Comparative validation of the HAS-BLED and the GARFIELD-AF bleeding risk scores in Asian patients with atrial fibrillation: a report from the COOL-AF registry

Autor: P Chichareon, A Winijkul, G Lip, R Krittayaphong
Rok vydání: 2022
Předmět:
Zdroj: European Heart Journal. 43
ISSN: 1522-9645
0195-668X
DOI: 10.1093/eurheartj/ehac544.625
Popis: Introduction There have been conflicting comparative data on the predictive value of the HAS-BLED and the GARFIELD-AF bleeding risk scores. Furthermore, there are limited data in Asian patients with atrial fibrillation (AF) who are at higher risk of bleeding compared to non-Asians. We compared the performance of the HAS-BLED score with the GARFIELD-AF score in predicting major bleeding in AF patients treated with OAC in the Cohort of antithrombotic use and Optimal INR Level in patients with non-valvular AF in Thailand (COOL-AF) registry. Since the sex and body weight seem to impact bleeding events in Asians, we explored an update of the HAS-BLED risk score by incorporating sex and body weight. Methods The COOL-AF registry was a nationwide study enrolling AF patients from 27 hospitals in Thailand between 2014 and 2017. The registry aimed to assess antithrombotic patterns, quality of oral anticoagulant (OAC) control and clinical outcomes. Data were collected every 6 months and until 3 years. AF patients treated with oral anticoagulants (OAC, either warfarin or non-vitamin K antagonist OAC (NOAC)) were included into the present study. We fitted the variables of the HA-SBLED and GARFILED-AF score to major bleeding in Cox regression model. We also explored a modified HAS-BLED by addition of two additional predictors (sex and body weight). Discrimination, calibration, net reclassification index and decision curve analysis were used to compare the predictive performance of the three models. Results Of 3,402 patients in the COOL-AF registry, 2,568 patients (mean age 68.4 years; 43.4% female) who received OAC at baseline were studied. The proportion of patients with prior bleeding was 10.6%. Majority of patients (68.7%) received warfarin. The incidence rate of major bleeding was 2.11 (1.79–2.48) per 100 person-years. The C-indexes of the GARFIELD-AF, HASBLED and modified HASBLED score were 0.65 (95% CI 0.63–0.67), 0.66 (95% CI 0.64–0.68) and 0.69 (95% CI 0.67–0.71) respectively. Calibration plot showed good agreement between predicted and observed bleeding in the deciles of original HAS-BLED and GARFIELD-AF scores, while the modified HAS-BLED score overestimated the risk in the last decile (Figure 1). The modified HASBLED score has superior NRI than the original HASBLED score (26.9%, 95% CI 9.7% to 42.2%). The NRI between the modified HASBLED and GARFIELD score was similar (20%, 95% CI −1.2% to 40.4%). The net benefit curve of the three models were overlapping among different risk thresholds (Figure 2). Conclusion The clinical utility for bleeding prediction GARFIELD-AF, HAS-BLED and modified HAS-BLED scores were similar in Asian patients with AF participating in the COOL-AF registry. Funding Acknowledgement Type of funding sources: Public Institution(s). Main funding source(s): This study was funded by the Health Systems Research Institute (HSRI) (grant no. 59-053), the HeartAssociation of Thailand under the Royal Patronage of H.M. the King.
Databáze: OpenAIRE
Externí odkaz: