Quality of Research on Treatment Harms in Cancer

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Informed decision making in health care requires a balancing of information on benefit and harms associated with treatments. If data on harms are not valid, not recorded, or suppressed in the published findings, then doctors will not have all data that are needed for optimal prescribing. Therefore, having accurate, reliable, and adequately reported data on both benefits and harms is an important research-integrity issue. The National Cancer Institute (NCI) via its cooperative group (COG) mechanism in the USA conducts the majority of the publicly funded randomized controlled trials (RCTs). Hence, we propose to evaluate a comprehensive population of RCTs conducted by the NCI-COGs to assess the adequacy of collecting and reporting data on harms. We propose to study the following specific aims related to treatment-associated harms in cancer randomized controlled trials: 1. We will determine the quality of reporting of data on harms from published reports by summarizing data on the most important quality domains related to harm. 2. We will assess the extent to which outcome reporting bias exists in the NCI-sponsored clinical research. We will measure outcome reporting bias by determining how often the assessment of harms was a priority specified in the original COG research protocols, and then comparing the protocols against the final publications. 3. We will evaluate whether investigators favored one treatment over another on the basis of simultaneous assessment of benefits and harms versus selective consideration of benefits only. The proposed research will provide the first systematic data on failures to adequately report harms in randomized clinical triasl, a significant questionable research practice. The findings will also inform considerations of research quality standards by both researchers and funding agency/policy-makers (including the NCI and the Agency for Healthcare Research and Quality, which funds systematic reviews on cancer treatments). If we find systematic inadequate reporting of treatment harms, the NCI will be able to undertake immediate corrective actions to improve methodological quality of RCTs and the standards of reporting of clinical trials it funds. Finally, the improvement of the reporting of harms would benefit physicians and above all patients, who need to understand the potential harms of the treatments they receive.