Surely the November presidential elections taught Americans that the same numbers, used differently, can tell quite different stories. Recent media reports on estimates of the rate of medical errors demonstrate yet again that basic arithmetic does not always add up, and that before making headlines out of graph lines, we should first be clear on what we're talking about.
Last year, the Institute of Medicine released an important study on patient safety and medical errors. That study, To Err Is Human, estimated that medical errors cause as many as 98,000 deaths each year, a finding that led to public outcry, federal hearings, and media coverage that ran from CNN to NPR to Family Circle.
In July, AP reported that a new study in the Journal of the American Medical Association cast doubt on the error figure. (For details, see QuickScan, in this issue.) The wire story opened, "A report that medical errors kill close to 100,000 U.S. hospital patients each year probably overestimated the problems, with the real total being perhaps 5,000 to 15,000." Dramatic copy at best - and misleading at worst. The actual JAMA article, by Drs. Rodney Hayward and Timothy Hofer, took pains to prevent just this kind of misrepresentation, or misunderstanding, of their study's results. Hayward and Hofer narrowed the scope of what events actually counted as an error, and which patients to include in those counts. The researchers excluded all terminally ill patients (more on that problem in a moment), and focused on errors that occurred in caring for patients who would have survived some period beyond their hospital discharge. Researchers concluded that about one-quarter of active-care patient deaths were "possibly preventable" by optimal care; but only 6 percent were rated as "probably or definitely preventable."
Dr. Hayward told ABCD, "Too much attention has been given to the 'numbers' controversy, and media reports on the paper have not included the careful caveats" the original article made. He added, "The most important points in the paper are that the 'errors' tend to relate to difficult decisions in very sick patients, rather than the outright blunders. Common problems that increase the risk of death are going to be more difficult to identify and avoid than is often appreciated."
Indeed, "blunders" in caring for people with serious and eventually fatal illness account for significant pain and suffering for these individuals and their loved ones. But such blunders are not quite as obvious and as dramatic as the ones commonly covered in the press: the operation on the wrong side of the brain, amputation of the wrong limb, incorrect prescriptions or misunderstood dosing instructions. Instead, for seriously ill people, errors may begin as simple oversights: a bed pan not readily available to an older, frail person, for instance, or an advance directive not included in records sent from a hospital to a nursing home. The errors can quickly multiply: Cramping and nausea leading to dehydration leading to confusion and falling and then to a broken bone (and then to death) can all result from the missing bedpan. Missing advance directives can result in unwanted resuscitation or invasive procedures. In such cases, at least one error occurred - but where is it counted, and at what point does it count?
In addition, it is not really legitimate to discount errors because the people harmed were already very sick. Suppose that in the ongoing strife between Ford Motor Company and Firestone Tires, the companies agreed to recount deaths and injuries, which now number in the thousands. For analysis' sake, they decide not to count every death related in some way to tire-failure. Instead, they do not count deaths among 18- to 25-year old males (they are, after all, notoriously likely to die in vehicle crashes) or deaths involving red SUVs (which seem to attract more reckless drivers), or deaths of anyone over the age of 70 (after all, they've had their time). Suddenly, the death and injury rate plummets to 100. Everyone cheers and stops worrying about the problem. But the families of the dead are sure to feel that those deaths counted.
Dr. Hayward told us that his article was not meant to dismiss the importance of medical errors, but to draw attention to the problem in classifying them. "Our study provides valuable information and some cautionary notes, but generally supports the notion that there is a big problem and we need to work harder to improve care. However, we should be careful not to over-simplify or sensationalize the extent and nature of the problem and possible solutions." In short, we need a care system that we can rely upon - one that is safe and effective.
The errors that shock the conscience and fuel the media are those in which a young person with an eminently curable condition ends up seriously harmed or dead, but the errors that inflict suffering or cut life short for those who were living with serious illness are important, too. Hayward and Hofer have shown that the pursuit of patient safety will be a complex undertaking, but their work does not diminish the urgency of building error-free systems just because many of the errors cut short the lives of those who are already sick enough to die.
<<< Previous Next >>> [ Go Up ]
