We present and evaluate a method for predicting individual treatment benefits based on random effects logistic regression models of binary outcomes that change over time. The method uses empirical Bayes predictors based on patients’ characteristics and responses to treatment. It is applicable to both 1-dimentional and 2-dimentional personalized medicine models. Comparisons between predicted and true benefits for simulated new patients using correlations, relative biases and mean-squared errors were used to evaluate prediction performance. The predicted benefits had relatively small biases and relatively high correlations with the true benefits in the simulated new patients. The predictors also captured estimated overall population trends in the evolution of individual benefits. The proposed approach can be used to retrospectively evaluate patients’ responses in a clinical trial, or to retrospectively or prospectively predict individual benefits of different treatments for new patients using patients’ characteristics and previous responses. The method is used to examine changes in the disorganized dimension of antipsychotic-naïve patients from an antipsychotic randomized clinical trial. Retrospective prediction of individual benefits revealed that more cannabis users had slower and lower responses to antipsychotic treatment as compared to non-cannabis users, revealing cannabis use as a negative prognostic factor for psychotic disorders in the disorganized dimension.
|Number of pages||25|
|Journal||Journal of Biopharmaceutical Statistics|
|State||Published - 2020|
Bibliographical noteFunding Information:
Dr. Zhang is a consultant at Boston Strategic Partners, Inc. (BSP), Boston, MA, and received no financial support from BSP for this study. In the last 36 months, Dr. Crespo-Facorro has received honoraria for his participation as a speaker at educational events from Otsuka, Lundbeck and Johnson & Johnson. He has been a consultant and/or advisor to or has received honoraria from Alkermes, Otsuka, Menarini and Teva. Drs. de Leon and Diaz declare no conflict of interest in the last 36 months.
The antipsychotic data was collected in a RCT supported by several grants described in Pelayo-Ter?n et al. (2014). Drs. Diaz, Zhang and de Leon were not involved in the funding, planning or operation of the RCT. Dr. Diaz was supported by an Institutional Clinical and Translational Science Award from the National Institutes of Health (NIH), NIH/NCATS Grant Number UL1TR002366 (KL2/TL1), awarded to the University of Kansas Medical Center. This manuscript may not reflect the opinions or views of the NIH;National Institutes of Health [UL1TR002366]; The authors thank Drs. Jonathan Mahnken, Jo Wick and Milind Phadnis for their comments and suggestions, and Lorraine Maw, M.A., who helped in editing this article. The authors also thank the two anonymous reviewers and anonymous Associate Editor for useful comments and suggestions that contributed to substantial improvements to the manuscript.
© 2020, © 2020 Taylor & Francis Group, LLC.
- Individual benefits
- empirical Bayesian prediction
- longitudinal binary outcomes
ASJC Scopus subject areas
- Statistics and Probability
- Pharmacology (medical)