PROFIL KLINIS, VALIDITAS-RELIABILITAS, DAN KINERJA SKRINING TUBERKULOSIS

Abstract

Tuberculosis screening in community settings requires a brief questionnaire with measurable validity, reliability, and screening performance. This descriptive observational study analyzed screening data from activities conducted on 29 April – 18 Mei 2026 in Metro, Bandar Lampung, and Pringsewu. A total of 221 respondents were included. Most respondents were of productive age (85.1%), female (68.8%), and residents of Lampung (96.8%). The most frequent symptoms were cough for more than two weeks (11.8%), night sweats (7.7%), weight loss (5.9%), hemoptysis (2.3%), fever for more than two weeks (1.4%), and lymph node swelling (0.5%). Cough for more than two weeks, hemoptysis, weight loss, and night sweats were significantly associated with the screening result (p<0.001), while fever and lymph node swelling were not significant. Using the program screening result as the comparator, any symptom had a sensitivity of 57.5% and specificity of 100.0%; symptom or risk-factor criteria had a sensitivity of 84.9% and specificity of 100.0%. All six symptom items were valid (r=0.427-0.782; r table=0.132), with Cronbach alpha of 0.612. The questionnaire is useful as an initial screening tool, but confirmatory examination remains necessary for respondents with positive results.