Identifying low-risk in patients with worsening heart failure with short hospital stay: A comparison of risk scores in predicting 30-day risk events

Abstract

Introduction

Heart failure (HF) is a leading cause of hospitalization worldwide, with high mortality rates and significant economic burden. To address the issue outpatient strategies (day-care diuretics) to avoid unplanned ED visits and reduce HF hospitalizations. However, the identification of low risk patients worsening heart failure (WHF) who could benefit from outpatient treatment remains poorly documented.

Objective

We aimed to evaluate the accuracy of multiple scores in predicting the risk of 30-day events in patients WHF who underwent brief hospitalizations.

Methods

We conducted a retrospective analysis of a prospective and consecutive cohort of WHF patients with hospitalizations of less than 72 h at a tertiary care hospital between 2015 and 2020. The risk of 30-day all-cause mortality was evaluated using the OPTIMIZE-HF, GWTG-HF, and ADHERE risk scores. And the secondary endpoint was the combined unplanned visit or readmission for worsening HF or death at 30 days. The risk of events in low-risk populations was analyzed by tertiles within the most predictive model.

Results

Among the 200 included patients (mean age: 75.5 ± 12 years; 62% male), 95.9% had a 30-day follow-up, with an overall mortality rate of 4% and a secondary composite endpoint of 14%. AUC-ROC for the prediction of 30-day all-cause mortality were 0.76 (95% CI 0.59-0.93), 0.66 (95% CI 0.46-0.86), and 0.64 (95% CI 0.44-0.85) for OPTIMIZE-HF, GWTG-HF, and ADHERE, respectively. For the secondary combined event, the AUC-ROC was 0.70 (95% CI 0.59-0.79) for OPTIMIZE-HF, GWTG-HF 0.67 (0.56-0.77) and ADHERE 0.67 (0.56 -0.77). The three scores had good calibration (Hosmer-Lemeshow goodness-of-fit test >0.05). Among the low-risk patients (n = 76, OPTIMIZE-HF score <32), the incidence of mortality and combined events at 30 days was 1.3% and 5.3%, respectively. Kaplan-Meier survival analysis showed that low risk patients had lower risk of the combined event (log rank p < 0.006).

Conclusion

Among WHF patients with short hospital stays, the OPTIMIZE-HF score exhibited superior predictive ability compared to other scores and may serve as a valuable tool for assessing the risk of death or combined events at 30 days. Utilizing the OPTIMIZE-HF risk score could aid in identifying low-risk patients who might benefit from outpatient management of AHF in a day-care diuretic clinic.