Machine learning prognosis model based on patient-reported outcomes for chronic heart failure patients after discharge.

BACKGROUND: Patient-reported outcomes (PROs) can be obtained outside hospitals and are of great significance for evaluation of patients with chronic heart failure (CHF). The aim of this study was to establish a prediction model using PROs for out-of-hospital patients. METHODS: CHF-PRO were collected in 941 patients with CHF from a prospective cohort. Primary endpoints were all-cause mortality, HF hospitalization, and major adverse cardiovascular events (MACEs). To establish prognosis models during the two years follow-up, six machine learning methods were used, including logistic regression, random forest classifier, extreme gradient boosting (XGBoost), light gradient boosting machine, naive bayes, and multilayer perceptron. Models were established in four steps, namely, using general information as predictors, using four domains of CHF-PRO, using both of them and adjusting the parameters. The discrimination and calibration were then estimated. Further analyze were performed for the best model. The top prediction variables were further assessed. The Shapley additive explanations (SHAP) method was used to explain black boxes of the models. Moreover, a self-made web-based risk calculator was established to facilitate the clinical application. RESULTS: CHF-PRO showed strong prediction value and improved the performance of the models. Among the approaches, XGBoost of the parameter adjustment model had the highest prediction performance with an area under the curve of 0.754 (95% CI: 0.737 to 0.761) for death, 0.718 (95% CI: 0.717 to 0.721) for HF rehospitalization and 0.670 (95% CI: 0.595 to 0.710) for MACEs. The four domains of CHF-PRO, especially the physical domain, showed the most significant impact on the prediction of outcomes. CONCLUSION: CHF-PRO showed strong prediction value in the models. The XGBoost models using variables based on CHF-PRO and the patient's general information provide prognostic assessment for patients with CHF. The self-made web-based risk calculator can be conveniently used to predict the prognosis for patients after discharge. CLINICAL TRIAL REGISTRATION: URL: http://www.chictr.org.cn/index.aspx ; Unique identifier: ChiCTR2100043337.

Identification of echocardiographic subgroups in patients with coronary heart disease combined with heart failure based on latent variable stratification.

BACKGROUND: The prognosis of chronic heart failure is poor, and it remains a challenge to classify patients for better personalized intervention. This study aimed to explore potential subgroups in patients with coronary heart disease and chronic heart failure using comprehensive echocardiographic indices. METHODS: 5126 patients with coronary heart disease with chronic heart failure were included. Latent class analysis was applied to identify the grouping patterns of patients based on echocardiographic indices. Network maps and radar charts of echocardiographic indices were drawn to visualize the distribution of echocardiographic findings. The incidence of adverse outcomes was presented on the Kaplan-Meier curve and compared using the log-rank test. The Cox regression model was used to analyze the relationship between subgroups and mortality. RESULTS: Three groups were identified: eccentric hypertrophy, concentric hypertrophy, and decreased diastolic function. Network plots showed a higher correlation between left atrial diameter, left ventricular mass index, and left ventricle ejection fraction in the eccentric hypertrophy group than in the other groups. The Kaplan-Meier curve showed a significant difference in mortality between the three subgroups (P < 0.001). Multivariate Cox analysis indicated that the eccentric hypertrophy group had the highest risk of death (HR = 1.586, 95% CI: 1.310-1.921, P < 0.001) compared with the other groups. CONCLUSION: Patients with coronary heart disease and chronic heart failure can be classified into three subgroups based on echocardiographic indices. This grouping has been shown to be an independent risk factor for mortality in these patients. Accurate subgrouping based on echocardiographic indices is important for identifying high-risk patients.

Social and therapeutic decline earlier than physical and psychological domains after discharge in heart failure patients: A patient-reported outcome measurements of latent transition analysis.

Background: Among patients with chronic heart failure (CHF), response shifts are common in assessing treatment effects. However, few studies focused on potential response shifts in these patients. Materials and methods: Data of CHF patient-reported outcome measures (PROMs) were obtained from three hospitals in Shanxi, China, from 2017 to 2019. A total of 497 patients were enrolled and followed up at 1 month and 6 months after discharge. Latent transition analysis (LTA) was employed to determine the longitudinal transition trajectories of latent subtypes in CHF patients in the physiological, psychological, social, and therapeutic domains. Results: The patients were divided into high- and low-level groups in the four domains according to the LTA. One month after discharge, the physiological and psychological domains improved, while the social and therapeutic domains remained unchanged. Six months after discharge, the former remained stable, but the latter deteriorated. The factors affecting the state transition in four domains were as follows. The influencing factor of the physiological domains are gender, age, tea consumption, smoking, alcohol consumption, physical activity, and light diet; those of the psychological domain are gender, occupation, smoking, alcohol consumption, and physical activity; those of the social domains are age; those of the therapeutic domains are education and income. Conclusion: The disease status of CHF patients has shifted over time. Risk factors accelerate the deterioration of patients' condition. Furthermore, the risk factors of social and therapeutic domains deteriorate patients' condition faster than those of physiological and psychological domains. Therefore, individualized intervention programs should be given for CHF patients who may be transferred to the low-level groups to maintain the treatment effect and improve the prognosis.

A clinical decision support system for predicting coronary artery stenosis in patients with suspected coronary heart disease.

Invasive coronary angiography imposes risks and high medical costs. Therefore, accurate, reliable, non-invasive, and cost-effective methods for diagnosing coronary stenosis are required. We designed a machine learning-based risk-prediction system as an accurate, noninvasive, and cost-effective alternative method for evaluating suspected coronary heart disease (CHD) patients. Electronic medical record data were collected from suspected CHD patients undergoing coronary angiography between May 1, 2017, and December 31, 2019. Multi-Class XGBoost, LightGBM, Random Forest, NGBoost, logistic models and MLP were constructed to identify patients with normal coronary arteries (class 0: no coronary artery stenosis), minimum coronary artery stenosis (class 1: 0 < stenosis <50%), and CHD (class 2: stenosis ≥50%). Model stability was verified externally. A risk-assessment and management system was established for patient-specific intervention guidance. Of 1577 suspected CHD patients, 81 (5.14%) had normal coronary arteries. The XGBoost model demonstrated the best overall classification performance (micro-average receiver operating characteristic [ROC] curve: 0.92, macro-average ROC curve: 0.89, class 0 ROC curve: 0.88, class 1 ROC curve: 0.90, class 2 ROC curve: 0.89), with good external verification. In class-specific classification, the XGBoost model yielded F1 values of 0.636, 0.850, and 0.858, for Classes 0, 1, and 2, respectively. The visualization system allowed disease diagnosis and probability estimation, and identified the intervention focus for individual patients. Thus, the system distinguished coronary artery stenosis well in suspected CHD patients. Personalized probability curves provide individualized intervention guidance. This may reduce the number of invasive inspections in negative patients, while facilitating decision-making regarding appropriate medical intervention, improving patient prognosis.

The Causal Effects of Anxiety-Mediated Social Support on Death in Patients with Chronic Heart Failure: A Multicenter Cohort Study.

Background: Chronic heart failure (CHF) affects more than 3.8 million people worldwide. There is a paucity of studies focusing on psychosocial issues in CHF patients. This study aimed to investigate the association of social support, mental adjustment and death to exploring whether mental adjustment could mediate the relationship. Methods: From May 2017 to June 2021, we conducted a multicenter clinical study to collect 1552 patients data. The Patient Report Outcome (PRO) scale were disseminated to collect information in the physical, psychological, social and therapeutic domains of patients. Marginal structural model was used to investigate the association of social support and CHF death, and the role of mental adjustment in their mediation. Results: The direct effect of social support accounted for 44.76% of the total effect. High social support (≥14 points) reduced mortality by 46.3% (RR=0.537, P=0.027), medium social support (11-14 points) reduced mortality by 45.3% (RR=0.547, P=0.042). Anxiety (effect percentage: 24.63%) and appetite-sleep quality (effect percentage: 30.61%) played a mediating role between social support and death in CHF patients. In women, aged >75 years, divorced or widowed patients were most prone to anxiety due to inadequate support (RR=0.519, P=0.019; RR=0.403, P=0.002; RR=0.413, P=0.041). Family care and socioeconomic assistance significantly reduced the risk of death (RR=0.689, P=0.040; RR=0.584, P=0.012). Conclusion: Social support can reduce patient mortality, especially family care and social economic assistance. The elderly, female, divorced or widowed patients are more likely to cause mental illness due to inadequate social support.

Dynamic Trajectory of a Patient-Reported Outcome and Its Associated Factors for Patients with Chronic Heart Failure: A Growth Mixture Model Approach.

Purpose: This study aimed to identify subgroups of chronic heart failure (CHF) patients with distinct trajectories of quality of life (QOL) and to identify baseline characteristics associated with the trajectories. Patients and methods: Two-year, prospective, cohort study including 315 patients with CHF was conducted from July 2017. Information on QOL assessed by CHF-patient-reported outcomes measure (CHF-PROM) was collected at baseline, 6, 12, 18, and 24 months. Demographic and clinical variables were recorded at baseline. Growth mixture model was used to identify distinct trajectories of CHF-PROM and its physical, psychological, social, and therapeutic domains. Single factor analysis was employed to assess the factors associated with development of CHF-PROM over time. Results: Two classes of overall score of CHF-PROM were identified: poorer (14.0%) and better (86.0%). Poorer class tended to be aged, have low diastolic blood pressure, have concomitant atrial fibrillation, diabetes, chronic obstructive pulmonary disease, cancers, and central nervous system diseases, and used nitrates. Three classes of physical scores were identified: unstable-poorer (5.2%), stable-poorer (29.4%) and better (65.4%). Age, NYHA grade, chronic obstructive pulmonary disease, combined with cancers and central nervous system diseases were related to the grouping. Poorer (8.6%) and better (91.4%) classes of psychological scores were identified. Poorer class tended to be female and had concomitant atrial fibrillation. Degenerate class (34.6%) and meliorate class (65.4%) of therapeutic scores were identified. Degenerate class tended to have concomitant chronic obstructive pulmonary disease and use less angiotensin converting enzyme inhibitors. Conclusion: We identified different classes with distinct trajectories of QOL that may help proper evaluate QOL and further improve its status for patients CHF.

Hierarchical management of chronic heart failure: a perspective based on the latent structure of comorbidities.

AIMS: Chronic heart failure (CHF) has an increasing burden of comorbidities, which affect clinical outcomes. Few studies have focused on the clustering and hierarchical management of patients with CHF based on comorbidity. This study aimed to explore the cluster model of CHF patients based on comorbidities and to verify their relationship with clinical outcomes. METHODS AND RESULTS: Electronic health records of patients hospitalized with CHF from January 2014 to April 2019 were collected, and 12 common comorbidities were included in the latent class analysis. The Fruchterman-Reingold layout was used to draw the comorbidity network, and analysis of variance was used to compare the weighted degrees among them. The incidence of clinical outcomes among different clusters was presented on Kaplan-Meier curves and compared using the log-rank test, and the hazard ratio was calculated using the Cox proportional risk model. Sensitivity analysis was performed according to the left ventricular ejection fraction. Four different clinical clusters from 4063 total patients were identified: metabolic, ischaemic, high comorbidity burden, and elderly-atrial fibrillation. Compared with the metabolic cluster, patients in the high comorbidity burden cluster had the highest adjusted risk of combined outcome and all-cause mortality {1.67 [95% confidence interval (CI), 1.40-1.99] and 2.87 [95% CI, 2.17-3.81], respectively}, followed by the elderly-atrial fibrillation and ischaemic clusters. The adjusted readmission risk of patients with ischaemic, high comorbidity burden, and elderly-atrial fibrillation clusters were 1.35 (95% CI, 1.08-1.68), 1.39 (95% CI, 1.13-1.72), and 1.42 (95% CI, 1.14-1.77), respectively. The comorbidity network analysis found that patients in the high comorbidity burden cluster had more and higher comorbidity correlations than those in other clusters. Sensitivity analysis revealed that patients in the high comorbidity burden cluster had the highest risk of combined outcome and all-cause mortality (P < 0.05). CONCLUSIONS: The difference in adverse outcomes among clusters confirmed the heterogeneity of CHF and the importance of hierarchical management. This study can provide a basis for personalized treatment and management of patients with CHF, and provide a new perspective for clinical decision making.