Triage Score: A Counterfactual Risk Assessment Instrument

2026-06-18

AI Summary

Traditional risk scores estimate the probability of adverse outcomes under no intervention, offering limited support for high-stakes decisions that require balancing multiple potential outcomes. This work proposes a “triage score” that explicitly incorporates counterfactual outcomes under different interventions into the decision framework through additive counterfactual utility modeling, with conventional risk scores emerging as a special case. By integrating counterfactual reasoning, additive utility models, and randomized controlled trial data, this approach is the first to embed counterfactual utilities directly into a scoring system. Applied to pretrial risk assessment, it effectively captures complex utility structures and yields policy evaluation and learning conclusions that markedly diverge from those of traditional methods, thereby overcoming the limitations inherent in single-outcome prediction.

Abstract

Risk assessment instruments, also known as "risk scores," are widely used in high-stakes decision-making settings such as medicine and the criminal justice system. A risk score predicts the likelihood of an undesired outcome if no intervention is made. Thus, a sufficiently high score is often interpreted as a recommendation to intervene. However, risk scores fail to account for what would happen if a decision-maker does intervene. This failure is problematic because effective decision making requires consideration of both or multiple potential outcomes. We propose "triage scores," which are based on additive counterfactual utilities and include risk scores as a special case. Unlike risk scores, triage scores can incorporate counterfactual outcomes under alternative decisions, enabling decision makers to incorporate a wide range of ethical and practical factors. We illustrate the use of triage scores with an application to our own randomized controlled trial evaluating a pretrial risk score. Our analysis demonstrates that triage scores are able to capture rich utility structures and yield substantively distinct results regarding policy evaluation and learning.

Problem

• risk assessment
• counterfactual reasoning
• decision-making
• intervention
• triage

Innovation

• triage score
• counterfactual utility
• risk assessment
• decision-making
• policy evaluation