Institute of Medicine Report: A Response

Richard E. Anderson, MD, FACP

Richard E. Anderson, MD, FACP, a medical oncologist, is chairman and chief executive officer of The Doctors Company. A member of the American Society of Clinical Oncology and a fellow of the American College of Physicians, Dr. Anderson was a clinical professor of medicine at the University of California, San Diego, and is past chairman of the Department of Medicine at Scripps Memorial Hospital, where he served as senior oncologist for 18 years. Dr. Anderson is the editor of a book on medical malpractice, and his commentaries on legal reform and defensive medicine have been widely cited. He is the 2004 recipient of the PLUS Foundation Award for Outstanding Leadership in Healthcare Professional Liability.

Originally published in The Physician Insurer, First Quarter, 2000

The Institute of Medicine monograph To Err Is Human has already generated extraordinary media commentary, a presidential mandate for federal agencies to implement methods for reducing medical error,1 Senate hearings, and a promise of legislation from Senator Edward Kennedy. In part, this reaction is due to the widely publicized statement made on the first page of the report that at least 44,000 and possibly as many as 98,000 Americans die each year as a result of medical errors.2 Is this statement true?

Both numbers arise from the Harvard Medical Practice Study, which utilized 1984 data from the State of New York.3 The same group then repeated the study using 1992 data from Colorado and Utah.4 These figures for deaths due to medical error are not real numbers at all, but represent extrapolations from extraordinarily small samples. For example, the figure of 98,000 deaths nationally is conflated from a controversial analysis of only 180 actual deaths that occurred among more than 30,000 acutely ill, hospitalized patients.

The authors noted in their original publication that in a subsequent review a second team of their own reviewers failed to identify the same set of adverse events as the first team, but they did find the same incidence of adverse and negligent adverse events. Satisfied, the authors declared their data reliable.5 This is roughly equivalent to saying it does not matter whether we convict the innocent or the guilty as long as the overall number of convictions matches the crime rate.

Even more remarkable, the Harvard reviewers agreed only 10% of the time on the simple presence or absence of medical negligence.6

Given these serious deficiencies in their test methods, it comes as no surprise that when the same group attempted to replicate their study in Colorado and Utah using the same methods, they came up with a calculation suggesting 44,000 deaths due to medical error. It is thus just as plausible to infer that preventable medical deaths actually fell by 55% (98,000 to 44,000) from 1984 to 1992.

When the Harvard Medical Practice Study was published, the authors noted that “[a] prime reason that modern medical care is so hazardous is that it so ambitiously attempts to cure the even greater hazards of nature.”7 In addition, they cautioned “against too quick a comparison of such fatality figures…a substantial proportion of patients were gravely ill, and many would have died from their underlying illnesses in months, days, perhaps hours[emphasis added], even absent the mishap in treatment.”8 It is unfortunate the authors have again chosen not to heed their own admonition against precisely the kind of extrapolations that have accompanied the release of the Institute of Medicine.

The Harvard study has been severely criticized for its failure to consistently identify adverse events, for improper definitions of medical negligence, for combining major and minor outcomes to generate large numbers, and for its uncritical evaluation of the tort system.9 For the Institute of Medicine editors to use the Harvard study as the basis for claiming there is an epidemic of medical death is simply wrong.

The Institute of Medicine makes a number of recommendations for reducing medical error. Putting the issue of numbers aside, the proposals still require consideration.

First is a call for a national focus on patient safety and improved medical outcomes. This has always been the physician’s primary goal, but it has recently been obscured by the excesses of managed care and the imperatives of our legal process. Everyone will gain from a renewed commitment to medical excellence in every facet of our health care system. Indeed, it is difficult to object to such a proposition. No one, with the very important exception of personal injury attorneys, benefits from medical error.

It would be a mistake, however, to miss opportunities for significant progress in patient safety and improved health outcomes which might come from a wider view. For example, consider these statistics:10

  • Moderate exercise would save 178,000 lives annually.
  • Lowering cholesterol by ten percentage points would save 133,000 Americans each year.
  • Smoking cessation would save 328,000 lives yearly in the United States.

Each of these estimates exceeds the Institute of Medicine figure for deaths due to medical error throughout the entire health care system in treating all the acutely ill, hospitalized patients in the United States. In addition, these improvements could be accomplished without either federal intervention or a new wave of litigation.

The point, of course, is not that medical error is acceptable, but rather that it must be appropriately identified for solutions to be developed. As we develop public policy to foster improved patient safety and improved outcomes, we should rely on a common commitment to the goal as opposed to dubious estimates of medical error.

Second, the Institute of Medicine advocates the creation of state and national databases for identifying and learning from errors that have occurred. Reporting of errors that caused patient harm would be mandatory and open to public scrutiny. A confidential, voluntary database would be maintained for evaluating problems that threatened patient safety. The effectiveness of this disclosure scheme is problematic in view of our current tort system and its foundation on blame.

It is essential that any program to improve patient safety afford practitioners the opportunity to discuss outcomes frankly. Identifying mistakes and developing techniques to avoid them becomes more difficult in proportion to the degree to which physicians suffer legal consequences for discussing poor outcomes with their peers.

Unfortunately, there is too little focus on confidentiality to allow physicians to reap the benefits of peer review. Courts in several states have opened peer review to discovery by plaintiff attorneys in malpractice claims. Some health facilities have been forced to advise their physicians against attending any regularly scheduled discussions of poor health outcomes because of the potential for litigation, as former Surgeon General C. Everett Koop noted in his response to the Institute of Medicine. Dr. Koop was explicit about the ill effects of the current attitude toward confidentiality: “The threat of malpractice litigation almost eliminates the opportunity to disclose error, learn from mistakes and receive support from colleagues and even patients when inevitable errors occur.”11

The Institute of Medicine takes note of this destructive pattern. Indeed, those who would misappropriate the report to stoke the fires of litigation should note its conclusion that “[t]he focus must shift from blaming individuals for past errors to a focus on preventing future errors by designing safety into the system….People must still be vigilant and held responsible for their actions. But when an error occurs, blaming an individual does little to make the system safer and prevent someone else from committing the same error.”12

Even more to the point is this observation: “Patient safety is also hindered through the liability system and the threat of malpractice, which discourages the disclosure of errors. The discoverability of data under legal proceedings encourages silence about errors committed or observed.”13

Yet despite clear recognition of the problem, the Institute of Medicine nonetheless calls for mandatory disclosure of errors into a centralized system of tracking and reporting. Such a system would inevitably become a database of evidentiary claims and potential clients for personal injury lawyers, and would further hamper efforts to identify and eliminate errors. The goal—improved outcomes and enhanced patient safety—becomes harder to reach under this scenario. Mandatory disclosure without adequate confidentiality contradicts the stated purpose of the Institute’s recommendations and would exacerbate an already deplorable excess of litigation.

Today, there are approximately 18 malpractice claims for every 100 practicing physicians in the United States every year. In high-risk specialties such as neurosurgery, obstetrics, emergency room medicine, trauma surgery, and orthopedics, that number exceeds four claims for every ten doctors every year.14 This existing avalanche of claims vastly exceeds every estimate of the number of “bad doctors” or the number of valid claims. Physicians can be forgiven for being leery of additional exposure to a system that requires three and a half years of litigation to close the average claim, 80% of which are found to be without merit.15

Any proposal for a national data base of medical error must assure complete confidentiality lest the entire effort be submerged in a new wave of litigation and an escalation of the already pervasive practice of defensive medicine.16 Stanford economists Kessler and McClellan have estimated a minimum cost to our healthcare system of $50 billion annually for defensive medicine.17 Their calculation is similar to that of the Office of Technology Assessment, which estimated that approximately 8% of all diagnostic testing in the United States is not medically necessary.18 Kessler and McClellan also demonstrated there is a readily accessible solution. National passage of existing tort reforms similar to those contained in California’s MICRA statutes (notably a cap on damages solely for pain and suffering, periodic payment of large awards, and exclusion of double recoveries) would eliminate much of the waste of these healthcare dollars.

Legislators contemplating implementation of some of the Institute of Medicine recommendations would thus be wise to link them with necessary reform of our liability-crazed system of medical care. New laws to improve patient safety must include confidentiality safeguards that will allow practitioners to discuss and disclose mistakes. Programs to improve outcomes must make use of proven strategies to limit the practice of defensive medicine. As a practical matter, medical liability reform is a prerequisite to furthering the goals of the Institute of Medicine.

To Err Is Human has focused national attention on healthcare in the United States. Though heavily weighed down by the methodological inadequacies of the Harvard Medical Practice Study, the report should be recognized for evolving beyond Harvard study terms like “negligent adverse event” and Harvard study assertions that the “real problem” is not too many but “too few” malpractice suits.19 If the Harvard study informs us at all about our medical-legal system, it is in the study’s most remarkable conclusion: There is no relationship between the presence or absence of medical negligence and the outcome of malpractice litigation.20

Most important, the Institute of Medicine urges progress toward a goal we all share: improved patient health and safety. To achieve it, however, we will need a far greater emphasis on medical systems and process than on lawyers and litigation. Whether this is possible remains to be seen.


  1. Pear R, Clinton to Order Federal Agencies to Reduce Medical Errors. The New York Times1999 Dec 7 National Politics.
  2. Kohn LT, Corrigan JM, Donaldson MS, editors, Committee on Quality of Health Care in America, Institute of Medicine, To Err Is Human: Building a Safer Health Care System, National Academy Press, Washington, DC, 1999.
  3. Weiler PC, Hiatt HH, et al., A Measure of Malpractice, Cambridge, Mass: Harvard University Press; 1993.
  4. Thomas, EJ, Studdert DM, et al., Incidence and Types of Adverse Events and Negligent Care in Utah and Colorado, Med Care, forthcoming Spring 2000.
  5. Localio AR, Lawthers AG, et al., Relation Between Malpractice Claims and Adverse Events Due to Negligence. N Engl J Med.1991;325:245–51.
  6. Localio AR, Weaver SL, Brennan TA, et al., Identifying adverse events caused by medical care: degree of physician agreement in a retrospective chart review. Ann Intern Med.1996;125:457–464.
  7. Weiler, 175.
  8. Weiler, 55–56.
  9. Anderson, RE, An Epidemic of Medical Malpractice? A Commentary on the Harvard Medical Practice Study. New York, NY: Manhattan Institute Civil Justice Memo No. 27.
  10. Woolf SA, The Need for Perspective in Evidence-Based Medicine, JAMA1999;282, No. 24:2359.
  11. Koop, CE, An Ounce of Error Prevention, The Washington Post 1999 Dec 23, A21.
  12. To Err Is Human, 4.
  13. To Err Is Human, 37.
  14. The Doctors Company data on file.
  15. The Doctors Company data on file.
  16. Anderson RE, Billions for Defense: The Pervasive Nature of Defensive Medicine, Arch Intern Med159;2399–2402.
  17. Kessler DP, McClellan M. Do doctors practice defensive medicine? Q J Econ,1996;111:353–390.
  18. Defensive Medicine and Medical Malpractice. Washington, DC: Office of Technology Assessment, US Congress; July 1994. Publication OTA-H602.

The guidelines suggested here are not rules, do not constitute legal advice, and do not ensure a successful outcome. The ultimate decision regarding the appropriateness of any treatment must be made by each healthcare provider considering the circumstances of the individual situation and in accordance with the laws of the jurisdiction in which the care is rendered.


Stay in the Know

Sign up for The Doctor’s Practice.

Our e-newsletter features timely articles, videos, and guides on a range of patient safety topics.