Background A previous audit by our group of Behçet's syndrome (BS) research, published in 2011 (1) had revealed a list of problems related to research methodology. They were mainly the relative lack of prospective studies, proper use of control groups, a marked under utilization of power calculations where needed and a paucity of studies reporting negative results.
Objectives We have reassesed the same items as in the previous survey in articles about BS since published. An additional item looked at was self criticism in manuscript preparation (2).
Methods Original articles from 15 highest impact factor journals of internal medicine, rheumatology, dermatology and ophtalmology between January 2010 and February 2016 were analyzed by two observers. Study designs, presence of necessary control groups, power calculations and reporting of negative outcomes were tabulated. Presence of self-criticism was assessed both by reading and specific word scanning. Discrepansies between the observers was reconciled in a joint session of all 3 authors.
Results A total of 188 articles, 149 (79%) clinical and 39 (21%) basic, were analyzed. Of 94 studies in which a time-element classification was appropriate; 14% were prospective, 57% cross-sectional and 29% retrospective studies. There were only 3/188 controlled trials. Out of 71 studies in which the study design necessitated control groups, 69 (97%) had healthy and 30% had diseased while 2 did not have any control groups. 50 of the same 71 studies were about genetic association and 13 (26%) had diseased controls in addition to the usual healthy controls. Out of 107 studies in which power calculations were necessary, only18 (16%) gave power calculations. Of these 13 were belonged to 50 (26%) of the genetic association studies and 3/3 to the controlled drug trials. Among 107 studies in which a negative outcome could be expected only 12 studies (11%) reported scuh outcomes. Finally, by electronic scannnig, a limitation acknowledgement was present in 92/188 (49%) of articles [76/149 (51) for clinical and 16/39 (41%) for basic]. When self critique was assesed by text reading these percentages increased to 113 (60%) for total, 93 (62%) clinical and 20 (51%) for basic science studies.
Conclusions Similar methodological problems seem to exist in current BS research as compared to what we had noted 6 years ago. The relative lack of basic science articles (21%) in a condition with a yet unknown cause (s) and the paucity of controlled clinical trials with the recent much increased avaliability of biologics are particularly worrisome. On the other hand; there was an increased inclusion of diseased controls in genetic association studies, 26% in the current and 13% in the former surveys. Similarly, an optimistic note might be that the currrent survey showed basic research in BS included more self-criticism (41–51%) as compared to what was noted among the general rheumatology manuscripts (15–20%) (2).
Esen F, et al. An Audit of Behcet's Syndrome Research: A 10 year Survey. J Rheumatol 2011;38:99–103.
Yazici H, et al. There was self critique among basic than in clinical Science articles in three rheumatology journals. J Clin Epid 2014;67: 654–657.
Disclosure of Interest None declared