Background: While prescription opioids have a role in treating certain pain conditions, their injudicious use has led to an epidemic with myriad adverse outcomes. We sought to evaluate the association of socioeconomic and racial disparities with healthcare utilization and outcomes in opioid overdose related hospitalizations using National Inpatient Sample (NIS) database from 2012 through 2014.
Methods: NIS is the largest all-payer inpatient care database sponsored by the Agency for Healthcare Research and Quality and represents a random, 20% stratified sample of all inpatient hospitalizations from 46 states (representing >97% of the total US population) comprising approx. 8 million hospitalizations/year. Patients aged >18years were screened for opioid (both prescription opiates and heroin) overdose (OD) related hospitalizations using valid ICD9 codes. National estimates were generated using discharge weights. Socioeconomic determinants (age, gender, income groups and primary payer) and race were used as independent variables to determine association with measures of healthcare utilization [skilled nursing facility placement (SNF), length of stay (LOS) and charges (CHG)] and adverse outcomes [inpatient mortality (IM) and discharge against medical advice (AMA)]. STATA 15.0 was used for all statistical analysis.
Results: Of 89,951,890 weighted hospitalizations screened, a total of 144,005 hospitalizations were identified with the principle diagnosis of OD from 2012 through 2014. Prescription OD and heroin overdose comprised 80% and 20% of total OD respectively. Patients had a mean age of 47.9 years, 50.5% were female, 81% Caucasians, 50% belonged to the medium income group, and Medicare was the primary payer for 36% of the hospitalizations. Mean LOS was 3.7 days, CHG $34527.6 and IM was 3.1%. On univariate regression analysis (Table 1), increased age and female gender was associated with lower IM and AMA but increased SNF and LOS. Caucasians had the highest odds of SNF vs African Americans and other races. The lowest income group had a higher likelihood of AMA and lower odds of SNF vs the highest income group. Uninsured patients and those with private insurance had the highest likelihood of IM, while Medicare patients had lower odds of IM and AMA, likely secondary to high chances of SNF.
Conclusions: OD impose a significant burden on the US healthcare system. Over 3 year study period, vast majority of patients in this sample overdosed on prescribed opioids. Furthermore, the socioeconomic characteristics were remarkably distributed across all major categories. Understanding marked socioeconomic and racial variation in OD patterns and associated outcomes will help guide care and could have important implications for population-level overdose prevention efforts.
.jpg)