Evaluating the use of generalized dynamic weighted ordinary least squares for individualized HIV treatment strategies

Dynamic treatment regimes (DTR) are a statistical paradigm in precision medicine which aim to optimize patient outcomes by individualizing treatments. At its simplest, DTR may require only single decision be made; this special case is called an individualized rule (ITR) and often used maximize short-term rewards. Generalized dynamic weighted ordinary least squares (G-dWOLS), estimation method that offers theoretical advantages such as double robustness of parameter estimators the rules, has been recently extended now accommodate categorical In work, G-dWOLS applied longitudinal data estimate optimal ITR, demonstrated simulations. This novel then population affected HIV whereby ITR for administration Interleukin 7 (IL-7) devised duration where CD4 load above healthy threshold (500 cells/$\mu$L) while preventing unnecessary injections.