Predicting Readmission or Death After Acute ST-Elevation Myocardial Infarction

Authors

  • Jeremiah R. Brown PhD, MS,

    1. The Dartmouth Institute for Health Policy and Clinical Practice, Geisel School of Medicine at Dartmouth,Lebanon, New Hampshire
    2. Section of Cardiology, Department of Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
    Search for more papers by this author
  • Sheila M. Conley BSN,

    1. Section of Cardiology, Department of Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
    Search for more papers by this author
  • Nathaniel W. Niles II MD

    1. Section of Cardiology, Department of Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
    Search for more papers by this author

  • Dr. Brown is supported by grant number K01HS018443 from the Agency for Healthcare Research and Quality. Funding for the STEMI registry is supported by the Section of Cardiology, Dartmouth-Hitchcock Medical Center.

  • The authors have no other funding, financial relationships, or conflicts of interest to disclose.

Address for correspondence: Jeremiah R. Brown, PhD, MS Dartmouth Hitchcock Medical Center, HB 7505 One Medical Center Drive Lebanon, NH 03756 jbrown@dartmouth.edu

Abstract

Background

Risk factors for emergent readmissions or death after acute myocardial infarction (AMI) are important in identifying patients at risk for major adverse events. However, there has been limited investigation conducted of prospective clinical registries to determine relevant risk factors.

Hypothesis

We hypothesize 30-day readmission or death could be predicted using patient, procedural, and process factors.

Methods

Patients presenting with ST-elevation myocardial infarction (STEMI) from 2006 to 2011 were prospectively enrolled in a STEMI registry (1271 patients). Thirty-day readmission was ascertained by administrative claims data. Death was determined by linking to the Social Security Death Master File. Univariate and stepwise multivariate logistic regression was conducted with Hosmer-Lemeshow goodness-of-fit statistics for model calibration and receiver operating characteristic (ROC) curve for model discrimination.

Results

The combined end point of 30-day readmission or postdischarge death included 135 patients (10.6%), including 109 emergent readmissions and 26 deaths. Factors associated with an increase risk of 30-day readmission or postdischarge death included age ≥80 years, diabetes, chest pain or cardiac arrest at presentation, and 3-vessel disease found at initial angiography. Factors associated with a decreased risk of 30-day readmission or postdischarge death included transfer to the catheterization lab from another emergency department, clopidogrel given during the procedure hypercholesterolemia, and receiving aspirin, β-blockers, and angiotensin-converting enzyme or angiotensin receptor blocker inhibitors at discharge. Index admission outcomes indicative of readmission or death postdischarge only included a new diagnosis of congestive heart failure. The model discriminated well with an ROC of 0.71 (95% confidence interval: 0.66-0.76).

Conclusions

Prehospitalization factors are overlooked and are important factors to incorporate in routine risk prediction models for readmission or death within 30 days following an AMI.

Ancillary