Mortality and its determinants during pre-treatment period among HIV adult patients in Tanzania.

Abstract

Background: HIV Care and Treatment services are guided by the National Care and Treatment Guideline. A recording and reporting system (paper and electronic) was developed to track those receiving services. Limited information is available on patients prior to the start of anti-retroviral therapy (ART), as most monitoring and evaluation of care and treatment programs concentrates on patients from ART initiation. Early initiation of antiretroviral therapy is desirable both in decreasing HIV transmission and having better treatment outcomes. In poor resource countries like Tanzania, delays in initiating ART is the reality. Greater understanding of the pre ART phase is crucial to help improve outcomes in HIV Care and determinants that affect the outcomes. Objective: The aim of this study was to assess the level of mortality and its determinants among HIV infected adults prior to ART initiation. Methods: A retrospective longitudinal analysis of more than 500,000 HIV infected adults enrolled in care was done from November 2004 and December 2011 using Stata 12. Patient characteristics at time of enrolment were described. Mortality rates and their 95% confidence intervals were estimated. Cox Proportional Hazard models were used to assess determinants of death before they access ART. Results: From 348 health facilities, a total of 526,059 clients (67% females) enrolled for HIV Care services were assessed in the pre ART phase. Majority (85%) had working functional status, 82% had done a CD4 test by 3rd month since enrolment and TB screening was done in 93% of all visits. Overall mortality among adult pre ART care patients was 37.6 deaths per 1000 person years (95% CI 36.9 - 38.3). Independent predictors of pre ART mortality were WHO stage 3 (adjusted hazard ratio [AHR] = 2.37; 95% CI 1.94-2.90), WHO Stage 4 (AHR= 4.53; 95 % CI (3.64-5.64). Female sex, CD4 ≥200 and weighing more than 45kg were significantly associated with a lower hazard of death. Conclusion and Recommendations: Mortality among patients who were eligible for ART before they accessed ART was high. Using routine electronic information to understand mortality and its predictors is resourceful for patient and program improvement. There is need to establish effective interventions targeting patients in the pre ART phase.