%0 Journal Article %T A systematic review and network meta-analysis of incentive- and non-incentive-based interventions for increasing blood donations. %A Irving AH %A Harris A %A Petrie D %A Mortimer D %A Ghijben P %A Higgins A %A McQuilten Z %J Vox Sang %V 115 %N 4 %D May 2020 %M 32043603 %F 2.996 %R 10.1111/vox.12881 %X OBJECTIVE: Blood services are tasked with efficiently maintaining a reliable blood supply, and there has been much debate over the use of incentives to motivate prosocial activities. Thus, it is important to understand the relative effectiveness of interventions for increasing donations.
METHODS: This systematic review used a broad search strategy to identify randomized controlled trials comparing interventions for increasing blood donations. After full-text review, 28 trials from 25 published articles were included. Sufficient data for meta-analysis were available from 27 trials. Monetary incentives were assumed to be equivalent regardless of value, and non-monetary incentives were assumed to be equivalent regardless of type. Non-incentive-based interventions identified included existing practice, letters, telephone calls, questionnaires, and the combination of a letter & telephone call. A network meta-analysis was used to pool the results from identified trials. A subgroup analysis was performed in populations of donors and non-donors as sensitivity analyses.
RESULTS: The best performing interventions were letter & telephone call and telephone call-only with odds ratios of 3·08 (95% CI: 1·99, 4·75) and 1·99 (95% CI: 1·47, 2·69) compared to existing practice, respectively. With considerable uncertainty around the pooled effect, we found no evidence that monetary incentives were effective at increasing donations compared to existing practice. Non-monetary incentives were only effective in the donor subgroup.
CONCLUSIONS: When pooling across modes of interventions, letter & telephone call and telephone call-only are effective at increasing blood donations. The effectiveness of incentives remains unclear with limited, disparate evidence identified.