DESIGN: Systematic review and meta-analysis.

SETTING: Primary care setting.

PARTICIPANTS: Of 1568 studies screened, 14 studies with 7035 PWID were included.

MEASURES: PubMed, Embase, Web of Sciences, CENTRAL and Cochrane review databases were searched without language restriction from their inception to 27 January 2016. All published study designs with control groups that reported the effectiveness of pharmacy-based NSP on outcomes of interest were included. Outcomes of interest are risk behaviour (RB), HIV/hepatitis C virus (HCV) prevalence and economic outcomes. The estimates of pooled effects of these outcomes were calculated as pooled odds ratio (OR) with 95% confidence interval (CI) using a random-effects model. Heterogeneity was assessed by I2 and χ2 tests.

FINDINGS: Most studies (nine of 14, 64.3%) were rated as having a serious risk of bias, while 28.6 and 7.1% were rated as having a moderate risk and low risk of bias, respectively. For sharing-syringe behaviour, pharmacy-based NSPs were significantly better than no NSPs for both main (OR = 0.50, 95% CI = 0.34-0.73; I2  = 59.6%) and sensitivity analyses, excluding studies with a serious risk of bias (OR = 0.52, 95% CI = 0.32-0.84; I2  = 41.4%). For safe syringe disposal and HIV/HCV prevalence, the evidence for pharmacy-based NSPs compared with other NSP or no NSP was unclear, as few of the studies reported this and most of them had a serious risk of bias. Compared with the total life-time cost of US$55 640 for treating a person with HIV infection, the HIV prevalence among PWID has to be at least 0.8% (for pharmacy-based NSPs) or 2.1% (for other NSPs) to result in cost-savings.

APPROACH: After developing standard operating procedures and agreement between its Prisons Department and Ministry of Health, Malaysia established pilot MMT programmes in two prisons in the states of Kelantan (2008) and Selangor (2009) - those with the highest proportions of HIV-infected prisoners. Community-based MMT programmes were also established in Malaysia to integrate treatment activities after prisoners' release.

LOCAL SETTING: Having failed to reduce the incidence of HIV infection, in 2005 Malaysia embarked on a harm reduction strategy.

RELEVANT CHANGES: STANDARD OPERATING PROCEDURES WERE MODIFIED TO: (i) escalate the dose of methadone more slowly; (ii) provide ongoing education and training for medical and correctional staff and inmates; (iii) increase the duration of methadone treatment before releasing prisoners; (iv) reinforce linkages with community MMT programmes after prisoners' release; (v) screen for and treat tuberculosis; (vi) escalate the dose of methadone during treatment for HIV infection and tuberculosis; and (vii) optimize the daily oral dose of methadone (> 80 mg) before releasing prisoners.

METHODS: Retrospective data pertaining to 713 patients from January 2009 to December 2013 were retrieved from hospital and disease notification records for analysis. The risk factors for hepatitis C virus (HCV) infection were grouped into IVDU and non-IVDU risk factors for analysis using multiple logistic regression.

RESULTS: Of the hepatitis C patients included in this study, the most common age group was 31 to 40 years (30.2%), and male patients (91.2%) made up the overwhelming majority. Ethnic Malays constituted approximately 80.4% of the patients, and IVDU was the main risk factor (77.8%) for HCV infection. Multiple logistic regression showed that male patients were 59 times more likely to have IVDU as a risk factor for HCV infection. Single patients were 2.5 times more likely to have IVDU as a risk factor. Patients aged ≥71 years were much less likely than patients aged ≤30 years to have IVDU as a risk factor for HCV infection.