Prevalence and Clinical Predictors of Pulmonary Tuberculosis Among Adults Diabetic Patients at Bugando Medical Centre, Mwanza Tanzania
- Mwanza, Tanzania: Catholic University of Health and Allied Sciences CUHAS - Bugando c2012
- xiii; 51 Pages Includes References and Appendices
Abstract:
Background: There is a rise in the incidence and prevalence of diabetes mellitus in low-income countries which have also increased burden of infectious diseases like TB. Many studies shown that diabetes mellitus increases susceptibility to TB and prevalence of TB among diabetic patients depends on TB burden in a given area. Despite of the evidence that diabetic patients are prone to develop active TB compared to general population there is no current study done in Tanzania to assess the burden of TB and also no screening protocol for pulmonary tuberculosis (PTB) in diabetic community.
Methodology: The objective of this study was to determine, presenting features and chest radiography finding of PTB in adults diabetic patient attending diabetic clinic or admitted at Bugando Medical Centre and intensively screened for PTB with sputum for AFB and mycobacterial culture. This study was a cross-sectional hospital based study whereby patients either attending at diabetic clinic or admitted at Bugando Medical Centre who were known diabetics and come with symptom of cough regardless of duration were screened for PTB.
Results: About 700 diabetic patients attended at Bugando Medical Centre during the study period. We managed to recruit 693 (99%) patients and 2 (0.2%) were already on anti-TB drugs. Of the 691 patients screened for cough, 121 (17.5%) had cough and among patients with cough 34 (28%) had productive cough. The prevalence of pulmonary tuberculosis in diabetic patients was found to be 1.3%. Inpatients and patients who reported current weight loss had a significant increased risk of PTB (OR=10.2[CI=1.07-97], p=0.035 and OR=10.5[CI=1.6-70.8], p=0.016 respectively). Patients with symptoms of TB like fever and drenching sweating were more likely to have PTB (OR= 1.8 and 1.4 respectively) but were not statistically significant (P-value=0.51 and 0.73 respectively). Presence of cavitation’s, upper and lower lobe opacities increased the possibility of PTB (OR=10.67 [CI=1.31-86.93, P=0.032], OR=10.67[CI=1.31-86.93, P=0.032] and OR=20 [CI=2.2-181, P=0.007] respectively) and lower lobe opacity was a strong predictor of PTB (OR=21, 1.41-314, P=0.025).
We found that the majority of patients had dry cough (n=87, 72%) and predictors which showed a significant increased risk of dry cough were female gender (OR=4.3, CI= [1.8-10.4], P=0.001), outpatients (OR=7.7, CI= [2.9-20.2], P <0.000) and BMI above 27 (OR=1.1, CI= [1.0-1.2], P=0.03). Use if ACE inhibitors increased risk of dry cough was not statistically significant (OR=2, CI= [0.8-4.2], P=0.131). Diabetic patients with fever, drenching night sweating and recent weight loss were more likely to have productive cough (OR=0.3 [CI=0.1-0.8 P=0.014], OR=0.3 [CI=0.1-1.0 P=0.043] and OR=0.3 [CI=0.1-0.9 P=0.027] respectively).
Conclusions: The prevalence of PTB among diabetic adults is higher than general population. Symptoms of PTB are very common among diabetics. It is not clear whether those symptoms were due to PTB or other complication of DM. Screening for PTB with sputum examination for AFB and culture did not address the problem of the large number of adult diabetic patients with non-productive cough and other symptoms of PTB.
Recommendations: Screening of PTB in DM patients in sub-Saharan Africa should not rely only on sputum examination and patients with symptoms of PTB (cough ≥2 wks, fever, weight loss and night sweats) should receive further investigations for PTB. Also we recommended the use of HIV/TB screening tool in diabetic clinic for PTB screening. A simple, reliable, sensitive and cost effective method of diagnosing PTB in diabetic patients at low-income setting should be developed. Newer diagnostic tests like urine LAM and/or Gene-Xpert should be evaluated among DM patients. Studies evaluating the causes of cough in adult diabetic population in our environment are required.