SPLINE-BASED STRATIFICATION FOR CONTINUOUS CONFOUNDERS IN OBSERVATIONAL STUDIES

In observational studies aimed at estimating the causal effect of an explanatory variable X on an outcome Y in the presence of confounding variables Z, we propose a novel spline-based stratification method for confounding adjustment when both X and Z are continuous. The proposed method utilizes inflection points of smoothing splines to define cut-off values for stratifying the data into relatively homogeneous subgroups. We demonstrate the utility of the approach using a stroke rehabilitation dataset, revealing essential findings that are overlooked by conventional regression-based adjustment methods.