Quality and reporting of trial design in scientific papers in <i>Anaesthesia</i> over 25 years

Although fewer human interventional trials are now published in Anaesthesia, the quality of these trials has improved in terms of study design, bias control and proper disclosure.

Abstract

We determined how the quality of trial design and its reporting in scientific papers published in Anaesthesia has changed in the last 25 years. All articles between the years 1983-87 and 2003-07 were reviewed and classified according to methodology. Reporting and trial design of all prospective, comparative clinical interventional trials were compared between the two time periods using 12 criteria. Fewer articles now originate from the United Kingdom and Ireland than 25 years ago. Although fewer human interventional trials are now published in Anaesthesia, the quality of these trials has improved in terms of study design, bias control and proper disclosure. Significant improvements were observed in all criteria of trial design except for the declaration of non-primary adverse outcomes and the minimisation of the risk of type I errors. Further improvements could still be made with respect to sample size calculation, description of the method of randomisation, and blinding.