Study objective: To evaluate the ability of multiple clinical variables to predict an increased or decreased probability of pelvic pathology, therefore determining when a pelvic examination is needed.
Design: Cross-sectional.
Setting: An urban emergency department.
Type of participants: 246 female patients presenting with abdominal pain and undergoing pelvic examination.
Measurements and main results: The patients were divided into two groups. Group 1 consisted of 99 patients without pelvic pathology (40%) and group 2 consisted of 147 patients with pelvic pathology (60%). Pelvic pathology was defined as presence of infective discharge (with confirmation on wet mount/potassium hydroxide smear), presence of adnexal mass and/or tenderness, cervical motion tenderness, uterine enlargement and/or tenderness, blood in the vaginal vault, and culture report positive for Neisseria gonorrhea. The following variables achieved statistical significance (P less than .05) using a logistic regression model: history of vaginal discharge (odds ratio, 2.30 [95% confidence interval, 1.23 to 4.32]); history of dysmenorrhea/menorrhagia (4.35 [1.52 to 12.40]); right upper quadrant pain on physical examination (0.33 [0.13 to 0.85]); and left lower quadrant pain on physical examination or history (1.73 [0.94 to 3.19]).
Conclusion: History of vaginal discharge, history of dysmenorrhea/menorrhagia, and left lower quadrant pain on physical examination act as risk variables predicting presence of pelvic disease. Right upper quadrant pain on physical examination acts as a protective variable to predict absence of pelvic disease. This information may aid the emergency physician in determining the need for a pelvic examination in women with abdominal pain.