OBJECTIVES: To investigate the diagnosis of pulmonary tuberculosis (PTB) and factors associated with a clinician's decision to stop anti-tuberculosis treatment before completion. DESIGN: The medical charts of all citizens of Taipei City, Taiwan, reported to have received treatment for PTB in 2003 were investigated. RESULTS: Of 1126 PTB patients, 512 (45.5%) started treatment immediately based solely on chest X-ray (CXR) findings; treatment for 214 (19.0%) was based on a positive sputum smear for acid-fast bacilli, for 261 (23.2%) it was based on other findings and for 139 (12.3%) it was based on a positive mycobacterial culture. Of the 1126 PTB patients, 156 (13.9%) had their diagnosis of TB changed by a clinician. Multivariate analysis shows that patients whose diagnosis was based on CXR or other findings, female patients, patients who interrupted treatment for 2 months, patients who continued care at other health facilities (transfer) and patients with lung cancer were significantly more likely to have their diagnosis changed than other groups. CONCLUSION: A substantial proportion of patients were prescribed anti-tuberculosis treatment based on CXR findings alone, and a considerable proportion were advised to stop treatment before completing a full course, findings that require the immediate attention of Taiwan's National Tuberculosis Programme.
OBJECTIVES: To investigate the diagnosis of pulmonary tuberculosis (PTB) and factors associated with a clinician's decision to stop anti-tuberculosis treatment before completion. DESIGN: The medical charts of all citizens of Taipei City, Taiwan, reported to have received treatment for PTB in 2003 were investigated. RESULTS: Of 1126 PTB patients, 512 (45.5%) started treatment immediately based solely on chest X-ray (CXR) findings; treatment for 214 (19.0%) was based on a positive sputum smear for acid-fast bacilli, for 261 (23.2%) it was based on other findings and for 139 (12.3%) it was based on a positive mycobacterial culture. Of the 1126 PTB patients, 156 (13.9%) had their diagnosis of TB changed by a clinician. Multivariate analysis shows that patients whose diagnosis was based on CXR or other findings, female patients, patients who interrupted treatment for 2 months, patients who continued care at other health facilities (transfer) and patients with lung cancer were significantly more likely to have their diagnosis changed than other groups. CONCLUSION: A substantial proportion of patients were prescribed anti-tuberculosis treatment based on CXR findings alone, and a considerable proportion were advised to stop treatment before completing a full course, findings that require the immediate attention of Taiwan's National Tuberculosis Programme.