Study Objective To assess diagnostic value of polymerase chain reaction (PCR) in endometrial aspirates (EAs) in comparison with conventional tests for diagnosis of female genital tuberculosis (TB) and to find agreement between EA PCR done for endometrial TB and laparoscopic findings of pelvic TB in women with unexplained infertility. Design Prospective observational cohort study. Setting Tertiary care hospital. Patients A total of 732 infertile females screened and 385 enrolled to undergo procedure to obtain EAs. Interventions EAs were tested by conventional tests (histopathology, acid-fast bacilli, Lowenstein-Jensen staining, liquid culture) and PCR for Mycobacterium tuberculosis. Patients with positive conventional tests were started on antitubercular treatment (ATT). Patients with negative conventional tests underwent laparohysteroscopy irrespective of PCR results to assess changes of tubercular infection in the pelvis. Peritoneal washings were also sent for liquid culture and PCR for TB, and suspicious lesions were biopsied at laparohysteroscopy. Findings at laparoscopy upgraded the diagnosis in these women. EAPCR results were analyzed to find agreement with the findings at laparoscopy. Measurements and Main Results Conventional tests were positive in 8 of 385 (2%) patients. PCR was positive in 58.1% (n = 224) of endometrial samples, with sensitivity of 62.5% (95% confidence interval [CI], 24.49–91.48), specificity of 41.91% (95% CI, 36.88–47.07), positive predictive value of 2.23% (95% CI, 1.31–3.78), negative predictive value of 98.14% (95% CI, 95.53–99.24), and a diagnostic accuracy of 42.34% (95% CI, 37.35–47.45) with conventional tests. A total of 265 patients underwent laparoscopy, of whom 165 were PCR positive and 100 were PCR negative. Laparoscopic findings suggestive of TB were found in 39.3% of patients who were PCR positive and 9% of patients who were PCR negative. Kappa agreement was 0.25, suggesting fair agreement between PCR and laparoscopy. Conclusion PCR as a stand-alone diagnostic test for endometrial TB is not justified to confirm diagnosis and initiate ATT. The addition of laparohysteroscopy improves diagnostic yield for genital TB. Referring patients with a suspicion of female genital TB to tertiary care for 1-time laparoscopy is better than initiating ATT solely on the basis of PCR results.
