Defining study designs in The Cochrane Collaboration: Do all Cochrane Review Groups talk the same language?

Article type
Authors
López-Alcalde J1, Calcerrada N1, Callejo D1, Reza M1, Gracia F1, Maeso S1, Escalona S1, Nieto B1, Fernández de Larrea N1, Valentín B1, Blasco J1
1LaínEntralgo Agency, Spain
Abstract
Background: Cochrane systematic reviews (SR) focus on randomized clinical trials (RCT) but some Cochrane Review Groups (CRG) also consider non-randomized studies (NRS) as eligible. However, NRS cover designs defined with different labels not consistently applied among researchers.

Objectives: To describe the study designs considered as eligible by the CRG and the definition provided for NRS.

Methods: We consulted CRG websites and their corresponding modules of The Cochrane Library (March 2011). At least two researchers assessed the information provided by each CRG. Any disagreement was resolved through discussion and, if necessary, with the involvement of a third researcher. The inter-rater reliability in the assessments of key items was rated by using the kappa statistic. We registered the following information for each CRG: 1) Are NRS eligible for inclusion? 2) If so, is this justified with any reason? 3) What 'study design label’ is used (if any) for each study design? 4) What definition (if any) is provided for each study design? 5) How are key 'study design features' described? We considered as key study design features those detailed in Tables 13.2a and 13.2b of The Cochrane Handbook: Which was the unit of allocation? Was there a comparison? How were participants allocated to groups? Which parts of the study were prospective? On which variables was comparability between groups assessed?

Results: We identified 53 CRGs. The preliminary results show that around 60% of the CRGs analysed considered NRS as eligible, although this decision was justified with explicit reasons by fewerthan 50% of the CRGs. The CRGs analysed did not use common study design labels and did not explicitly describe allstudy design features suggested inChapter 13 of The Cochrane Handbook.

Conclusions: Consistent definitions of key study design features for NRS are needed in The Cochrane Collaboration.