Table 1.
Characteristics and risk of bias of included studies.
| Author | Design | Sex | Age (years) | n | HCV/ 100 py | New HC/py | Interventions | Effect estimate | Confounders included in analysis | Risk of bias | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Aitken, 2015 55 Australia | Cohort | 31.7% | 29.4 (median) | 98 | 8.6 | 17/196 | OST current Use of OST in last month | HR | 0.8 (0.3–2.0) | – | C |
| Bruneau, 2015 [Bruneau unpublished] Canada | Cohort | NA | 37.4% < 30 | 285 | 17.3 | 102/589.35 | OST current use of OST in last 6 months 3 | HR | 0.74 (0.47–1.16) AOR = 0.74 (0.47–1.16) | Years injecting, numbers of injection in past month | S |
| NSP (high; high NSP (>100% coverage) | HR | 0.77 (0.5–1.19) AOR = 0.7 (0.45–1.09) | Years injecting, numbers of injection in past month | ||||||||
| Combined NSP/OST OST in last 6 months and high (> 100% coverage) | HR | 0.63 (0.37–1.07) AOR = 0.59 (0.35–1.01) | Years injecting, numbers of injection in past month | ||||||||
| OST othera (high dose ≥ 60 mg) | HR | 0.37 (0.17–0.8) AOR = 0.39 (0.18–0.86) | Years injecting, numbers of injection in past month | ||||||||
| OST othera (low dose < 60 mg) | HR | 1.15 (0.7–1.89) AOR = 1.06 (0.64–1.77) | Years injecting, numbers of injection in past month | ||||||||
| Craine, 2009 24 Wales | Cohort | 29.0% | 27.2 (mean) | 286 | 5.9 | 17/287.33 |
OST current In OST at interview |
IRR | 0.27 (0.095–0.77) AOR = 0.34 (0.12–0.99) | Region, homelessness, sharing equipment, sharing needles (all in last 12 months) | |
| Crofts, 1997 25 Australia | Cohort* | 41.9% | 29.2 (mean) | 73 | 22.2 | 13/85.4 | OST current Continuous OST during follow‐up | IRR | 1.8 (0.37–4.77) | – | S |
| OST othera (interrupted use) | IRR | 0.66 (0.55–0.79) | – | ||||||||
| Hagan, 1995 26 USA | Case–control | 45.0% | 24% < 25 | 46 | 20/NA |
NSP (low)a
Ever used NSP |
OR | 0.12 (0.026–0.55) AOR = 0.14 (0.03–0.62) | Sex, age, ethnicity, and duration of injection | S | |
| Hagan, 1999 27 USA | Cohort | 38.0% | 19% < 25 | 187 | 20.8 | 26/209 |
NSP (high) current regular NSP use |
RR | 1.42 (0.64–3.13) AOR = 1.31 (0.79–2.2) | Onset of injection < 1 year from baseline interview, sharing at baseline | S |
| NSP (low)a | 1.72 (0.71–4.19) AOR = 2.59 (0.79–8.5) | Onset of injection < 1 year from baseline interview, sharing at baseline | |||||||||
| Holtzman, 2009 28 USA | Cohort | 38.0% | 28% < 21 | 1288 | 139/NA |
NSP (low)a
NSP use in the last 6 months |
OR | 1.22 (0.86–1.74) AOR = 1.49 (0.96–2.29) | Duration of injection, shared needles/ paraphernalia; daily injecting; injecting with others | S | |
| Hope, 2011 29 UK | Cross‐sectional | 23.0% | 17% < 25 | 119 | 40 | 14/35 | NSP alone (high); high NSP (≥ 100% coverage | IRR | 0.11 (0.024–0.53) | – | S |
|
NSP/OST combined OST in last 6 months and high (> 100% coverage) |
IRR | 0.17 (0.12–1.54) | – | ||||||||
|
NSP/OST combined OST in last 6 months and low (< 100% coverage) |
IRR | 1.08 (0.31–3.82) | – | ||||||||
| Hope, 2015 [Hope (unpublished)] UK | Cross‐sectional | 25.0% | 919 | 9.9 | 30.3 | OST current | OR | – | S | ||
| Use of OST in last 4 weeks | |||||||||||
| Birmingham: | 1.55 (0.14–17.4) | ||||||||||
| Bristol: | 0.24 (0.05–1.16) | ||||||||||
| Leeds | 1.31 (0.08–21.5) | ||||||||||
| NSP high NSP coverage (≥ 100% coverage) (low NSP, no OST) | OR | – | |||||||||
| Bristol: | 0.99 (0.21–4.59) | ||||||||||
| Leeds: | 0.73 (0.04–11.98) | ||||||||||
| Birmingham: | 0.55 (0.05–6.26) | ||||||||||
| Judd, 2015 [Judd (unpublished)] UK | Cohort | 29.0% | 27.4 | 149 | 42 | 49/116.7 |
OST current Use of methadone in last 6 months |
RR |
0.47 (0.16–1.33) AOR = 0.49 (0.17–1.47) |
Duration and frequency of injecting | C |
| Lucidarme, 2004 30 France | Cohort | 17.6% | 26.9 (mean) | 165 | 11 | 16/178.4 |
OST current No definition |
RR |
0.34 (0.11–0.99) AOR = 0.41 (0.12–1.4) |
Sex, geographical region, condom use, daily injection of cocaine, duration of injecting, sharing paraphernalia (cotton, cup, water) | S |
| Maher, 2015 [Maher (unpublished)] Australia | Cohort | 38.0% | 24 (median) | 368 | 24.9 | 53/212.86 |
OST current OST in last 6 months |
HR | 0.43 (0.24–0.75) AOR = 0.46 (0.25–0.84) | Duration and frequency of injecting | S |
| NSP (low)a | HR | 1.86 (1.05–3.28) AOR = 0.15 (0.88–2.78) | Duration and frequency of injecting | ||||||||
| Mehta, 2015 [Mehta (unpublished)] USA | Cohort | 34 (median) | 324 | 17.8 | 27/166.5 |
OST current OST in last 6 months |
IRR | 0.6 (0.14–2.51) AOR = 0.82 (0.19–3.54) | Years injecting, daily injection, age (for whole cohort) | N/I | |
| NSP (low)a | IRR | 1.38 (0.17–11.5) AOR = 0.76 (0.1–5.67) | Years injecting, daily injection, age (for whole cohort) | ||||||||
| Nolan, 2014 31 Canada | Cohort | 30.4% | 23–34 | 1004 | 6.32 | 184/2108.4 |
OST current MMT in last 6 months |
OR | 0.67 (0.45–0.99) AOR = 0.47 (0.29–0.76) | Unstable housing, cocaine, heroin or methamphetamine injection, cohort and year of recruitment, follow‐up time | S |
| Page, 2015 [Page (unpublished)]USA | Cohort | 21.7% | 33.7 (mean) | 552 | 25.1 | 171/681.3 |
NSP (low)a
NSP use in the last 3 months |
HR | 2.82 (1.84–4.34) AOR = 2.62 (1.71–4.02) | Years injecting, age, sex, race, homeless, and recent jail | N/I |
| Palmateer, 2014 32 | Cross‐sectional | 27.5% | 34 (mean) | 2788 | 7.3 | 392/602.7 | OST current; OST at time of survey; low NSP (< 200%) | OR | 0.51 (0.29–0.9) AOR = 0.52 (0.23–1.18) | Survey year, homelessness or stimulant injection in last 6 months, time since onset of injecting | S |
| Scotland |
NSP (high) high NSP (> 200%) coverage and not on OST; |
OR | 0.26 (0.08–0.88) AOR = 0.18 (0.04–0.87) | Survey year, homelessness in last 6 months, stimulant injection in last 6 months, time since onset of injecting | |||||||
|
OST/NSP combined; high NSP (> 200%) coverage |
OR | 0.24 (0.1–0.6) AOR = 0.05 (0.01–0.18) | Survey year, homelessness in last 6 months, stimulant injection in last 6 months, time since onset of injecting | ||||||||
|
OST/NSP combined; low NSP (< 200%) coverage |
OR | 0.48(0.27–0.95) AOR = 0.59 (0.26–1.35) | Survey year, homelessness in last 6 months, stimulant injection in last 6 months, time since onset of injecting | ||||||||
| Patrick, 2001 33 Canada | Cohort | 30.3% | 34 (median) | 155 | 29.1 | 62/207.95 |
NSP (high) Attendance at least 1 per week at NSP in last 6 months |
HR | 3.69 (2.12–6.43) AOR = 2.56 (1.37–6.79) | Sex, injection of cocaine or speedballs, frequency of injection | S |
| Rezza, 1996 34 Italy | Case–control | 2.8% | 21% > 28 | 106 | 28.6 | 21/73.4 |
OST current OST in last 6 months |
OR | AOR = 0.34 (0.10–1.11) | Female gender, age, duration of drug use, injection of cocaine | C |
| Ruan, 2007 36 China | Cohort | 44% < 28 | 86 | 33.0 | 47/258 |
OST othera
Ever used OST |
RR | 0.5 (0.2–1.3) | – | C | |
| Roy, 2007 35 Canada | Cross‐sectional | 27.0% | 31.8 (mean) | 359 | 27.1 | 94/267 |
NSP (low)a
Use of NSP in last 6 months |
HR | 3.02 (95% CI NA) | Age, injection experience, injection with used needle, drug most often injected, sex work, district of recruitment | C |
| Spittal, 2012 37 Canada | Cohort | 53.4% | 23 (median) | 148 | 11.6 | 45/338.6 |
OST current In OST at time of survey (yes/no) |
HR | 2.11 (0.83–5.37) | – | S |
| Thiede, 2000 38 | Cohort | 48.9% | 5.4% < 25 | 80 | 8.75 | 7/80 |
OST current Continuous use during follow‐ up |
OR | 0.3 (0.01–3.6) AOR = 0.4 (0.0–4.2) | Injected at follow‐up, pooled money to buy drugs, injected with used needles, backloading | M |
| USA |
OST othera
Interrupted left treatment at least 1 during follow up |
OR | 0.8 (0.01–3.6) AOR = 1.2 (0.2–7.3) | Injected at follow‐up, pooled money to buy drugs, injected with used needles, backloading | |||||||
| Thorpe, 2002 39 USA | Cohort | 39.7% | 52% 18–22 years | 353 | 10 | 29/327.2 |
NSP (low)a
Use of NSP in last 6 months |
HR | 1.29 (0.6–2.79 | Receptively sharing syringes, sharing cookers, cotton filters, rinse water or backloading | S |
| Tsui, 2014 40 USA | Cohort | 31.9% | 16% 15–18 | 552 | 25.1 | 145/680 |
OST current and other Opiate agonist therapy maintenance treatment in last 3 months; |
HR | 0.31 (0.14–0.65) AOR = 0.39 (0.18–0.87) | Age, duration of injection drug use, sex, ethnicity, homelessness or incarceration in past 3 months | M |
| OST othera Opiate agonist detoxification in last 3 months | HR | 1.45 (0.8–2.69) | Age, duration of injection drug use, sex, ethnicity, homelessness or incarceration in past 3 months | ||||||||
| Vallejo, 2015 41 Spain | Cohort | 27.3% | 40% ≥ 25 | 137 | 39.8 | 42/105.4 |
OST othera
Life‐time use of OST |
IRR | 0.9 (0.5–1.6) | – | S |
| Van Den Berg, 2007 12 | Cohort | 33.0% | 31.4 (median) | 168 | 6.78 | 57/598.56 |
OST current; OST ≥ 60 mg methadone daily |
IRR | 0.67 (0.39–1.13) | – | S |
| Netherlands | NSP alone; High NSP (100% coverage) | IRR | 0.62 (0.3–1.3) | – | |||||||
| Combined OST/NSP high NSP | RR | 0.15 (0.05–6‐0.4) AOR = 0.36 (0.13–1.03) | Duration of injection, HIV status of steady partner | ||||||||
| Combined OST/NSP low NSP | RR | 1.04 (0.53–2.05) AOR = 1.17 (0.59–2.31) | Duration of injection, HIV status of steady partner | ||||||||
| OST othera (high dose) | RR | 0.68 (0.39–1.13) | |||||||||
| OST othera (low dose) | RR | 0.58 (0.3–1.15) | |||||||||
| Van Beek, 1998 42 Australia | Cohort* | 55.9% | 61.5% < 20 | 152 | 20.9 | 26/148.2 |
OST othera
Ever used OST (yes/no) |
OR | 1.08 (0.37–3.17) | C | |
| White, 2014 43 Australia | Cohort | 25.0% | 27 (median) | 127 | 7.9 | 20/215.2 | OST: OST last 6 months | HR |
Heroin users 0.65 (0.15–2.94) AOR = 0.56 (0.12–2.56) Amphetamine users 0.14 (0.04–0.51) AOR = 0.51(0.18–0.04) |
Sex, ethnic background, age, daily or more frequent injecting, receptive syringe sharing and not receiving OST while reporting heroin or other opioids as the main drug injected | S |
| White, 2014 43 Australia | NSP (low)a Accessed NSP in last 6 months; | HR | 1.0 (0.36–2.86) | – |
HR = hazard ratio; OR = odds ratio; AOR = adjusted odds ratio; RR = risk ratio; IRR = incident rate ratio; HCV = hepatitis C virus; OST = opioid substitution therapy; py = person‐years; NSP = needle and syringe programme; NA = not available.
a
Not reported in the primary analysis. Risk of bias: C = critical; S = serious; M = moderate; L = low; N/I = no information. HCV incidence for cross‐sectional surveys calculated as I = [(365/T)n]/[(N–n) + (365/T)n], where I = Incidence, T = estimated mean duration of the HCV antibody‐negative/RNA‐positive ‘window period’ = 75 days, n = number of HCV incident infections (HCV antibody‐negative and HCV RNA‐positive) and N = number of susceptibles (HCV antibody‐negative); NA = not available.
*
Signifies retrospective follow‐up.