An evaluation of the diagnostic value of clinical and radiological manifestations in patients attending the Addis Ababa Tuberculosis Centre

Our investigation is based on 1,000 consecutive patients attending the rout ine outpatient services of the Addis Ababa Tuberculosis Demonstration and T raining Centre (ATBDTC) in Addis Ababa during the period November 1996-Marc h 1997. By using sputum microscopy for acid-fast bacilli (AFB), chest radio graphy (CXR) and clinical assessment these patients had been diagnosed as h aving either AFB-positive pulmonary tuberculosis (PTB) (n = 139), AFB-negat ive PTB (n = 61) or non-tuberculosis (non-TB) (n=800), These three diagnost ic groups were subsequently re-assessed by us with regard to selected demog raphic and clinical parameters, including CXR, in order to identify and wei ght markers significantly linked to proven PTB. The sum of the individual w eights provided diagnostic scores (DS); the average DS for AFB-positive pat ients was 653.5 +/- 174 and that for non-TB patients was 219.1 +/- 138.7. T he calculated cut-off value between these two groups was 444. Ten (7.2%) AF B-positive PTB patents had a DS below the calculated cut-off, while 46 (5.8 %) of the 800 non-TB cases had diagnostic scores exceeding this value. Our DS system achieved 92.8% sensitivity and 94.3% specificity, with positive a nd negative predictive values of 73.7% and 98.7%, respectively. When plotti ng the individual DS values of the 61 AFB-negative TB patients, 24 (39.3%) fell below the cut-off. It is most likely that these patients did not have PTB, We conclude that there is a need for improved and standardized diagnos tic approaches for TB-suspected patients who depend upon clinical and CXR c riteria for correct diagnosis. Our score system may be helpful in this cont ext.