Page 136 - SAHCS HIVMed Journal Vol 20 No 1 2019

Page 3 of 7 Original Research

patients who were identified as having participated in both summarised using medians and interquartile ranges (IQRs)
programmes. Patients were excluded if they were enrolled in for continuous variables and proportions for categorical
an AC before the ROTF intervention, enrolled in a family variables. Cross-sectional retention outcomes are reported at
AC (utilised for children and their caregivers), missing study closure. Kaplan–Meier methods were used to estimate
from the AC register or confirmed to have never joined an the survival probabilities of retention, AC retention and
AC (indicating EMR AC participation incorrect), never viral suppression, and are reported at 3-monthly intervals to
suppressed after ROTF or if they never had a VL greater 18 months with 95% confidence intervals (CIs).
than 400 copies/mL (indicating EMR ROTF participation
incorrect) (Figure 1). One AC register could not be found, and Data were analysed using Stata 13.0 software (STATA
all patients referred to that AC were excluded from analysis. Corporation, College Station, TX, US).
This left only high-risk patients confirmed to have joined
ACs directly after participation and successful suppression Ethical consideration
following the ROTF intervention.
Because of the nature of the study, individual patient consent
Data collection was not obtained, consistent with the Declaration of Helsinki.
All participants and data were drawn from an ongoing cohort
Data for each patient in the analysis cohort were collected
from patient visit and laboratory data from the EMR and AC study of routine ART outcomes in Khayelitsha, Cape Town,
registers. Missing VL results were obtained from the National approved by the Human Research Ethics Committee of the
Health Laboratory Service database. Patient clinic folders Faculty of Health Sciences at the University of Cape Town
were consulted for patients whose most recent status was (HREC 395/2005). Only routine clinical service data were
missing from the AC registers to confirm their current AC used and no identifying patient information was entered into
status. Key variables collected included ART regimen, ART the database.
start date, ROTF enrolment date, last unsuppressed VL and
date, first suppressed VL and date, all VLs and dates after Results
club enrolment and all clinic and club visits after suppression. Patient characteristics
From February 2012 to February 2014, 165 high-risk patients
Statistical analysis
who completed the ROTF intervention and suppressed were
Patients entered the analysis on their first AC date (between immediately enrolled in an AC. The cohort was predominantly
February 2012 and February 2014) and were followed until female (81.8%) with a median age at ART start of 31 years
March 24, 2015. We assessed three outcomes: retention in (IQR: 28–37). Current treatment regimens were available for
care, retention in club care and viral suppression. Retention 133 patients, and of those 105 (79%) were on second-line
in care was defined as having contact with the clinic or AC therapy (Table 1) at the time of AC enrolment. The median
between March 24 and June 21, 2015, with retention in club time from ART initiation to enrolment in the ROTF
care defined as attending an AC in the same period. Patients intervention was 3.4 years (IQR: 2.1–5.5), and the median
were classified as virally suppressed if their last VL before time from ROTF intervention to AC enrolment was 1.2 years
analysis closure was less than 400 copies/mL. We define viral (IQR: 1.0–1.5).
rebound as an elevated VL above 400 copies/mL after having
achieved viral suppression. Known deaths and transfers
contributed retention time until they were censored at the Cross-sectional outcomes
time of death or transfer. During the study period, two patients (1.2%) died, 15 (7.8%)
were lost to follow-up and 40 (24.0%) experienced viral
Patient characteristics at enrolment into an AC (gender, age rebound. At the closure of the study, 148 patients (89.0%)
at ART start, age at AC start, year of ART start, treatment were retained in care and 119 patients (72.0%) were virally
regimen) and time from ART initiation to ROTF participation suppressed. When stratified by known ART regimen, 26
and from ROTF participation to AC enrolment were patients (93.0%) on first line and 97 patients (92.0%) on
second line were retained in care, while 20 patients (71.0%)
Iden fied in both AC and on first line and 83 patients (79.0%) on second line were
ROTF database virally suppressed.
N = 321
Assessed for eligibility Excluded n = 165
• Never re-suppressed n = 5 Time to event outcomes
• Family clubs n = 16
• Missing club number n = 8
• Never in club n = 11 Retention in care was estimated to be 98.8% (95% CI,
• Not ROTF (no high VL) n = 18 94.4–99.4), 94.8% (95% CI, 89.8–97.4) and 89.3% (95% CI,
• Missing register n = 10
• Missing from EMR and register n = 10 81.8–93.8) at 6, 12 and 18 months after AC enrolment,
Final analysis • ROTF aŠer AC enrolment n = 78
n = 165 respectively (Table 2, Figure 2a). Retention in AC care was
estimated to be 98.2% (95% CI, 94.4–99.4), 92.0% (95% CI,
AC, adherence club; ROTF, risk of treatment failure; EMR, electronic monitoring records;
VL, viral load. 86.3–95.4) and 80.5% (95% CI, 72.0–86.6) over the same time
FIGURE 1: Flow chart of analysis inclusion criteria applied to obtain study sample. periods (Table 2, Figure 2b). Eighteen months after enrolment
http://www.sajhivmed.org.za 129 Open Access