Adverse Events Reporting Suboptimal in Oncology Publications
Transparent and comprehensive reporting of AEs in published results of oncology-related clinical trials is crucial for the treatment of patients with cancer. In efforts to improve reporting of clinical trials results, the Consolidated Standards of Reporting Trials (CONSORT) extension group developed 10 recommendations in 2003 for reporting AEs. A new study of 175 publications assessed the degree to which the publication of phase 3 trials in oncology adhered with CONSORT recommendations (Sivendran S, et al. J Clin Oncol. 2014;32:83-89).
The findings from this analysis showed that the reporting of AEs is suboptimal in published studies and is characterized by significant selectivity of the data. The researchers reviewed PubMed, Medline, and EMBASE citations to identify randomized, phase 3 trials in metastatic solid malignancies published between January 2009 and December 2011. Publications were assessed for 14 AE reporting elements taken from the CONSORT harms extension statement, with a completeness score of 0 to 14 being calculated. Data on 96,125 patients were included in the analysis. The median completeness score was 8 (range, 3-12). The majority of publications (96%) only reported AEs occurring above a particular threshold rate or severity. Overall, 37% did not specify the criteria for determining which AEs were reported, and 88% grouped AEs of varying severity.
The findings showed that 91% of the articles analyzed had a title or an abstract stating whether AEs were addressed, and 58% had introductions stating whether AEs were addressed in the study. For methods, 93% specified an instrument or scale used to categorize AEs, 65% specified whether the reported AEs encompassed all recorded events or a select sample, and 62% specified a surveillance time frame. For the results, 77% of articles reported absolute numbers of AEs, 75% reported if there were deaths related to AEs, 71% reported whether patients were evaluable for toxicity, and 65% reported reasons for treatment discontinuation.
Regression analysis revealed that trials without a specified funding source and with an earlier year of publication were associated with significantly lower completeness scores.
This analysis shows substantial selectivity and heterogeneity in the reporting of AEs in oncology publications. The researchers cited potential factors for the observed variability in AE reporting and poor adherence with the recommendations, including the authors’ awareness of these recommendations and a lack of compliance with guidelines imposed by journals as a prerequisite.
This study further illustrates the importance of developing oncology-specific standards for AE reporting to ensure consistency of reporting of AEs to provide oncologists with the information required to determine treatment recommendations, and to facilitate shared decision-making.