First African validation study of HARMS2-AF and C2HEST scores to predict the onset of atrial fibrillation in patients with acute coronary syndrome treated by PCI: A report from the Moroccan data

Abstract

Introduction

New-onset atrial fibrillation (NOAF) during acute coronary syndrome (ACS) represents a complex association given the high thrombo-embolic risk.

Objective

We analyzed the validity of the HARMS2-AF and C2HEST score in predicting NOAF in patients with ACS treated by PCI in an African population using data from a Moroccan population.

Method

We examined the medical data of patients admitted to our CICU over a 4-year period. We excluded patients with AF prior to admission. Our primary outcome was an episode of NOAF in-hospital. Discrimination for NOAF was assessed using area under the receiver characteristic curve (AUC) values, and a comparison between the two scores was made using the De-long test. We also performed a multivariable logistic regression analysis to predict NOAF using the C2HEST score for a cut-off at 3 and the HARMS2-AF score for a cut-off at 7.

Results

We included 763 patients, 76.3% men, with 75% of patients admitted for STEMI. NOAF was diagnosed in 8.4% with 93% within the first 48 h after admission. Patients with NOAF were older (70.27 ± 11.64 vs. 64.24 ± 11.55, p < 0.001), with a longer length of hospital stay (5.7 ± 3.86 vs. 3.8 ± 2.05, p < 0.001), a lower ejection fraction (40, 96 ± 12.25 vs. 46.6 ± 12.06, p < 0.001), a higher C2HEST score (2.45 ± 1.79 vs. 2.27 ± 1.43, p = 0.003), and a higher HARMS2-AF score (9.22 ± 3.50 vs. 6.72 ± 2.94, p < 0.001). In multivariate logistic regression analysis, a C2HEST score > 3 was independently associated with the occurrence of NOAF (odds ratio 3.54; 95% CI 2.07–6.07, p < 0.001), as was a HARMS2-AF score > 7 (odds ratio 5.56; 95% CI 2.91–10.61, p < 0.001). ROC curve analysis showed good accuracy of the C2HEST score (AUC 0.69, 95% CI 0.665; 0.731; p < 0.001) and HARMS2-AF score (AUC 0.72, 95% CI 0.689; 0.754; p < 0.001) in predicting NOAF, with no significant difference between the two scores (De long test p value = 0.326) (Fig. 1)

Conclusion

As demonstrated by our results, HARMS2-AF and C2HEST scores can predict NOAF after ACS treated with PCI, which will allow selection of patients with a higher risk of stroke.