Background: There is ample discussion as to whether the current way of identifying topics and producing Cochrane reviews is the best possible compromise among feasibility, comprehensiveness and clinical relevance. We empirically explored this issue by looking at the effect of broadening the inclusion criteria of two reviews looking at the effectiveness of antispastic treatment for, respectively, spinal cord injury and multiple sclerosis patients.

Methods: Inclusion criteria of the two reviews have been compared to determine whether: a) informative studies had been excluded by either or both reviews; b) a single review concentrating on the drug effect rather than the disease would have been more informative.

Results: Almost the same population of studies have been evaluated by the two reviews and 19 studies with a mixed population have been excluded from the Multiple Sclerosis review and 23 from the Spinal Cord Injury.

Conclusions: Our preliminary findings suggest some merit in running a large drug- rather than drug specific review. Several of the excluded studies would contribute to a review with less narrow objectives thus increasing the evidence base for the therapeutic question at issue. While individual patient data may be needed to come up with an optimal analysis it remains to be seen whether clinicians would value more a narrow vs. a broader type of review. The paper will discuss the results as well as present the judgement elicited from treating physicians.