Share
Title
The effect of unplanned care interruptions on the mortality of adults resuming antiretroviral therapy in South Africa: a survival analysis
Presenter
Haroon Moolla
Authors
H. Moolla1, M.-A. Davies1, C. Davies2, J. Euvrard1, H.W. Prozesky3, M.P. Fox4, C. Orrell5, P. von Groote6, L.F. Johnson1
Institutions
1Centre for Infectious Disease Epidemiology and Research, University of Cape Town, Cape Town, South Africa, 2Division of Epidemiology and Biostatistics, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa, 3Stellenbosch University, Department of Medicine, Cape Town, South Africa, 4Health Economics & Epidemiology Research Office, Department of Internal Medicine, University of the Witwatersrand, Johannesburg, South Africa, 5Desmond Tutu HIV Centre, Institute of Infectious Diseases and Molecular Medicine & Department of Medicine, University of Cape Town, Cape Town, South Africa, 6Institute of Social and Preventive Medicine (ISPM), University of Bern, Bern, Switzerland

BACKGROUND: Interrupting antiretroviral therapy (ART) is associated with adverse outcomes such as viral resistance, AIDS-defining illnesses, and mortality. However, little is known about the mortality of those resuming ART after unplanned interruptions. The objective of this study is to estimate the relative rate of mortality among adults resuming ART after unplanned interruptions compared to those who do not interrupt care.
METHODS: We included data from 44386 adults with HIV initiating ART between 2014 and 2019 at four South African cohorts of the International epidemiology Databases to Evaluate AIDS. We defined care interruption as a gap in contact longer than 180 days. Observation time prior to interruption was allocated to the “no interruption” group. Observation time after the first interruption was allocated to one of two groups based on whether the first interruption occurred before or after 6 months of ART. We determined vital status from clinic records and linkage to the National Population Register. Cox regression was used to estimate hazard ratios, which were adjusted for the effects of sex, baseline age, baseline CD4 count, year of ART initiation, and cohort.
RESULTS: There were 79762 person-years of observation and 797 deaths. 12601 people interrupted care, of whom 7038 interrupted within the first 6 months of ART. Those resuming ART experienced increased mortality compared to those without an interruption: those whose first interruption occurred within 6 months of starting ART had a 208% (95% CI: 158%-268%) increase in mortality, and those whose first interruption occurred after 6 months had a 147% (95% CI: 96%-211%) increase in mortality. Table 1 shows adjusted hazard ratios for selected variables.

Table 1: Adjusted hazard ratios for predictors of mortality

Variable

Person-years of observation
Deaths
Crude mortality rate per 1000 person-years (95% CI)
Adjusted hazard ratio (95% CI)

Interruption status
No ART interruption
62 228
484
7.78 (7.11 - 8.50)
1.00
Interruption within the first 6 months10 83520619.01 (16.59 - 21.79)3.08 (2.58 – 3.68)
Interruption after the first 6 months

6 69910715.97 (13.21 - 19.30)2.47 (1.96 – 3.11)
SexFemale54 7244107.49 (6.80 - 8.25)1.00
Male

25 03838715.46 (13.99 - 17.08)1.47 (1.27 - 1.70)
CD4 count at ART initiation (cells/mm3)500+11 851554.64 (3.56 – 6.04)1.00
350-49918 6561105.90 (4.89 – 7.11)1.31 (0.94 – 1.81)
200-34924 5471927.82 (6.79 – 9.01)1.60 (1.18 – 2.17)
<20024 70844017.81 (16.22 – 19.55)3.00 (2.25 – 4.00)

CONCLUSIONS: The substantially increased mortality of those resuming ART after an interruption highlights the need to prioritise and support retention in care, particularly during the first 6 months of ART.