A comparative study of Mycobacterium Tuberculosis in hospitalised adult HIV infected patients with normal and abnormal renal function at the University Teaching Hospital, Lusaka, Zambia

Abstract

Background: Mycobacterium tuberculosis (TB) remains a leading cause of mortality and morbidity worldwide, including Zambia, especially among those infected with the Human Immunodeficiency Virus (HIV). Both kidney dysfunction and TB have been shown to be highly prevalent among hospitalised HIV infected patients. Little is known about how TB and kidney dysfunction impact each other in HIV patients, and whether there is any association between the occurrence of kidney dysfunction and active TB infection in this population. This study was aimed at determining the prevalence and risk factors of active TB infection in HIV positive patients with and without kidney dysfunction.

Methods: This was an analytical cross-sectional study. Using simple random sampling, HIV positive patients on the medical wards were recruited in two arms (74 with & 59 without kidney dysfunction). Urine Lipoarabinomannan (LAM), TB blood culture, sputum culture and genexpert MTB/Rif were used for TB diagnosis. Data was analysed using STATA version 13.

Results: TB prevalence in all HIV positive hospitalized patients was 45%, and more prevalent in the kidney disease than non-kidney disease group (54% vs 35.59%; p=0.034). TB diagnosis pick up was comparable in the kidney disease and the non-kidney disease group using urine LAM and blood culture at 31.1% vs 22.2% and 8.9% vs 3.1% respectively, but lower using sputum culture; 12.5% vs 24.1%. Among kidney disease patients, a higher CD4 count > 200cells/μl was protective for active TB (P = 0.011). Severe immunosuppression (CD4 count < 200cells/μl) was 18.64% higher in the kidney dysfunction group compared to the non-kidney disease group (P=0.026). The only factor associated with active TB was male gender (P = 0.029); while Proteinuria in a TB patient was strongly associated with kidney disease (P < 0.001). Patients with WHO stage III/IV were likely to present with TB in both groups (P=0.004, 95% CI 1.47 - 7.20). Kidney dysfunction severity (measured by estimated glomerular filtration rate), age, antiretroviral therapy status and duration on combination antiretroviral therapy, history of TB contact and current cough, had no significant association with active TB in the two groups.

Conclusion: Patients with kidney disease are more likely to present with active TB infection than HIV infected  patients with no kidney disease. Among patients with active TB, urinalysis can help predict renal dysfunction; while determinants of active TB in the general population are not similar to those in kidney disease patients.