Self-reported conflicts of interest (sfCOI) of authors and the interpretation of randomized phase III trials (RCT) and related editorials (REd) in cancer research
Nenhuma Miniatura disponível
Citações na Scopus
Tipo de produção
conferenceObject
Data de publicação
2012
Editora
AMER SOC CLINICAL ONCOLOGY
Indexadores
Título da Revista
ISSN da Revista
Título do Volume
Autor de Grupo de pesquisa
Editores
Coordenadores
Organizadores
Citação
JOURNAL OF CLINICAL ONCOLOGY, v.30, n.15, suppl.S, 2012
Resumo
Background: Growing participation from industry in cancer research has resulted in increased reporting of COI. We aimed to test any association between author’s conclusion and sfCOI in cancer studies. Methods: All RCT and REd published in 6 major cancer journals in a 3.5 year period were selected. Two investigators blinded to COI disclosure independently analyzed each RCT and REd, classifying authors’ conclusions as highly positive, positive, neutral, negative, and highly negative with respect to author’s opinion on the experimental therapy. The agreement rate between investigators for conclusion classification was 90% (consensus was achieved for the remaining 10%). We also collected data on study results, COI and sponsorship. COI was defined as any self-reported financial tie between author and industry except for research funds. Predictors of positive/highly positive conclusions of RCT and of REd were tested separately in logistic regression multivariable models. Results: From Jan 2008 to Oct 2011, 1,485 articles were retrieved: 150 RCT and 140 REd were eligible. Among the RCT, 82 (55%) were positive, and 78 (52%) were entirely or partially funded by industry. Any sfCOI was present in 103 (69%) RCT and in 71 (47%) REd. Conclusions of REd and RCT were: 7.3% and 11.3% highly positive, 42.7% and 57.3% positive, 8.0% and 2.0% neutral, 29.3% and 18.7% negative, and 12.7% and 10.7% highly negative, respectively. Multivariable analysis showed that RCT positive result was the only significant predictor for positive conclusion by RCT authors (OR=109, 95% CI: 21-567; p<0.001). The only factor associated with positive conclusions of REd authors was a positive conclusion by RCT author (OR=42, 95% CI: 7-244; p<0.001). While 64 (43%) RCT reported negative results, 103 (68.7%) RCT authors interpreted studies positively. Logistic regression for discordance between RCT result and RCT conclusion did not find any association with COI. Conclusions: The interpretation of RCT results by authors was not influenced by sfCOI or trial sponsorship. Authors of REd were not influenced by study results or by their sfCOI when discussing cancer RCT.