The 1999 Institute of Medicine (IOM) publication of To Err is Human: Building a Safer Health System raised many hopes: would hospitals finally be forced to make patient safety a priority? Unfortunately, in most cases the answer is no. As noted by quality improvement expert Don Berwick, MD, in the June 19, 2003, issue of the New England Journal of Medicine, the progress most institutions have made has been unimpressive, with error data largely unchanged since the IOM report. Berwick wrote, "If the Institute of Medicine is right, then at the very least, 100 patients will die in hospitals in the United States today because of injuries from their care, not from their diseases. How many will die tomorrow? . . . I think it will still be about 100-and 100 the day after tomorrow and 100 the day after that."
An April 6, 2004, article in the New York Times highlighted one example of institutional inattentiveness to the issue. Journalist Milt Freudenheim reported that "only a few dozen medical centers across the country are making full use of the latest computerized patient safety systems" and that just 40 of 4,900 nongovernmental hospitals meet the safety standards set by the Leapfrog Group, a coalition of businesses concerned about medical errors.
FIGURE
But there may be a more fundamental reason for some institutions' negligence. I realized after reading a study we're publishing in this month's issue that there is a deep-seated problem in how some institutions view errors. Ann Cook, PhD, and colleagues' multimethod study of patient safety in 29 rural hospitals in nine states presents some stunning findings (see page 32). For example, health care error-which the IOM defined as "the failure of a planned action to be completed as intended or the use of a wrong plan to achieve an aim"-is viewed by the health care personnel they surveyed (nurses, physicians, administrators, pharmacists, and others) as more often involving nursing practice than medical practice. The researchers note that physicians consider "delays in treatment, the use of outmoded treatments, the failure to employ needed diagnostic tests, the failure to act on the results of testing, errors in administration of treatment, and the failure to communicate with staff and patients not as errors but as 'practice variances,' 'suboptimal outcomes,' or examples of differences in 'clinical judgment.'"
I was saddened to read that many nurse participants reported a reluctance to challenge physicians' definitions of medical error; some nurses who did so said physicians rebuffed them. Many administrators deferred to physicians in this regard, as well, saying they lacked the clinical expertise to define medical error. The researchers concluded that disagreements among health care workers over what constitutes an error lead to disagreements about how often errors are occurring, as well as personnel being hesitant to report and correct errors.
In the hospitals they surveyed, Cook and colleagues found that nurses, administrators, and physicians viewed patient safety as a nursing responsibility, saying that a mere "22% of respondents to one survey said that physicians, nurses, pharmacists, and administrators should share responsibility equally for patient safety." And yet, in a prior study, Cook found that only 8% of physicians saw nurses as part of the decision-making team, and the present study reported that only 37% of nurse respondents said they participated in analyses of the causes of errors or potential errors.
As hospital executives start to comprehend the importance of supportive, collegial work environments to attracting and retaining nurses, they must also acknowledge that adequate nurse staffing is not the sole solution to the problem of hospital errors. Institutions should initiate discussions about the findings of this study and begin to grapple with defining error consistently; promote the equal sharing of responsibility for errors among physicians, nurses, and others; and support nurses' authority in ensuring that patients are safe.