Menu Back to Poster-Presentations-Details

PP04-38: Assessment of Peer Review Comments and Publication Metrics From Comparative Effectiveness Research (CER) Studies

Poster Presenter

Michelle Kissner

Director, Publications Management
Pfizer Inc
United States

Objectives

In order to improve acceptance rates and publication times, we analyzed peer reviewers’ comments on industry-sponsored CER manuscripts, identified areas of most and least interest to reviewers, and compared publication metrics for submitted manuscripts to publicly available estimates.

Method

Records of manuscripts accepted and rejected by peer-reviewed journals from Mar 3, 2016–Aug 27, 2019 were evaluated. Numbers of peer-review comments, time to journal response/publication, and other publication metrics were assessed. Comments were categorized to 16 pre-defined types.

Results

Forty manuscripts, representing 57 submissions to 30 journals, were included. The type of peer review varied across these journals: single blind, 40%; double blind, 17%; open, 7%; not reported, 37%. Overall, 78% of manuscripts were accepted by 16 unique journals: 58% at first submission, 10% at second submission, and 10% at subsequent submissions. The same journals accepted manuscripts with different comparison types and statistical techniques, and from different countries. Overall, there were 636 comments from 57 submissions (median: 9 comments per submission; range: 1–41) from a median of 2 peer reviewers (range: 1–4). The median number of comments was similar for accepted and rejected manuscripts overall (8 vs 9.5) and for first submissions (6 vs 8); accepted manuscripts received more comments for second submissions than rejected manuscripts (17.5 vs 10), and fewer comments for subsequent submissions (range: 6–10 vs 14–22). Among 30 journals to which manuscripts were submitted, the median impact factor (IF) was 3.171 (range: 1.485–23.239). The median (range) IF (2019 values) of journals that rejected manuscripts was higher than that of journals that accepted manuscripts: 4.5575 (2.345–23.239) vs 2.776 (1.485–7.091). The median time from submission to journal response was 3.7 weeks (range: 0.1–28.9). The median time from submission to publication was 17.7 weeks (range: 4.7–45.1) and was shorter for first than subsequent submissions (17.4 vs 21.6 weeks). The top five comment categories were methodology (19%; endpoints, data source, eligibility), results (19%; adherence, discontinuation, components of composite endpoints), discussion (14%; limitations, comparison with published data), study design (13%), and statistical analysis (9%); overall, these categories represented 73% of all comments. Comments on the abstract, priority of topic, and journal guidelines contributed to <5% of comments; comments on industry involvement were infrequent (3%).

Conclusion

Overall, the majority of submitted CER observational study manuscripts were accepted by peer-reviewed scientific journals, including by those not specializing in outcomes research. Of the accepted manuscripts, most were accepted at first submission. All journals to which manuscripts were submitted had a peer-review process; however, the type of review differed between journals, with most accepted manuscripts undergoing single-blind review. Journals with higher IFs were more likely to reject manuscripts. Assessing peer-review comments has provided guidance on areas of most and least interest in CER observational study publications; peer reviewers commented most often on the description of methodology, results, discussion, study design, and statistical analysis. The overall manuscript development process for CER observational studies may be improved by accurately describing the study methodology, results, and discussion during manuscript development.