Table 1 Characteristics and outcomes of studies on the impact of task-shfting in HIV/AIDS care
From: A systematic review of task- shifting for HIV treatment and care in Africa
Study | Setting | Study design | Study size | Intervention | Outcomes |
|---|---|---|---|---|---|
Uganda (rural) | Cohort | 2522 | 'Field officers' provide home-based ART | Cumulative outcomes at 4 years showed excellent adherence (96.8% were > 95% adherent) and < 1% defaulting. Social improvements: reduced stigma, stronger family and community relationships | |
Arem et al, 2009 [69]. | Uganda (rural) | Qualitative Survey | --- | Peer adherence supporters | Peer health workers successfully understood ART regimens and physical danger signs; 97% of clinic staff reported that peer health workers improved patient outcomes. |
Bedelu et al, 2007 [40]. | South Africa (rural) | Cohort | 1025 | Decentralized, nurse-initiated ART | Task-shifted, decentralised care increases access and is more acceptable to patients loss-to-follow-up was clinics 2% at clinics compared to 19% at hospital for comparable virological and immunological outcomes. |
Bolton-Moore et al, 2007 [50] | Zambia (urban) | Cohort (paediatric) | 2938 | Nurse- and clinical officer-initiated paediatric ART | Decentralization allows for dramatically scaled-up rollout; cumulative 3-year mortality (8.3%) and defaulting (5.4%) comparable to other programmes. |
Chang et al, 2008 [74] | Uganda (rural) | Cohort | 360 | Patients trained as 'peer health workers' to monitor ART adherence by mobile phone | Extremely cost effective. 72% retention and 86% virological suppression at 2 years |
Chiambe et al, 2009 [42]. | Kenya (urban and rural) | Cohort | 39,900 | Lay health care workers supporting basic clinic tasks and adherence counselling | Enrollment increased from 1,176 to 39,900 patients within 3 years |
Chung et al, 2008 [25] | Rwanda (rural) | Modelling | 3194 | Nurse-initiated ART | Substantial time savings: nurse-initiated ART reduces physician HIV-related workload by 78%, saving up to 56 hours physician time/month. |
Cohen et al, 2009 [55]. | Lesotho (rural) | Cohort | 4,347 | Nurse-initiated ART | Favourable outcomes at 12 months among adults (9.3% mortality, 2.5% defaulting) and children (5% mortality, 2% defaulting) |
Gimbel-Sherr et al 2008 [48]. | Mozambique | Cohort | 6,006 | ART initiated by mid-level workers (2.5 years training) vs doctors | Patients seen by NPCs (69.4% of cohort) were 44% less likely to be lost to follow up; no difference in mortality |
Jaffar et al, 2009 [59]. | Uganda (rural) | RCT | 859 | Home vs clinic-based ART delivery | Similar outcomes of mortality and viral suppression in home-based and faculty-based ART |
Koenig et al 2004 [35]. | Haiti (rural) | Cohort | 2300 | Decentralized, CHW-monitored ART | Approach increases access, reduces defaulting, and delays resistance to first-line medication |
McGuire et al, 2008 [29]. | Malawi (rural) | Cohort | 1676 | Nurses/medical assistants initiating and managing ART | More rapid time to initiation (21.5 days for nurses/medical assistants vs 35 days for clinical officers); no difference in outcomes and retention rates |
Sanjana et al, 2009 [73]. | Zambia | Cross-sectional survey | --- | Assessment of record-keeping errors among lay vs health care workers | Error rate for lay counsellors was less (6.44/1,000 field) than health care workers (16.81/1,000 fields) |
Shulman et al, 2009 [50]. | Malawi (rural) | Cohort | --- | Lay workers trained as pharmacist assistants | Expanded pharmacy capacity (500 prescriptions per day) and reduced errors (30% to 5%) |
Shumbusho et al, 2008 [47]. | Rwanda (rural) | Concordance study | --- | Nurses trained in ART initiation | Discordance between eligibility and initiation < 1% (n = 343) |
Shumbusho 2008 [47]. | Rwanda (rural) | Cohort | 3194 | Nurse-initiated ART | Mortality at defaulting < 5% at 12 months. |
Tweya et al, 2008 [64]. | Malawi (rural) | Cohort | 1,617 | Lay-workers to pre-screen for adult ART eligibility | Symptom screening checklist had high sensitivity (91.8%) but low specificity (28%) |
Tootla et al 2007 [53]. | South Africa (urban) | Cohort | 2,084 | Nurse/pharmacist managed ART | 75% of clients had undetectable viral load at 12 months |
Torpey et al 2008 [27]. | Zambia | Cohort (quantitative and qualitative analysis) | 500 | Lay-workers used as 'adherence supporters' | Lay adherence supporters reduced loss-to-follow-up from 15% to 0%; reduced wait times |
Udegboka et al, 2009 [28]. | Nigeria | Cohort | --- | Nurse ART treatment and peer support | Task shifting reduced waiting times by 4 hours |
Van Rie et al 2009 [46]. | DRC (urban) | Blinded concordance study | 339 | Nurse vs doctor decisions to initiate ART | 95% agreement |
Van Griensven et al, 2008 [57]. | Rwanda (urban) | Cohort | 315 | Nurse-initiated and monitored paediatric ART | 84% retention and 83% virological suppression at 2 years |
Van Griensven et al, 2009 [58]. | Rwanda (urban) | Cohort | 435 | Nurse-initiated and monitored Adult ART | 0.3% attrition and 8.5% mortality at 1 year |
Wood et al, 2009 [45]. | South Africa (urban) | RCT | 812 | Doctor vs nurse-initiated ART | Non-inferiority according to virological failure, toxicity, adherence, and mortality. |
Zachariah et al, 2007 [62]. | Malawi (rural) | Cohort | 1634 | Community support vs no support | 26% increase in survival; 98% reduction in loss to follow up. |