Opioid, cocaine, and amphetamine use disorders are associated with higher30-day inpatient readmission rates in the United States

Abstract

Background: Patients with substance use disorders (SUDs) are more likely to experience serious health problems, high healthcare utilization, and premature death. However, little is known about the contribution of SUDs to medical 30-day readmission risk. We examined the association between SUDs and 30-day all cause readmission among non-pregnant adult in-patients in the US. Methods: We conducted a retrospective study using 2010–2014 data from the Nationwide Readmissions Database. Our primary focus was on opioid use compared to stimulant use (cocaine and amphetamine) identified by ICD-9-CM diagnosis codes in index hospitalizations. Multivariable logistic regression models were used to estimate adjusted odds ratios and 95% CI representing the association between substance use and 30-day readmission, overall and stratified by the principal reason for the index hospitalization. Results: Nearly 118 million index hospitalizations were included in the study, 4% were associated with opioid or stimulant use disorder. Readmission rates for users (19.5%) were higher than for nonusers (15.7%), with slight variation by the type of substance used: cocaine (21.8%), opioid (19.0%), and amphetamine (17.5%). After adjusting for key demographic, socioeconomic, clinical, and health system characteristics, SUDs and stimulant use disorders increased the odds of 30-day all-cause readmission by 20%. Conclusions: Reducing the frequency of inpatient readmission is an important goal for improving the quality of care and ensuring proper transition to residential/outpatient care among patients with SUDs. Differences between groups may suggest directions for further investigation into the distinct needs and challenges of hospitalized opioid- and other drug-exposed patients.

Keywords:

• Opioid use disorder; substance use disorder; 30-day readmission; inpatient hospitalization; stimulant use disorder

Additional information

Notes on contributors

Maria C. Mejia de Grubb

Maria C. Mejia de Grubb contributed to conception/design, data analysis, and manuscript writing. Jason L. Salemi contributed to collection/assembly of data, data analysis, and manuscript writing. Sandra J. Gonzalez contributed to data analysis and manuscript writing. Charles C. Chima contributed to data analysis and manuscript writing. Alicia A. Kowalchuk contributed to manuscript writing and provided expert opinion. Roger J. Zoorob contributed to conception/design, data analysis, and manuscript writing.