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

Effective Clinical Practice


Media Mistakes in Coverage of the Institute of Medicine's Error Report

Effective Clinical Practice, November/December 2000.

Susan Dentzer

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

Perhaps it was only fitting that news coverage of the Institute of Medicine's (IOM) report on medical errors should start with a slip-up. When the IOM distributed its meaty 223-page report last year, it imposed a routine press embargo to give journalists sufficient time to review the document carefully before the contents were made public. But NBC's health correspondent, Robert Bazell, obtained a copy ahead of time, outside of the normal IOM distribution process. Feeling free to ignore the embargo, he broke the story on November 29, 1999, two days before the embargo was to have been lifted. As a result, the report's dramatic conclusions--especially the projection that anywhere from 44,000 to 98,000 Americans would die in 1999 from errors in hospitals--led that evening's NBC and ABC evening news broadcasts, even though the news conference at which the IOM was planning to unveil the document was still two days away.

By the next morning, November 30, coverage of the report had crested to tsunami-like heights. The IOM report got front-page treatment in The New York Times, The Washington Post, USA Today, and other major daily newspapers. It was featured on all three network morning-news shows, in multiple reports on wire services, and in all three major national news magazines the following Monday. Numerous news outlets abroad, such as The Sunday Times of London, also picked up the story. Within days, perhaps as many as 100 million or more readers and viewers in the United States alone were exposed to the rash of pieces about medical errors (Table 1).

This saturation coverage helped pave the way for the many official responses that followed--including congressional hearings and a push by the Clinton administration to incorporate error-reduction efforts into public health insurance programs like Medicare. But although the news coverage was widespread, little of it went deeply into the report's analysis of the sources of medical errors or some of the report's highly controversial recommendations. As a result, many news accounts were seriously flawed in several important ways.

First, in the vast majority of the coverage, there was an undue focus on the numbers of likely deaths from medical errors, and this tended to give the projections a misleadingly totemic significance. Second, much of the news media simply equated medical errors with malpractice--perpetuating the notion that most of the bad stuff going on in medicine can be attributed to the negligence of incompetent professionals. Third, many journalists never moved beyond this blame game to a broader understanding that many errors result from systems failures, which are amenable to systemic solutions. Collectively, these media mistakes and misjudgments may have led the public to draw false or simplistic conclusions about a serious problem that is already proving difficult to fix.

Cult of Numerology

Probably the single most important factor in the saturation coverage was how the media conveyed the sources and method used in the IOM report to estimate the number of Americans killed by medical errors in hospitals. These numbers were derived from data from two large previous studies of adverse events--a then-unpublished study of adverse events in Colorado and Utah and the well-known Harvard Medical Practice Study of adverse events in New York hospitals published in The New England Journal of Medicine in 1991. As explained in the executive summary of the IOM report:

When extrapolated to the over 33.6 million admissions to U.S. hospitals in 1997, the results of the study in Colorado and Utah imply that at least 44,000 Americans die each year as a result of medical errors. The results of the New York Study suggest the number may be as high as 98,000. (1)

When this paragraph was distilled in the popular press, however, all subtleties were lost. First to go was the lower-range estimate of 44,000; in follow-up stories, even reporters for The New York Times and The Washington Post referred only to the high-end estimate of 98,000. Only a handful of news stories clarified that both of these estimates were based on extrapolations from the Colorado-Utah and New York studies, at least one of which was 15 years old. Nor was it widely noted that these numbers at best refer only to estimated deaths in hospitals, rather than to total deaths from errors throughout the health care system--a significant omission now that more than half of all surgical procedures takes place outside of a hospital setting. (2)

Why did the numbers, and particularly the higher estimate of 98,000, become the definitive truth as far as the press was concerned? In part because the IOM's own language sometimes conjured an unreasonable sense of certainty. For example, the executive summary asserted without qualification, "More people die in a given year as a result of medical errors than from motor vehicle accidents (43,458), breast cancer (42,297) or AIDS (16,516)." Yet it's also true that the news media can be all-too-willing customers for anything that smacks of numerical precision, especially if the numbers are bad. After all, it's a lot easier to sell an editor on a story about something that kills almost 100,000 people a year than on something that kills or injures a tiny fraction of that.

What seems especially strange about the news media's willingness to embrace the numbers as definitive truth is that in individual interviews, IOM panel members readily admitted that no one knows just how widespread deaths or injuries from errors really are. For example, Janet M. Corrigan, director of the IOM's Quality of Health Care in America Project, has pointed out that the enormous rise in prescription-drug use suggests that injuries from medication mix-ups could be increasing in ways that would drive the overall injury numbers even higher. On the other hand, writing in the Journal of the American Medical Association in July 2000, researchers from the Indiana University Center for Aging Research dismissed the IOM's estimates of error-related deaths on the basis of a critique of the original Harvard Medical Practice Study. (3) This debate over the accuracy of the numbers shows no sign of abating. For example, IOM panel member Dr. Donald Berwick has analyzed the Indiana group's critique and has concluded that this critique itself is erroneous (Personal communication).

Having earlier swallowed whole-hog the IOM's projections, the news media just as quickly jumped on the story of numbers backlash as soon as the Journal of the American Medical Association article appeared. "Medical errors not so deadly?" read part of the headline of a July 6 piece in USA Today. (4) Meanwhile, IOM panel members voiced frustration that the controversy was drawing attention away from discussions of the causes of medical errors and the potential solutions. As panel member Lucian Leape, MD, an adjunct professor at the Harvard School of Public Health, responded in the same USA Today article, "Is it somehow better if the number is only 20,000 deaths? No, that's still horrible, and we need to fix it." (4)

Bloopers, Blunders, and Bungles

One subject that tripped up much of the news media was just what was meant by the word error. The IOM tried to put some definition on an admittedly hazy area, making it clear in its report that the word was merely shorthand for a broad range of potential adverse events. These included events that were deemed preventable and others that weren't: diagnostic errors, including everything from delays in diagnosis to failure to act on test results; treatment errors, ranging from mistakes made in very sophisticated surgeries to administering wrong doses of a drug; and miscellaneous other types of errors, such as communication or equipment failures. Adding to the confusion was the use of still other terms, such as references to patient safety. To the uninitiated media consumer, this might have implied that a medical error could include a patient falling on the wet floor of a hospital bathroom.

The confusion took its toll on the news media, which tended to deal with the muddle by reducing the word error down to its most basic connotation--a screw-up. As Troyen A. Brennan, MD, noted in The New England Journal of Medicine, synonyms for error in the Merriam-Webster Thesaurus include "blooper, blunder, boner, bungle, goof, lapse, miscue, misstep, mistake, and slip-up." (5) Those words may have haunted many news reporters' and editors' consciousness as they wrote stories and headlines about the report. Many must have been picturing the unfortunate Willie King; surgeon Roberto Sanchez had amputated the diabetic King's right leg, instead of his more diseased left one, in a notorious early 1990s incident mentioned in the IOM report. "It's hardly news that medical professionals make mistakes--even dumb, deadly mistakes," observed Time staff writer Michael D. Lemonick in a December 13, 1999, story in the magazine. (6) A December 1, 1999, editorial in the Chicago Tribune was even more blunt. Headlined, "The operation was a success, but. . .," the article read:

Never mind those tight-fisted HMO's. It's your doctor who's more likely to kill you. Or your hospital. Or the nurse or the pharmacist. . .How else to interpret a report from the prestigious National Academy of Science. . . .Why, the worst care-denying HMO probably kills only a patient or two in a bad year. (7)

It clearly wasn't much of a leap for journalists to move fully into the blame game or malpractice-mongering mode--the inadvertent categorization of many types of errors as individual acts of negligence or incompetence. Few journalists seemed aware that in framing the errors issue this way, they were only reinforcing the medical profession's fears about some of the recommendations in the IOM report. The report's call for mandatory systems of error reporting, for example, has run into fierce opposition on the ground that it could expose providers to even greater individual legal liability.

Slighting the System

Only a handful of journalists seemed to absorb the essential message of the report: All too frequently, errors in health care were the result of systems problems rather than of individual acts of malfeasance. In other words, to err really is human; at the same time, health care, like any other system in which we operate, is devised by and composed of humans. As a result, like any system that aims to minimize or eliminate error, health care must be designed to compensate for our inevitable human shortcomings.

As the IOM report noted, almost every other major industry in the United States seems to have figured this out decades before health care did. The result has been a broad spectrum of initiatives in system safety, error reduction, and quality control that have built many of the state-of-the-art industrial systems. For example, the report noted that the risk for dying in a domestic jet flight decreased from 1 in 2 million in 1967 through 1976 to 1 in 8 million by the 1990s as a result of efforts to minimize pilot error and improve overall air safety.

Clearly, not all analogies to industrial quality control or safety programs are relevant to health care--a system in which people are often called on to think and act in complex ways and make decisions amid considerable uncertainty about medical outcomes. At the same time, as the IOM report noted, health care is also rife with relatively simple errors--such as when a pharmacist or nurse misreads a physician's handwriting. That's why the report emphasized the need for systems-based improvements in health care, such as not stocking look-alike products on hospital shelves that could readily be confused for one another--or having health care organizations adopt computerized prescription-drug order-entry systems to mitigate any medication mix-ups. But little of this apparently struck many reporters as especially newsworthy or sexy, so the systems recommendations got relatively short shrift in most of the coverage.

A happy exception was a December 8, 1999 piece by Dateline NBC. In it, the NBC correspondent, Bob Arnot, MD, referred to auto manufacturers, and even the Domino's pizza chain, as examples of companies or industries that had adopted error-reduction systems far ahead of health care. Arnot showed how Brigham and Women's Hospital in Boston had installed a computerized prescription system 6 years earlier to minimize the risks for adverse drug interactions and other prescribing errors. The piece concluded with Dr. Donald Berwick saying, "We can beat this problem. The reduction of errors and improvement of safety is absolutely achievable." The report itself couldn't have said it better.

Lessons To Be Learned

This saga clearly illustrates how the news media struggles to digest and convey the nuances of medical stories. Reporting these stories is much like combating medical errors: Just as in trying to figure out how a cancer patient got the wrong chemotherapy drug, journalists must often analyze a complex chain of causation in writing about any pervasive problem in medicine. Then, we must convey this complexity to our readers and viewers in a way that doesn't bore their socks off--and still does justice to a problem that is usually far thornier, and often less sexy, than the headlines of our stories suggest.

The gratifying conclusion that can be drawn from the Dateline NBC piece is that it really is possible to take a deeper and more balanced look at issues like medical errors. That's an especially hopeful sign in light of the fact that the IOM's Quality of Health Care in America Project will be releasing still more reports in the future. Some of the reports are likely to address issues of potentially even greater import that may be still more difficult for journalists and members of the public to digest. These include the well-documented variations in rates of medical and surgical procedures across the country and the stunningly wide disparities in patterns of health care delivery. As IOM panel member Mark R. Chassin, MD, fretted to reporter Lawrence K. Altman of The New York Times, "Talk about overuse and underuse of health care, issues that are as important as errors, and everybody goes to sleep." (8) That may mean it's not too early for the IOM--or anybody else concerned about health-quality issues--to begin better educating the news media well in advance of the next blockbuster report. That way, even if another reporter bucks a future embargo, the flood of coverage will stand a better chance of being both broad and deep.


1. Kohn LT, Corrigan JM, Donaldson MS, eds. To Err Is Human: Building a Safer Health System. Institute of Medicine (U.S.) Committee on Quality Health Care in America. Washington, DC: National Academy Pr; 1999.

2. Hall MJ, Lawrence L. Ambulatory Surgery in the United States, 1996. Hyattsville, MD: National Center for Health Statistics; 1998.

3. McDonald CJ, Weiner M., Hui SL. Deaths due to medical errors are exaggerated in Institute of Medicine report. JAMA. 2000;284:93-5.

4. Vergano D. Medical errors not so deadly? "Shrill" estimate not supported by data, doctors say. USA Today. 2000;5 Jul:8D.

5. Brennan TA. The Institute of Medicine report on medical errors-could it do harm? New Engl J Med. 2000;342:1123-5.

6. Lemonick MD. Doctors' deadly mistakes. Time. 1999;154:24.

7. Unsigned editorial. Chicago Tribune. 1999;1 Dec:28.

8. Altman LK. The doctor's world: Getting to the core of mistakes in medicine. The New York Times. 2000;29 Feb:D.


Susan Dentzer, The NewsHour with Jim Lehrer, 2700 S. Quincy Street, Suite 240, Arlington, VA 22206; telephone: 703-998-1841; fax: 703-998-1857; e-mail: