摘要：The A-DROP scoring system was originally designed to assess clinical severity of community acquired pneumonia using the following parameters: advanced Age, Dehydration, Respiratory failure, Orientation disturbance (confusion); and, low blood Pressure. Total A-DROP score ranges zero to five assigning one point for each component, wherein five indicates the poorest prognosis. The purpose of this single-center retrospective study was to determine whether A-DROP could predict the risk for death in patients with pulmonary tuberculosis. We reviewed consecutive HIV-negative, non-multidrug-resistant smear-positive adult pulmonary tuberculosis patients. The cohort consisted of 134 men (38.8%), 211 women (61.2%), 272 who discharged alive (28.8%), and 73 who died in-hospital (21.2%) with a median age of 72 (IQR: 54-82) years. A one-point increase in the A-DROP score was associated with a higher risk for in-hospital mortality with odds ratio of 3.8 (95% confidence interval 2.8-5.2, P < 0.001). The area under receiver operating characteristics curve was 0.86. The total score cutoff of 1.5 provided the best Youden Index of 0.61. Using this criteria, total score >1.5, sensitivity was 85% and specificity was 76%. Kaplan-Meier curve clearly indicated that in-hospital mortality increased with higher A-DROP scores (Log-rank test <0.001). In conclusion, A-DROP score clearly indicate pulmonary tuberculosis in-hospital mortality.
