Preventing medical errors by designing benign failures

Jt Comm J Qual Saf. 2003 Jul;29(7):354-62. doi: 10.1016/s1549-3741(03)29043-2.

Abstract

Background: One way to successfully reduce medical errors is to design health care systems that are more resistant to the tendencies of human beings to err. One interdisciplinary approach entails creating design changes, mitigating human errors, and making human error irrelevant to outcomes. This approach is intended to facilitate the creation of benign failures, which have been called mistake-proofing devices and forcing functions elsewhere. USING FAULT TREES TO DESIGN FORCING FUNCTIONS: A fault tree is a graphical tool used to understand the relationships that either directly cause or contribute to the cause of a particular failure. A careful analysis of a fault tree enables the analyst to anticipate how the process will behave after the change. EXAMPLE OF AN APPLICATION: A scenario in which a patient is scalded while bathing can serve as an example of how multiple fault trees can be used to design forcing functions. The first fault tree shows the undesirable event--patient scalded while bathing. The second fault tree has a benign event--no water. Adding a scald valve changes the outcome from the undesirable event ("patient scalded while bathing") to the benign event ("no water")

Limitations: Analysis of fault trees does not ensure or guarantee that changes necessary to eliminate error actually occur. Most mistake-proofing is used to prevent simple errors and to create well-defended processes, but complex errors can also result.

Conclusions: The utilization of mistake-proofing or forcing functions can be thought of as changing the logic of a process. Errors that formerly caused undesirable failures can be converted into the causes of benign failures. The use of fault trees can provide a variety of insights into the design of forcing functions that will improve patient safety.

MeSH terms

  • Causality
  • Decision Trees*
  • Humans
  • Medical Errors / prevention & control*
  • Organizational Innovation
  • Probability
  • Process Assessment, Health Care / methods*
  • Risk Assessment / methods*
  • Safety Management / methods*
  • Systems Analysis
  • Task Performance and Analysis*