Testimony of the American Academy of Family Physicians

Before the National Committee on Vital and Health Statistics

On Personal Health Records

Presented by
David C. Kibbe, MD MBA
Director, AAFP Center for Health Information Technology

April 26, 2005


Thank you, Dr. Cohn, for the opportunity to testify before the National Committee on Vital and Health Statistics’ Subcommittee on the National Health Information Infrastructure and the role of Personal Health Records (PHR).  I am here on behalf of the 93,700 members of the American Academy of Family Physicians (AAFP).  The Academy is very interested in issues of health information technology, particularly in seeing software and hardware that is compatible and interoperable across health care settings, and our members have a particular interest in the relationship between electronic health records (EHRs) used by them in their practices and electronic patient health records (PHRs) accessible to or in the actual possession of their patients.   In preparation for today’s testimony I related many of the questions that I was sent by NCVHS on this subject to the nearly 1,000 family physicians on our EHR email listserv, and received a large number of responses, which will, I hope, help inform the Committee’s deliberations.

I have also drawn upon the EHR and PHR vendors who are among the eighty-five (85) companies in the AAFP’s Partners for Patients program, which is an innovative forum based on the four ACID test principles agreed upon by the AAFP and these vendors as underlying further adoption of health information technology (HIT) at the medical practices level, namely affordability, compatibility, interoperability, and data stewardship.  Many of these vendors either already offer some form of PHR, or are in the process of developing a PHR feature to their products and services, largely in response to market demand.

I have organized my testimony today into three parts.  The first will deal with a consideration of the definition of PHRs.  The second will lay out what we believe are several of the underlying important technical issues for PHRs to be of value to family physicians and their patients, and provide examples of how PHRs might be used by patients.  And the third will discuss the readiness of AAFP members to make use of and deploy PHRs as part of their medical care processes.


As you are certainly aware, the term “electronic patient health record” means different things to different people, a situation that is a function of their novelty and the rapid pace of change in the HIT environment.  In my experience as Director of the Center for Health Information Technology at the AAFP, the term PHR is used to describe at least the following items:

  1. PHR as stand-alone software.  Patient Health Record software residing on a laptop or desktop computer, and capable of functioning in a stand-alone mode, which assists patients to collect, organize, and assemble reports on their personal health information, is one common meaning applied to the term PHR.  There are several vendors of this kind of product.
  2. PHR as web-based software application.  This is probably the most common form of PHR in use today.   The idea is that patients gain access to the application from a web browser, using an ID and password, and then enter and store their personal health information into a remote database application, which they later call up and view, add to, and amend at any time.   This is sometimes referred to as the ASP or Application Service Provider model.
  3. PHR as a view of medical information.  A Patient Health Record that exists as a view of health care data, most often in a web browser, is another meaning applied to the term PHR.  In this case, the web-based file or record may be a subset of data assembled from a database or databases containing diagnoses, medications and dosages, allergies, etc., by some entity, such as a hospital, health plan, or regional health information organization. The patient may or may not be able to add to or amend the viewed data: in most cases not.
  4. PHR as file or record.  Finally,  PHR may also refer simply to a file, or computer record, either in electronic form or on paper, of a summary nature and consisting of patient health data from one or more EHRs or databases that contain patient health information, e.g. encounters, lab results, medications.   The ASTM Continuity of Care Record, CCR,  is an example that forms the basis of several PHRs, and is the output of numerous EHRs.   Similarly, the printed or file output of any of the PHRs in categories 1-3 above might be considered “a PHR.”  In this case, the PHR is a “snapshot” of the patient’s health at any given moment in time, with data elements conforming to logical categories such as “patient identifying information,”  “insurance information,”  “medication list,” etc., which can be updated each time a new instance of the PHR file is created.

Of course, the real world doesn’t line up in neat categories, and there is considerable sharing of these characteristics in the instances of PHRs that currently exist.  I am certain that the Committee members will be familiar with these instances of PHRs, and perhaps other categories that I’ve not identified here.

Uses of PHRs and Several Technical Issues

What all these categories of PHR have in common is the potential benefit they offer to patients and their care givers should the information they contain and its use allow us to make better clinical decisions, to avoid duplication and error, and to provide better self-management of diseases or conditions, or their prevention.

Let me make the point, one that we at the AAFP feel is paramount in importance, that this potential benefit is most likely to be realized if the clinical content of PHRs is standardized, and if the various types of PHRs are made transportable and interoperable, not only with each other but with electronic health records used in doctors’ office and hospitals.    In the following comments, I will lay out several characteristics regarding an  “ideal PHR,” and then illustrate how these might lead to patient benefits in terms of quality, cost, and convenience of medical care.

First, let us assume that the ideal Personal Health Record, or PHR, consists of important and summary health information about the individual, information such as problems and medical diagnoses, medications with their dosages and frequencies, allergies and adverse reactions, advance directives in case of impending death, and, of course, personal demographics and insurance information.  Content is an important issue, because the PHR can’t be useful if is just any information about a patient, or if it is all  the health information about a patient.  We must be selective and base our selection on what physicians, nurses, and patients have as the highest priorities for the content of a PHR, based on reasonable uses and use cases for that information:  it must be the most relevant health information.

Secondly, let us assume that this information is available to patients or health consumers in standard format, either electronically or on paper, and can be easily obtained from their personal physicians or their usual setting for receiving medical care, the clinic or practice they consider as their medical home.

And finally, the ideal PHR will be expressed in XML, an important Internet standard markup language, used extensively throughout the banking, financial services, and retail trade industries for interoperability between multiple computerized systems, allowing for efficient exchange of information across different technologies and networks, all of which accept the Internet protocols such as TCP/IP.

One thing that patients in possession of their PHR might do immediately is seek out caregivers on the basis of their ability to use this new and convenient method of transfer of key personal health data from one place to another.  Why would patients tolerate the traditional wait and bother associated with doctors and nurses, administrators and assistants, who ask the same basic questions over and over again before transferring this information to paper, when they could instead locate health care personnel and institutions capable of instantly “knowing and recognizing me” by use of the PHR?      Patients and health consumers cannot always judge medical quality of a technical nature, but they are more than able to appreciate service that is respectful of their time and a willingness to offer convenience.  What could be more demonstrative of such respect and customer-service in health care than the ability to instantly accept and be ready to manage the patient/customer’s personal health information?    Who among us would open a new bank account with an institution that did not offer the convenience of ATM machines or online banking services, a chief component of which is nearly instantaneous access to our summary personal financial information?    If doctors and hospitals begin to use electronic health record systems that can accept and exchange PHRs that are truly standardized and interoperable, does anyone doubt that this would be a valuable differentiator in the market for health care services?

But there is much more that patients could do with their PHRs.   Patients with multiple chronic illnesses would use the PHR and its informational contents to access personalized alerts, reminders, and news pertinent to their conditions – all from their home computers, even hand-held devices.  My de-identified information from the PHR could be uploaded to a “health information server” (maintained by NIH or the AAFP) that would then scan the diagnoses, medication list, allergies, and other pertinent information, sending me back suggestions for ways to best manage my conditions and citing resources in the literature that provide the evidence behind such recommendations.  Some of this could be very simple, for example notification that a hemoglobin A1C test is overdue for a patient with diabetes; that a mammogram ought to be scheduled within the next three months; or that two medications on my list carry the risk of drug-drug interactions of which I might not have been made aware, perhaps because the medications were prescribed by two different providers. (!)    Some uses of the information may be more complex, requiring opt-in by the patient, such as notification of research studies and clinical trials that are seeking participants.

If all this sounds futuristic, let me remind you that computer systems perform this kind of analysis routinely between remote users and Internet-based servers in non-health care fields.  My home computer regularly “checks in” with other computers for the purposes of updating software programs, performing scans for virus and spyware code, and tracking new musical offering.  The tools are there.  Why not use them for better health care for the individual?

Imagine, also, the following scenario:  as my patient, you’ve come to me with symptoms of joint pain and swelling.  After a couple of office examinations and appropriate lab testing, we’ve come to a working diagnosis of rheumatoid arthritis, a chronic illness that is both highly treatable but carries some risk of future debilitation.   As a busy family physician, I’ve got about 20 minutes to discuss the illness and our treatment plan, providing you with a new prescription, next appointment, and educational materials, before sending on your way.  At home you are almost certain to turn to the Web and perform a Google search to learn more about your new chronic condition.   You type in “rheumatoid arthritis” and – voila! – you get 1,830,000 pages in return.  Some of it is useful, some of it is simply bogus, and much of it is advertising.  Is this really helpful?  Is it really personal, or relevant to you?

Now, rewind the video camera.  In addition to the appointment and new prescription, you leave my office with an updated PHR which contains your new diagnosis, coded appropriately, and the names of the medication you’re about to start taking.  When you get home, you upload your PHR (de-identified, of course) into the new Google Health and Medical search engine, a joint venture between the creative geniuses at Google, NIST (National Institute of Standards and Technology, and the National Library of Medicine.  Nowyou do your search for “rheumatoid arthritis” using some of your personal health information as a filter.   Your age, gender, medications, and allergies – perhaps even a laboratory test indicating genomic predispositions towards ineffectiveness of certain classes of medications – provide the search engine with elements for inclusion and exclusion of retrieval.  And voila!  This time you get back 8 pages of highly personalized, highly relevant, and very helpful information.  Completing the loop, you now have a new set of questions to ask me, via email, or course, before your next visit.

Please note that in my remarks I am repeatedly stressing what the individual patient can do with his or her own PHR and the information it contains, with the assistance of his or her professional caregivers.  Note that I am not emphasizing what doctors, hospitals, and health care corporate institutions – nor any of their suppliers and partners — can achieve, or what will help them to meet their own agendas relative to information technology.

Readiness of Family Physicians to Create and Use PHRs

Are family physicians ready for PHRs?   And how do family physicians expect to utilize PHRs?  To answer these and other questions, we solicited comments from family physicians around the country through our email listservs, the largest of which is the EMR Discussion List with over 700 members.   Many of these family physicians currently use EHRs in their practices, and many others are looking to purchase EHRs.  I received over fifty responses, and I will attempt to summarize them here, with some specific examples.

These family physicians generally expect PHRs  will reach wide-spread use over the next few years.   For one thing, they recognize that the HIPAA Privacy Rule gives patients the right to request and receive a copy of their medical records under most circumstances, and the PHR, particularly if standardized with respect to content and format, may make responding to patient requests for their records both easier and more affordable.  “PHRs are coming, and will be the next big thing the public demands,” was the way one AAFP member expressed his expectation.  Another said this:

In anticipation of eventually sharing PHI electronically, I am now in the habit of printing out a copy of the visit and a medical summary at the end of each encounter and physically handing it to the patient. I encourage everyone to consider doing this now as it is an interesting transition (knowing the patient is now in regular, direct ownership of their record via HIPAA). Overall, I prefer this approach, as it tends to encourage more patient accountability.

Family physicians already experience widely varying demand from patients for email el access to them and their practices, and expect to experience the same variability in demand for PHRs.  A growing number of family physicians offer patients the use of web portals which typically provide secure email and additional electronic services, such as online scheduling and billing, and several of these patient portals now include a view of the patient’s health information.  Some of these physicians find that patients can pre-populate online forms for registration, demographic, and insurance information, along with a description of symptoms or reasons for a visit, and that doing this prior to the appointment saves times for both doctor and patient.  However, demand for these services is growing slowly, perhaps because patients have very little direct experience with these tools.  One AAFP member put it this way:

My patients demanded several years ago that I use the Internet and World Wide Web, because they did.  I don’t think we’ll see widespread offering by doctors of PHRs until patients start coming into our offices with their health information on USB drives saying, “here, you use this instead of asking me the same questions over and over.”   We’re not there yet, but I can see it coming.

In our informal poll, family physicians expressed mixed feelings, even forebodings, about patient control of the information in a PHR originating from the doctor’s notes.   The following was a typical response:

I think it is imperative that the patient be in possession of his/her Pertinent health information, and software allowing management of the diagnoses, medications, allergies, immunizations, family history, hospitalizations, surgery, lab values, etc would be most helpful.  However, there is a portion of the record indicating the logic used to arrive at certain diagnoses, conclusions, and opinions which is the functional equivalent of “intellectual property” of the physician and does not “belong” to the patient. This might include privileged information, data provided in confidence by family members, subjective observations, psychiatric musings, etc. Therefore, it would be a very bad idea to allow unfettered access to the entire record. The trial lawyers would love that – it would open up a whole new arena for “mining for dollars!”

And from another family physician, this opinion:

We all have patients that would be quick to alter their PHR to fit their agenda.  If a patient uses an independent PHR to track medical information, then it is only as good as the information the patient enters (which they can change at any time).  However, a PHR hosted by a physician’s practice is a different story, especially if the patient prints it out and takes it to another physician.  In this case it would be best that the PHR reflect medical information entered and/or altered only by the provider, with the patient having viewing and printing privileges.  I think it would be very important for any printed report or viewable screen to clearly identify the source of the information.

Family physicians also see value in PHRs for reasons associated with interoperability.  For example:

I think PHRs are a very good idea. In fact, I think this is a better pathway to the goals of interoperability than asking all of the different EMRs [vendors] to be able to use a common language to talk to each other. I say this because I think the best way to achieve the stated goals of interoperability such as improved safety and saving billions of dollars is to make the information patient centric, not doctor centric.

I see PHRs as the way to do this.  EHRs can contribute to and read from a PHR that follows the patient. If we have a few PHRs with standardized communication language, then the EHRs can write and read to those. Asking them to do this with each other is ridiculously problematic and runs the risk of wiping out the smaller, more affordable priced vendors if EHR interoperability becomes part of any CCHIT certification. Plus, it’s a political issue in that some vendors see EHR interoperability as a way to allow for the swapping of vendors and other vendors don’t want it because they want to protect their turf.

Clearly, then, at least a portion of the AAFP’s active membership anticipate the widespread use of PHRs, attach value to this trend, and consider PHRs an extension of existing e-health services such as email, EHRs, and telemedicine.    There remain many unanswered questions:

  • Who will update, renew, or obsolete the information contained in a PHR that is in the possession of a patient?
  • What privacy and security challenges will the widespread adoption of PHRs present for patients, their doctors, and data aggregation entities?
  • Will we be able to agree on a single standard for the expression of summary health information in a PHR, or will there be multiple technologies?
  • What is the relationship between the PHR and new “consumer driven” health care innovations such the Health Savings Account?