There is a paucity of data in the literature regarding negative articles concerning surgery in orthopedics and trauma. Knowledge pertaining to treatments or techniques which confer a beneficial effect remains important, as does knowledge regarding those which have a null or pejorative effect. Thus, this study was carried out on negative articles in order to: (1) determine their proportion in the ten predominant journals concerning orthopedic and trauma surgery; (2) assess variability in their publication rate depending on the journal and the year, and compare their citation rate to that of positive articles; (3) specify whether a positive article was more likely to be cited compared to a negative article; (4) carry out the same bibliometric analysis with the “Orthopedics & Traumatology: Surgery & Research (OTSR)” journal, and detect possible selection bias for negative articles during the review.

There are fewer negative articles than positive articles in the literature relating to orthopedic and trauma surgery.

The study was carried out using the ten orthopedic and trauma surgery journals with the highest impact factors for the year 2021. Two periods were compared, 2009–2010 and 2019–2020. Among the 17,812 publications obtained, 11,962 publications were retained to carry out the analysis (technical notes, meta-analyses, editorials and letters to the editor were all excluded). An analysis using the same method was carried out on the 3,727 articles submitted to OTSR from 2015 to 2021, which made it possible to compare the rejected articles to the accepted articles.

Negative articles represented 11% (1,342/12,023) of the literature relating to orthopedic and trauma surgery. There were differences in the rate of publication of negative articles depending on the journals (from 4.04% to 17.14%) (p<0.0001). The negative article publication rate did not change between the two periods studied: 534/4963 articles (10.76%) in 2009–2010 versus 802/6999 (11.46%) in 2019–2020 (p=0.23). Positive articles were not cited more often than negative ones: no significant difference between the Category Normalized Citation Impact (CNCI) classes (respectively for classes 0;1[/[1;2[/≥ 2 with 45.66% 28.22% and 26.12% for negative articles versus 44.90% 27.46% and 27.64% for positive articles [p=0.4]) and the Top10% (with 18.86% for negative articles versus 20.10% for positive ones [p=0.28]). The OTSR journal had a rate of negative articles of 9.46% which was within the average range of the journals studied. A selection bias (p<0.02) for negative articles during the review of the OTSR journal was identified with fewer negative articles accepted (115/1216 [9.46%]) than positive articles (164/1330 [12.33%]).

The publication of negative articles varies according to the journals and although it is modest, at only 11%, it is essential because it allows us not to repeat errors but also not to bias the carrying out of meta-analyses, and among other things to avoid useless studies.

III; case control study from the literature.