Table 2. Characteristics of studies not included in the individual participant data meta-analysis.
Study Number and Country/Region | Short Study Name | Primary Publications [Reference] | Reason Not Included in IPD Meta-Analysis | Study Design | Number of Women with HIV/Total Number of Women | Number of Women with HIV in Contraceptive Groups | Comparison Group a | Study Results: Adjusted Effect Measure (95% CI) unless Otherwise Specified b |
---|---|---|---|---|---|---|---|---|
Studies that did not meet inclusion criteria (n = 8) | ||||||||
19. Rwanda | Not known | Bulterys et al., AIDS 1994 [55] | Did not meet inclusion criteria: follow-up visits >6 mo apart | Cohort | 31/1,524 | 12 | No contraception | HC use: OR 1.9 (0.8–4.6) |
20. Thailand | Not known | Ungchusak et al., JAIDS 1996 [51] | Not sub-Saharan Africa | Cohort | 15/365 | NR | No contraception | COC: IRR 0.22 (0.03–1.9); injectable HC: IRR 3.8 (1.0–14.4) |
21. Tanzania | Not known | Kapiga et al., AIDS1998 [53] | Did not meet inclusion criteria: follow-up visits >6 mo apart | Cohort | 75/2,471 | 7 (COC); 2 (injectable HC) | No COC use; no injectable use | COC: HR 1.0 (0.5–2.3); injectable HC: HR 0.3 (0.07–1.3) |
22. Thailand | Not known | Kilmarx et al., AIDS 1998 [56] | Not sub-Saharan Africa | Cohort | 30/340 | 20 (COC); 5 (DMPA) | No COC use; no DMPA use | COC: RR 1.8 (0.8–4.0); DMPA: unadjusted RR 1.5 (0.6–4.0) |
23. Uganda | Rakai Study | Kiddugavu et al., AIDS 2003 [52] | Did not meet inclusion criteria: follow-up visits >6 mo apart | Cohort | 202/5,117 | 12 (COC); 16 (injectable HC) | No contraception, no condoms | COC: IRR 1.1 (0.5–2.6); injectable HC: IRR 0.8 (0.4–1.7) |
24. Benin, Ghana, Nigeria, South Africa, Uganda, India | SAVVY/CS Trials | Feldblum et al., Sex Transm Dis 2010 [50] | Did not meet inclusion criteria: no longitudinal data on injectable contraception | RCT | 114/7,364 | 13 (COC); 15 (injectable HC) | No contraception or emergency contraception only | COC: unadjusted RR 1.8 (0.8–4.1); injectable HC: unadjusted RR 2.5 (1.1–5.6) |
25. South Africa | CAPRISA 050/051 | Abdool Karim et al., Int J Epidemiol 2011 [46] | Did not meet inclusion criteria: longitudinal data on injectable contraception >6 mo apart | RCT | 39/594 | NA | NA | No published HC—HIV results |
26. South Africa, Zambia, Zimbabwe | HPTN 039 | Reid et al., JAIDS 2010 [54] | Did not meet inclusion criteria: >5% missing data for exposure | RCT | 72/1,358 | NR | No contraception, no condoms | COC: HR 0.9 (0.5–1.8); injectable HC: HR 0.9 (0.5–1.9) |
Studies not included because agreement was not obtained from investigators (n = 3) | ||||||||
27. South Africa, west Africa, Southeast Asia | COL 1492 | Van Damme et al., Lancet 2002 [45] | No agreement/dataset | RCT | 99/552 | NA | NA | No published HC—HIV results |
28. Malawi, South Africa, Zambia, Zimbabwe, US | HPTN 035 | Abdool Karim et al., AIDS 2011 [58];; Chirenje et al., Int’l Microbicides Conf. 2012 [61] | No agreement/dataset | RCT | 192/2,887 | NR | No HC | COC: HR 0.6 (0.3–1.2); injectable HC: HR 1.4 (0.9–2.0) |
29. Kenya, Uganda | Partners Prep | Baeten et al., N Engl J Med 2012 [47] | No agreement/dataset | RCT | 28/1,584 | NA | NA | No published HC—HIV results |
Studies with results published after September 2012 (n = 2) | ||||||||
30. Uganda | Rakai Study | Lutalo et al., AIDS 2013 [60] | Does not meet inclusion criteria: follow-up visits >6 mo apart; published after meta-analysis dataset closed | Cohort | 30/190 | 3 (COC); 7 (DMPA) | No contraception, no condoms | COC: IRR 2.7 (0.82–8.6); DMPA: IRR 1.4 (0.6–3.4) |
31. South Africa, Uganda, Zimbabwe | VOICE Study | Noguchi et al., CROI 2014 [62] | Published after meta-analysis dataset closed; no non-hormonal-contraceptive comparison | RCT | 207/3,141 | 204 | NET-EN use | DMPA: HR 1.4 (1.0–2.0) |
aComparison group, based on information reported by the authors; for study 31 there was no comparison with non-hormonal-contraceptive users, so the comparison group is women using NET-EN.
b“Injectable” reported if this was the term used by the authors and the specific progestin was not mentioned; all effect sizes rounded to one decimal place.
IRR, incidence rate ratio; NA, not applicable; NR, not reported; OR, odds ratio; RR, risk ratio.