Dopamine agonists in animal models of Parkinson's disease: a systematic review and meta-analysis

Background: Parkinson's disease (PD) can be a severely disabling condition in spite of therapies currently available. Systematic review and meta-analysis can provide an overview of a field of research and identify potential sources of bias and limits to efficacy. In this study we use these tools to describe the reported efficacy of dopamine agonists in animal models of PD. Methods: Publications were identified by electronic searching of three online databases. Data were extracted for neurobehavioural outcome, for study design and for the reporting of measures to avoid bias. Standardised mean difference meta-analysis was used to provide summary estimates of efficacy, with the effects of study quality and study design explored using stratified meta-analysis. Results: 253 publications reported the use of a dopamine agonist in an animal model of PD; of these 121 reported data suitable for inclusion in meta-analysis. 47 interventions were tested in 601 experiments using 4181 animals. Overall, neurobehavioural outcome was improved by 1.08 standard deviations (SD; 95% Confidence Interval (CI) 0.97-1.19). Reporting of measures to reduce bias was low and publications which reported the blinded assessment of outcome had significantly smaller effect sizes (0.85, 95% CI 0.64 to 1.07) than those which did not (1.18, 95% CI 1.05 to 1.31, p < 0.005). Conclusions: While dopamine agonists do appear to have efficacy in animal models of PD the low prevalence of reporting of measures to avoid bias is of concern. Systematic review of individual interventions may be helpful in the design of future preclinical and clinical trials.