Maria Elena Flacco1, Lamberto Manzoli2, Stefania Boccia3, Anna Puggina3, Annalisa Rosso4, Carolina Marzuillo4, Giacomo Scaioli5, Maria Rosaria Gualano5, Walter Ricciardi6, Paolo Villari4, John P A Ioannidis7. 1. Department of Medicine and Aging Sciences, University of Chieti, Chieti, Italy; Regional Healthcare Agency of the Abruzzo Region, Pescara, Italy; Local Health Unit of Pescara, Pescara, Italy. 2. Department of Medicine and Aging Sciences, University of Chieti, Chieti, Italy; Regional Healthcare Agency of the Abruzzo Region, Pescara, Italy; Local Health Unit of Pescara, Pescara, Italy. Electronic address: lmanzoli@post.harvard.edu. 3. Institute of Public Health, Catholic University of Sacred Heart, Rome, Italy. 4. Department of Public Health and Infectious Diseases, Sapienza University of Rome, Rome, Italy. 5. University of Turin, Torino, Italy. 6. Institute of Public Health, Catholic University of Sacred Heart, Rome, Italy; Italian National Institute of Health, Rome, Italy. 7. Stanford Prevention Research Center, Department of Medicine and Department of Health Research and Policy, Stanford University School of Medicine, Stanford, CA; Department of Statistics, Stanford University School of Humanities and Sciences, Stanford, CA; Meta-Research Innovation Center at Stanford (METRICS), Stanford, CA.
Abstract
OBJECTIVES: To evaluate the research agenda of registered randomized trials comparing generic and brand-name drugs in terms of who sponsors them, whether they are published promptly, and whether they find favorable results. METHODS: We included randomized trials comparing the safety or efficacy of brand-name vs generic medications that were registered in ClinicalTrials.gov or other registries from January 1, 2000, through July 31, 2015. To identify published articles or results generated from such trials, we searched PubMed, Scopus, Google, and registry databases. Data were compared across sponsorship categories ("inbred" if the compared drugs were owned by the same company or its partners/subsidiaries, "competitive" if the compared drugs were owned by competing companies, and "apparently nonprofit"), and time to publication was evaluated with Cox analysis. RESULTS: We found 207 registered protocols reporting on 186 completed trials. Among those trials, 37 had published their results and another 56 had posted results in registries, for a total of 93 trials with available results. Four years after trial completion, results were available for 64 of 138 trials (46.4%), with substantial differences by sponsor: 70.8% (34 of 48), 28.1% (18 of 64), and 46.2% (12 of 26) of the inbred, competitive, and nonprofit trials, respectively. In multivariate modeling, inbred trials had a 1.73-fold risk of having results available compared with competitive trials (P=.04). Almost all trials reported favorable results, with the exception of 4 (4.3% of the 93 trials with results). CONCLUSION: Despite the importance of generic drugs, relatively few registered randomized trials have compared the health effects of generic vs brand-name medicines, and there is an associated unsatisfactory publication rate and almost ubiquitous favorable results. The overall literature on the topic is at high risk of bias, possibly in favor of generic drugs. Higher nonprofit funding and stronger pressure to register trials and publish results are needed.
OBJECTIVES: To evaluate the research agenda of registered randomized trials comparing generic and brand-name drugs in terms of who sponsors them, whether they are published promptly, and whether they find favorable results. METHODS: We included randomized trials comparing the safety or efficacy of brand-name vs generic medications that were registered in ClinicalTrials.gov or other registries from January 1, 2000, through July 31, 2015. To identify published articles or results generated from such trials, we searched PubMed, Scopus, Google, and registry databases. Data were compared across sponsorship categories ("inbred" if the compared drugs were owned by the same company or its partners/subsidiaries, "competitive" if the compared drugs were owned by competing companies, and "apparently nonprofit"), and time to publication was evaluated with Cox analysis. RESULTS: We found 207 registered protocols reporting on 186 completed trials. Among those trials, 37 had published their results and another 56 had posted results in registries, for a total of 93 trials with available results. Four years after trial completion, results were available for 64 of 138 trials (46.4%), with substantial differences by sponsor: 70.8% (34 of 48), 28.1% (18 of 64), and 46.2% (12 of 26) of the inbred, competitive, and nonprofit trials, respectively. In multivariate modeling, inbred trials had a 1.73-fold risk of having results available compared with competitive trials (P=.04). Almost all trials reported favorable results, with the exception of 4 (4.3% of the 93 trials with results). CONCLUSION: Despite the importance of generic drugs, relatively few registered randomized trials have compared the health effects of generic vs brand-name medicines, and there is an associated unsatisfactory publication rate and almost ubiquitous favorable results. The overall literature on the topic is at high risk of bias, possibly in favor of generic drugs. Higher nonprofit funding and stronger pressure to register trials and publish results are needed.