A self-assessed questionnaire can help in the diagnosis of pelvic inflammatory disease.

BACKGROUND

Pelvic inflammatory disease (PID) is often responsible for acute pelvic pain, yet its clinical diagnosis is difficult. The aim of this study was to develop and validate prediction rules for the diagnosis of PID in patients seen for acute pelvic pain, based solely on a self-assessed questionnaire.

METHODS

From September 2006 to April 2008, 499 consecutive patients presenting at the gynecology emergency departments of 5 hospitals for acute pelvic pain completed a Self-Assessment Questionnaire for Gynecological Emergencies. Seventy-three were identified as having a PID. Two-thirds of the database was randomly selected for the derivation of the prediction rules, and the other third was used for internal validation. We developed 2 scores, one that rules out a diagnosis of PID and one that predicts PID, based on multiple logistic regression with jackknife estimates. These scores were then validated with the validation data set.

RESULTS

Four variables were independently associated with PID: scattered pain radiation and/or diffuse pain, insidious pain, peritoneal irritation, and abnormal vaginal discharge. They were used to create a sensitive prediction model that rules out PID. Four other variables were used to build another model that predicted PID with high specificity: abnormal vaginal discharge, bilateral pelvic pain, constipation, and presence of an intrauterine device. The probability of PID for the patients in the low-risk group was 1.1% (95% confidence interval, 0.03-5.8), and the probability of PID in the high-risk group was 55% (95% confidence interval, 31.5-76.9).

CONCLUSION

These 2 prediction rules that classify patients in low risk or high risk of PID, based on simple items collected by a self-assessed questionnaire that are composed only of case history and autodescription of the pain, may prove useful for diagnosing or ruling out PID in patients with acute pelvic pain.