Today, returning to this blog after something of a gap, I find reason to reflect on the many flavours of evidence review which now exist.
In the RiSC project we’ve been using the EPPI-Centre approach, developed by people working at the Institute of Education in London. This framework has a number of desirable features, including the combination of a phase 1 mapping with a more in-depth phase 2 involving quality appraisal. Deciding the focus of phase 2 in an EPPI-Centre review involves discussion with stakeholding collaborators. That’s all to the good, proving that in evidence syntheses, as in primary data-generating studies, it is possible for researchers to work jointly with service user, carer and practitioner colleagues.
Last week, with other members of the Wales chapter of the COCAPP team, I spoke on our meta-narrative mapping of care planning and care coordination at a Swansea University seminar. Amongst other things meta-narrative mapping traces the different research traditions found within a given field. And, today, I mock-examined a delightful doctoral thesis containing a scoping review, which lays out what’s there but does not include formal quality appraisal. Then there are realist syntheses, where reviewers look across multiple studies for evidence of the generative mechanisms underpinning change in policy, services or practice. The list goes on, encompassing thematic literature reviews and, of course, Cochrane-style systematic reviews. This latter approach has been very important in driving the evidence-based practice movement, but personally I’ve always been a little disappointed at its insistence on hierarchies with randomised trials as the gold standard.
So how might decisions be made on selecting one approach over another? Practical considerations have a bearing, but perhaps more important are commitments to certain intellectual or other principles. We chose the EPPI-Centre approach in RiSC because we valued user, carer, practitioner and manager perspectives and wanted a way of hearing these and using them to inform our project. Realist reviewers sign up to particular sets of ideas on how programmes work, and meta-narrative mappers embrace the idea, and seek out examples, of paradigmatic differerence. Perhaps the key thing is to be aware of, and articulate, these in justifying the choices which inevitably have to be made.
Ben Hannigan's blog 
Not coming from an academic or professional background, and in wonderment at all these sophisticated analytical trend tools working the old oracle with information, which one’s the best for predicting life-role outcomes for service users? Post codes still the best grist for the technological mill?
Using existing evidence to predict things for people, Dave? Sounds like a challenge whatever the approach. Perhaps casting a wide net using some kind of scoping, just to capture what’s been written in the first instance? I’d want to pin the aim down, too: would this be about the very long-term prospects (broadly put: health, social, employment, education, and so on) for people who have used mental health services?
Yes, you’re on the money, Ben. The whole socio-economic shooting fish in a barrel scenario.
But on a more inspirational note, your piece led me to this: “Health Thinkers: Man of the Rhondda” http://www.cf.ac.uk/insrv/resources/scolar/HSSJ-article-on-AC.pdf – Made my Saturday morning.
Good spot: a pioneer, without a doubt.
‘policy-friendly’, Ben?! (realist syntheses). Are there penalties for how many methodologies miss that particular target? ‘Wicked’, mate!