Drug-Associated Endocarditis Admissions Fell as Stimulant Share Rose

Key Takeaways

• Estimated hospitalizations among US adults with DUA-IE declined each year after a 2020 peak.
• The opioid-only share of DUA-IE hospitalizations fell from 49.4% in 2016 to 39.9% in 2022.
• The stimulant-only share rose from 14.7% in 2016 to 18.3% in 2022.
• The combined opioid–stimulant share rose from 22.6% in 2016 to 28.6% in 2022.

Drug use-associated infective endocarditis (DUA-IE) is life-threatening, and changing patterns of drug use complicate the picture of who is hospitalized with the condition. Shifts in access to harm reduction and addiction treatment also raise questions about the national admission burden. Across the United States, the mix of drugs associated with these hospitalizations may differ by region, making geography as well as time relevant to investigators’ assessment of inpatient hospitalizations.

Investigators in a Clinical Infectious Diseases analysis used a representative sample of US adult hospitalizations from 2016 through 2022. They accounted for survey design and applied weights to generate national estimates, stratifying hospitalizations by year, census division, and drug category: opioid-only, stimulant-only, combined opioid–stimulant, or other drugs. Logistic regression estimated the marginal probability of stimulant-associated DUA-IE while accounting for patient and hospital characteristics. Hospitalizations, rather than distinct individuals, were the analytical unit.

An estimated 118,190 US adult DUA-IE hospitalizations occurred during the study period. Annual hospitalizations reached a high of 18,735 in 2020, then declined each year to 15,345 in 2022. The overall decrease coincided with a changing mix of drug categories.

In regression-adjusted estimates across the modeled hospitalizations, the marginal probability of stimulant involvement was 0.59 (95% CI 0.56–0.62) in the Pacific division versus 0.34 (95% CI 0.31–0.38) in New England. These are modeled probabilities, not crude regional shares or observed proportions for a particular year. The investigators also found that stimulants surpassed opioids in several Western census divisions.

Trends in hospitalizations cannot establish why admissions fell or measure the effect of any particular intervention. The authors suggested that existing harm reduction and addiction treatment efforts may have contributed to the decline, but they did not present that possibility as a demonstrated treatment effect. The changing drug mix describes which categories accounted for admissions, not why the national count changed.

Declining overall admissions accompanied increasingly prominent stimulant involvement, especially in the Western United States, according to the investigators.

Clinician Questions

Which patient and hospital characteristics were considered in estimating stimulant involvement in US DUA-IE hospitalizations?

For US adults hospitalized with drug use-associated infective endocarditis, the logistic regression assessing the marginal probability of stimulant-associated disease considered patient age, sex, race/ethnicity, and insurance, along with hospital location/teaching status, bed size, and census division.

Does a higher stimulant share of US drug use-associated endocarditis admissions prove that the absolute number of stimulant-related admissions rose?

No. Stimulant-only and combined opioid–stimulant shares are proportions of US adult drug use-associated infective endocarditis hospitalizations; rising shares alone do not establish an increase in absolute stimulant-associated admission counts, particularly when the overall annual hospitalization total fell after its peak. Category-specific annual counts would be needed to determine the absolute trend.