Table 3.
Comparison of SVR12 outcomes in direct-acting antiviral regimen studies with or without comparative groups, according to population group and level of decentralisation
| Full decentralisation, % (95% CI) | Partial decentralisation, % (95% CI) | No decentralisation, % (95% CI) | p value | |||
|---|---|---|---|---|---|---|
| People who inject drugs | ||||||
| Full decentralisation vs none | ||||||
| Comparative | 100% (78–100) | .. | 100% (72–100) | 0·688 | ||
| Wade et al (2018), RCT | 100% (78–100) | .. | 100% (72–100) | .. | ||
| Middleton et al (2018), RCS | .. | .. | .. | .. | ||
| Non-comparative | 95% (93–97) | .. | 94% (90–97) | 0·627 | ||
| Number of studies | 28 | .. | 3 | .. | ||
| Full vs partial decentralisation | ||||||
| Comparative | 91% (84–95) | 93% (90–95) | .. | 0·422 | ||
| Radley et al (2018), RCT | .. | .. | .. | .. | ||
| Alavi et al (2018), PCS | .. | .. | .. | .. | ||
| Cooper et al (2017), RCS | .. | .. | .. | .. | ||
| McClure et al (2017), PCS | Nurse 88% (78–95); GP 93% (82–99) | 93% (90–95) | .. | .. | ||
| Non-comparative | 95% (93–97) | 90% (83–96) | .. | 0·150 | ||
| Number of studies | 28 | 6 | .. | .. | ||
| General population | ||||||
| Full vs partial decentralisation | ||||||
| Comparative | 95% (74–100) | 95% (91–97) | .. | 0·780 | ||
| Cooper et al (2017), RCS | 95% (74–100) | 95% (91–97) | .. | .. | ||
| Non-comparative | 93% (90–96) | 93% (86–98) | .. | 1·000 | ||
| Number of studies | 10 | 4 | .. | .. | ||
| People living with HIV | ||||||
| Full decentralisation vs none | ||||||
| Comparative | 94% (88–99) | .. | 100% (93–100) | 0·867 | ||
| Doyle et al (2018), NRS | 94% (88–99) | .. | 100% (93–100) | .. | ||
| Non-comparative | 99% (97–100) | .. | NA | NA | ||
| Number of studies | 3 | .. | 0 | .. | ||
A full list of studies is provided in the appendix (pp 2–25). p values for comparative studies were from pairwise meta-analyses of dichotomous outcomes. p values for non-comparative studies were from two-sample z-tests comparing the results of meta-analyses for proportions using double arcsin transformations. SVR12=sustained virological response at 12 weeks after completion of treatment. GP=general practitioner. RCT=randomised controlled trial. RCS=retrospective comparative study. PCS=prospective comparative study. NRS=non-randomised study. NA=not applicable (tests could not be performed due to a paucity of data).