Patient-level Risk Identifier Models for a Multifactor Opioid Abuse Risk Assessment Strategy

Principled Strategies, Inc., Del Mar CA

Investigators

Abstract

ABSTRACT For this Phase I project, Principled Strategies will develop innovative, patient-level models for opioid risk identification and integrate them into the SafeUseNow managed care system ?an actionable solution for combating prescription drug abuse that currently operates at the prescriber level only. Incorporating patient-level risk identifier models will strengthen an already powerful and demonstrably effective program and constitutes a critical step in generating a first-in-class, multifactor risk assessment strategy that is truly holistic. Using a variety of data sources, advanced analytics, and multiple empirically validated risk identification models, the ground- breaking advancement in SafeUseNow technology will enable healthcare stakeholders to identify combinations of prescribers, patients, and pharmacies whose behaviors may contribute to prescription drug abuse. Expanding the capabilities of the system in this way will permit earlier identification of potential high-risk patients, enabling new approaches to eliminate high-risk prescription activities. To achieve these objectives, our Phase I project entails the following Aims 1) Request, receive and assess Centers for Medicare & Medicaid Services (CMS) analytic dataset; 2) Supplement CMS datasets with alternatively sourced descriptive data for dimension tables; 3) Build national patient level risk identification models for all outcomes of interest. Successful completion of this Phase I project will yield one or more patient-level opioid risk assessment of models that meet target performance levels (e.g., sensitivity, specificity, positive prediction value). This will enable Phase II research and development to incorporate regional and state-level datasets, determine the preferred scale of implementation for production (local vs. national), and evaluate risk identification, which is expected to be significantly improved. The completed next-generation SafeUseNow prototype, with its combination of prescriber- and patient-level risk models, will then be ready for pilot studies to demonstrate the improved capacity of the managed care system to eliminate excessive and higher risk prescription activity, ultimately reducing opioid use, opioid overdoses, overdose-related mortality, and the prevalence of opioid use disorder.