Occurence and determinants of treatment faiure in antiretroviral therapy at Tshwane District Hospital

Sokoya, Temitope
Journal Title
Journal ISSN
Volume Title
Objective: To determine the proportion of HIV+ patients who fail treatment on a yearly basis in a 5-year treatment cohort in Tshwane District Hospital and to determine the correlation of treatment failure with variables routinely measured at the clinic namely WHO stage, CD4 count, HIV viral load, age, gender, presence of concomitant diseases, concomitant medication and distance travelled to clinic. Design: A retrospective study with an analytical component was conducted using the hospital records of adult patients receiving antiretroviral therapy in 2004 and followed for 5 years (until 2009) at the Tshwane District Hospital. Methods: All adult patients receiving antiretroviral therapy in 2004 were identified and followed for the next 5 years till 2009 at Tshwane District Hospital. The proportion of patients that failed treatment yearly was calculated. Univariate analysis was used to compare all patients who failed at any time point with the patients who did not fail at all for all variables. A repeated measures logistic regression model was developed to determine the variables that impacted on the binary outcome, namely treatment failure or not. Results: Of the 1104 adult patients who were attending the TDH Immunology clinic in 2004, 870 adults were receiving ARVs. 333 patients (38.28 %) experienced treatment failure throughout the study period. 6.9 % (60/870) of the study population failed virologically. 307 of the 870 patients (35.29 %) failed treatment immunologically. 102 patients (11.72 %) experience treatment failure at the 12 month time point, 37 patients(4.49 %) at the 24 month time point, 57 patients(6.93 %) at 36 month time point, 101 patients(12.27 %) at the 48 month time point and 140 patients (7.01 %) failed treatment at 60 month time point. Univariate analysis showed significant correlation between treatment failure and non-adherer, interrupting treatment, defaulted treatment, viral load at baseline, 12, 24, 36, 48, 60 months, and CD4 count at baseline, 12, 24, 36, 48, 60 months. In the multivariate analysis, there was a significant association between short term stoppage of treatment (STSTOP) (coefficient ratio = 1.41; p<0.001), long term stoppage of treatment (LTSTOP) (coefficient ratio = 3.24; p<0.001), transfers from other health institutions (coefficient ratio = 1.96; p<0.001), regimen (coefficient ratio = -0.1734) and treatment failure. The change in log viral load at 12 months from baseline (LOGVLBL12) (coefficient ratio =-1.7145; p<0.001) was highly significant for reaching the end point - treatment v failure. Older patients were less likely to fail treatment (coefficient ratio = -0.0517, p<0.001) and patients with an advance stage of the disease (WHO stage 3 or 4) were at a lower risk of failing treatment (coefficient ratio = -0.4175; P=0.008). The CD4 count was significant in the univariate analysis P<0.01) and XTGEE (coefficient ratio =- 0.0001; p<0.001). There was no significant correlation between gender, place of residence, employment status and treatment failure. Conclusion: More than one–third of the patients receiving treatment in TDH failed treatment within the 5 year study period. The determinants of treatment failure are age, WHO stage, transfer from other institutions, short term stoppage of treatment, long term stoppage of treatment, CD4 cell count and the level of viral suppression within the first year of treatment (LOGVLBL-12). This study reinforces the need for identifying high risk patients earlier in treatment in order to implement strategies that might strengthen adherence to treatment.
A research report submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master of Science in Pharmaceutical Affairs Johannesburg, 2012