What we need to do from the standpoint of improving care is to develop methods of analyzing errors to identify the contributing factors and devise ways to prevent them. At the annual meeting of the American Association of Clinical Endocrinologists this year, we conducted a workshop on ways to analyze and prevent medical errors in patients with diabetes. To my knowledge, this was the first workshop of this kind conducted during a major medical society meeting.
We adopted some of the theories on errors that are well known to the Federal Aviation Administration and the Nuclear Safety Commission. In their areas, errors have a devastating effect, and so they are very interested in finding ways to analyze errors that have occurred and how to prevent them. The technique they both use involves assessing multiple contributing factors.
In our study, we found that the most important factor associated with a patient’s
death was a physician’s diagnostic error. But that was not enough. It appeared that the most important other contributing factor was a system that was unsafe. That is, there was a defective culture of safety. A culture of safety exists when there is a cooperative unit of care in which the members work well together, back each other up, and communicate well with each other (Table 2).
Table 2—What Is a Culture of Safety?
Preventing medical errors and, more important, preventing medical injury from errors, requires a cooperative effort.
The culture of safety in diabetes must extend to the home, the school, and the workplace.
Education of all involved is essential and must be realistic for the patient.
• Realistic educational goals
• Clear and legible handouts
• Assessment of patients’ cognitive and emotional status
• Follow-up to objectively measure patients’ retention and accuracy of their knowledge and understanding
In the acute care setting, the following factors are essential:
• The most skilled health care providers must work together during the clinical crisis.
• In-service training, written protocols, clinical simulations, and multidisciplinary conferencesare valuable tools.
• The goals must be clear and understood and agreed upon by all major parties involved.
In almost all the circumstances, when the physician made a major diagnostic error that led to death, it was in a setting in which there was absolutely no backup. For instance, if the physician wrote a prescription at a hospital, there was no pharmacist who doublechecked it. If the physician gave an order for insulin, there was no nurse who felt empowered to question it. We found that was a striking feature. It was so important that we felt we needed to look at how the system worked because it is our belief—and that of
many others—that improving the system of care is probably more likely to reduce morbidity and mortality than is just focusing on individual error. We considered all the factors contributing to medical errors. The lack of treatment algorithms was a common feature.
When a team works together well, and they have explicit rules that everyone uses, it is much easier to do things right. Although, in diabetes, the patient’s level of knowledge is, of course, a major contributing factor, we found that it was not a disproportionate factor for one type of problem. In fact, in the most severe injuries, level of patient knowledge was
not an unusually prominent factor. In other words, the catastrophic errors that led to death did not occur disproportionately in the least knowledgeable patients.
We also found that the training of the physician was not as important as many other factors. We feel, therefore, that it is more important to develop a system and to encourage cooperation among the health care team rather than to recommend more training for the individual physician.