Effectiveness of implementing a decentralized delivery of hepatitis C virus treatment with direct-acting antivirals: A systematic review with meta-analysis

Direct-acting agents (DAAs) for hepatitis C virus (HCV) treatment are safe and highly effective. Few studies described the sustained virologic response rates of treatment conducted by non-specialists. We performed a systematic review and meta-analysis to evaluate the effectiveness of decentralized strategies of HCV treatment with DAAs. PubMed, Embase, Scopus and LILACS were searched until March-2019. Studies were screened by two researchers according to the following inclusion criteria: HCV treatment using DAAs on real-life cohort studies or clinical trials conducted by non-specialized health personnel. The primary endpoint was the sustained virologic response rate at week 12 after the end-of-treatment (SVR12), which is binary at the patient level. Data were extracted in duplicate using electronic-forms and quality appraisal was performed with the NIH Quality Assessment Tool. Heterogeneity was assessed by I-2 statistics. Random-effects meta-analysis models were used for pooling SVR12 rates. Publication bias was assessed using funnel plots. Among the 130 selected studies, nine papers were included for quantitative synthesis. The quality-appraisal was good for two, fair for three and poor for four studies. The pooled relative risk (RR) of SVR12 was not statistically different between decentralized strategy and treatment by specialists [RR = 1.05; 95% confidence interval (95% CI): 0.98-1.1; I-2 = 45% (95% CI: 0-84%), p = 0.145]. SVR12 rate for decentralized HCV treatment was 81% [SVR12 95% CI: 72-89%; I-2 = 93% (95% CI: 88-96%)] and 95% [SVR12 95%CI: 92-98%; I-2 = 77% (95% CI: 52-89%)] with intention to treat analysis and per-protocol analysis, respectively. SVR12 rates using DAAs managed by non-specialized health personnel were satisfactory and similar to those obtained by specialists. This new delivery strategy can improve access to HCV treatment, especially in resource-limited settings.