3003 — Defining and Exposing Contextual Errors in Medical Decision Making

Author List:
Weiner SJ (Jesse Brown VA Medical Center)
Schwartz A (University of Illinois at Chicago)
Yudkowsky R (University of Illinois at Chicago)
Schiff GD (John H. Stroger Jr Hospital of Cook County)
Weaver FM (Midwest Center for Health Services and Policy Research, Hines VA Hospital)
Golberg J (University of Illinois at Chicago)
Weiss KB (Midwest Center for Health Services and Policy Research, Hines VA Hospital)

Objectives:
The Institute of Medicine defines medical error as “the failure of a planned action to be completed as intended or the use of a wrong plan to achieve an aim.” Current taxonomies of medical error are predominantly biomedical, overlooking inattention to factors in a patient’s life context that are essential to their care. The purpose of this study was to develop and test a methodology for examining physicians’ propensity to make contextual errors compared with their propensity to make biomedical errors.

Methods:
We employed a quasi-randomized, blinded experimental design using standardized patients (SPs). 54 clinicians were presented with opportunities to make pre-identified biomedical or contextual errors, and we then tabulated their success or failure at eliciting and then incorporating essential information into their plan of care. Physician communication behavior was also coded using an SP completed instrument. The cases were first validated in a pre-study phase on a separate group of 16 physicians to confirm that inattention to specific information met IOM criteria for medical error.

Results:
In the pre-study phase, physicians reached consensus on the criteria for biomedical and contextual error for each case. In the main study, 57% of subjects made biomedical errors and 60% made contextual errors. All biomedical errors were due to failures to elicit critical information. 56% of contextual errors, however, were due to failures to incorporate critical information successfully elicited. Controlling for the information in each case, better patient/physician communication also correlated with fewer errors (OR = 4.5, p<.05).

Implications:
Physicians can reach the same high consensus regarding what constitutes essential contextual information as they can regarding what constitutes essential biomedical information relevant to clinical decision making. Case narratives, validated through consensus, can be instruments to assess physicians’ propensity to overlook or disregard essential patient context. In this pilot study, both biomedical and contextual errors were about equally likely to occur, but not for the same reasons: Physicians were significantly less likely to incorporate elicited contextual information into their plan of care. Errors were inversely associated with good communications behavior.

Impacts:
This study demonstrates an empirical method for incorporating contextual as well as biomedical variables into assessments of the quality of physician practice.