Table 3.

Inclusion and quality rating of publications and databases across systematic reviews and meta-analyses assessing the association of injectables (versus nonuse of hormonal contraception) with risk of HIV acquisition in women.

		Current systematic review and meta-analysis	Ralph et al. [19] meta-analysis	Brind et al. [20] meta-analysis (longitudinal data)	Morrison et al. [26] IPD meta-analysis
	References	Inclusion, quality rating	Inclusion	Inclusion	Inclusion/exclusion rationale, quality rating and rationale
Published manuscripts with risk estimates for injectables and HIV acquisition	Bulterys et al. (1994) [40]	○	○	●	Did not meet inclusion criteria; follow-up visits >6 months apart
	Ungchusak et al. (1996) [62]	○	○	●	Did not meet inclusion criteria; not sub-Saharan Africa
	Kilmarx et al. (1998) [46]	○	○	○	Did not meet inclusion criteria; not sub-Saharan Africa
	Kapiga et al. (1998) [44]	○	○	○	Did not meet inclusion criteria; follow-up visits >6 months apart
	Kiddugavu et al. (2003) [45]	○	●	●	Did not meet inclusion criteria; follow-up visits >6 months apart
	Baeten et al. (2007) [39] (update of [77], [54])	●	●	●	○ A
	Myer et al. (2007) [56]	● (6 mo estimates)	●	●	○ A, B, C
	Kleinschmidt et al. (2007) [47]	●	●	●	○ A, B
	Kumwenda et al. (2008) [48]	○	○	●	○ B
	Watson-Jones et al. (2009) [60]	○	○	●	○ B
	Morrison et al. (2010) [52] (reanalysis of [53])	●	●	●	●
	Feldblum et al. (2010) [41]	○	○	●	Did not meet inclusion criteria; no longitudinal data on contraception
	Reid et al. (2010) [58]	●	●	●	Did not meet inclusion criteria; >5% missing data for exposure
	Heffron et al. (2012) [42,43]	●	●	●	●
	Morrison et al. (2012) [55]	●	●	●	●
	Wand and Ramjee (2012) [31]	Replaced by [21]	● Duplicates [21]	● Duplicates [21]	Replaced by [21]
	McCoy et al. (2013) [51]	●	●	●	●
	Lutalo et al. (2013) [50]	○	○	δ	Did not meet inclusion criteria; follow-up visits >6 months apart; published after meta-analysis dataset closed
	Kapiga et al. (2013) [24]a	○	δ	δ	●
	Crook et al. (2014) [21]a	●	●	●	○ B
	Wall et al. (2015) [27]a	●	†	†	†
	McKinnon et al. (2015) [25]a	○	†	†	†
	Byrne et al. (2016) [28]a	○	†	†	†
	Balkus et al. (2016) [29]a	●	†	†	†
Data in Morrison et al. [26] with no associated publication specifically assessing HC-HIV	Kaul et al. (2004) [32]a	●	†	†	○ A, B
	Vallely et al. (2007) [33]a	●	†	†	○ A
	Delany-Moretlwe and Rees (2010) [34]a	●	†	†	○ A, B
	McGrath et al. (2014) [35]a	●	†	†	○ C, D
	Vandepitte et al. (2011) [36]a	●	†	†	○ A
	Abdool Karim et al. (2010) [37]a	●	†	†	○ D
	Van Damme et al. (2012) [17]a	●	†	†	○ D

● Included in analysis and ranked as: ‘informative but with important limitations’ (in current systematic review) or ‘lower risk of bias’ (in Morrison et al. [26]); indicates ‘Included’ in Ralph et al. [19] or Brind et al. [20]. ○ Included in analysis and ranked as ‘unlikely to inform the primary question’ (in current systematic review) or ‘higher risk of bias’ (in Morrison et al. [26]); indicates ‘not included’ in Ralph et al. [19] or Brind et al. [20]. δ Reason for noninclusion not specified. † Likely published after meta-analysis dataset closed (in Morrison 2015) or subsequent to search strategy (Brind et al. [20] and Ralph et al. [19]). A: <80% retention rate. B: Did not measure one or more of the following variables: pregnancy status, coital frequency, marital status/living with partner, or transactional sex. C: Contraceptive method measurement occurred less frequently than every 3 months. D: <10% in no–hormonal contraception comparison group.

^aNew information identified for this systematic review.