Testing selective responses of Cochrane groups to the request of conducting a Cochrane systematic review: A crossover randomized controlled trial

Article type
Authors
Ciapponi A1, Glujovsky D1, Rey Ares L1, García Martí S1, Reveiz L2
1IECS, Institute for Clinical Effectiveness and Health Policy Buenos Aires. Argentina
2Andean Cochrane Branch of the Ibero-American Cochrane Center
Abstract
Background: Thousands of people from around the world contribute to The Cochrane Collaboration. Although there are a number of steps that need to be followed to register a review, titles are accepted at the discretion of each Cochrane Review Group (CRG) and criteria for accepting those titles are not explicit [1]. Previous studies found that authors reported having problems in getting titles registered [2,3].

Objectives: To assess if the first language or country location of 'simulated newcomers’ influence the response patterns to register titles in CRGs.

Methods: Two hypotheses were addressed: 1) native English speakers (NESs) receive more positive responses (+R) than non-NESs (NNESs); 2) responses will not differ by country income (low income countries (LMC) versus High Income Countries (HIC)). After ethics committee approval, we emailed substantially identical versions of a title request (Table 1) to 53 CRGs written by: 'NES-LIC’, 'NES-HIC’, 'NNES-LIC’ and 'NNES-HIC’. In this crossover RCT (one month of washout period), CRGs were randomized centrally by computer software. The answers were independently classified as 'positive’ (offer review titles/updates, specific topics or mentors); 'negative’ (explicit R or requirements without help offer) or 'neutral-R’ (the others) by two assessors blinded to interventions and e-mail senders/recipients.

Results: No response was obtained in 23/106 (21%) of emails. The patterns of responses were: 24% positive, 40% neutral and 36% negative. NES received significant more +R than NNES ( P=0.031) as well as HIC than LIC ( P=0.031), being worst for NNES-LIC and best for NES-HIC (Table 2; Figure 1). Besides no response, frequent barriers included requesting curriculum vitae and newcomer’s publication list. Some CRGs had very positive, warm and collaborative responses.

Conclusions: Although CRGs should focus on potential authors that will produce high quality and timely protocols and reviews, criteria for registering titles should be explicit and standardized. The Cochrane Collaboration should consider consistent ways of supporting authors from NNES-LIC.

References

1. The Cochrane Collaboration. Registering Titles. Available at: http://www.cochrane.org/cochrane-reviews/registering-titles. Access in April 2011.

2. Ciapponi A RL, Martí-Carvajal A, Ortiz Z, Sanchez-Gomez LM, Delgado-Ramirez MB, Manterola-Delgado CG, Martínez-Pecino F. How difficult is to register a Cochrane title? An lberoamerican authors sample. XIV Cochrane Colloquium Dublin, Ireland 2006: [P003].

3. D Gillies HM, K New, V Pennick, Z Fedorowicz, J Van Der Wouden, J Oliver, R. Scholten, A Ciapponi, J Verbeek. A collaboration-wide survey for Cochrane authors. XVI Cochrane Colloquium. Freiburg, Germany 2008: [P04-33].