Articles published in Prescrire are prepared collectively, using a systematic procedure that has been refined over the years (a). To ensure the reliability and relevance of the information they contain, the drafts are subjected to numerous internal quality controls, including controls with the references to hand. The drafts are also evaluated by external reviewers independent of the editorial team (1,2).

Who are these reviewers? How are they chosen? And how are their views and suggestions taken into account?

The following article is a brief description of Prescrire's external review process, along with a few data for the period 2013-2014.

A multidisciplinary panel of reviewers specific to each draft article
Draft articles are sent to a group of reviewers from various backgrounds, including general practitioners, specialists in various medical disciplines, hospital and community pharmacists, pharmacologists, methodologists, nurses and other professionals in the community or in university hospitals, as well as end-users (1,2).

Each draft is read by about 10 to 40 individuals, depending on its nature, scope, category, length, complexity and focus (specific patient populations, public health) (1,2).

The reviewers are specifically chosen for each draft, from a continually evolving in-house roster (1,2). New reviewers are continuously recruited, through various channels including specialised working groups, subscribers' correspondence, Readers' test, unsolicited applications, and authors of the references cited in the draft (2).

Reviewers' comments: both substance and style
Reviewers are asked to explain the rationale for their comments on the substance and style of the text, and may, if they wish, provide supporting evidence (documents, published articles or references). Each reviewer's comments are then examined by several members of the editorial team (3,4).

The reviewers' comments allow us to correct or qualify certain statements, complete the information provided in the article, clarify certain points, and sometimes even modify our conclusions (3,4).

Some reviewers mention possible conflicts of interest, but absence of conflicts of interest is not a precondition because our aim is to obtain as diverse a panel as possible. In addition, any biases due to conflicts of interest or personal opinion would have a limited impact, as we interpret comments not accompanied by adequate supporting arguments as encouragement to check the strength of our analysis (2).

More than 850 active reviewers in 2013
In early 2014, the Prescrire's roster of reviewers contained over 3000 contacts. In comparison, there were about 2000 contacts in 2003, 93% of whom resided in France, while the others lived mainly in Belgium or Switzerland. Specialist physicians accounted for 64% of reviewers, GPs 10%, and pharmacists 9%. The other reviewers specialised in nursing, physiotherapy, the environment, health economics, sociology, etc.

In 2013, more than 850 reviewers contributed to improvement of one or more of the draft versions of articles eventually published in Prescrire. In comparison, there were about 600 reviewers in 2003, for a total of more than 270 texts. In 2013, nearly half of the 350 newly recruited reviewers agreed to participate when asked.

In 2013, the list of contributors identified in the masthead each month included an average of 150 to 180 reviewers, compared to about 100 in 2003 (b).

An original, optimized process designed to ensure quality
The involvement of numerous and diverse groups of external reviewers in the editorial process is specific to Prescrire. Our approach differs from that of so-called primary journals that publish research reports and call on only a small number of "peer" reviewers (often just two) specialising in the relevant subject area (2). Peer review has the disadvantage of being highly vulnerable to conflicts of interest. 

The experience Prescrire has accumulated over more than 30 years shows that draft review by large, multidisciplinary panels of reviewers can both refine and consolidate the information contained in the final article, thus helping to ensure it promotes quality of care and patients' interests.

Heartfelt thanks to all Prescrire reviewers, past and present! 

©Prescrire 1 January 2015

"Ensuring the quality of articles published in Prescrire: the important role of reviewers" Prescrire Int 2015; 24 (156): 26-27. (Pdf, subscribers only).

Notes:
a- The procedure used to prepare articles for Prescrire is described at english.prescrire.org "How we work") > Available here
b- The names of all reviewers who evaluated the drafts of articles published in a given issue of Prescrire are printed on the inside front cover, in the list of contributors. On publication of the article, the reviewers receive either the complete issue or the relevant text, accompanied by a letter explaining the main modifications and also the principal suggestions that were not followed.

Selected references from Prescrire's literature search.
1- Prescrire Editorial Staff "How a review in Prescrire is produced", "How we work?" english.prescrire.org.
2- Prescrire Rédaction "Les groupes de relecture de la revue Prescrire" Rev Prescrire 2003; 23(245): 864-866.
3- Prescrire Rédaction "Quelle est l'utilité des groupes de relecteurs de la revue Prescrire ?" Rev Prescrire 2006; 26 (273): 476 (full text: www.prescrire.org: 1 page).
4- Prescrire Rédaction "Comment les avis des relecteurs sont-ils pris en compte ?" Rev Prescrire 2012; 32 (339): 76.