Table 1.
Characteristics of the studies and their Design, nHFBC model and key findings
| Study | Setting | Non-facility based model | Comparator | Sample Size | Length of follow-up | Outcomes and key findings |
|---|---|---|---|---|---|---|
| RANDOMIZED CONTROL TRIALS | ||||||
| Fox 2019 [37] | South Africa | Adherence clubs | Health care facility |
N = 596 AC n = 275 HCF n = 294 |
18 months |
Viral Suppression – comparable 12 months viral suppression between the intervention (80%) and control (79.6%) arms (aRD: 3.8%; 95% CI: −6.9 to14.4%). Retention – AC’s had a higher 1-year retention (89.5% vs 81.6%, aRD:8.3%; 95% CI: 1.1 to 15.6%) |
| Hanrahan 2019 [40] | South Africa | Community Adherence clubs | Health care facility clubs (Standard of care) | N = 775 | 24 months |
Loss from the club – proportion of patients who dropped out of clubs in both community and facility clubs or were transitioned to standard of care. Overall, 47% [95%CI 44–51%] of patients were returned to health care facility. Among community-based club participants, the cumulative proportion lost from club-based care was 52% (95% CI: 47–57%), compared to 43% (95% CI: 38–48%, p = 0.002) among clinic-based club participants. Virological failure - Documented viral rebound was higher among participants assigned to facility-based clubs (21, 95% CI 13–27%) than those assigned to community-clubs (13, 95% CI 8–18%, p = 0.051). But this was not significant. All-cause mortality – no mortality observed in both arms Loss from ART care -during follow up, 77 (10%) overall. No significance between the two arms. Among community club participants, the proportion lost from any ART care was 12% (95% CI 9–16%), compared to 7% (95% CI 5–10%, p = 0.024) among facility- club participants, corresponding to a difference of 5% (95% CI 1–9%, p = 0.018). In a univariate Cox proportional hazards model, the risk of loss to any ART care was non-significantly increased among participants assigned to community clubs as compared with those assigned to facility clubs (HR 1.69, 95% CI 0.98–2.91, p = 0.057). |
| Geldsetzer 2018 [41] | Tanzania | Home ART delivery | Health care facility |
N = 2172 HD n = 1163 HCF n = 1009 |
326 days |
Virological failure – 10.9% (95/872) in the control arm and 9.7% (91/943) in the intervention arm were failing at the end of the study period. Risk ratio demonstrated non-inferiority of the HBC to HCF (RR 0.89 [1-sided 95% CI 0.00–1.18]) Lost to follow-up – 18.9% in HBD versus 13.6% in HCF. No P value or CI reported. Mortality – 0.09% in HBD versus 0.2% in HCF. No P value or CI reported. |
| Woodd 2014 [42] | Uganda | Home ART delivery | Health care facility |
N = 1453 HD n = 859 HCF n = 594 |
28 months |
Home delivery of ART and support leads to similar survival rates as clinic-based care. Mortality – One hundred and ninety-seven participants died over a median follow-up time of 28 months (IQR 15–35) giving an overall mortality rate of 6.36 deaths per 100 person-years [95% confidence interval (CI) 5.53–7.32]. 110 (25%) deaths in participants with baseline CD4 < 50 cells and 87 (9%) in those with higher baseline CD4.Among participants with baseline CD4+ count < 50 cells/μl, mortality rates were similar for the home and facility-based arms; adjusted mortality rate ratio 0.80 [95% confidence interval (CI) 0.53–1.18] compared with 1.22 (95% CI 0.78–1.89) for those who presented with higher CD4+ cell count. In CD4 counts < 50 cells – crude mortality RR 0.81 and In CD4 counts higher - crude mortality RR 0.55 Lost to follow up – 1.8% among those with CD4 < 50 and 2.6% among those with CD4 at least 50. |
| Amuron 2011 [43] | Uganda | Home deliveries | Health care facility |
HD n = 594 HCF n = 859 |
42 months | Mortality – in the facility there were 117 deaths (mortality rate 6.3 per 100 persons per yrs.) whereas in HBD, 80 deaths (mortality rate 6.5 per 100 person yrs.). The one, two and three year survival probabilities (95% CI) were 0.89 (0.87–0.91), 0.86 (0.84–0.88) and 0.85 (0.83–0.87) respectively |
| Selke 2010 [44] | Kenya | Home ART delivery | Health care facility |
HD n = 96 HCF n = 112 |
28 months |
Home delivery of ART and support resulted in similar clinical outcomes as clinic care but with half the number of clinic visits. Task-shifting and mobile technologies can deliver safe and effective community-based care to PLHIV. LTFU – 4.5% in the HCF and 5.2% in Home delivery [95% CI: 0.24 to 3.03; p = 1.0] Mortality – 0 in both arms Viral rebound – no significant difference between the two groups (10.5% in HBD and 13.5% in HCF, 95%CI: 0.54 to 3.31, p = 0.65) |
| OBSERVATIONAL COHORT STUDIES | ||||||
| Fox 2019 [37] | South Africa | Decentralized medication delivery (DMD) | Health care facility |
N = 578 DMD n = 232 HCF n = 346 |
18 months | |
| Tun 2019 [45] | Tanzania | Community Based ART distribution (CBPDs) | Health care facility |
CBPD n = 309 HCF n = 308 |
6 months |
Retention in the CBDP – 82.8% vs 82.1% in the HCF at 6 months LTFU – 53 in the intervention and 55 in the HCF arms |
| Pasipamire 2018 [46] | Swaziland |
1. Community Adherence groups (CAGs) 2. Facility Based clubs 3. Treatment outreach |
No comparator |
N = 918 CAGs n = 531 FBC n = 289 Outreach n = 98 |
12 months |
Retention in the models – The overall care model retention was 90.9 and 82.2% at 6 and 12 months. Retention in the care models differed significantly by model type, being lowest in CAGs at all time points (p < 0.001). Only 70.4% of patients were retained in CAGs at 12 months compared with 86.3% in comprehensive outreach and 90.4% in clubs. Retention in care model was significantly higher in eligible patients compared with non-eligible patients (85.0 and 76.4% at 12 months, p = 0.017. Retention to ART – over 90% from all three models and no difference noted (p = 0.52).Patients in CAGs had a higher risk of disengaging from the care model (aHR 3.15, 95%CI: 2.01–4.95, P < 0.001) compared with treatment clubs. Note: disengagement defined as LTFU, Death, return to clinical care) |
| Myer 2017 [47] | South Africa | Adherence clubs [post-partum women] | Health care facility |
N = 110 AC n = 77 HCF n = 33 |
6 months post-partum follow-up |
Viral suppression - overall no difference in viral suppression between the two groups. 86% of women remained in the evaluation through 6 months postpartum; in this group, there were no differences in VL < 1000 copies/mL at six months postpartum between women choosing HCFs (88%) vs. adherence clubs (92%; p = 0.483. |
| Vogt 2017 [48] | Democratic Republic of Congo (DRC) | Community based refill centers | No comparator | N = 2259 | 24 months |
Attrition increased steadily after decentralizing services such as drug pick up points. Low attrition throughout follow-up LTFU – 9.0% at 24 months Mortality – 0.3% at 24 months overall attrition was 5.66/100 person years (95% CI: 4.97 to 6.45) |
| Tsondai 2017 [49] | South Africa | Adherence clubs | No comparator | N = 3216 | 24 months |
Stable patients on ART can safely be offered differentiated care as they overall had good outcomes. Adherence clubs scaled up at large scale had had high levels of retention and viral suppression. Retention – Retention was 95.2% (95% CI: 94.0–96.4) at 12 months and 89.3% (95% CI: 87.1–91.4) at 24 months after AC enrolment. Viral suppression - Of the 88.1% who had a viral load assessment, 97.2% (95%CI, 96.5–97.8) were virally suppressed < 400 copies/ml LTFU – 4.2% (135). Cumulative incidence of LTFU was 2.6% (95% CI, 2.1–3.2) at 12 months, rising to 6.9% (95%CI, 5.7 to 8.1) at 24 months after AC enrolment. Mortality – 0.1% (95% CI, − 0.01 to 0.2) at 12 months and 0.2% (95%CI, − 0.01 to 0.4) |
| Decroo 2017 [50] | Mozambique | Community ART groups (CAGs) | Health care facility |
CAGs n = 901 HCF n = 1505 |
24 months |
LTFU – overall 12% [11.2% in HCF and 0.8% in CAGs]. CAG members had a greater than fivefold reduction in risk of dying or being LTFU (adjusted HR: 0.18, 95% CI 0.11 to 0.29). Retention - 12-month and 24-month retention in care from the time of eligibility were 89.5 and 82.3% respectively among patients in individual care and 99.1 and 97.5% among those in CAGs (p < 0.0001). |
| Auld 2016 [51] | Mozambique | Community support ART groups (CASG) | Health care facility |
N = 306,335 CASG n = 6766 HCF n = 299,569 |
4 years |
Mortality – similar rates in both groups [0.3% among CASG at 2 yrs. and 1.4% at 4 yrs.] CASG patients were associated with a 35% lower LTFU rates [AHR 0.65; 95% CI:0.46, 0.91] but similar mortality. |
| Grimsrud 2016 [52] | South Africa | Adherence clubs | Health care facility |
N = 8150 AC n = 2113 HCF n = 6037 |
12 months |
Viral suppression – high rates of VLS among those who had a VL result, but no comparison made between the two cohorts. LTFU – clubs were associated with a decreases risk of LTFU compared to facility in all crude and adjusted models. Clubs were associated with a 67% reduction in LTFU compared with facility (aHR 0.33, [95% CI, 0.27–0.40]). |
| Okoboi 2016 [53] | Uganda | Community based distribution points (CBDP) | Health care facility |
CDDP n = 476 HCF n = 752 |
5 years |
Overall retention rates were above 80% in both HCF and CBDP Retention rates – 83.9% in the facility and 82.9% retained in the community distribution model of delivery (p = 0.670) |
| Jobarteh 2016 [54] | Mozambique | Community ART support groups (CASG) | Health care facility (non-CASG) |
CAGs n = 6760 HCF n = 123,178 |
12 months |
LTFU – LTFU among CASG and non-CASG members was 7.2 and 15.9%, respectively. Compared with CASG participants, non-CASG participants had significantly higher LTFU (hazard ratio [HR]: 2.36; 95% confidence interval [CI]: 1.54–3.17; p = .04] Mortality -no significant mortality differences between CASG and non-CASG members (1.4% vs 1.2%) (HR:0.98; 95%CI, 0.14 to 1.82; p = 0.96) |
| Okoboi 2015 [36] | Uganda | Community distribution points (CDDP) | No comparator | CDDP n = 3340 | 5 years |
Community-based ART distribution systems are capable of overcoming barriers to ART retention and result in good rates of virologic suppression. Viral suppression- of the 870 patients who had a VL measured, 87% were suppressed Mortality- mortality rate was low (3.22 per 100 person-years) LTFU- 1.59 per 100 person-years Retention- more than 69% of patients who initiated ART from 2004 to 2009 were retained in care after more than 5 years of treatment. |
| Decroo 2014 [32] | Mozambique | Community ART groups (CAGs) | No comparator | CAGs n = 6158 | 4 years |
Long-term retention in CAG was exceptionally high [91.8% at 4 years of follow-up (95% CI, 90.1 to 93.2)]. LTFU – event rate was 0.1% per 100-person yrs. Mortality – event rate was 2.1 per 100-person yrs. Retention among CAG members at 1 year on ARTwas 97.7% (95% CI 97.4–98.2); at 2 years, 96.0% (95% CI 95.3–96.6); at 3 years, 93.4% (95% CI 92.3–94.3); and at 4 years, 91.8% (95% CI 90.1–93.2). Overall, the attrition rate was 2.2 per 100 person-years among the 5729 adult members. |
| Study | Setting | Non-facility based model | Comparator | Sample size | Length of follow-up | Key outcomes |
| Luque-Fernandez 2013 [55] | South Africa | Community Adherence clubs | Health care facility |
ACs n = 502 HCF n = 2372 |
3 years |
Outcomes less frequent in patients participating in the clubs. Viral rebound – 214 patients had viral failure at study end in the HCF (90.4 event rates per 1000 person yrs. [95%CI: 79.1–103.4). In the clubs 14 had viral rebound 31.8 event rates per 1000 person yrs. Retention - 97% of club patients remained in care compared with 85% of other patients. In adjusted analyses club participation reduced loss-to-care by 57% (hazard ratio [HR] 0.43, 95% CI = 0.21–0.91). Mortality + LTFU - 12.8% of patients were LTF or had died (323 LTF and 40 deaths). Both outcomes were less frequent for patients participating in the clubs (29.8 vs 116.8 per 1000 person-yrs. for LTFU/death, crude rate ratio [RR = 0.25, 95% CI 0.14–0.41] |
| Kipp 2012 [56] | Uganda | Home based ART delivery | Health care facility |
HBD n = 185 HCF n = 200 |
24 months |
ART outcomes such as viral suppression in community models were equivalent to those receiving care in the facility. Viral suppression – patients in the home delivery model were 2.47 times more likely to achieve viral suppression compared to those in the facility based [95% CI for OR 1.02–6.04 p = 0.046]. Mortality – 32(17%) in Home delivery vs 23 (12%) in HCF. This had limitations as the LTFU in both groups includes unknown number of deaths. Crude mortality was higher in the HBD cohort compared to the HCF cohort, though this difference was not statistically significant (17.3% vs. 11.5%, p = 0.10). Retention − 70% in home model vs 71% in facility |
| CROSS-SECTIONAL STUDY | ||||||
| Chimukangarta 2017 [57] | Zimbabwe | Outreach ART delivery | No comparator | N = 143 | 18 months | Viral suppression- over the course of the study period, 94% were virally suppressed |