To the Editor:

Although I commend your journal for highlighting the importance of computerized patient records (CPRs), I believe your vendor survey (“Computerized Patient Record Systems: A Survey of 28 Vendors,” November/December 1997) does not help physicians understand what's really important when selecting a CPR system. I am a user of HealthPoint ACS, and I chose this system for reasons not adequately represented in your survey.

The composite ratings in the article are a major concern to me because they are based on an equal weighting given to a long list of CPR features. For example, a key long-term benefit in implementing a CPR system is the ability to do outcomes reporting to find efficiencies and improve care. This is impossible without a standardized data dictionary and a structured database. Yet in your survey these critical components are weighted equally with far less important features, such as drawing tools.

As a clinical physician in family practice for 12 years, I know it is critical to see products firsthand for proper evaluation, as the authors stated they did not do, and to look at each company — its stability and its ability to continue developing and upgrading its products. Although HealthPoint has one of the newer products, it is supported by two strong partners (Glaxo Wellcome and Physician Computer Network) and has 12 years of product development. Its product is strong now, and I'm betting it will be around awhile.

The survey covered lots of bells and whistles that would be nice to have in a CPR, but let's focus on what's most important.

To the Editor:

Having undertaken survey work myself regarding CPR systems, I have a keen sense of the tremendous commitment of effort and energies that clearly went into the preparation of your CPR survey.

I do, however, have two critiques. First, the survey offered no hierarchy of importance for weighting its criteria. This implicitly flattens the field by equating the importance of one attribute with the importance of every other attribute. And second, far more rigorous criteria needed to be included and applied. Such criteria are readily available through the National Academy of Science's Institute of Medicine (IOM) landmark study, The Computer-Based Patient Record: An Essential Technology for Health Care, for which I served as study director.

For the past three years, William F. Andrew and I have published surveys using just a few of the criteria we have established to measure the key attributes of nearly 100 CPR systems. This year, we have sent the survey to more than 200 vendors and have focused considerable attention on two areas:

• The importance of utilizing at least one of the emerging controlled-vocabulary systems or tools,

• The importance of having structured databases that can support the new demands for documentation, including E/M codes. Without a structured database, it is impossible to design a system that will assess appropriate E/M codes for patient encounters. Physicians who do not have robust CPR systems will find it difficult to comply with Medicare's rigorous documentation guidelines and may be undercoding to such an extent that they will be leaving substantial sums of money on the table. Nowhere in your survey was this crucial issue foregrounded.

I fully agree that system assessments are necessary and important; however, given the gravity and consequences of the decision to procure one particular vendor's system and not another, such surveys must be adequate to the task.

Authors' response:

We expect Dr. Van, as a user of HealthPoint ACS, to be an advocate for this product. Naturally he would select a product that has features he believes are most crucial in a CPR.

The structure of our survey was based in part on the IOM report and subsequent additional criteria that Dr. Dick and Andrews developed. We acknowledge their seminal work in the field and look forward to reading the results of their forthcoming survey. Our survey was focused on systems for primary care physicians in office-based practices and by design was limited in its scope of questions and vendors surveyed.

Composite or hierarchical ratings are always a matter of debate. For this reason, we provided extensive tables containing vendor responses to each question in the survey. This format allows health care providers and managers to select systems based on their individual beliefs about the relative importance of system attributes.

Given the low penetration of CPR systems in clinical practice and the fact that many health care providers have very limited knowledge about them, we believe it is important to disseminate this information. As we stressed in our article, we believe that our survey is a starting point for those looking for a CPR system, not a definitive recommendation of which systems to choose.