Demystifying Optimal Dynamic Treatment Regimes

• Published in: Biometrics Vol. 63; no. 2; pp. 447 - 455
• Main Authors: Moodie, Erica E.M, Richardson, Thomas S, Stephens, David A
• Format: Journal Article
• Language: English
• Published: Malden, USA Blackwell Publishing Inc 01.06.2007; International Biometric Society; Blackwell Publishing Ltd
• Subjects: Acquired Immunodeficiency Syndrome - drug therapy; Acquired Immunodeficiency Syndrome - immunology; AIDS; Algorithms; Anti-HIV Agents - administration & dosage; Anti-HIV Agents - therapeutic use; Biometrics; Biometry; Biostatistics; CD4 Lymphocyte Count; Clinical trials; Clinical Trials as Topic - statistics & numerical data; Cohort Studies; Contrafactuals; Decision support systems; Estimation methods; Humans; Mathematical functions; Models, Statistical; Multicenter Studies as Topic - statistics & numerical data; Musical intervals; Optimal dynamic regimes; Optimal structural nested mean models; Optimization; Parametric models; Random Allocation; Randomized controlled trials; Randomized Controlled Trials as Topic - statistics & numerical data; Research methodology; Sequential randomization; Time Factors; Treatment algorithms; Zidovudine - administration & dosage; Zidovudine - therapeutic use
• ISSN: 0006-341X; 1541-0420
• DOI: 10.1111/j.1541-0420.2006.00686.x

A dynamic regime is a function that takes treatment and covariate history and baseline covariates as inputs and returns a decision to be made. Murphy (2003, Journal of the Royal Statistical Society, Series B65, 331-366) and Robins (2004, Proceedings of the Second Seattle Symposium on Biostatistics, 189-326) have proposed models and developed semiparametric methods for making inference about the optimal regime in a multi-interval trial that provide clear advantages over traditional parametric approaches. We show that Murphy's model is a special case of Robins's and that the methods are closely related but not equivalent. Interesting features of the methods are highlighted using the Multicenter AIDS Cohort Study and through simulation.