ACP American College of Physicians - Internal Medicine - Doctors for Adults

Effective Clinical Practice


Is This Issue a Mistake?

Effective Clinical Practice, November/December 2000.

Elliott S. Fisher, H. Gilbert Welch

For author affiliations, current addresses, and contributions, see end of text.

This issue of ecp is focused on the current hot topic in medicine: medical error. Although there have been physicians concerned with medical errors for years, the release of the Institute of Medicine report, "To Err Is Human," triggered a great deal of publicity and launched a medical movement. This movement has garnished sufficient attention to warrant critical review (see Primer). We hope that the papers in this issue help highlight both the opportunities and the complexities involved. In this editorial, we touch on two mistakes made by many in the error movement: loose language and missing the mark.

Loose Language

To most of us, an error is a screw-up. The word connotes unambiguous culpability: Someone is to blame. In the error movement, however, individuals are not to blame--it's the system that's at fault. The term adverse event, on the other hand, probably has no immediate meaning to most people. It is perhaps best translated as "something bad has happened." Although error and adverse event mean different things, the terms are often muddled in the error movement, and the public may be led to believe that eliminating errors will eliminate adverse events. Let's be clear: It won't.

The reason is that many adverse events are expected--that is, the use of diagnostic tests, medications, and surgical interventions all lead, at a predictable rate, to false-positive results, adverse reactions, complications, and death. It is true that some of these events--the so-called preventable adverse events--might have been prevented if something different been done. But this kind of preventability often depends on both hindsight and a model of single-factor causation (e.g., "If I hadn't anticoagulated him, then he wouldn't have bled" or "If I had anticoagulated him, then he wouldn't have had a stroke"). Most of the preventable adverse events in medicine bear no resemblance to the big screw-ups that are typically held up as examples of errors (amputating the wrong leg in surgery, administering 10 times the normal chemotherapy dose).

Finally, there is the word death. All deaths are not the same. Talking about death in the context of airplane crashes or car accidents engenders particularly poignant images--active, healthy people experiencing sudden, unexpected death. This is how the publicists communicate the magnitude of the medical error problem. Talking about death in the medical care system this way, however, is at best misleading. Many people die, most after contact with the health care system. Sometimes their death is hastened by medical care (unintentionally), and how often this happens is a subject of legitimate debate. But the impact of deaths from errors on overall life expectancy is small. To equate this cause of death with the traumatic death of healthy individuals is disingenuous.

Missing the Mark

Despite the loose language, the error movement has at its core some powerful ideas about how to improve care and has achieved demonstrable success in some areas. In particular, two old ideas are receiving well-deserved attention: increasing the use of computerized order entry and concentrating complex procedures in high-volume centers. These are pragmatic and effective approaches to improving medicine's "production lines." But restricting the notion of safety simply to errors of execution (failure of a planned action to be completed as intended, such as failure to give the appropriate drug dose) misses the mark. We need to expand the focus to errors of decision making (referred to in the IOM report as "errors in planning") and consider the possibility that the error movement is distracting us from other important problems.

Errors in decision making are widespread. Consider the overuse of increasingly sensitive diagnostic tests: Minute elevations in troponin become heart attacks, subtle changes in density become strokes, small collections of atypical cells become cancer. If, in response, we increasingly use the hospital, the cath lab, the operating room, the chemotherapy suite, and the radiation facility, we will subject people to real risk with little, if any, opportunity for benefit. Also consider that many surgical and medical decisions involve complex trade-offs between risks and benefits, particularly as patients near the end of life. If we fail to adequately incorporate patient preferences we will pursue the wrong course. Paying careful attention to the proper delivery of a test or treatment that is either unnecessary or unwanted misses the mark.

The error movement also risks distracting us from other tasks. It feels like we have less and less time during a clinic visit to talk about what matters to patients. To the extent that the patient safety movement encourages us to check every potential drug interaction and ensure that every diabetic gets an annual eye, foot, and urine examination, it risks distracting us from the core of our work--caring for patients. At the policy level, we risk ignoring other important challenges. Are errors more important than figuring out how to provide health coverage to the 44 million uninsured? What about conflicts of interest within the profession?

A skeptic might ask, indeed, whether the error movement doesn't serve the interests of established stakeholders very well. It reinforces a public belief that everything is fine except for a few technical glitches that the system can virtually eliminate. Once the system is safe, everything else will be OK. If only medicine were as simple as building a cell phone.


Elliott S. Fisher, MD, MPH, Center for the Evaluative Clinical Sciences, Dartmouth Medical School, Hanover, NH 03755; telephone: 603-650-1822; fax: 603-650-1225; e-mail: