|
Abstract
|
BACKGROUND Violence and injury prevention interventions conducted in community settings are complex due to multiple reasons, including: (1) they address a complex situation by addressing many risk factors simultaneously, (2) they are often overlapped in their implementation, (3) they are implemented in phases over time, (4) implementation must be consistent with community priorities and budgetary limitations, and (5) they are not conducted in randomised, controlled studies. Thus, evaluating their impact is not statistically straightforward.
METHODS A methodological analytic framework is proposed that uses random effects meta-regression methodology and incorporates a taxonomy that allows for complex interventions to be 'disentangled' into their active components. The framework is illustrated with pilot study data from communities in Japan participating in the Safe Communities global effort, using motor vehicle related deaths and hospitalizations as the indicators of effectiveness.
RESULTS Information from 8 communities in Japan for 2008-2011 (4 years each) provided information. Being designated implied a reduction of 1.6/100,000 MV related deaths annually, and that being designated implied a statistically significant reduction in 81/100,000 MV related hospitalizations! We also found that media campaigns were somewhat effective in explaining some of the reduction in hospitalizations in these communities in Japan, but that education was not.
CONCLUSIONS The proposed statistical framework is very useful to understand the effectiveness of community-based, multi-component, dynamic interventions. The framework may help researchers and policy-makers evaluating the effectiveness and impact of complex intervention programs.
Abstract from Safety 2016 World Conference, 18-21 September 2016; Tampere, Finland.
Copyright © 2016 The author(s), Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://group.bmj.com/group/rights-licensing/permissions
Language: en |