Focused review is more effective than random review in discovering errors in surgical pathology cases
Research Activities, April 2009
Clinicians have debated whether taking a second look at randomly selected pathology samples is more likely to find diagnostic errors than reviewing all of the samples in specific pathology subspecialties. A new study that compared a targeted, random review of pathology case materials with focused review of materials in five difficult-to-diagnose situations found that focused review detected a much higher incidence of errors, including a much higher incidence of major errors. Most importantly, the researchers suggest that focused review of subspecialty pathology helps rule out or identify aspects of surgical pathology as topics where disagreement between pathologists is common.
Over a 5-year period, pathologists at one clinical institution performed secondary reviews on nearly 7,500 randomly-selected cases, representing 3.5 percent of all surgical pathology cases seen at this locality. Errors were detected in 195 cases, representing 2.6 percent of the reviewed cases. This included 27 major errors (0.4 percent of the cases reviewed)—such as a change in diagnosis from benign to malignant—which were likely to affect the patient's treatment. Following the end of the targeted random review, the clinical institution chose to review a predetermined number of consecutive surgical pathology cases in five subspecialties over a 1-year period. Assuming an error rate of up to 5 percent, there was focused review of 380 cases (0.7 percent of all cases in these subspecialties), resulting in discovery of 50 errors (13.2 percent of the examined cases). The focused reviews turned up 12 major errors, representing 3.2 percent of the reviewed cases. Among the subspecialties, the major error rates varied from 0 percent for two of the subspecialties to 6.0 percent for the highest, the researchers reported.
Follow-up found harm in 11 of the targeted random review cases (5 cases of minimal harm, 4 cases of mild harm, and 2 cases of moderate harm). Among the focused review cases, harm occurred in 18 cases (46 percent) of major errors, but only 2 cases (5 percent of reviewed cases) were associated with moderate harm.
This study was funded in part by the Agency for Healthcare Research and Quality (HS13321). More details are in "Effectiveness of random and focused review in detecting surgical pathology error," by Stephen S. Raab, M.D., Dana M. Grzybicki, M.D., Ph.D., Laura K. Mahood, C.T. (A.S.C.P.), and others, in the December 2008 American Journal of Clinical Pathology 130(6), pp. 905-912.