when we were last talking, we were up to the last paragraph
here.  Let’s read the last paragraph and see if it meets
everyone’s needs.

The paragraph now reads, NCVHS recommends that the Department allow
continued use of UPC and HRI codes within the NCPDP telecommunications
standard, and investigate rulemaking if necessary.  This is important for
current business.  However, NCVHS acknowledges that there are limitations
in how the UPC and HRI currently are assigned and managed.  We recommend
the Department further investigate these issues.

NVCHS appreciates the opportunity to provide these recommendations.
Sincerely.

Comments from the subcommittee, questions?  I think we need to specify
telecommunications standard 5.1.

MR. REYNOLDS:  This is important for current business.  You could
add maybe a couple of words.

DR. COHN:  Maybe it’s important for current pharmacy operations?

MR. REYNOLDS:  Yes, I guess current business just kind of says you
know – if I was reading it, sitting in an office somewhere, that could say
they don’t want to change, and we’re going to let them be okay.
It’s really sustaining current business practices.  I don’t
know.

MS. FRIEDMAN:  Why aren’t you saying sustaining current business
practices?

DR. COHN:  I don’t know that that’s a – okay.

MR. REYNOLDS:  Did industry accept it?  How about industry
accepts it as business operations?

DR. COHN:  Okay, so this is important for sustaining current business
operations?

MR. REYNOLDS:  Well, the reason I say industry is –

DR. COHN:  Well, I guess the question is do we need this
sentence?  I don’t think we need the sentence.

MS. FRIEDMAN:  Do you want it gone?

DR. COHN:  I think so.  I’m looking to see if anybody thinks
it a –

MS. PICKETT:  Simon, I agree, because the information is in the
previous paragraph.  It’s already there in the previous paragraph.

DR. COHN:  Okay, any further changes, modifications,
wordsmithing?  Hopefully, the full committee will tear it apart in
mid-June, but you never know.

Do you want to take one last look at it here?

MS. PICKETT:  Well, just one question.  The word limitations, is
it limitations or something else?  And I pose that to Lynn and Tom and
Karen.

DR. COHN:  Well, what would you –

MS. PICKETT:  NCVHS is acknowledging that there are limitations in how
the codes are currently managed.

DR. COHN:  Are there issues?

MS. PICKETT:  I think it’s more like issues, as opposed to
limitations.

DR. COHN:  I guess maybe I should ask you, these two last sentences,
where it says acknowledges that there are problems with how the UPC and HRI
codes are assigned and managed.  We recommend that the Department further
investigate these issues.  Is that helpful to the department?

MR. LEVIN:  At least it says to me that current codes for how you deal
with supplies are not sufficient, not adequate, and need some help.  And
so, we have to look into ways to enter the codes for supplies.

MR. BIZZARO:  Because I wouldn’t want to give the impression here
that we are talking about UPC codes in general.  I would think that John
Roberts from UCC would probably argue the fact that there are problems with the
way that is assigned, because they do a really good job in managing those UPC
codes.

Our issue comes with abuse of the UPC codes, and representing bar codes as
UPC codes when they really aren’t, so I think there are some issues
there.  I think if you notice now on some linear bar codes on products,
you will see that there is either a UPC copyright, or UPC mentioned, that it is
a UPC code.  There are a lot of bar codes out there that don’t
necessarily represent themselves as actual UPC codes.

MR. LEVIN:  Then how about they are currently assigned and managed by
companies?

MS. FRIEDMAN:  As opposed to the government?

MR. LEVIN:  As opposed to the UCC.

DR. COHN:  Steve, did you have a comment?  Is it about the same
sentence?

DR. STEINDEL:  Yes, it’s about the same subject, because I
don’t see any other place in the letter where we mention anything
regarding limitations, or anything related to whatever word we want to use
there.  So, we are introducing a whole new topic in the end.  So, I
think we should either eliminate that sentence, or we have to add another
paragraph.

MS. AMATAYAKUL:  I guess my question is are you recommending that
further investigation be done after they accept your recommendation that
it’s okay to use these for now?  That this is really a recommendation
sort of in the future.  This is a temporary fix.

So, if that’s the case, I would separate those, as Steve is
suggesting, and then indicate that it is recognized there are issues associated
with this, and it is recommended that the department investigate these issues
in the future.

DR. COHN:  Though I think Steve has brought up a good point that we
say nothing about the problems prior.

MR. BLAIR:  Grant permission now.  Investigate whether rulemaking
is necessary after.

DR. STEINDEL:  All we are saying is that for the time being, we
recommend that the present process continue. If it doesn’t continue,
it’s going to be harmful for business practices, and we have already
stated that.

Then we go on and we say that the department should investigate this
further.  And at that point, the issue may come up concerning usage
limitations, or whatever word we want to use regarding these codes.

MS. FRIEDMAN:  Why don’t we move this sentence and say NCVHS
acknowledges, and move that up here with this other stuff related?  It
would get the idea that there are these issues, and then leave it as it is.

DR. STEINDEL:  What issues?

MR. BLAIR:  Yes, if you don’t say what the issues are, it’s
just like there’s a problem here, but we don’t know what it is.

DR. STEINDEL:  If I read this letter, and I’m saying that there
are issues in using this code, and then I read the next paragraph and they say
go ahead and use it, the question that is going to come to my mind, why should
I continue using it if there are problems?

DR. COHN:  Well, I guess there are many ways for us to handle this
one.  I’m sort of struggling, because I’m not sure that even
despite Donna’s wonderful 10 minute summary, that I feel that we have
enough to really talk very substantively about any limitations or issues.

Maybe we need to go back to where we talked about earlier, which was a
sentence saying we recommend continued use.  That we are going to go do
further investigation of the larger issues of supply codes, and leave it at
that.  Or we could just leave it even without that.

DR. STEINDEL:  Simon, my feeling is we can just leave it with we
recommend the department investigate this further, because I know the first
thing that the department is going to say is NCVHS hold hearings on it.

DR. COHN:  Oh, that’s true.  Knowing who responds to us,
that’s a likely outcome.

MS. EBHART:  I think again, the problems we were trying to acknowledge
here is more the business practices of how UPCs are assigned and managed by the
manufacturer organization.  There is not a problem with the Uniform Code
Council, of how they administer their portion of it, but the fact that UPCs can
be reused.  Like we have all talked about NDCs, and we want to get rid of
that aspect.

And what constitutes when you can reuse and when you can’t reuse, how
far of a change there is.  As well as clarifying to make sure that we can
never have duplicates between HRIs and UPCs and NDCs, to make sure that we
don’t have those number collisions that we were all very concerned about
with the NDC.  This was more the business practices that was in that
problem scenario; just trying to clarify why that sentence has been added.

DR. STEINDEL:  I feel that there is an issue, that there is something
we need to discuss.  The question is if you want to resolve being
recommended as continuing, I think that’s something to leave out at this
point, and let them make a decision on continuing, because they will come back
to NCVHS, and we’re going to have to confront the issue.

MR. REYNOLDS:  I move we approve it as it is.

DR. COHN:  Okay, is there a second?

MR. BLAIR:  Second.

DR. COHN:  Further discussion?  All in favor say aye.
Opposed, nay.  Abstentions?  Okay, the letter passes.

[Whereupon, the letter was unanimously approved.]

Thank you all for your help on this.

It’s 5:30 pm at this point.  We meet again tomorrow at 8:30
am.  Thank you for your help.  It’s been a long two days, and
we’ll see you in the morning.

[Whereupon, the meeting was adjourned at 5:30 pm, to reconvene the
following day, Thursday, May 26, 2004, at 8:30 am.]