Introduction This paper describes and analyses the editor's decision-making process at the British Journal of Psychiatry (BJP), and investigates the association between reviewers' assessments and editorial decisions.
Method Four hundred consecutive manuscripts submitted over a six-month period to the BJP were examined prospectively for assessors' comments and editorial decisions on acceptance or rejection. Interrater reliability of assessments was calculated and a logistic regression analysis investigated the effect of the rank allocated by assessors and the comprehensiveness of the assessments on the editor's decision.
Results The editor sent 248/400 (62%) manuscripts to assessors for peer review.Kappa for reliability of assessors' rankings was 0.1 indicating poor interrater reliability. Assessors agreed best on whether to reject a paper. A ranking of five (indicating rejection) had the greatest association with editor's rejection (P < 0.001, odds ratio 0.079), and the mean ranking of assessments was also significantly associated with editorial acceptance or rejection (P=0.004, odds ratio 0.24).
Conclusion Assessors and editors tend to agree on what is clearly not acceptable for publication but there is less agreement on what is suitable.
Declaration of interest The second author is Editor of the BJP.