| Summary: | A research report submitted in partial fulfillment of the requirements for the degree of
Master of Science in Epidemiology and Biostatistics
University of the Witwatersrand,
Faculty of Health Sciences, School of Public Health
May, 2014 === Introduction
HIV-associated TB is curable with standard TB therapy and yet it is the leading cause of illness and death in patients infected with HIV. Coinfection with HIV poses considerable challenges to early diagnosis of TB in HIV-infected people and diagnostic delay and the rapidly progressive TB associated with HIV results in rapid clinical deterioration and increased mortality. There is an urgent need for research to identify risk factors for TB in HIV-infected people in order to refine diagnostic algorithms for the early and accurate diagnosis of tuberculosis in HIV-positive patients.
The aim of this study was to determine the prevalence of TB, and identify factors associated with prevalent TB in HIV-infected adults paying a subsidized, all-inclusive monthly fee for HIV care in a private setting in downtown Johannesburg – a unique sub-population.
Material and Methods
This study was a retrospective, cross-sectional, secondary analysis of data extracted from the routine electronic medical records of HIV-infected adults who attended ZuziMpilo from August 2009 to December 2011. The outcome of interest was prevalent TB at the time of enrollment into care and exposures included age, sex, ethnicity, CD4 count, WHO Clinical Stage, BMI, alcohol and tobacco use history, level of education, employment status, monthly income, monthly cell phone expenditure, medical insurance status, source of funding for HIV care and source of knowledge about ZuziMpilo Medical Centre.
Multivariable logistic regression modeling was used to determine risk factors for prevalent TB at the time of enrollment.
Results
Approximately 8 out of every 100 HIV-infected adults enrolling at ZuziMpilo from August 2009 to December 2011 had prevalent TB disease (8.24%). Significant predictors for prevalent TB included BMI categorised as non-obese, CD4 count <350 cells/mm3 and duration on HAART of less than six months. With respect to BMI, individuals who were not obese had greater risk of prevalent TB, the risk increasing in a dose response fashion as the BMI decreased. Compared to obese patients, overweight patients were 2.8 times as likely (aOR = 2.83, CI 1.06 – 7.52, p = 0.037), normal weight patients were more than 3.7 times more likely (aOR = 3.72, CI 1.44 – 9.60, p = 0.007) and underweight patients were more than 6.4 times more likely to have prevalent TB (aOR = 6.42, CI 2.33 – 17.70, p = 0.000). A CD4 count of < 200 cells/mm3 predicted an 11.3 times increased risk of prevalent TB relative to CD4 count greater than 350 cells/mm3 (aOR = 11.27, CI 4.84 – 26.28, p = 0.000). Patients treated with HAART for longer than 6 months were significantly less likely to have prevalent TB than HAART-naïve patients (aOR = 0.47, CI 0.23 – 0.98, p = 0.043).
Conclusion and Recommendations
This study corroborates the growing body of evidence that underpins several key recommendations that have the potential to reduce mortality from TB in those people infected with HIV; vigilant and regular routine TB screening in HIV-infected patients at all CD4 counts and especially in those with profound immunosuppression and in the first three
to four months following HAART initiation, the urgent development and distribution of more sensitive and point of care diagnostic tests for TB in HIV-infected patients at all levels of health care (most especially primary health care facilities) and the importance of initiation of HAART before CD4 counts drop below 350 cells/mm3. This study highlights that BMI is a useful proxy marker of TB risk among HIV-positive individuals. Height and weight are easily assessed anthropomorphic measures and should be conducted routinely in all patients at regular intervals.
This study has described a unique population with the capacity to pay a subsidised monthly fee for their HIV care and thus the results may not be generalisable to the large population of HIV-infected adults in South Africa, who receive free-of-charge health care in public sector facilities. However, they may certainly be generalisable to other clinics that provide services for a fee and this information may be especially important if this model is replicated and scaled up in private and semi-private facilities around the country.
|
|---|
