METHODS: This was a cross-sectional study conducted at the Aga Khan University clinical laboratory, Karachi, Pakistan. One hundred and seventy-eight ETSA samples sent for routine bacteriological cultures were processed quantitatively as part of regular specimen processing method and semi-quantitatively. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy was calculated for both methods using clinical diagnosis of pneumonia as reference standard. Agreement between the quantitative and semi quantitative methods was assessed via the kappa statistic test. Pathogen yield between the two methods was compared using Pearson\'s chi-square test.
RESULTS: The quantitative and semi-quantitative methods yielded pathogens in 81 (45.5%) and 85 (47.8%) cases respectively. There was complete concordance of both techniques in 155 (87.1%) ETSA samples. No growth was observed in 45 (25.3%) ETSA specimens with quantitative culture and 37 (20.8%) cases by semi-quantitative culture. The diagnostic accuracy of both techniques were comparable; 64.6% for quantitative and 64.0% for semi-quantitative culture. The kappa agreement was found to be 0.84 (95% CI, 0.77-0.91) representing almost perfect agreement between the two methods. Although semi-quantitative cultures yielded more pathogens (47.8%) as compared to quantitative ETSA cultures (45.5%), the difference was only 2.3%. However, this difference achieved statistical (chi-square p-value < 0.001) favoring semi-quantitative culture methods over quantitative culture techniques for processing ETSA.
CONCLUSIONS: In conclusion, there is a strong agreement between the performances of both methods of processing ETSA cultures in terms of accuracy of LRTI diagnosis. Semi-quantitative cultures of ETSA yielded more pathogens as compared to quantitative cultures. Although both techniques were comparable, we recommend processing of ETSA using semi-quantitative technique due to its ease and reduced processing time.