Influence of the history on physicians' interpretations of girls' genital findings

Objective. Because physicians customarily obtain histories before examining children in cases of possible sexual abuse, and because the resulting diag nostic opinions can influence important social and legal decisions, we inve stigated whether clinical histories influence physicians' interpretations o f girls' genital findings. Design. In mailed questionnaires, 1387 randomly selected Fellows of the Ame rican Academy of Pediatrics and all 802 members of four professional groups concerned with child abuse or pediatric gynecology were asked to interpret seven simulated cases. Respondents were asked to interpret seven additiona l cases in separate questionnaires mailed 4 months later. Both sets of case s involved the same seven photographs of girls' external genitalia. However , in six of the seven case pairs, the histories in the two questionnaires d iffered in the extent to which they suggested sexual abuse. In the remainin g (control) pair, the same history was presented in both questionnaires. Results. Of 2189 physicians, 1114 (50.9%) responded. Responses from 604 phy sicians (54.2%) were eligible for analysis. Overall, the genital findings w ere interpreted most consistently by the most experienced physicians and le ast consistently by the least experienced physicians. The proportion of phy sicians whose interpretations of a photograph reversed in the direction sug gested by the change in the associated history from "no indication of abuse " to "probable abuse," or vice versa, ranged for experienced physicians fro m none to 5.6%; for moderately experienced physicians from 1.6% to 19.8%; a nd for inexperienced physicians from 3.6% to 27.2%. This difference between the experience groups was statistically significant in four case pairs. Me an interpretation scores for genital findings changed significantly when th e histories changed in two case pairs for the experienced physicians, in fi ve pairs for the moderately experienced physicians, and in all six pairs fo r the inexperienced physicians. Conclusions. In some cases and especially for less experienced physicians, diagnostic expectation appears likely to influence physicians' interpretati ons of girls' genital findings. Physicians should be alert to the possibili ty of diagnostic expectation bias and its potentially serious social and le gal consequences.