Purpose: To investigate the accuracy of clinical diagnosis of TB in Japan in recent years and to compare them with previous studies.
Method: Data (sex, age, clinical diagnosis, pathological diagnosis as cause of death) on deceased cases clinically or pathologically diagnosed ante mortem as having tuberculosis was collected from annual reports of the pathological autopsy cases in 1984, 1989, 1994, and 1999-2004. Information on TB death from population statistics in those 9 years also was collected and compared with data of autopsied cases.
Result: Autopsy rate in these years was stably around 10 %. Comparison of gender ratio and mean age between the two surveys showed similar numbers. During 1999-2004, 1725 death cases were diagnosed as TB clinically or pathologically. Number of pathologically proven pulmonary TB cases was 429 and that of miliary TB was 283. 55.7% of pulmonary tuberculosis and only 21.9% of miliary tuberculosis were correctly diagnosed before death. Out of 156 cases clinically diagnosed as non-TB diseases but proven as TB pathologically, 30.8% of clinical diagnosis was pneumonia and/or bronchitis, followed by diagnoses of interstitial pneumonia, respiratory failure, pneumoconiosis and lung cancer. However, the main clinical diagnoses of 175 miss-diagnosed miliary TB cases were diseases other than pulmonary diseases such as renal failure, malignant diseases and sepsis.
Conclusion: In order to reduce undiagnosed pulmonary TB cases and to prevent nosocomial TB infection, differential diagnosis among pneumonia and/or bronchitis cases should be done. In case of miliary TB, not only pneumonia but also diseases other than pulmonary diseases such as renal failure, malignant diseases and sepsis should be included in the list differential diagnosis.