Instant overviews, and beyond: the case for standardizing reporting in the characteristics of included studies tables

Article type
Authors
Struthers C1
1Education and Training Manager, EQUATOR Network, UK
Abstract
Background:
There will be many cases where several single intervention reviews are written for the same patient population. Currently patients, carers and the professionals advising them often have to use several Cochrane Reviews to help them make a decision about a range of treatment options for their particular circumstances.

Objectives:
To investigate ways of reporting and linking population, outcome, and other information consistently across reviews to allow the creation of instant overviews, and facilitate easier evidence-informed decision-making between two or more interventions used to treat the same patient population.

Methods:
I chose two published single intervention reviews looking a pharmacological and non-pharmacological treatment for people with behavioural and psychological symptoms of dementia. I created a matrix of outcomes reported in the included studies in both reviews to highlight differences between the characteristics of research conduct and reporting in the included studies. An interesting added dimension was including numbers of participants and funding source for each study. If the funding source was not reported in the characteristics of included studies table I went back to the original study references to find it.

Results:
Information in the matrix was reported graphically to allow a clear comparison between the reviews in terms of numbers of participants, outcomes reported, and funding source, which could be weighed up alongside the main conclusions of each review.

Conclusions:
Where a network meta-analysis is not possible, as when comparing pharmacological, non-pharmacological and complex interventions for the same patient population, standardising the reporting of the characteristics of included studies, and storing this information in the Cochrane Register of Studies for interrogation could allow users to compare reviews of treatments for the same patient population at the touch of a button.