Abstract:

Objective To evaluate the effect of different active screening strategies on the detection of pulmonary tuberculosis (PTB) patients. Methods Cluster random sampling method was used to select 2 units from 11 communities (villages) in Lianjiangkou Town, Yingde City, Qingyuan City, Guangdong Province as the study site in December 2019. Basic information was collected from 2800 permanent residents aged 5 and older at the study site. Screening for tuberculosis symptoms,screening for Mycobacterium tuberculosis infection (interferon-gamma release test (IGRA)) and chest digital radiography (DR) were performed. Sputum samples from patients with positive results from any of the above screenings were examined by smear, culture and molecular biology tests. Using clinical diagnosis as a reference standard, performance of different screening strategies were evaluated by sensitivity, specificity and area under the receiver operating characteristic curve (AUC). Results Among the 2800 people undergoing active screening, there were 272 cases (9.71%) with suspected symptoms of tuberculosis, 301 cases (10.75%) with abnormal chest DR, and 617 cases (22.04%) with positive IGRA results. Among them, 67 cases (22.26%) with abnormal chest DR were presumptive tuberculosis patients. A total of 8 active pulmonary tuberculosis patients were diagnosed and the discovery rate was 285.71/100000.Except for 1 patient who had been registered and receiving treatment, the other 7 patients were all newly detected from aged 15 and older population with abnormal chest DR. With clinical diagnosis as the reference standard, chest DR screening had the best sensitivity and specificity (7/8 and 88.28% (2207/2500), respectively), and the highest AUC (0.88). Conclusion Among different active screening strategies, chest DR has the highest diagnostic value and can be used as the priority choice of active screening strategy

Key words: Tuberculosis, pulmonary, Communicable disease control, Multiphasic screening, Population surveillance, Data interpretation, statistical