Introduction: Wake Forest Baptist Medical Center has perceived a rise in the number of injection drug use associated endocarditis cases. Many of these patients report injection of the prescription opioid oxymorphone, a drug specifically linked to a microangiopathic hemolytic anemia when injected and large HIV outbreak in Indiana.
Hypothesis: 1) The number of IDU-IE cases at our institution is increasing 2) The characteristics of these patients will differ from the classically taught phenotype of a mostly male, right sided predominant, low morbidity and mortality disease process 3) Those who injected oxymorphone will have differing clinical characteristics and clinical outcomes than those who did not inject oxymorphone.
Methods: Retrospective analysis of IE cases at an academic medical center from 1/1/2004 to 9/30/2015 stratified by IDU and reported intravenous oxymorphone use.
Results: Three hundred seven cases of IE were identified which included 92 IDU cases: 46 oxymorphone-exposed and 46 oxymorphone-unexposed . As compared to the non-IDU group, the IDU group was younger (31.3 +/-9.2 vs 54.2 +/-15.8, p<0.0001), predominantly Caucasian (95.7% vs 78.1%, p=0.0002), and female (58.7% vs 35.4%, p=0.02). When comparing the oxymorphone-exposed to oxymorphone-unexposed group, age (31.3+/-9.2 vs. 33.2+/-9.2, p=.32), race (95.7% Caucasian both groups), and gender (58.7% female vs. 41.3% p=0.1) were equivalent. As compared to the oxymorphone-unexposed group, the oxymorphone exposed group had significantly more MRSA (63% vs 48.9%, p=.007), tricuspid valve involvement (52.2% vs 30.4%, p=.03), and hepatitis C antibody seropositivity (78.3% vs 52.2%, p=.01). The oxymorphone unexposed group had significantly more aortic valve disease (23.9% vs 2.2%, p=.004), cardiac surgery (37% vs 13%, p=.008), and ICU admission (73.9% vs 54.4%, p=.05) than the oxymorphone exposed group. Inpatient mortality was comparable (6.5% exposed vs 8.7% unexposed, p=1.0) whereas one-year mortality was 19.6% in the oxymorphone exposed group and 13% in the oxymorphone unexposed group (p=.06).
Conclusions: Our results support the hypothesis that there may be differences in clinical characteristics and outcomes of IDU-IE which are associated with the specific drug of abuse.