Contemporary predictors of readmission outcomes in patients with infective endocarditis: analysis from a national readmission database. (25th November 2020)

• Record Type: Journal Article
• Authors: Khayata, M; Addoumieh, A; Alkharabsheh, S; Collier, P; Klein, A; Griffin, B; Xu, B
• Is Part Of: European heart journal. Volume 41:(2020)Supplement 2
• Publication Date: 2020-11-25
• Subjects: Health Economics; Cardiology -- Periodicals; Heart -- Diseases -- Periodicals; 616.12005
• DOI: 10.1093/ehjci/ehaa946.3534
• Languages: English
• ISSNs: 0195-668X

Abstract

Abstract: Background/Introduction: Infective endocarditis (IE) remains a serious illness that is associated with remarkable morbidity and mortality in the United States (US). There are limited studies that investigated predictors of 30-day readmission risk in this population. Purpose: We aimed to perform a contemporary analysis to investigate predictors of 30-day readmission in IE patients in the US. Methods: We used the 2017 national readmission database to identify index admissions among adults (age ≥18) with the diagnosis of IE. Appropriate International Statistical Classification (ICD-10) codes were used to identify patients with IE. Primary outcome of interest was 30-day readmission, and hospital cost was the secondary outcome. Results: Out of 49, 692 admissions for IE, 5, 743 (11.6%) patients were readmitted within 30 days. Patients who had 30-day readmission were younger (55±20 vs 61±19 years, P<0.001), 44.8% were females (P=0.08), 27.9% had diabetes mellitus (DM), 56.8% had hypertension, 37.9% had heart failure (HF), 31.3% had chronic kidney disease (CKD), 12.2% had end stage renal disease, and 47.8% had Medicare insurance. Patients who had readmissions within 30 days were more likely to have non-elective index admission (96.9% vs 93.4%, P<0.001), more likely to have hepatitis C (19.5% vs 12.2%, P<0.001), human immunodeficiency virus (HIV) (1.7% vs 1.2%, P<0.001), substance abuse (8.6% vs 5.4%, P<0.001), opioid abuse (24% vs 14.7%, P<0.001), and cocaine use (7.4% vs 4.3%, P<0.001). Overall, 5, 393 (10.9%) patients died during index admission. Median cost for readmissions within 30 days was $83, 217 [$41, 457-$165, 487], compared to the index admission cost of $90, 257 [$41, 945- $208, 851] (P<0.001). After adjusting for age, DM, HF, and CKD, substance abuse (odds ratio (OR): 1.19 [1.07–1.33]; P 0.001), opioid abuse (OR: 1.37, [1.26- 1.48]; P<0.001), cocaine use (OR: 1.33 [1.18- 1.48]; P<0.001), HIV (OR: 1.25 [1.01–1.56]; P=0.04), and hepatitis C (OR: 1.34 [1.24–1.45]; P<0.001) correlated with higher odds of readmissions within 30 days (Figure 1). Conclusion: Approximately 1 in 4 patients admitted for IE in the US had a history of opioid abuse and almost one fifth had hepatitis C. The 30-day readmission rate remains significant in IE with high financial burden on the health system. Both opioid abuse and hepatitis C were among the highest predictors of readmission within 30 days. Identifying modifiable predictors of readmission in this population may reduce readmission risk and healthcare cost. Funding Acknowledgement: Type of funding source: None