Abstract
BACKGROUND: Direct Healthcare Professional Communications (DHPCs) are an important risk minimisation measure. Their effect has been shown to be variable and has been measured using different outcomes and study populations. Depending on the content of the message, the optimal outcome to measure a direct effect of the DHPC can differ. This systematic review investigates whether the effects of DHPCs differ according to the use of proximal outcomes and the inclusion of the targeted population. METHODS: EMBASE and MEDLINE were searched for European DHPC effectiveness studies performed up to April 6, 2022, evaluating the impact of DHPCs issued from 2008. Outcomes and their impact were extracted, together with a classification of the message. The outcomes were categorised as knowledge/awareness, self-reported behaviour (prescribing/monitoring), prescribing of medication (including dosage changes), monitoring, or adverse events/other health outcomes, including hospitalisation. The outcomes closest to the message of the DHPC were defined as proximal. Outcomes were coded 1 when effective and 0 if not. If multiple outcomes were reported in a study, a composite outcome was created ranging from 0 to 1. Chi-square or Fisher exact tests were performed. RESULTS: From 7063 (scientific) publications identified in our literature search, 60 publications evaluating 31 different DHPCs were selected for our review. As publications could study multiple messages with an outcome, from the 60 scientific publications, 103 outcomes were generated for the messages, of which 30 had a high impact on the composite outcome, with the proportion of analyses with a significant association between 0.75 and 1. When taking the target population into account, some messages were studied in more than one population, resulting in 115 outcomes, of which 33 had a high impact, that is, a composite outcome between 0.75 and 1. CONCLUSION: Neither the use of proximal outcomes nor the restriction of the analysis to the targeted population significantly influenced the impact observed of the DHPC. These results stress the need for improving drug safety communication.