A CROSS SECTIONAL SURVEY OF PULMONARY TUBERCULOSIS AMONG ELDERLY DIABETICS ATTENDING PRIMARY CARE CLINICS IN PENANG, MALAYSIA

摘要

Latent tuberculosis (TB) is more likely to convert to active TB with increasing age and with poorly controlled diabetes mellitus (DM). We aimed to determine the prevalence of and factors associated with pulmonary tuberculosis (PTB) among elderly patients with DM attending public primary care clinics in Penang, Malaysia in order to inform TB control efforts among elderly with DM. Study subjects were selected from 15 primary care clinics. A total of 4,209 subjects were calculated to be needed for the study assuming a PTB prevalence among elderly diabetics of 3.89 per 1,000 among the study population obtained from the National Diabetic Registry (NDR) of patients aged >= 60 years consisting of 22,980 patients. Study subjects were randomly selected from the NDR of patients aged >= 60 years. Inclusion criteria were being aged >= 60 years and having DM. The exclusion criteria was being treated for TB. A total of 4,209 subjects were included in the study. Each subject was interviewed and asked about PTB symptoms defined as cough > 2 weeks, weight loss, night sweats or fever > 4 weeks and other past medical problems. Each subject had a chest x-ray (CXR) and blood test for serum creatinine and glycated hemoglobin (HbA1c). Subjects with a cough were asked to give a sputum sample for acid-fast bacilli (AFB) stain. Subjects were classified as either having sputum smear positive PTB, sputum smear negative PTB or not having PTB. Criteria used to classify a subject as having sputum smear positive PTB were having a sputum smear positive for AFB and having an abnormal chest radiograph (CXR) with findings typical of active PTB. Sputum smear negative PTB was defined as having a sputum smear negative for AFB, having symptoms of PTB and having an abnormal CXR with findings typical of active PTB. Sputum smear positive PTB and smear negative PTB were grouped as having PTB in this study. The socio-demographic and clinical factors associated with PTB were determined using logistic regression analysis. A total of 4,209 subjects were included in the study. The mean (+/- standard deviation) age of study subject was 68.8 (+/- 6.4) years. Eight out of 4,209 subjects (0.19%) were diagnosed with having PTB; 6 (0.14%) were classified as having smear positive PTB and 2 (0.05%) as having smear negative PTB. This gives a PTB prevalence of 1.9 per 1,000 population. This is far lower than the 3.9 per 1,000 prevalence used to calculate the number needed for the study. The mean HbA1c among subjects with PTB was significantly higher than among subjects without PTB (OR=1.3; 95% CI: 1.01-1.76; p=0.040). There were no other factors significantly associated with having PTB. Diabetic patients aged >= 60 years with an elevated HbA1c having symptoms of PTB in the study population should be screened for PTB.