Adverse events reported in randomized clinical trials of drug therapies: the information is still insufficient

Abstract

Eight years after the publication of “Better Reporting of Harms in Randomized Trials: An Extension of the CONSORT Statement” [[1]Ioannidis J.P. Evans S.J. Gotzsche P.C. O'Neill R.T. Altman D.G. Schulz K. et al.Better reporting of harms in randomized trials: an extension of the CONSORT statement.Ann Intern Med. 2004; 141: 781-788Crossref PubMed Scopus (1064) Google Scholar], little is known about its impacts on the quality of safety reporting in randomized controlled trials (RCTs). Although some authors analyzed RCTs of specific drug therapies [2Chowers M.Y. Gottesman B.S. Leibovici L. Pielmeier U. Andreassen S. Paul M. Reporting of adverse events in randomized controlled trials of highly active antiretroviral therapy: systematic review.J Antimicrob Chemother. 2009; 64: 239-250Crossref PubMed Scopus (45) Google Scholar, 3Lee P.E. Fischer H.D. Rochon P.A. Gill S.S. Herrmann N. Bell C.M. et al.Published randomized controlled trials of drug therapy for dementia often lack complete data on harm.J Clin Epidemiol. 2008; 61: 1152-1160Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar], others evaluated journals specialized in otolaryngology [[4]Bibawy H. Cossu A. Cogan S. Rosenfeld R. Reporting of harms and adverse events in otolaryngology journals.Otolaryngol Head Neck Surg. 2009; 140: 241-244Crossref PubMed Scopus (16) Google Scholar], pediatrics [[5]de Vries T.W. van Roon E.N. Low quality of reporting adverse drug reactions in paediatric randomised controlled trials.Arch Dis Child. 2010; 95: 1023Crossref PubMed Scopus (49) Google Scholar], and urology [[6]Breau R.H. Gaboury I. Scales Jr., C.D. Fesperman S.F. Watterson J.D. Dahm P. Reporting of harm in randomized controlled trials published in the urological literature.J Urol. 2010; 183: 1693-1697Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar]. The studies that evaluated RCTs published in the general medical journals of greater impact [7Pitrou I. Boutron I. Ahmad N. Ravaud P. Reporting of safety results in published reports of randomized controlled trials.Arch Intern Med. 2009; 169: 1756-1761Crossref PubMed Scopus (197) Google Scholar, 8Haidich A.B. Birtsou C. Dardavessis T. Tirodimos I. Arvanitidou M. The quality of safety reporting in trials is still suboptimal: survey of major general medical journals.J Clin Epidemiol. 2011; 64: 124-135Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar] reviewed RCTs published on a date close to that of the publication of the harm extension of the Consolidated Standards of Reporting Trials (CONSORT) statement. To test the hypothesis that the process of incorporating recommendations supplied by the harm extension of the CONSORT takes longer to become evident in publications, we have evaluated its incorporation in 122 studies published in 2009 in the British Medical Journal, the Journal of the American Medical Association, The New England Journal of Medicine, and The Lancet. We had assumed that we would find better results on the quality of information on safety than Haidich et al. [[8]Haidich A.B. Birtsou C. Dardavessis T. Tirodimos I. Arvanitidou M. The quality of safety reporting in trials is still suboptimal: survey of major general medical journals.J Clin Epidemiol. 2011; 64: 124-135Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar], but this has not been confirmed. A substantial proportion of the RCTs of drug therapies analyzed failed to report methodological aspects such as how the information on harms had been collected (Table 1). Less than half of the studies reported absolute risk per study harm and type of adverse effect, and few studies reported a balanced discussion on the harms and benefits of drug therapy. On the other hand, we have found that studies focusing on harms (n=37) presented better information on safety, but a small proportion of them met all the recommendations of the harm extension of the CONSORT statement.Table 1Safety reporting according to the recommendations of the harm extension of the CONSORT statement [1]Ioannidis J.P. Evans S.J. Gotzsche P.C. O'Neill R.T. Altman D.G. Schulz K. et al.Better reporting of harms in randomized trials: an extension of the CONSORT statement.Ann Intern Med. 2004; 141: 781-788Crossref PubMed Scopus (1064) Google Scholar in total sample (n=122) and in RCTs studies focusing on harms (n=37)Paper sectionRecommendation of the harm extension of the CONSORT statementTotal (n=122), %RCTs that mentioned harms in title/abstract/introduction (n=37), %Title and abstractFirstTitle or abstract mentioned harms Yes72.1100 No27.90IntroductionSecondIntroduction mentioned harms Yes32.8100 No67.20MethodologyThirdListed definitions for adverse events monitored Yes15.621.6 Partially32.037.8 No52.440.5FourthClarified how the information related to harms was collected Yes10.75.4 Partially72.189.2 No17.25.4FifthDescribed plans for the presentation and analysis of the information on harms Yes53.373.0 No46.727.0ResultsSixthDescribed for each arm, the participant withdrawals that are because of harms (including reasons) Yes55.770.3 Partially2.52.7 No41.827.0SeventhProvided the denominators for the analysis of harms Yes30.337.8 Partially58.251.4 No11.510.8EighthPresented the absolute risk for each adverse event Yes43.448.7 Partially44.345.9 No12.35.4NinthaThis recommendation was not evaluated for few studies presented this type of analysis.DiscussionTenthBalanced discussion on harms and benefits Yes18.035.1 Partially48.451.4 No33.613.5Abbreviations: CONSORT, Consolidated Standards of Reporting Trial; RCT, randomized controlled trial.a This recommendation was not evaluated for few studies presented this type of analysis. Open table in a new tab Abbreviations: CONSORT, Consolidated Standards of Reporting Trial; RCT, randomized controlled trial. Furthermore, we had expected that the trials financed exclusively by the pharmaceutical industry would avoid reporting adverse events for marketing reasons, but the opposite has occurred and the percentage of almost all recommendations increased when this subgroup was analyzed. This finding is consistent with Haidich et al. [[8]Haidich A.B. Birtsou C. Dardavessis T. Tirodimos I. Arvanitidou M. The quality of safety reporting in trials is still suboptimal: survey of major general medical journals.J Clin Epidemiol. 2011; 64: 124-135Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar], and it may be attributed to the fact that these studies were designed to meet the rules of regulatory agencies. Similar to Haidich et al. [[8]Haidich A.B. Birtsou C. Dardavessis T. Tirodimos I. Arvanitidou M. The quality of safety reporting in trials is still suboptimal: survey of major general medical journals.J Clin Epidemiol. 2011; 64: 124-135Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar], we believe that journals should explicitly require the adoption of the harm extension of the CONSORT statement, rather than generically mentioning the CONSORT statement. Additionally, once the objective of most RCTs remains efficacy oriented, we suggest that all data on safety must be interpreted with more caution regarding patient selection, limited follow-up duration, limited power, and other methodological issues.