Reliability and validity of self-reported symptoms for predicting vulvodynia

To evaluate the reliability and validity of self-reported symptoms to predict vulvodynia, compared with examination-based confirmation. Between August 5, 2004, and December 13, 2004, 1,046 members of the University of Michigan Women's Health Registry were surveyed regarding the presence of symptoms suggestive of vulvodynia. Diagnoses of vulvodynia and of control status based on survey responses were made, and a subset of these respondents was evaluated in the office. One thousand forty-six of 1,447 (72.3%) eligible women, aged 19 to 92 years, completed the survey. Seventy-nine (7.6%) of the survey respondents who reported ongoing vulvar pain lasting more than 3 months were predicted to have vulvodynia, while women reporting no current pain with intercourse and no history of prolonged vulvar pain were predicted to be controls (N = 543). Agreement between the history taken at the office and that reported on the survey was very good (reliability: Cohen's kappa = 0.86, 95% confidence interval 0.73-0.99). Of the 28 women predicted to have vulvodynia who were examined in the office, 27 (96.4%) were confirmed to have vulvodynia, and 28 of the 34 (82.4%) asymptomatic women examined did not have increased vulvar sensitivity (Cohen's kappa = 0.78, 95% confidence interval 0.64-0.92). Excellent reliability and validity of survey responses for predicting vulvodynia were demonstrated. II-2.