The review addressed a well-defined question in terms of study design, participants, interventions and relevant outcomes. Only one relevant database was searched and studies needed to be published and in English, so some relevant studies may have been missed. Publication bias was not assessed. Study quality was assessed, but no relevant information was provided and the study design of the included studies was unclear and so it was impossible to assess study quality. Some efforts were made to avoid error and bias in study selection, but validity assessment and data extraction were each performed by only one reviewer, which risked errors and bias. Some relevant study details were reported. Outcome results were extracted only for intervention groups and not for controls. No comparisons were made between results for different interventions and placebo groups. Outcomes were extracted for only patients who took anti-tumour necrosis factor therapy. No relevant data were available for two of the seven included studies and one of these was the only study of adalimumab treatment. A narrative synthesis was provided due to the reported heterogeneity in study design.
One author had been a speaker, consultant and advisory board member for Abbott Laboratories, Shire, Schering-Plough and Centocor and had received funding from Abbott Laboratories and Schering-Plough.
The authors commented that two of the included studies were of responders to infliximab and the results should be considered with caution. In view of the limitations of the review process and evidence, and uncertainties about the quality of included studies, the authors’ conclusions should be treated with caution.