Quality assessment of reports on clinical trials in the Journal of Hepatology

Abstract

Background/aims: Electronic searches on databases for randomised clinical trials and controlled clinical trials do not identify as many trials as handsearches, and trial reporting may be flawed. The aims were to identify all fully reported randomised clinical trials in the Journal of Hepatology and to make a qualitative assessment of the reporting.

Methods: The publications were identified by systematically handsearching the full text of the journal and searching MEDLINE. Central dimensions of trial quality were used to assess the reporting quality of the trials.

Results: Randomised clinical trials represented 8.4% of the original articles (171/2028). Ten original articles (0.5%) could not be classified. A search on MEDLINE identified 81.3% of the randomised clinical trials, i.e., 139 out of the 171 identified by the handsearch. A total of 166 randomised clinical trials could be quality assessed. Forty-seven (28.3%) of them reported adequate generation of allocation sequence; 22 (13.3%) adequate allocation concealment; 95 (57.2%) allowed intention-to-treat analysis with only a few losses to follow-up; 50 (30.1%) were double-blind; 33 (19.9%) reported sample-size calculations; 13 trials (7.8%) employed the crossover design; and the median number of subjects per intervention arm in parallel group trials was 19 subjects (interquartile range: 11-31; range: 5-519). The quality of reporting was significantly better in regular issue articles than in supplement articles.

Conclusions: Many important randomised clinical trials are published in the Journal of Hepatology, but there seems to be ample room for improvement of quality of reporting.