[Skip to Navigation]
Sign In
Invited Commentary
February 2014

Rescuing Failures: Can Large Data Sets Provide the Answer?

Author Affiliations
  • 1Department of Surgery, University of Texas Health Science Center at Houston, Houston
JAMA Surg. 2014;149(2):124. doi:10.1001/jamasurg.2013.3674

Secondary analyses of large data sets including that by Gonzalez et al1 have repeatedly reported on an association between failure to rescue after surgery, or death after a major complication, and outcome.2-4 While few provide insight as to how to change this relationship, these analyses can generate hypotheses about why high-mortality hospitals have higher rates of failure to rescue. For example, studies have identified higher rates with lower levels of intensive care unit facilities,5 poor nurse practice environments,6 and less aggressive treatment style.7 However, these data sets are not granular enough to determine whether these are actual causes of failure to rescue or just surrogates for other factors such as better teamwork and communication, which Gonzalez and colleagues have suggested as a potential solution. Identifying the correct targets for improvement is essential as interventions are likely to be multifaceted, resource intensive, and costly.

Add or change institution