Medicine and Information

Some health care reformers and most presidential candidates put electronic medical records high on their list of needed improvements for the U.S. health care system. Last year, when I switched primary care physicians, I had to pay $35 to get the assistants in my old doctor's office to make photocopies of my records in the manila folder, which I then hand-carried to my new doctor's office, where they were placed in a new manila folder. Moving to electronic recordkeeping would certainly make the process easier and my records more easily accessible.

But there's an old saying from the early days of the computer era when mainframes dominated the landscape that may be applicable here. Garbage in, garbage out, programmers used to say about the data poured into their hulking machines, and that may be worth contemplating as we move toward electronic medical record keeping.

Two new reports highlight the information deficit in our medical system that electronic medical recordkeeping may give us a tool to address, but doesn't in and of itself resolve. The Institute of Medicine last month called for a new national organization to conduct systematic reviews of medical evidence and write objective clinical practice guidelines (CPGs) for use by the nation's medical practitioners. "Knowing What Works in Healthcare: A Roadmap for the Nation" said a centralized authority for writing guidelines was necessary because:

With thousands of new clinical studies published every year, the amount of medical data has become so vast that it is essentially unmanageable for providers, patients, health plans, and others. Most people, including many health professionals, lack the scientific training necessary to evaluate and interpret such clinical findings by themselves. Moreover, research has shown that when evidence reviews are financed by manufacturers or vendors — as a significant proportion are — they are more likely to show effectiveness, which leads some to question whether, or to what extent, the cumulative body of evidence for any given health care product or service is biased.

But when this new agency sets out to gather the available evidence to write its CPGs, what will it find? The problem of generating real-time information from our clinical trial system was one issue addressed by the business-backed Committee for Economic Development in its new report, "Harnessing Openness to Transform American Health Care." Its perspective on the clinical trial information system, which ultimately is what clinical practice guideline writers must depend upon, is worth noting:

Most important for improved healthcare research is to make the trial results and the data underlying trials more accessible more quickly in a form that is searchable and computable using common standards. In addition, applicants for FDA approval should be required to submit all studies they have conducted on the intervention with any safety-related results being made publicly available. . . The Council joins a long list of groups recommending far more comparative testing of drugs to supplement clinical trials that simply compare an intervention’s effects with those of a placebo.

In theory, the recently enacted Food and Drug Administration reform law dealt with the first part of this problem. All clinical trials beyond the earliest safety trials are now going to be registered, and their results publicly posted. But will they be in a form that is easy accessible to desktop researchers across the country, indeed, across the globe? And when will the FDA publish its standards to bring clinical trialists, whether working in the public or private sector, into compliance with the new law's public posting requirements? I haven't seen any proposed regulations from the agency. Hopefully, it will get off the dime and publish them soon.