In this issue of JCH, Kinn and colleagues1 provide a brief insight into two important issues whose interrelationship has been ignored for too long: the inability of the majority of practitioners to meet national standards in chronic disease management and the barrier caused by suboptimal record-keeping in this failure. While not a strong study from a scientific perspective, what makes Kinn's report potentially significant is the fact that it arose from a “real-world” group practice environment. It describes a system that has potential to improve care quality simply through better information formatting.

The principal conclusion, that hypertension management was better with physicians who chose to employ electronic medical records (EMR) compared to those who did not, could be questioned on scientific grounds because there are major flaws in the design and implementation of the study. The largest flaw is the probable selection bias that arose in the small number (three) of self-selected physicians who actually used the EMR. It is typical that “early adopters” of change become more invested in the process and the outcomes and tend to work harder at perfecting the process. Based on the numbers supplied, the three EMR physicians may also have routinely treated a greater number of patients with hypertension than their traditionalist colleagues, and thus may have had greater expertise in this area. This hypothesis is borne out by the greater diuretic use in the EMR group. A second major flaw is the study design, which employed a historical control population rather than a randomized trial design. It is likely that care patterns and attitudes drifted during the observation periods. Third, the specific issues of how the EMR system was employed were not adequately described, so it is difficult to assess the specific contribution of the EMR in daily management. Under most circumstances, the severe structural deficiencies in the current report would be grounds for editorial rejection. But the near-complete absence of studies in the area of record-keeping and quality improvement confers special-case status. It could be argued that the study is particularly useful because of the choice of hypertension as the model disease.

Hypertension is an ideal model disease in that it meets virtually all criteria for effective quality-assurance monitoring: it is extremely common; there is a ubiquitous, trackable end point (blood pressure); there is a clearly defined, widely accepted, evidence-based practice standard (JNC VI); the impact of good care is clearly proven; and the existing level of care is substandard. Given these characteristics, hypertension may be the best single disease model in health care delivery research. The most promising research in the area of hypertension records is occurring within the Veteran's Administration system where investigators at Stanford University have begun a comprehensive study of an EMR-based hypertension management system.2

Ultimately, results of the present study may be simply a manifestation of the well known “Hawthorne effect,” which states that close observation of a process automatically tends to improve its outcome. But even if the present study is simply a Hawthorne-type effect, the results are potentially important because what the EMR may have actually provided is a way to sustain the Hawthorne effect, which otherwise would be expected to wane with time. Can the present results be generalized? There is definitely a need to try because the ancient medical “progress note” system still employed by virtually all physicians is itself an extremely formidable barrier to effective chronic disease management. Progress notes are time-consuming and typically do not provide trackable information that automatically guides decision-making. Most importantly, in a multiple-provider environment, critical information is often missed as “coverage” providers sort through endless unintelligible progress notes (or simply give up).

I have achieved better care in my own patients using a standardized “spreadsheet” that tracks recurring prescriptions (rows) by date of service (columns). In a second part of the spreadsheet, values for blood pressure, cholesterol, glycosylated hemoglobin, albumin excretion, or other critical indicators (rows) are also charted by date of service (columns). This spreadsheet appears in the first section of the chart opposite the problem list. In this system, standard progress notes are still included in a separate section of the record, largely to explain the rationale behind any changes in medications or disease status. I have tested the spreadsheet system in the training of internal medicine residents in ambulatory care clinics and have found lower blood pressure and cholesterol levels in the patients of the residents who actually use the spreadsheets, compared to those who do not. Our experience with paper-based flow charts suggests that the critical issue is the ability to visualize data and analyze trends in real time. In this way, the likelihood of continued inattention to the level of blood pressure is strongly diminished and therapeutic changes are made earlier. This system is also easily adopted to an EMR environment.

Medical record-keeping has changed in the past few years, but largely as a response to potential litigation or loss of reimbursement. The potential for improved records to allow improvements in ambulatory care quality for chronic diseases such as hypertension has simply not been addressed. It seems eminently logical to propose that simple, cost-effective measures, such as improved records, can improve care quality and efficiency (and therefore cost). Given the current poor blood pressure control rates and the staggering cost of care of conditions such as hypertension, the most interested parties should be governments, insurers, and health maintenance organizations. These institutions have done very little to date to foster effective health care delivery research, especially in medical records. The problem starts with the recognition that a record-keeping problem exists. Will the wake-up call to the responsible parties begin here?


  1. Top of page
  2. References
  • 1
    Kinn JW, Marek JC, O' Toole MF, et al. Effectiveness of the electronic medical record in improving the management of hypertension. J Clin Hypertens. 2002;4:415419.
  • 2
    Goldstein MK, Hoffman BB, Coleman RW, et al. Implementing clinical practice guidelines while taking account of changing evidence: ATHENA DSS, an easily modifiable decision-support system for managing hypertension in primary care. Proc AMIA Annual Symposium. 2000;300.