How to formulate research recommendations

Brown et al (BMJ 14th Oct p 804) set out a useful framework for
guiding future research in clinical trials, but the title of their paper
(“How to formulate research recommendations”) contains the inherent
assumption that all research can be reduced to the EPICOT acronym. In many
areas of health services research (take diabetes, obesity, mental health,
or sexual health, for example), and particularly for complex interventions
aimed at behaviour change or incorporating new service models, the most
pressing unanswered research questions are qualitative. There is another
paper to be written about how to formulate these qualitative questions,
which are likely to include

(a) What are the priorities of patients, clinicians and policymakers
for further research in this field?

(b) What is the mechanism by which particular complex interventions work,
and how might existing interventions be modified to optimise impact?

(c) What factors explain the gap between the effect size typically shown
in research trials and that demonstrated in real-life practice?

Whilst the evidence-based medicine movement has many strengths, and
the systematic review of randomised trials with a clear definition of
population, intervention, comparison and outcome is rightly seen as the
gold standard in the evaluation of simple interventions, there is a danger
that the research agenda will be impoverished rather than enriched if we
sign up to a ‘framework for future research’ that focuses exclusively on
this slice of the pie.

Competing interests:
None declared