Establishment of a Predictive Model for Inpatient Sudden Cardiac Death in a Chinese Cardiac Department Population: A Retrospective Study
Sudden cardiac death (SCD) is a critical global health issue, defined as an unexpected natural death due to cardiac causes, typically occurring within one hour of symptom onset. This condition is a significant public health concern worldwide, including in China, where the annual incidence rate is approximately 41.8 per 100,000 people. SCD accounts for a substantial proportion of global mortality, contributing to around 3.7 million deaths annually, or 15% to 20% of all deaths. Despite its prevalence, the risk factors for SCD in hospitalized cardiac department populations remain poorly understood, particularly in Asia. This study aimed to identify these risk factors and develop a predictive model for in-hospital SCD using conventional and low-cost clinical data.
The study was conducted as a retrospective analysis of patients admitted to the cardiac department of the First Affiliated Hospital of Xinjiang Medical University from June 2015 to February 2017. The research team collected clinical data from medical records, focusing on demographic characteristics, lifestyle factors, medical history, physical examinations, 12-lead electrocardiograms (ECGs), 24-hour Holter monitoring, two-dimensional echocardiography, and blood laboratory tests. The study included 262 patients who experienced in-hospital SCD and 4,485 control patients who did not. Exclusion criteria included incomplete clinical data, implantation of an implantable cardioverter-defibrillator (ICD) or cardiac resynchronization therapy-defibrillator, and severe systemic organ diseases such as infectious diseases or malignant tumors.
The primary outcome of the study was in-hospital SCD, defined as death within one hour of symptom onset. Symptoms included sudden loss of consciousness, a rapid drop in blood pressure below 90/60 mmHg, sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) observed on ECG monitoring, cardiac arrest, and escape rhythms. The cause of death was verified by a team of experienced cardiologists to ensure accuracy.
The study identified several significant risk factors for in-hospital SCD through multiple stepwise logistic regression analysis. These factors included age, main admitting diagnosis, diabetes mellitus (DM), corrected QT interval (QTc), QRS duration, ventricular premature beat (VPB) burden, left ventricular ejection fraction (LVEF), and estimated glomerular filtration rate (eGFR). Each of these factors was assigned a score based on its contribution to the risk of SCD, with the total risk score ranging from 0 to 11 points.
Age was a significant predictor, with patients aged 65 years or older receiving one point. The main admitting diagnosis also played a crucial role, with heart failure exacerbation, post-myocardial infarction (MI) or unstable angina, and acute MI within 30 days of admission receiving two, one, and three points, respectively. Diabetes mellitus added one point to the risk score, while a QRS duration greater than 150 milliseconds and a QTc interval exceeding 450 milliseconds for men or 460 milliseconds for women each contributed one point. LVEF values between 25% and 39% added one point, while values below 25% added two points. A VPB burden greater than 20% and an eGFR below 40 milliliters per minute per 1.73 square meters each added one point to the risk score.
The predictive model demonstrated good discrimination and calibration, with an area under the receiver operating characteristic (AUROC) curve of 0.774, indicating a robust ability to distinguish between patients at high and low risk of SCD. The Hosmer-Lemeshow goodness-of-fit test yielded a chi-square value of 2.527 with a p-value of 0.640, suggesting that the model fit the data well. The incidence of in-hospital SCD increased significantly with higher risk scores, with rates of 1.3%, 4.1%, and 18.6% for scores of 0 to 2, 3 to 5, and 6 or more, respectively.
The study highlighted the importance of several key factors in predicting in-hospital SCD. Coronary heart disease (CHD) was the most common underlying condition among SCD cases, accounting for 68.7% of the cohort. Patients with a history of CHD, particularly those who had experienced an acute MI within the previous 30 days, were at the highest risk. LVEF, a measure of cardiac systolic function, was also a critical predictor, with lower values indicating a higher risk of SCD. Abnormal cardiac electrical activity, as reflected by prolonged QTc intervals, QRS durations, and high VPB burdens, further increased the risk of SCD.
Diabetes mellitus and reduced eGFR were additional significant predictors. DM contributes to cardiac ischemia, myocardial damage, and scar formation, increasing the risk of SCD. Impaired kidney function, as indicated by a reduced eGFR, was also associated with a higher risk of SCD, consistent with findings from previous studies.
The study’s findings have important clinical implications. The predictive model provides a practical, non-invasive, and low-cost tool for identifying hospitalized patients at high risk of SCD. This model can be particularly valuable in primary medical institutions, where access to advanced diagnostic tools may be limited. By stratifying patients into low, intermediate, and high-risk groups, clinicians can tailor their treatment and monitoring strategies to reduce the likelihood of SCD. For example, high-risk patients may benefit from more intensive monitoring or consideration of ICD therapy, even if they do not meet the standard criteria for such interventions.
The study also addressed several limitations inherent in its retrospective design. The lack of autopsy data for deceased patients and the single-center nature of the study cohort may limit the generalizability of the findings. Additionally, the study relied on a single measurement of clinical data, which may not capture changes over time. Future research should aim to validate the predictive model in a multi-center, prospective study to confirm its applicability across different populations.
In conclusion, this study established a predictive risk score for in-hospital SCD in a Chinese cardiac department population. The model incorporates eight key risk factors: age, main admitting diagnosis, diabetes mellitus, QTc interval, QRS duration, VPB burden, LVEF, and eGFR. By providing a simple and effective tool for risk stratification, this model can help clinicians identify and manage high-risk patients, ultimately reducing the incidence of in-hospital SCD. The findings underscore the importance of comprehensive risk assessment and tailored treatment strategies in improving patient outcomes.
doi.org/10.1097/CM9.0000000000000010
Was this helpful?
0 / 0