Clinician judgments of functional outcomes: How bias and perceived accuracy affect rating

Objective: To evaluate the accuracy of clinician judgments of patient function, the susceptibility of judges to bias, and the relation between a judge's degree of belief in his/her accuracy of classification to observed accuracy when using the FIM(TM) instrument. Participants: Fifty rehabilitation professionals. Setting: 3 urban medical centers. Design: Four randomized experiments among subjects to examine the effect of potentially biasing information on FIM ratings of patient vignettes. Participants answered 60 true/false questions regarding patient function and FIM score and indicated confidence in the accuracy of their answers. Interventions: Manipulation of patient information. Main Outcome Measures: The standard FIM 7-point scale, observed proportion of correct responses to the 60 true/false questions, and a 6-category confidence scale for each of the 60 questions were used as dependent measures. Results: FIM ratings assigned to others biased participants' FIM ratings of patient vignettes. Functional ability was over estimated when ratings in other domains were high and underestimated when they were low. Participants were overconfident in their ability to answer FIM questions accurately across all professional disciplines. Conclusion: Bias and poor judgment of level accuracy play a significant role in clinician ratings of patient functioning. Blind ratings and training in debiasing are potential solutions to the problem.