Emergency Heart Failure Mortality Risk Grade score performance for 7-day mortality prediction in patients with heart failure attended at the emergency department: validation in a Spanish cohort

Eur J Emerg Med. 2018 Jun;25(3):169-177. doi: 10.1097/MEJ.0000000000000422.

Abstract

Objective: The Emergency Heart Failure Mortality Risk Grade (EHMRG) scale, derived in 86 Canadian emergency departments (EDs), stratifies patients with acute-decompensated heart failure (ADHF) according to their 7-day mortality risk. We evaluated its external validity in a Spanish cohort.

Patients and methods: We applied the EHMRG scale to ADHF patients consecutively included in the Epidemiology of Acute Heart Failure in Emergency departments (EAHFE) registry (29 Spanish EDs) and measured its performance. Patients were distributed into quintiles according to the original and their self-defined score cutoffs. The 7-day mortality rates were compared internally among different categories and with categories of Canadian cohorts.

Results: The EAHFE group [n: 1553 patients; 80 (10) years; 55.6% women] had a 5.5% 7-day mortality rate and the EHMRG scale c-statistic was 0.741 (95% confidence interval: 0.688-0.793) compared with 0.807 (0.761-0.842) and 0.804 (0.763-0.840) obtained in the Canadian derivation and validation cohorts. The mortality rate of the EAHFE group mortality increased progressively as the quintile categories increased using intervals defined by either the Canadian or the Spanish EHMRG score cutoffs, although with more regular increments with the EAHFE-defined intervals; using the latter, patients at quintiles 2, 3, 4, 5a and 5b had (compared with quintile 1) odds ratios of 1.77, 3.36, 4.44, 9.39 and 16.19, respectively.

Conclusion: The EHMRG scale stratified risk in an ADHF cohort that included both palliative and nonpalliative patients in Spanish EDs, showing an extrapolation to a higher mortality risk cohort than the original derivation sample. Stratification improved when the score was recalibrated in the Spanish cohort.

Publication types

  • Evaluation Study

MeSH terms

  • Aged
  • Canada
  • Cohort Studies
  • Emergency Service, Hospital*
  • Female
  • Heart Failure / mortality*
  • Hispanic or Latino / statistics & numerical data*
  • Humans
  • Male
  • Middle Aged
  • Risk Adjustment
  • Risk Assessment / methods*
  • Risk Factors
  • Severity of Illness Index*