Models of integration of TB and HIV services and factors associated with perceived quality of TB-HIV integrated service delivery in O. R Tambo District, South Africa.

BACKGROUND

Tuberculosis is the leading infectious cause of death among people living with HIV. Reducing morbidity and mortality from HIV-associated TB requires strong collaboration between TB and HIV services at all levels with fully integrated, people-centered models of care.

METHODS

This is a qualitative study design using principles of ethnography and the application of aggregate complexity theory. A total of 54 individual interviews with healthcare workers and patients took place in five primary healthcare facilities in the O.R. Tambo district. The participants were purposively selected until the data reached saturation point, and all interviews were tape-recorded. Quantitative analysis of qualitative data was used after coding ethnographic data, looking for emerging patterns, and counting the number of times a qualitative code occurred. A Likert scale was used to assess the perceived quality of TB/HIV integration. Regression models and canonical discriminant analyses were used to explore the associations between the perceived quality of TB and HIV integrated service delivery and independent predictors of interest using SPSS® version 23.0 (Chicago, IL) considering a type I error of 0.05.

RESULTS

Of the 54 participants, 39 (72.2%) reported that TB and HIV services were partially integrated while 15 (27.8%) participants reported that TB/HIV services were fully integrated. Using the Likert scale gradient, 23 (42.6%) participants perceived the quality of integrated TB/HIV services as poor while 13 (24.1%) and 18 (33.3%) perceived the quality of TB/HIV integrated services as moderate and excellent, respectively. Multiple linear regression analysis showed that access to healthcare services was significantly and independently associated with the perceived quality of integrated TB/HIV services following the equation: Y = 3.72-0.06X (adjusted R2 = 23%, p-value = 0.001). Canonical discriminant analysis (CDA) showed that in all 5 municipal facilities, long distances to healthcare facilities leading to reduced access to services were significantly more likely to be the most impeding factor, which is negatively influencing the perceived quality of integrated TB/HIV services, with functions' coefficients ranging from 9.175 in Mhlontlo to 16.514 in KSD (Wilk's Lambda = 0.750, p = 0.043).

CONCLUSION

HIV and TB integration is inadequate with limited access to healthcare services. Full integration (one-stop-shop services) is recommended.