Objective: To develop an individualized treatment rule (ITR) for early steroid use in hospitalized patients with CAP.

Methods: Using a single center cohort of hospitalized patients with CAP from 2009 to 2019, we developed a single decision ITR to initiate or not initiate steroids early (within 24 hours) after admission. The outcome of interest was hospital-free days. Demographic, physiologic, and laboratory data represented tailoring predictor variables: data within 6 hours of admission for delayed steroid use (>24 hours) or no steroid use, and data prior to steroid administration for early steroid use (within 24 hours) respectively. Regression-based learning with LASSO selected a model estimating expected outcomes of potential intervention with steroids individualized to predictors. The ITR is the strategy maximizing expected hospital-free days. Missing data for predictor variables were multiply imputed.

Results: A total of 4383 patients were included in this cohort with median age 73.6, 54% male and 94% white race. 1413 (32.2%) patients received steroids within 24 hours of hospital admission. Our ITR identified one interaction with steroids: COPD (present in 36% of patients). Mean hospital-free days in our observed data was 21.68. Under a hypothetical treatment rule where no patients receive steroids, we estimated mean hospital-free days of 21.65. Whereas under the optimal ITR, mean hospital-free days was 21.71.

Conclusion: In this study, an ITR for early steroid use was not clinically meaningful at increasing the number of hospital-free days.