Concentration of Benefit index: A threshold-free summary metric for quantifying the capacity of covariates to yield efficient treatment rules

TL;DR

The paper introduces the Concentration of Benefit index, a threshold-free metric that measures how well covariates can identify individuals who will benefit most from treatment, aiding personalized treatment decisions.

Contribution

It proposes a novel, threshold-free index for quantifying covariate capacity to identify treatment benefit, with estimators and practical implementation demonstrated using clinical trial data.

Findings

The index can be estimated with simple regression models.

It provides an intuitive and theoretically sound measure of covariate utility.

The method is applicable for out-of-sample validation and correction for optimism.

Abstract

When data on treatment assignment, outcomes, and covariates from a randomized trial are available, a question of interest is to what extent covariates can be used to optimize treatment decisions. Statistical hypothesis testing of covariate-by-treatment interaction is ill-suited for this purpose. The application of decision theory results in treatment rules that compare the expected benefit of treatment given the patient's covariates against a treatment threshold. However, determining treatment threshold is often context-specific, and any given threshold might seem arbitrary when the overall capacity towards predicting treatment benefit is of concern. We propose the Concentration of Benefit index (Cb), a threshold-free metric that quantifies the combined performance of covariates towards finding individuals who will benefit the most from treatment. The construct of the proposed index is comparing expected treatment outcomes with and without knowledge of covariates when one of a two randomly selected patients are to be treated. We show that the resulting index can also be expressed in terms of the integrated efficiency of individualized treatment decision over the entire range of treatment thresholds. We propose parametric and semi-parametric estimators, the latter being suitable for out-of-sample validation and correction for optimism. We used data from a clinical trial to demonstrate the calculations in a step-by-step fashion, and have provided the R code for implementation (https://github.com/msadatsafavi/txBenefit). The proposed index has intuitive and theoretically sound interpretation and can be estimated with relative ease for a wide class of regression models. Beyond the conceptual developments, various aspects of estimation and inference for such a metric need to be pursued in future research.