Patterns of health care utilization and cost before and after opioid overdose: findings from 10-year longitudinal health plan claims data

Abstract

Objective

To describe the longitudinal pattern of health care utilization and cost of care before and after opioid overdose (OD) over a 10-year period using health plan claims data.

Methods

Patients who had experienced opioid ODs between April 2005 and March 2015 were identified from Geisinger Health System’s electronic health records. Among these patients, a subgroup of patients who were Geisinger Health Plan (GHP) members at any point between January 2006 and December 2015 were also identified. From the corresponding GHP claims data, their all-cause health care utilization (inpatient admissions, emergency department [ED] visits, and physician office visits) and total medical costs, excluding prescription medication cost, were obtained. Per-member-per-month estimates for each month before and after the index date of opioid OD were calculated, adjusting for age, gender, plan type, year, and comorbidity via multivariate regression models.

Results

A total of 942 opioid OD patients with an average GHP enrollment period of 41.4 months were identified. ED visit rates rose rapidly starting around 19–24 months prior to the opioid OD date. Acute inpatient admission rates and total medical cost also rose rapidly starting around 12 months prior. After the OD date, the utilization rates and cost declined but tended to remain above those of the pre-OD period.

Conclusion

Opioid OD is preceded by sharp increases in utilization of acute care and cost well before the actual OD. These findings therefore suggest that early signals of OD may be detected from patterns of acute care utilization, particularly the ED visits.

Keywords:

• opioid
• overdose
• utilization
• cost of care
• electronic health records
• claims data

Supplementary materials

Table S1 Full regression output for total cost and acute care

Table S2 Full regression output for physician office visits

Acknowledgments

Support for this study included research funding from Indivior, Inc., Richmond, VA, USA.

Disclosure

Dr Maeng conducted this study as a part of his employment with Geisinger Health System. All authors are current employees of Geisinger Health System. The authors report no other conflicts of interest in this work.