Wednesday, November 02, 2022 A generalized single-index linear threshold model for identifying treatment-sensitive subsets based on multiple covariates and longitudinal measurements Ge X, Peng Y, Tu D. A generalized single-index linear threshold model for identifying treatment-sensitive subsets based on multiple covariates and longitudinal measurements (ONLINE). Can J Statistics 2022. https://doi.org/10.1002/cjs.11737 Abstract: Identification of a subset of patients who may be sensitive to a specific treatment is an important step towards personalized medicine. We consider the case where the effect of a treatment is assessed by longitudinal measurements, which may be continuous or categorical, such as quality of life scores assessed over the duration of a clinical trial. We assume that multiple baseline covariates, such as age and expression levels of genes, are available, and propose a generalized single-index linear threshold model to identify the treatment-sensitive subset and assess the treatment-by-subset interaction after combining these covariates. Because the model involves an indicator function with unknown parameters, conventional procedures are difficult to apply for inferences of the parameters in the model. We define smoothed generalized estimating equations and propose an inference procedure based on these equations with an efficient spectral algorithm to find their solutions. The proposed procedure is evaluated through simulation studies and an application to the analysis of data from a randomized clinical trial in advanced pancreatic cancer. Unified estimation for Cox regression model with nonmonotone missing at random covariates. Thiessen DL, Zhao Y, Tu D. Unified estimation for Cox regression model with nonmonotone missing at random covariates. Statist Med 41: 4781-90, 2022. https://doi.org/10.1002/sim.9512 Abstract: This article investigates a unified estimator for Cox regression model (Cox, 1972) when covariate data are missing at random (Rubin, 1976). It extends the idea of using parametric working models (Zhao and Liu, 2021) to extract the partial information contained in the incomplete observations. The working models are flexible and convenient to deal with nonmonotone missing data patterns. It can also incorporate auxiliary variables into the analysis to reduce estimation bias and improve efficiency. The unified estimator is consistent and more efficient than the (weighted) complete case estimator. Similar to multiple imputation (MI) method (Rubin, 1987 and 1996), the proposed method is also based on standard (weighted) complete data analysis and can be easily implemented in standard software. Simulation studies comparing the unified estimator with the substantive model compatible modification of the fully conditional specification MI (SMC-FCS) estimator (Bartlett et al., 2015) in various settings indicate that the unified estimator is consistent and as efficient as SMC-FCS estimator. Data from a clinical trial in patients with early breast cancer are analyzed for illustration.
