Summary
A randomised trial at BMJ Open found that asking authors to use a SPIRIT-based template during revision modestly improved the completeness of randomised trial protocols. Revision times were similar between groups, although some responding authors found the template time-consuming.
A randomised controlled trial conducted within the journal BMJ Open found that asking authors to use a structured SPIRIT template during manuscript revision modestly improved how completely randomised trial protocols reported key information. The adjusted difference between groups was 0.61 points on a 0-10 reporting-completeness scale (95% confidence interval 0.20 to 1.01); median revision times were similar.
SPIRIT, the Standard Protocol Items: Recommendations for Interventional Trials, is a reporting guideline for trial protocols. Clear protocols make planned methods easier to assess and provide a record against which a completed trial can be understood.
How the trial tested the template
Researchers screened manuscripts reporting randomised trial protocols that had received a decision of major or minor revision after their first round of peer review. Of 74 manuscripts screened between 29 April and 27 May 2025, 70 were eligible and randomly allocated: 35 to the template group and 35 to usual peer review. All submitted a first revision and were included in the analysis.
Editors asked authors in the intervention group to use a template organised around SPIRIT 2025 items. It supplied suggested section headings and guidance on the information to include. Use was requested but was not a condition of editorial decisions. Control manuscripts followed the journal’s standard process, which already asked authors to submit a completed SPIRIT checklist.
The primary outcome was a score based on 11 SPIRIT items that are often inadequately reported, including recruitment, data collection, statistical methods and access to data. Researchers assessed the original and revised manuscripts, scoring subitems as not reported, partially reported or fully reported, then rescaling the result to 0-10.
A modest gain without longer revisions
The intervention group’s mean score rose from 5.42 to 6.70, while the control group’s rose from 5.77 to 6.32. The adjusted between-group difference was 0.61 points. The largest item-level gains were for access to data, recruitment and data collection methods. The researchers describe the overall change as less than one additional SPIRIT item being more completely reported per manuscript, on average.
The median time to submit a first revision was 27.10 days in the template group and 29.86 days in the control group. Among 35 authors invited to a survey, 14 responded; 13 said they had used the template. Of those 13, 9 (69%) considered it helpful for improving the manuscript and 10 (77%) found it easy to use. Eight (62%) said it took too long, and 5 (39%) found it frustrating.
The results suggest that structured guidance can improve some aspects of protocol reporting within a journal’s existing revision process. The measured gain was modest, and the trial took place at a single journal. The outcome assessed completeness across selected reporting items, not the quality of the underlying trials; the authors also note that template headings may have made information easier for assessors to identify. Whether introducing templates earlier or using them at other journals would produce different results was not tested.