How soon should patients be eligible for differentiated service delivery models for antiretroviral 1 treatment? Evidence from Zambia 2

Introduction : Attrition from HIV treatment is high during patients’ first 6 months after antiretroviral 42 therapy (ART) initiation and patients with less than 6 months on ART are systematically excluded 43 from most differentiated service delivery (DSD) models, which are intended to reduce attrition. 44 Despite eligibility criteria requiring greater than 6 months on ART, some patients enroll earlier. Using 45 routinely-collected medical record data in Zambia, we compared loss to follow-up (LTFU) among 46 patients enrolling in DSD models early (<6 months on ART) to LTFU among those who enrolled 47 according to guidelines ( ≥ 6 months on ART) in order to assess whether the ART experience eligibility 48 criterion is necessary. 49 50 Methods : We extracted data from electronic medical records for adults ( ≥ 15 years) who initiated 51 ART between 01/01/2019 and 31/12/2019 and evaluated LTFU, defined as >90 days late for last 52 scheduled medication pickup, at 18 months for “early enrollers” (DSD enrolment after <6 months on 53 ART) and “established enrollers” (DSD enrolment after ≥ 6 months on ART). We used a log-binomial 54 model to compare LTFU risk between groups, adjusting for age, sex, urban/rural status, ART refill 55 interval and DSD model. 56 57 Results : For 6,340 early enrollers and 25,857 established enrollers there were no important differences between the groups in sex (61% female), age (median 37 years), or setting (65% urban). 59 ART refill intervals were longer for established vs early enrollers (72% vs 55% were given 4–6- month refills). LTFU at 18 months was 3% (192/6,340) for early enrollers and 5% (24,646/25,857)


Introduction
We defined patients who enrolled into a DSD model with <6 months of ART as "early enrollers", 1 2 5 while a comparison group of patients who enrolled into a DSD model with ≥ 6 months of ART as 1 2 6 "established enrollers". Patients on second-line ART (defined as those dispensed protease inhibitors 1 2 7 such as lopinavir, atazanavir or ritonavir) were excluded from this analysis, as they are already known 1 2 8 to be at high risk of attrition [18,19]. For both early and established enrollers, we assessed loss to 1 2 9 follow-up (LTFU) at 18 months post-ART initiation, with LTFU defined as patients who were 1 3 0 reported as "lost to follow-up" or "inactive" in the SmartCare database between 15 and 21 months 1 3 1 after ART initiation date. "Inactive" was defined as having missed a scheduled visit by more than 30 1 3 2 days. Rates of LTFU were calculated for early and established enrollers and stratified by DSD model 1 3 3 type and ART dispensing duration. DSD models, which had multiple names in the SmartCare 1 3 4 database, were grouped into the following categories: 1) adherence groups (community adherence 1 3 5 groups, rural/urban adherence groups); 2) extended clinic hours (DSD models designed for clinic 1 3 6 access before/after hours or weekends, including scholar models); 3) fast-track (procedures to 1 3 7 accelerate dispensing at clinics); 4) home ART delivery; 5) multi-month dispensing (MMD); and 6) 1 3 8 community pick-up point (central dispensing units, community retail pharmacies, community ART 1 3 9 distribution points, health posts, mobile ART distribution models) (Table 1). 1 4 0 1 4 1 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 13, 2022. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 13, 2022. ; https://doi.org/10.1101/2021.08.25.21262587 doi: medRxiv preprint 1 4 4 Statistical analysis 1 4 5 We described the demographics of our study population using descriptive statistics. We compared 1 4 6 loss to follow-up risk between early enrollers and established enrollers and Wilson's score interval 1 4 7 was used to calculate 95% confidence intervals around proportions. We used a log-binomial 1 4 8 regression to calculate risk ratios for loss to follow-up, adjusting for age, sex, urban/rural status, DSD 1 4 9 model type and ART dispensing duration. We also conducted an age-stratified analysis and a sub- The full SmartCare data set included 1,520,125 unique patients on ART over 2018-2021, of which 1 6 2 32,197 patients had enrolled into a DSD model after ART initiation and had an 18-month outcome 1 6 3 reported within the 15-to-21-month window ( Figure 1). Of these, 6,340 patients were reported to have 1 6 4 been enrolled in DSD models <6 months after ART initiation during the study period (early enrollers). 1 6 5 The remaining 25,857 patients comprised the comparison group of established enrollers. For early 1 6 6 enrollers, median time enrolled in a DSD model at the time of outcome evaluation was 14.7 months 1 6 7 (IQR 13.0-16.5); majority (81%, n=20,856) of established enrollers were on DSD models at outcome 1 6 8 evaluation at a median of 5.8 months (interquartile range (IQR) 2.9-8.9) ( Table 2). Early enrollers and 1 6 9 established enrollers were similar with respect to age, sex and urban/rural location. Across both 1 7 0 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) groups, the median age was 37 years (IQR 29 -44), a majority (61%, 19,580/32,197) were female 1 7 1 and most patients resided in urban settings (64%, n=20,618). . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 13, 2022. ;https://doi.org/10.1101https://doi.org/10. /2021 In an analysis adjusting for age, sex, location, and ART dispensing duration or DSD model type 2 0 4 (where applicable),, early enrollers in all DSD model types and dispensing durations were 41% less 2 0 5 likely to be lost to follow-up than established enrollers (adjusted risk ratio (aRR) 0.59 [0.50-0.68]) 2 0 6 ( Figure 2). The reduced adjusted risk of being lost to follow-up were similar for patients in adherence . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 13, 2022. ;https://doi.org/10.1101https://doi.org/10. /2021  An age-stratified analysis produced similar results to the main analysis, with early enrollers in each 2 2 0 age group being less likely to be lost to follow-up than established enrollers in the same age group. 2 2 1 However, the effect of earlier enrollment in DSD on reduced loss to follow-up appeared less 2 2 2 pronounced in patients on 4-6 months' ART dispensing for those aged 25 to 49 years (Appendix 2 2 3 Figure S1). In facilities where a larger proportion of all DSD patients enrolled in DSD models early, 2 2 4 the trend towards early enrollers performing better persisted with respect to loss to follow-up 2 2 5 compared to outcomes for established enrollers (Appendix Figure S2). In nearly all of sub-Saharan Africa, DSD model eligibility criteria require that patients be on ART for 2 2 9 a minimum of six months (and in some countries a minimum of 12 months) prior to DSD model 2 3 0 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 13, 2022. ; https://doi.org/10.1101/2021.08.25.21262587 doi: medRxiv preprint enrollment [21]. We present novel data from Zambia highlighting good outcomes when newly 2 3 1 initiated ART patients (those with less than 6 months' ART experience) are referred early to DSD 2 3 2 models. Those referred early to DSD appear to have good outcomes across different DSD models and 2 3 3 age categories. While ART patients in Zambia have historically been lost to follow-up at high rates in the first few 2 4 5 months after ART initiation [3], in our DSD patient population this was less likely to be the case. Our 2 4 6 results provide evidence to support the recent revision of WHO guidelines that reduce time on ART 2 4 7 from 12 to six months on treatment as part the definition of "established" on ART [14]. These 2 4 8 findings offer reassurance and evidence to countries that have expanded eligibility as they scale up 2 4 9 DSD models [21,23], particularly to support uninterrupted access to HIV treatment during the 2 5 0 COVID-19 pandemic, that earlier referral to DSD is possible without compromising patient care.

5 1
Even if many, or most, of the patients in our "early enrollment" sample were selected deliberately 2 5 2 because they were considered at low loss to follow-up risk, our results demonstrate that early 2 5 3 eligibility for DSD models should be considered for at least some patients before they reach six 2 5 4 months on ART. Loss to follow up at 18 months after ART initiation for early and established enrollers averaged 1-2 5 7 11% for all six categories of DSD models studied. We did not observe any programmatically 2 5 8 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 13, 2022. ;https://doi.org/10.1101https://doi.org/10. /2021