Grant: Fellowship - Priority Announcement: Knowledge Translation

The goal of health research is to understand the causes and ultimately develop treatments for human disease. The foundation of health research for humans rests with animal research and the reporting of animal research is an essential part of the health research system. Recent research has found the quality of reporting of biomedical animal experiments to be poor, and evidence of publication bias has been found with the growth of systematic reviews and meta-analyses of animal experiments. If animal research is not reported correctly or reported at all, then the scientific and ethical integrity of human health research are jeopardized. Although criticisms of reporting biases are not new, there is growing evidence that it undermines the ability of animal models to reliably inform us about clinical interventions. For instance, Sena and colleagues examined systematic reviews of animal models for acute ischaemic stroke and found statistical evidence that publication bias caused an exaggerated estimate of effect for treatment of 30%. There are economic costs from publication bias in animal research such as bringing falsely positive treatments to clinical trials, and incorrect reporting or failure to report research findings cannot justify such waste. There are also ethical costs to volunteers in clinical trials and patients awaiting treatments who depend on a health research system that protects them and ensures their welfare is based on sound science, not biased reporting. Similarly, the ethical use of animals in research requires that their use not be wasted due to biases. This application is to conduct the first direct assessment of publication biases of biomedical animal research.