Merging statewide data in a public/university collaboration to address opioid use disorder and overdose

William C Becker,Alex M Secora,Catherine M Dormitzer,Lauretta E Grau,David A Fiellin,Robert Heimer,Gail D’Onofrio,Hsiu-Ju Lin,Kathryn Hawk,Molly Doernberg

doi:10.1186/s13722-020-00211-9

Abstract

ObjectiveDescribe methods to compile a unified database from disparate state agency datasets linking person-level data on controlled substance prescribing, overdose, and treatment for opioid use disorder in Connecticut.MethodsA multidisciplinary team of university, state and federal agency experts planned steps to build the data analytic system: stakeholder engagement, articulation of metrics, funding to establish the system, determination of needed data, accessing data and merging, and matching patient-level data.ResultsStakeholder meetings occurred over a 6-month period driving selection of metrics and funding was obtained through a grant from the Food and Drug Administration. Through multi-stakeholder collaborations and memoranda of understanding, we identified relevant data sources, merged them and matched individuals across the merged dataset. The dataset contains information on sociodemographics, treatments and outcomes. Step-by-step processes are presented for dissemination.ConclusionsCreation of a unified database linking multiple sources in a timely and ongoing fashion may assist other states to monitor the public health impact of controlled substances, identify and implement interventions, and assess their effectiveness.

Highlights

Connecticut has experienced a fivefold increase in its annual opioid overdose rate over the past decade; it had the 10th-highest overdose death rate among the 50 states in 2018 [1]
Based on related work in Maryland [4] adapted to our local context, the Connecticut Opioid REsponse (CORE) team proceeded through six steps to build the data analytic system: (1) stakeholder engagement, (2) articulation of key metrics, (3) funding to establish the system, (4) determination of the data sources needed to track metrics, (5) accessing data and (6) mergingand matching identified patient-level data across the datasets
Stakeholder engagement To succeed in the data sharing goal, stakeholders were selected for their access to and knowledge of important datasets and held the necessary authority or influence to facilitate signing of memoranda of understanding (MOUs) to enable cross-agency data sharing agreements

Summary

Introduction

Connecticut has experienced a fivefold increase in its annual opioid overdose rate over the past decade; it had the 10th-highest overdose death rate among the 50 states in 2018 [1]. Connecticut’s overdose crisis has become a major focus of public and political discourse and, increasingly, action. To help inform the state’s actions, in May 2016 the Connecticut Governor requested creation of a strategic initiative from a team of experts from Yale University’s Schools of Medicine and Public Health [2]. After a three-month effort that included a review of regional and national data, scientific literature, and consultation with stakeholders, six strategies emerged, including one that is the subject of this report—increasing data sharing across agencies and organizations.

Methods

Results

Conclusion