Mohammad Ahmad Alamgir, Muhammad Fayyaz Atif, Shajeel Akhtar, Mamoon Akbar Qureshi, Wajiha Rahat.
Diagnostic Accuracy of Three Consecutive Sputum Smear for AFB Among TB Patients Visiting Chest OPD.
Proceeding Shaikh Zayed Postgrad Med Comp Jan ;24(1):91-5.

Background: Sputum smear examination for acid-fast bacilli (AFB) can diagnose up to 50-60% of cases of pulmonary tuberculosis in well-equipped laboratories. In low-income countries, poor access to highquality microscopy services contributes to even lower rates of AFB detection. Objective: To assess diagnostic accuracy of three consecutive sputum smear for tuberculosis taking sputum culture as gold standard. Study Design: Cross Sectional retrospective study. Study Setting: Study conducted in a tertiary care, teaching hospitals out patient department of Tuberculosis and Chest. Study Duration: Three months from June ? August 2008. Material and Methods: Data was collected retrospectively after fulfilling the inclusion criteria in a structured questionnaire of 88 patients who presented to out door department for diagnosis of pulmonary tuberculosis and was investigated for presence of pulmonary tuberculosis. Reports of three consecutive sputum examinations were taken along with sputum culture for and diagnostic accuracy of sputum for Acid Fast Bacilli was calculated taking sputum culture as Gold standard. Demographic information and symptoms were also taken into account. Results: 88 patient?s reports were recruited for the study. 55.6% of the respondents were between age group of 15 ? 44 years. 51.1% were male in our study groups and 48.9% were females. Diagnostic accuracy of three consecutive sputum for AFB was calculated. Sensitivity turned out to be 93.2%. Specificity was 80.0%. Positive predictive value of sputum turned out to be 95.8% and negative predictive value was 70.6%. Conclusions: Sputum smear examination for AFB has appreciable sensitivity, specificity and predictive values and three sputum smears is sufficient for the early detection of AFB in outpatient setting.

PakMediNet -Pakistan's largest Database of Pakistani Medical Journals - http://www.pakmedinet.com