Performance of the Meta-Analysis Global Group in Chronic Heart Failure Score in Black Patients Compared With Whites

Circ Cardiovasc Qual Outcomes. 2019 Jul;12(7):e004714. doi: 10.1161/CIRCOUTCOMES.118.004714. Epub 2019 Jul 3.

Abstract

Background: Risk stratification is critical in heart failure (HF) and the Meta-Analysis Global Group in Chronic HF (MAGGIC) score is a validated tool derived from ~40,000 patients. However, few of these patients self-identified as black, raising uncertainty regarding performance in blacks with HF.

Methods and results: This study analyzed a racially diverse group of 4046 patients (1646 black and 2400 white) from a single center from 2007 to 2015. Baseline characteristics were collected to tabulate MAGGIC score and test its discrimination and calibration within race groups. The primary end point was all-cause mortality. Death was detected using system records and the social security death master file. Discrimination was tested using Cox models of MAGGIC score stratified by race, and combined analysis including MAGGIC, race, and MAGGIC×race. Calibration was assessed using linear regression models and plots of observed versus predicted data. Overall, 901 (21%) patients died during 1-year follow-up. MAGGIC score discrimination was similar in both race groups in terms of C statistic (0.707±0.027 versus 0.725±0.014, for black versus white; P=0.556) and the hazard ratio (HR) per MAGGIC point was 1.12 in black patients (95% CI, 1.10-1.14) and 1.13 in white patients (95% CI, 1.12-1.14). Race was a significant correlate of survival, with better survival in black patients compared with white (HR, 0.66; 95% CI, 0.56-0.78), but the interaction of MAGGIC×race was not significant (β=-0.013; P=0.16), and adding race to the model did not improve discrimination (C statistic for MAGGIC versus MAGGIC+race, 0.721 versus 0.722; P=0.79). In calibration testing, the slope was not significantly different from 1 in either group, but the groups differed from each other, and it was closer to unity among black patients (0.94 versus 1.4; P=0.004).

Conclusions: These data support the use of the MAGGIC score to risk stratify black patients with HF.

Keywords: heart failure; linear models race; risk.

Publication types

  • Meta-Analysis
  • Research Support, N.I.H., Extramural
  • Validation Study

MeSH terms

  • Aged
  • Aged, 80 and over
  • Black or African American*
  • Chronic Disease
  • Decision Support Techniques*
  • Female
  • Health Status Disparities*
  • Health Status Indicators*
  • Healthcare Disparities
  • Heart Failure / diagnosis*
  • Heart Failure / ethnology
  • Heart Failure / mortality
  • Heart Failure / therapy
  • Humans
  • Male
  • Michigan / epidemiology
  • Middle Aged
  • Predictive Value of Tests
  • Prognosis
  • Reproducibility of Results
  • Retrospective Studies
  • Risk Assessment
  • Risk Factors
  • White People*