This study aimed to develop and validate deep-learning-based artificial intelligence algorithm for predicting mortality of AHF (DAHF). 12,654 dataset from 2165 patients with AHF in two hospitals were used as train data for DAHF development, and 4759 dataset from 4759 patients with AHF in 10 hospitals enrolled to the Korean AHF registry were used as performance test data. The endpoints were in-hospital, 12-month, and 36-month mortality. We compared the DAHF performance with the Get with the Guidelines–Heart Failure (GWTG-HF) score, Meta-Analysis Global Group in Chronic Heart Failure (MAGGIC) score, and other machine-learning models by using the test data. Area under the receiver operating characteristic curve of the DAHF were 0.880 (95% confidence interval, 0.876–0.884) for predicting in-hospital mortality; these results significantly outperformed those of the GWTG-HF (0.728 [0.720–0.737]) and other machine-learning models. For predicting 12- and 36-month endpoints, DAHF (0.782 and 0.813) significantly outperformed MAGGIC score (0.718 and 0.729). During the 36-month follow-up, the high-risk group, defined by the DAHF, had a significantly higher mortality rate than the low-risk group(p<0.001). DAHF predicted the in-hospital and long-term mortality of patients with AHF more accurately than the existing risk scores and other machine-learning models.

Park JJ, et al. Int J Heart Fail. 2021 Oct;3(4):224-236. https://doi.org/10.36628/ijhf.2021.0023

The application of a simple blood test to predict prognosis in acute heart failure (AHF) patients is not well established. Neutrophil-lymphocyte ratio (NLR) is inexpensive and easy to obtain in hospitalized patients using a routine blood test. We evaluate the prognostic implications of NLR as an independent predictor of in-hospital and long-term mortality in AHF patients. Among 5625 patients enrolled in the Korean Acute Heart Failure registry, 5580 patients were classified into quartiles by their NLR level, and analyzed for in-hospital and post-discharge three-year mortality. Patients in the highest NLR quartile had the highest in-hospital and post-discharge three-year mortality. The same results were seen by dividing the aggravating factor into the infection or ischemia group and the non-infection or non-ischemia group. For patients aggravated from infection or ischemia, a cut-off NLR value was 7.0 that increase the risk of in-hospital and post-discharge three-year mortality. In subgroups of patients not aggravated from infection or ischemia, a cut-off NLR value was 5.0 that increase the risk of in-hospital and post discharge three-year mortality. Elevated NLR in AHF patients at the index hospitalization is an independent predictor for in-hospital and post-discharge three-year mortality. Taken together, NLR is a marker for risk assessment of AHF patients.

Kim MS, et al. Korean Circ J. 2017 Sep;47(5):555-643. https://doi.org/10.4070/kcj.2017.0009

In clinical practice, a risk prediction model is an effective solitary program to predict prognosis in particular patient groups. B-type natriuretic peptide (BNP)and N-terminal pro-b-type natriuretic peptide (NT-proBNP) are widely recognized outcome-predicting factors for patients with heart failure (HF).This study derived external validation of a risk score to predict 1-year mortality after discharge in hospitalized patients with HF using the Meta-analysis Global Group in Chronic Heart Failure (MAGGIC)program data. We also assessed the effect of adding BNP or NT-proBNP to this risk score model in a Korean HF registry population. We included 5625 patients from the Korean acute heart failure registry (KorAHF) and excluded those who died in hospital. The MAGGIC constructed a risk score to predict mortality in patients with HF by using 13 routinely available patient characteristics (age, gender, diabetes, chronic obstructive pulmonary disorder (COPD), HF diagnosed within the last 18 months, current smoker, NYHA class, use of beta blocker, ACEI or ARB, body mass index, systolic blood pressure, creatinine, and EF). We added BNP or NT-proBNP, which are the most important biomarkers, to the MAGGIC risk scoring system in patients with HF. The outcome measure was 1-year mortality. In multivariable analysis, BNP or NT-proBNP independently predicted death. The risk score was significantly varied between alive and dead groups (30.61 ± 6.32 vs. 24.80 ± 6.81, p < 0.001). After the conjoint use of BNP or NT-proBNP and MAGGIC risk score in patients with HF, a significant difference in risk score was noted (31.23 ± 6.46 vs. 25.25 ± 6.96, p < 0.001).The discrimination abilities of the risk score model with and without biomarker showed minimal improvement (C index of 0.734 for MAGGIC risk score and 0.736 for MAGGIC risk score plus BNP or NT-proBNP, p = 0.0502) and the calibration was found good. However, we achieved a significant improvement in net reclassification and integrated discrimination for mortality (NRI of 33.4%,p < 0.0001 and IDI of 0.002, p < 0.0001). In the KorAHF, the MAGGIC project HF risk score performed well in a large nationwide contemporary external validation cohort. Furthermore, the addition of BNP or NT-proBNPto the MAGGIC risk score was beneficial in predicting more death in hospitalized patients with HF.

Youn JC, et al. Int J Heart Fail. 2023 Apr;5:e9. https://doi.org/10.36628/ijhf.2023.0011