Table 3

Distribution of reviewer comments—by direction of trial results

Mean percentage of comments on a manuscript (95% CI)*
Type of commentNegative results (N=96)Positive results (N=150)p Value
Importance
 1. Research question not clinically relevant4.3 (1.1 to 7.5)5.9 (3.1 to 8.7)0.469
Originality
 2. Lack of novelty3.5 (0.6 to 6.3)7.4 (4.4 to 10.4)0.066
Background and rationale
 3. Incorrect background information20.3 (14.9 to 25.8)18.9 (14.5 to 23.3)0.670
 4. Poor justification for conducting study4.2 (1.3 to 7.2)3.0 (1.2 to 4.9)0.497
Methods
 5. Poor experimental design64.9 (56.9 to 73.0)60.0 (54.5 to 65.5)0.285
 6. Methods inadequately reported55.4 (48.2 to 62.6)56.3 (50.7 to 61.9)0.844
 7. Statistical analysis methods inappropriate23.7 (17.3 to 30.1)20.5 (15.7 to 25.3)0.453
Results
 8. Study outcome data incomplete61.4 (53.5 to 69.2)66.4 (61.0 to 71.8)0.266
 9. Flow of participants through study unclear5.5 (2.3 to 8.8)7.4 (4.5 to 10.3)0.411
Discussion and conclusion
 10. Meaning results inadequately discussed47.8 (40.0 to 55.7)48.5 (42.1 to 54.8)0.889
 11. Study insufficiently related to literature12.6 (7.7 to 17.5)13.7 (9.7 to 17.7)0.748
 12. Limitations not sufficiently discussed14.4 (9.4 to 19.3)18.4 (14.3 to 22.4)0.203
 13. Conclusions inappropriate29.3 (22.6 to 36.0)18.9 (14.0 to 23.9)0.010
Abstract
 14. Abstract does not correctly reflect paper16.4 (11.4 to 21.4)15.5 (11.8 to 19.2)0.790
 15. Discrepancies between the abstract and the main text1.9 (0.1 to 3.6)0.8 (0.0 to 1.8)0.320
References
 16. References missing/irrelevant references used14.9 (9.8 to 20.0)8.6 (5.8 to 11.5)0.079
 17. Errors in reference citation2.3 (0.3 to 4.3)2.8 (1.1 to 4.4)0.731
Presentation
 18. Title not representative of study6.7 (2.5 to 10.9)3.6 (1.7 to 5.5)0.191
 19. Poor writing38.4 (31.4 to 45.4)38.4 (32.5 to 44.3)0.997
 20. Inaccurate tables or figures35.1 (28.6 to 41.6)41.4 (36.1 to 46.7)0.156
Ethics
 21. Ethics committee approval not clear1.4 (0.0 to 3.0)2.2 (0.7 to 3.7)0.463
 22. Other ethical issues related to study2.0 (0.1 to 3.9)3.8 (1.4 to 6.2)0.267
Trial registration, protocol, CONSORT
 23. Registration/protocol/CONSORT missing1.9 (0.1 to 3.6)3.0 (1.3 to 4.8)0.352
 24. Deviations from registry or protocol1.8 (0.0 to 4.0)1.4 (0.2 to 2.7)0.758
COI
 25. Bias by author COIs/contribution funder unclear1.9 (0.1 to 3.6)3.3 (1.5 to 5.1)0.254
 26. Systematic bias or spin in favour of sponsor0.0 (0.0 to 1.1)1.0 (0.0 to 2.0)0.120
  • *The mean percentage of comments on a manuscript is controlled for the journal to which a manuscript was submitted.

  • COI, conflicts of interest; N, number of submitted manuscripts.