NINDS awards $3.14 million grant to Drs. Ertefaie, McDermott, and Venuto to advance personalized medicine in Parkinson’s disease using harmonized multi-site clinical data. Parkinson’s disease (PD) manifests as a heterogeneous clinical syndrome and the variability in the clinical phenotype highlights the need to tailor the type and/or the dosage of treatment to the specific and changing needs of individuals. However, the relative lack of comparative evidence for different classes of drugs and the timing of their initiation has created challenges in devising recommendations to follow any specific therapeutic strategy. This two-phase study, funded by NINDS, will attempt to fill this important gap. The first phase (R61) focuses on creating a harmonized and curated data set by integrating data from six clinical trials and an observational study. In the second phase (R33), the harmonized data set will be leveraged to develop high quality individualized treatment strategies for PD with respect to several clinical outcomes. A robust marginal structural model will be developed that has better convergence properties than existing methods and leverages a non-parametric regression approach to mitigate the chance of misspecification of the nuisance parameters while providing valid inference (p-values and confidence intervals) for the parameters of interest.
Robust Q-learning
Ertefaie A, McKay J. R., Oslin D., and Strawderman R. L. (2020). Journal of the American Statistical Association, DOI: 10.1080/01621459.2020.1753522 Q-learning is a regression-based approach that is