Demographic Characteristics and Diagnostic Performance of Pathologists who Enjoy Interpreting Melanocytic Skin Lesions

Andrea Radick | 2017

Advisor: Amanda Phipps

Research Area(s): Clinical Epidemiology


Background: Diagnostic errors among pathologists when interpreting melanocytic skin lesions (MSL) is an ongoing concern for patient safety and quality of care. The number of skin biopsies has increased annually over the past decade, subsequently applying a higher demand on pathologists’ work performance. Given that physician job satisfaction can plausibly impact upon patient care, we aimed to estimate the association between enjoyment of MSL interpretation and diagnostic performance in pathologists, and to characterize the attributes of pathologists who do vs. do not enjoy interpreting MSL. Materials and Methods: A cross-sectional survey was conducted among pathologists from 10 U.S. states who interpret MSL. Characteristics of pathologists’ demographics, training, experience, and perceptions of MSL interpretation were gathered and the associations with self-report of enjoyment when interpreting MSL were estimated by Mantel-Haenszel chi-square tests, or Spearman’s correlation tests for ordinal variables. Pathologists subsequently reviewed a set of 48 MSL cases, and their interpretations were compared to a reference standard diagnosis to determine diagnostic accuracy. A multivariable logistic regression model was fit to determine if the characteristics associated with enjoyment were also independently associated with diagnostic accuracy. The unadjusted and adjusted associations between enjoyment and diagnostic accuracy were evaluated by generalized estimating equations (GEE) models. Results: Of 207 pathologists who enrolled in the study, 187 completed a baseline survey followed by histological interpretations (90%). Seventy percent agreed at least slightly that interpreting MSL is enjoyable. Pathologists who enjoyed interpreting MSL were more likely to interpret an average of ≥50 benign MSL cases per month (p=<0.001), report that their colleagues consider them an expert in MSL pathology (p=<0.001), found MSL cases more challenging to interpret (<0.001), were more nervous about MSL compared to other types of pathology (0.002), and had a higher degree of confidence in their MSL assessment (p=<0.001). In multivariable analyses, expertise and number of benign MSL interpreted per month remained statistically significantly associated with diagnostic accuracy; however, the adjusted GEE model showed no association between enjoyment and diagnostic performance. Conclusions: Most pathologists agreed that interpretation of MSL is enjoyable. The number of benign MSL cases interpreted per month, and perceived expertise and confidence in MSL interpretation were highly associated with enjoyment. Due to the annual increase in skin biopsies and a greater demand for pathologists to enter the workforce, it is reassuring to know that there is no association between enjoyment and diagnostic accuracy.