First line antiretroviral treatment failure and second line treatment outcomes among HIV patients in Southern Africa
Rohr_bu_0017E_11372.pdf (2.740Mb) Main dissertation
Rohr_bu_0017E_429/Study 1 Risk Group Calculator.xlsx (21.12Kb) View Study 1 Risk Group Calculator.xlsx
Rohr, Julia Katherine
MetadataShow full item record
Southern Africa has the highest prevalence of HIV worldwide, and South Africa has the highest number of HIV infected people. South Africa and other resource-limited countries provide antiretroviral therapy (ART) for people with HIV, with limited, standardized regimens for first line and second line. Patients who fail first line treatment are put on second line regimens, yet options for third line are very limited. The first study looks at predictors of first line treatment failure in South Africa and develops a predictive model that can estimate absolute risk of treatment failure over 5 years on ART, given a baseline profile of clinical and demographic factors. The model was developed with accelerated failure time models, using predictors that maximized discrimination between patients. The model can be used to identify patients who need adherence interventions, and to estimate how changes in baseline parameters in the population influence long-term need for second line ART. The second study explores whether delays from detection of first line treatment failure until second line treatment initiation, which are widespread in South Africa, decrease the effectiveness of second line ART. Marginal structural models were used to include patients who never switched to second line after failure in analysis. This study shows that, despite potency of second line drugs, short delays in second line among very sick patients can lead to worse outcomes. These findings may be due to drug resistance, immune system damage, and/or lack of adherence to medication. The third study examines whether switch in type of NRTI (nucleoside reverse transcriptase inhibitor, which is a drug class used in both first and second line regimens) from first line to second line improves outcomes on second line ART. While a switch in NRTI is recommended by treatment guidelines, it cannot always occur due to contraindications to some NRTIs. Using clinical data from South Africa and Zambia and adjusting for propensity scores, we see that switching from zidovudine in first line to a different NRTI in second line leads to less treatment failure on second line, suggesting that NRTI resistance may play a role in second line outcomes.