Table 1 Descriptions of quantitative studies included in the systematic review
Location; study period [Ref] | Study type | Population [comparison group] | Time on HAART | Relevant reported HAART adherence outcome | Relevant reported treatment outcome | Adjusted analysis (outcome; factors associated with outcome with p<0.05) |
|---|---|---|---|---|---|---|
Lacor Hospital, Gulu District, Uganda; June 2005-Jan 2008 [12] | Prospective cohort study by structured questionnaire | IDPs, n=1625; 14% residing in IDP camps, 86% residing in outlying areas; >14 years-old | Cumulative patient-years follow-up=1981 | Composite of pharmacy monitored drug possession ratio, pharmacy refill records, and 3-day self-reported recall by patients or caregivers: | Mortality incidence=3.48 (95% CI 2.66-4.31) per 100 person-years, log rank p-value<0.01; | Lower all-cause mortality:-Sex, female vs. male (HR=0.7 95% CI 0.55-0.91, p=0.02) |
[No comparison group] | ≥95% doses taken as prescribed=92.2% | Median CD4 change (IQR)= 0 (0-0) | -Baseline CD4 count, per 100 cell increase (HR=0.14, 95% CI 0.06-0.34, p<0.001) | |||
Lacor Hospital, Gulu District, Uganda; Jan-Feb 2008 [13] | Cross-sectional survey by semi-structured questionnaire | IDPs, n=200; 29% residing in IDP camps, 71% residing in outlying areas; ≥18 years-old | ≤12 months=33.0%13-24 months= 29.5%>24 months=37.5% | Mean 4-day self-reported adherence recall, ≥95% doses taken as prescribed= 99.5% | NA | <95% adherence:-First line vs. second line treatment [OR=22.22, 95% CI 1.48-333.33, p=0.03] |
[No comparison group] | -Staff were condemning, yes vs. no (OR=22.22, 95% CI 1.5-333.33, p=0.02) | |||||
Nyanza province, Kenya; Dec 2007-July 2008 [14] | Retrospective cohort study by review of demographic surveillance data | IDPs, n=28 (proportion on HAART unknown); rural; ≥5 years-old | Not known | NA | 53% (28/53) HIV mortality in IDPs vs. 25% (235/936) HIV mortality in 2008 DSS residents, p<0.001 | NA |
[IDP HIV mortality compared with prior DSS residents] | ||||||
Kinkala/Mindouli, Republic of Congo; May 2006-Dec 2007 [15] | Retrospective cohort study by chart review | Conflict-affected, n=222; rural; adults ≥15 years-old | Mean follow-up time on HAART=9 months | NA | Probabilities of survival-(n=129) at 6 months=0.94 95% CI 0.89-0.96 | NA |
[No comparison group] | -(n=70) at 12 months=0.89 95% CI 0.82-0.93 | |||||
12 MSF programs; Oct 2003- [2] | Retrospective cohort study by chart review | Conflict-affected, n=2572; rural; adults ≥15 years-old [No comparison group] | Median follow-up time on HAART=11.8 months (IQR 3.9-22.7) | NA | -Median probability of survival at 12 months 0.89, 95% CI 0.88-0.91 | NA |
-Proportion lost to follow-up 0.11, 95% CI 0.09-0.12 | ||||||
-Median 6-month CD4 gain=129 cells/mm3 | ||||||
Location; study period [Ref] | Study type | Population [comparison group] | Time on HAART | Relevant reported HAART adherence outcome | Relevant reported treatment outcome | Adjusted analysis (outcome; factors associated with outcome with p<0.05) |
MSF Project, Bukavu, Democratic Republic of the Congo; May 2002-Jan 2006 [16] | Retrospective cohort study by chart review | Conflict-affected, n=494; urban and outlying areas [Compared with 18 low-income setting cohorts in Africa, Asia, and South America and 12 high-income setting cohorts in Europe and North America] | Person-years follow-up=235 | >95% pills taken as prescribed as of last clinic visit, measured by pill counts=99% | 6-month median CD4 gain (IQR): 163 (82-232)12-month mortality (95% CI)=7.9% (3.6-12.1) | NA |
Equatorial region of southern Sudan; July 2009-March 2010 [17] | Retrospective cohort study by chart review | Refugees and IDPs, n=159 (69% living in refugee camps, 12% internally-displaced at time of HAART start); rural; adults age-cut-off not reported [No comparison group] | 64% on HAART for ≥6 months | >95% adherence by self-report over past month=88% (of those on HAART for ≥6 months) | NA | NA |
Nazareth House, Johannesburg, South Africa; April 2004-March 2007 [18] | Retrospective cohort study by chart review | Foreigners, n=568 (% refugees or IDPs unknown); urban; age≥16 [Compared with local citizens (n=431) and persons of unknown citizenship (n=298)] | Median (IQR) person-years on HAART: | NA | Viral failure (ART cessation, patient death, viral load >1000 copies/mL, any decrease in CD4 from pre-ART levels): | Viral failure (ART cessation, patient death, viral load >1000 copies/mL, any decrease in CD4 from pre-ART levels): |
-Foreigners=0.5 (0.1-0.9)-Local citizens=0.6 (0.2-1.1) | -Foreigners=24%-Local citizens=42% | -Citizenship status, foreigner vs. local citizen (OR=0.45, 95% CI 0.23-0.87, p= 0.017) | ||||
-Unknown citizenship=0.5 (0.2-1.1) | -Unknown=53%p-value (foreigners vs. local citizens) = 0.001 | -Opportunistic infections, TB before ART vs. none (OR=2.5, 95% CI 1.4-4.5, p=0.002) | ||||
Coptic Hope Centre for Infectious Diseases, Nairobi, Kenya; December 2006-February 2007 [19] | Retrospective cohort by chart review | Local population, n=2,534; Kenyan residents; age ≥18 [Compared with n=2,167 one year earlier, before post-election violence] | Median duration on treatment (IQR) 19.5 months (9.8-28.5) Comparison group was “similar” | Proportion interrupting treatment (visiting pharmacy ≥48 hours after ARTs completed) | NA | -Odds of TI during PEV increased by 71% 95%CI 34 to 118] |
-16.1% in PEV group-10.2% in comparison group | -During post-election violence, odds of TI increased for men (OR=1.37, 95%CI 1.07 to 1.76, p=0.01) and clients travelling ≥ 3 hours to clinic (OR=1.86, 95% CI 1.28 to 2.71, p=0.001) | |||||
Location; study period [Ref] | Study type | Population [comparison group] | Time on HAART | Relevant reported HAART adherence outcome | Relevant reported treatment outcome | Adjusted analysis (outcome; factors associated with outcome with p<0.05) |
Miriam Hospital, Providence, Rhode Island, USA; 2000–2006 [20] | Matched case-control study by retrospective chart review | Refugees, n=52 (29 started ART); non-refugees, n=52 (41 started ART); urban [Controls were non-refugees matched on gender] | NR | Adherence to scheduled appointments:-Refugees=75%-Non-refugees=86%, p=0.17-Initiation of HAART: Refugees=56%-Non-refugees=79% (OR=0.37, 95% CI 0.13-0.92, p=0.03) | Not reported | NA |
Mangere Refugee Resettlement Centre, Auckland, New Zealand; June 1993-June 2004 [21] | Retrospective cohort study by chart review | Refugees from Africa and Asia, n=98 (n=60 started HAART); urban [No comparison group] | NR | NA | Undetectable viral load 1 year after HAART start= 61% (36/59) | NA |
Boston Medical Centre, USA; June 2000-June 2001 [22] | Retrospective cohort study by chart review | Refugees, n=34, n=15 on HAART; urban [No comparison group] | NR | “Reported adherence with medications”= 87% | Undetectable viral load (not defined)=87% | NA |
Southern Alberta, Canada; Jan 2001-Jan 2007 [23] | Retrospective cohort study by chart review | Sub-Saharan African, n=126 (68% refugees); Other foreign-born, n=72 (14% refugees) [Canadian-born, n=455] | NR | “Good adherence within foreign-born patients to HAART” (data not shown) | 80% viral suppression (no comparison between groups reported) | NA |
Miriam Hospital, Providence, Rhode Island, USA; 2000–2006 [24] | Retrospective cohort study by chart review | Pregnant, resettled refugee women, n=14; rural [No comparison group] | NR | Lost to follow-up=1/14 (7%) | Median viral load at time of pregnancy=3.36 log10 copies/mL | NA |
Median viral load at time of delivery=1.88 log10 copies/mL | ||||||
Great Lukole camp, Tanzania; Oct 2002-Sept 2004 [25] | Retrospective cohort study by chart review | Women delivering in camp, n=189 [No comparison group] | NA | Single dose nevirapine uptake at labour=98% (185/189) excluding repatriated women and 62% (185/301) including refusals and repatriations | NA | NA |