Systematic reviews of health effects of social interventions: 2. Best available evidence: how low should you go?
- 1MRC Social and Public Health Sciences Unit, University of Glasgow, UK
- 2Development and Alumni Office, University of Glasgow
- Correspondence to: Dr D Ogilvie MRC Social and Public Health Sciences Unit, University of Glasgow, 4 Lilybank Gardens, Glasgow G12 8RZ, UK;
- Accepted 19 April 2005
Study objective: There is little guidance on how to select the best available evidence of health effects of social interventions. The aim of this paper was to assess the implications of setting particular inclusion criteria for evidence synthesis.
Design: Analysis of all relevant studies for one systematic review, followed by sensitivity analysis of the effects of selecting studies based on a two dimensional hierarchy of study design and study population.
Setting: Case study of a systematic review of the effectiveness of interventions in promoting a population shift from using cars towards walking and cycling.
Main results: The distribution of available evidence was skewed. Population level interventions were less likely than individual level interventions to have been studied using the most rigorous study designs; nearly all of the population level evidence would have been missed if only randomised controlled trials had been included. Examining the studies that were excluded did not change the overall conclusions about effectiveness, but did identify additional categories of intervention such as health walks and parking charges that merit further research, and provided evidence to challenge assumptions about the actual effects of progressive urban transport policies.
Conclusions: Unthinking adherence to a hierarchy of study design as a means of selecting studies may reduce the value of evidence synthesis and reinforce an “inverse evidence law” whereby the least is known about the effects of interventions most likely to influence whole populations. Producing generalisable estimates of effect sizes is only one possible objective of evidence synthesis. Mapping the available evidence and uncertainty about effects may also be important.
Funding: the review was funded by the Chief Scientist Office of the Scottish Executive Health Department and by the ESRC Evidence Network. DO is now funded by a Medical Research Council fellowship. The funding sources played no part in the design, analysis, interpretation, or writing up of the study or in the decision to publish.
Competing interests: none known.
Ethical approval: not required.
A list of references to the studies excluded from the systematic review is available on request from the first author.