Table 1.
Study characteristics of 37 studies, of which eight report data for HIV and HCV incidence, included in systematic review and meta-analysis
| Study period | Location (city, country) | Study design | Effect measured and definition | HCV or HIV estimate | Sample size | Incidence (per 100 person-years) | Effect estimates | Confounding factors included in adjusted estimates | Newcastle-Ottawa Scale score | |
|---|---|---|---|---|---|---|---|---|---|---|
| Alary (unpublished estimate) | 2003–18 | Province of Quebec and Ottawa, Canada | Retrospective cohort46 | Unstable housing in the past 6 months | HIV and HCV | HIV: 1930 HCV: 814 | HIV: 1·08 HCV: 20·2 | HIV: HR 1·26 (0·83–1·93);aHR 0·98 (0·62–1·53) HCV: HR 1·64 (1·34–2·00);aHR 1·44 (1·16–1·78) | Living in jail within the past 6 months, OST exposure, using syringes used by someone else, cocaine being most often injected drug, injecting drugs every day, age ≥25 years, male gender, prostitution, urban sites | HIV: 7 HCV: 7 |
| Artenie et al (2019)47 | 2004–17 | Montreal, QC, Canada | Prospective cohort | Unstable housing in the past month | HCV | 513 | 11·8 | HR 2·34 (1·72–3·17);aHR 2·14 (1·54–2·96) | OAT dose and perceived adequacy, sex, duration of injection drug use, cocaine injection in the past month, incarceration in the past 3 or 6 months, previous HCV infection | 8 |
| Bruneau et al (2011)48 | 1992–2008 | Montreal, QC, Canada | Prospective cohort | Unstable housing in the past 6 months | HIV | 2137 | 3·3 | HR 3·08 (2·22–4·28);aHR 2·07 (1·47–2·90) | Age ≥30 years, gender, cocaine use in the past month, heroin use in the past month, sharing syringes with a person known to be HIV positive, “booting”, having sex with a person known to be HIV positive, period of recruitment, NEP participation, obtaining 100% of syringes from a safe source | 7 |
| Craine et al (2009)49 | 2004–06 | Newport and Calidicot, Cardiff and Barry, Bridgend, Neath & Porth Talbot, Swansea, Merthyr Tydfil, Pontypridd, Rhydfelin, Treorchy, Aberdare, Abergavenny, UK | Prospective cohort | Homelessness in the past 12 months | HCV | 286 | 5·9 | IRR 4·41 (1·60–12·5);aIRR 2·9 (1·02–8·28) | In OST at follow-up, any equipment sharing in the past year, sharing needles and syringes in past year, population size of region | 7 |
| Debeck (unpublished estimate) | 2005–16 | Vancouver, BC, Canada | Prospective cohort50 | Homelessness in the past 6 months | HIV and HCV | HIV: 476 HCV: 405 | NA | HIV: HR 1·88 (0·27–13·03);aHR 1·96 (0·31–12·27) HCV: HR 1·69 (1·1–2·6);aHR 1·45 (0·92–2·28) | Recent incarceration (past 6 months), MMT coverage, crack injecting (past 6 months) | HIV: 8 HCV: 8 |
| Dumchev (unpublished estimate) | 2013–15 | Ukraine | Prospective cohort51 | Current homelessness | HIV and HCV | HIV: 2157 HCV: 2157 | HIV: 1·8 HCV: 21·5 | HIV: IRR 1·16 (0·07–18·86) HCV: HR 1·80 (0·80–4·07);aHR 1·57 (0·69–3·54) | Ever been in prison, age (continuous), injecting drug use duration (continuous) | HIV: 6 HCV: 6 |
| Hagan et al (2001)41 | 1994–97 | Seattle, WA, USA | Prospective cohort | Homelessness in the past 12 months | HCV | 317 | 16·7 | RR 1·08 (0·59–1·97) | NA | 5 |
| Hagan et al (2010)52 | 2002–04 | Baltimore, MA; Chicago, IL; Los Angeles, CA; New York, NY; and Seattle, WA USA | Prospective cohort | Homelessness in the past 6 months | HCV | 483 | 17·2 | HR 0·93 (0·68–2·29) | NA | 5 |
| Hayashi (unpublished estimate) | 1996–2016 | Vancouver, BC, Canada | Prospective cohort32 | Homelessness in the past 6 months | HIV and HCV | HIV: 1763 HCV: 387 | NA | HIV: HR 0·78 (0·54–1·14);aHR 0·73 (0·50–1·06);HCV: HR 1·57 (1·11–2·22);aHR 1·62 (1·14–2·29) | Recent incarceration (past 6 months), MMT coverage, crack injecting (past 6 months) | HIV: 8 HCV: 9 |
| Hope et al (2018)33 | 2011–13 | England, Wales, and Northern Ireland, UK | Cross-sectional | Homelessness in the past 12 months | HCV | 2816 | 12·3 | RR 1·40 (1·02–1·92) | NA | 3 |
| Hope (unpublished estimate) | 2006–09 | Birmingham, Bristol, Glasgow, and Leeds, UK | Cross-sectional53 | Homelessness in the past 12 months | HCV | 1247 | 16·9 | RR 1·85 (0·72–4·73);aRR 1·62 (0·55–4·56) | Recent incarceration (past 12 months), current OST status, cocaine use, duration of injecting | 6 |
| Judd (unpublished estimate) | 2001–03 | London and Brighton, UK | Prospective cohort54 | Unstable housing in the past 12 months | HIV and HCV | HIV: 263 HCV: 149 | HIV: 3·5 HCV: 39·7 | HIV: HR 0·94 (0·23–3·76) HCV: HR 1·53 (0·84–2·77) | NA | HIV: 5 HCV: 5 |
| Kåberg (unpublished estimate) | 1987–2020 | Stockholm, Sweden | Retrospective cohort55 | Homelessness in the past 3 months | HCV | 832 | 17·5 | HR 2·12 (1·62–2·78) | NA | 5 |
| Kral et al (2001)56 | 1986–98 | San Francisco, CA, USA | Case control | Current homelessness | HIV | 6115 | 1·2 | OR 1·24 (0·71–2·17) | NA | 4 |
| Kurth (unpublished estimate) | 2012–15 | Nairobi and Coastal region of Kenya | Retrospective cohort57 | Current homelessness | HIV | 978 | 2·6 | IRR 3·45 (1·48–7·62) | NA | 5 |
| Lucidarme et al (2004)58 | 1999–2001 | Northern and eastern France | Prospective cohort | Unstable housing in the past 3 months | HCV | 165 | 9·0 | IRR 2·20 (0·51–7·22) | NA | 5 |
| Maher (unpublished estimate) | 1999–2002 | New South Wales, Australia | Prospective cohort59 | Unstable housing in the past 6 months | HCV | 258 | 26·1 | HR 1·01 (0·31–3·23) | NA | 6 |
| Maher (unpublished estimate) | 2008–14 | Sydney, NSW, Australia | Prospective cohort60 | Unstable housing in the past 6 months | HCV | 169 | 6·6 | HR 1·17 (0·58–2·36) | NA | 7 |
| Mehta (unpublished estimate) | 1993–2019 | Baltimore, MD, USA | Prospective cohort61 | Homelessness in the past 6 months | HIV and HCV | HIV: 2456 HCV: 1731 | HIV: 1·1 HCV: 0·9 | HIV: IRR 1·58 (1·15–2·17);aIRR 1·16 (0·84–1·60);HCV: IRR 1·74 (1·11–2·73);aIRR 1·66 (1·01–2·74) | Injected cocaine in past 6 months; incarcerated in past 6 months; OST or MAT | HIV: 8 HCV:9 |
| Mehta (unpublished estimate) | 2013 | India | Cross-sectional62 | Current homelessness | HIV | 9440 | 5·2 | IRR 1·56 (0·90–2·70);aIRR 1·52 (0·88–2·63) | Injected stimulants in past 6 months, participated in OST programme in past 6 months, incarcerated in past 6 months | 6 |
| Morris (unpublished estimate) | 2000–19 | San Francisco, CA, USA | Prospective cohort63 | Homelessness in the past 3 months | HCV | 712 | 24·9 | HR 1·95 (1·44–2·64);aHR 1·65 (1·21–2·25) | Gender, age, injecting frequency, recent unsafe injecting behaviours, number of injecting partners | 5 |
| Niccolai et al (2011)64 | 2005–08 | St Petersburg, Russia | Cross-sectional | Homelessness in the past 12 months | HIV | 438 | Estimate 1: 18·7* estimate 2: 20·0* | RR 0·70 (0·33–1·52) | NA | 4 |
| Palmateer et al (2014)65 | 2008–12 | Scotland, UK | Cross-sectional | Homelessness in the past 6 months | HCV | 7951 | 10·0 | RR 3·80 (2·20–6·57) | NA | 5 |
| Sacks-Davis (unpublished estimate) | 2005–10 | Melbourne, VIC, Australia | Prospective cohort66 | Current unstable housing | HCV | 89 | 15·4 | HR 1·63 (0·72–3·70):aHR 1·58 (0·66–3·79) | OST (any pharmacotherapy in the past 3 months), type of infection (primary, reinfection), correlation within individuals | 6 |
| Samo et al (2013)67 | 2009–11 | Karachi, Pakistan | Prospective cohort | Current homelessness | HIV | 474 | 12·4 | IRR 1·70 (1·20–2·50);aIRR 1·70 (1·10–2·50) | Sharing of syringes, non-Muslim religion, daily frequency of injecting drugs, source of registration (registered with drop-in centres through outreach compared with other methods), physical disability, monthly income and sources of syringes or needles | 5 |
| Schulkind et al (2019)68 | 2012–16 | Dundee, UK | Prospective cohort | Current unstable housing | HCV | 94 | 21·5 | IRR: 0·42 (0·06–3·23) | NA | 6 |
| Spittal et al (2012)69 | 2003–09 | Vancouver, BC, Canada | Prospective cohort | Homelessness in the past 6 months | HCV | 148 | 11·6 | HR 1·26 (0·83–1·90) | NA | 5 |
| Strathdee (unpublished estimate) | 2006–10 | Tijuana, Mexico | Prospective cohort70 | Unstable housing in the past 6 months | HIV | 812 | 0·9 | HR 1·50 (0·55–4·07) | NA | 7 |
| Strathdee (unpublished estimate) | 2011–20 | Tijuana, Mexico | Prospective cohort71 | Unstable housing in the past 6 months | HIV | 472 | 2·5 | HR 2·10 (1·13–3·90) | NA | 5 |
| Sypsa et al (2017)31 | 2012–13 | Athens, Greece | Retrospective cohort | Current homelessness | HIV | 3320 | 4·5 | HR 1·75 (1·30–2·36);aHR 1·96 (0·98–3·85) | Age, sex, country of origin, history of any imprisonment, size of participant's network of PWID, currently on OST programme, main substance of use, injecting drug use in past 1 month, frequency of injecting drug use, sharing syringes, use of drugs divided with a syringe that someone else had already used for injection | 8 |
| Sypsa (unpublished estimate) | 2012–13 | Athens, Greece | Retrospective cohort31 | Current homelessness | HCV | 63 | 64·5 | IRR 2·31 (0·86–6·19) | NA | 5 |
| Thorpe et al (2002)72 | 1997–99 | Chicago, IL, USA | Prospective cohort | Homelessness in the past 6 months | HCV | 353 | 10·0 | HR 0·76 (0·31–1·86);aHR 0·63 (0·25–1·58) | Injection related risk exposures (sharing cookers, sharing cotton filters, sharing rinse water, sharing syringes), demographic covariates (high-school diploma, suburban residence), drug use covariates (daily injection in the past 6 months, cocaine injection in the past 6 months) | 6 |
| Todd (unpublished estimate) | 2007–09 | Kabul, Afghanistan | Prospective cohort73 | Homelessness in the past 6 months | HIV and HCV | HIV: 316 HCV: 191 | HIV: 1·5 HCV: 40·4 | HIV: HR 0·45 (0·05–3·91) HCV: HR 0·76 (0·45–1·29) | NA | HIV: 6 HCV: 6 |
| Valencia (unpublished estimate) | 2003–16 | Madrid, Spain | Prospective cohort74 | Current homelessness | HCV | 127 | 60·4 | HR 3·82 (0·80–16·9);aHR 4·90 (1·07–23·1) | Crack injecting, OST exposure | 7 |
| Vallejo et al (2015)75 | 2001–06 | Barcelona, Madrid, and Valencia, Spain | Prospective cohort | Unstable housing in the past 12 months | HCV | 513 | 39·8 | IRR 1·71 (0·9–3·25) | NA | 5 |
| Van Santen (unpublished estimate) | 1989–2014 | Amsterdam, Netherlands | Prospective cohort76 | Homelessness in the past 6 months | HIV and HCV | HIV: 690 HCV: 174 | HIV: 1·2 HCV:3·9 | HIV: HR 2·02 (1·01–4·02);aHR 2·02 (1·01–4·01) HCV: HR 2·95 (1·39–6·23);aHR 3·04 (1·42–6·52) | Methadone dosing (no methadone vs <60 mg/day vs ≥60 mg/day) | HIV: 6 HCV: 6 |
| Wijnand (unpublished estimate) | 2011–19 | Melbourne, VIC, Australia | Prospective cohort77 | Current homelessness or unstable housing | HCV | 125 | 5·7 | Homelessness: HR 1·18 (0·16–8·86);aHR 1·09 (0·14–8·59) Unstable housing: HR 1·19 (0·41–3·48);aHR 1·21 (0·40–3·68) | Methamphetamine (ice, crystal, or shabu) injected in the past month, OST | 8 |
Data in parentheses are 95% CIs. For unpublished estimates, studies are listed by the name of the investigator who provided the data or the unpublished estimate and we cited the most recently published article that was based on the same cohort. Unpublished estimates are those that have been calculated from raw data for this study and have not been presented in previously published papers. aHR=adjusted hazard ratio. aIRR=adjusted incidence rate ratio. aRR=adjusted relative risk. BED EIA=BED capture enzyme immunoassay. HCV=hepatitis C virus. HR=unadjusted hazard ratio. IRR=unadjusted incidence rate ratio. MAT=medication-associated treatment. MMT=methadone maintenance treatment. NA=not applicable. NEP=needle exchange programme. OAT=opioid agonist therapy. OST=opioid substitution treatment. RR=unadjusted relative risk.
*
Two estimates for HIV incidence were available because of use of two different formulas for incidence estimation to adjust for misclassification due to sensitivity and specificity characteristics of the BED EIA.