Risk prediction of heart failure in patients with ischemic heart disease using network analytics and stacking ensemble learning.

BACKGROUND: Heart failure (HF) is a major complication following ischemic heart disease (IHD) and it adversely affects the outcome. Early prediction of HF risk in patients with IHD is beneficial for timely intervention and for reducing disease burden. METHODS: Two cohorts, cases for patients first diagnosed with IHD and then with HF (N = 11,862) and control IHD patients without HF (N = 25,652), were established from the hospital discharge records in Sichuan, China during 2015-2019. Directed personal disease network (PDN) was constructed for each patient, and then these PDNs were merged to generate the baseline disease network (BDN) for the two cohorts, respectively, which identifies the health trajectories of patients and the complex progression patterns. The differences between the BDNs of the two cohort was represented as disease-specific network (DSN). Three novel network features were exacted from PDN and DSN to represent the similarity of disease patterns and specificity trends from IHD to HF. A stacking-based ensemble model DXLR was proposed to predict HF risk in IHD patients using the novel network features and basic demographic features (i.e., age and sex). The Shapley Addictive exPlanations method was applied to analyze the feature importance of the DXLR model. RESULTS: Compared with the six traditional machine learning models, our DXLR model exhibited the highest AUC (0.934 ± 0.004), accuracy (0.857 ± 0.007), precision (0.723 ± 0.014), recall (0.892 ± 0.012) and F1 score (0.798 ± 0.010). The feature importance showed that the novel network features ranked as the top three features, playing a notable role in predicting HF risk of IHD patient. The feature comparison experiment also indicated that our novel network features were superior to those proposed by the state-of-the-art study in improving the performance of the prediction model, with an increase in AUC by 19.9%, in accuracy by 18.7%, in precision by 30.7%, in recall by 37.4%, and in F1 score by 33.7%. CONCLUSIONS: Our proposed approach that combines network analytics and ensemble learning effectively predicts HF risk in patients with IHD. This highlights the potential value of network-based machine learning in disease risk prediction field using administrative data.

Gaseous air pollutants and hospitalizations for mental disorders in 17 Chinese cities: Association, morbidity burden and economic costs.

The short-term morbidity effects of gaseous air pollutants on mental disorders (MDs), and the corresponding morbidity and economic burdens have not been well studied. We aimed to explore the associations of ambient sulfur dioxide (SO2), nitrogen dioxide (NO2), ozone (O3) and carbon monoxide (CO) with MDs hospitalizations in 17 Chinese cities during 2015-2018, and estimate the attributable risk and economic costs of MDs hospitalizations associated with gaseous pollutants. City-specific relationships between gaseous pollutants and MDs hospitalizations were evaluated using over-dispersed generalized additive models, then combined to obtain the pooled effect. Concentration-response (C-R) curves of gaseous pollutants with MDs from each city were pooled to allow regional estimates to be derived. The morbidity and economic burdens of MDs hospitalizations attributable to gaseous pollutants were further assessed. A total of 171,939 MDs hospitalizations were included. We observed insignificant association of O3 with MDs. An interquartile range increase in SO2 at lag0 (9.1 µg/m³), NO2 at lag0 (16.7 µg/m³) and CO at lag2 (0.4 mg/m³) corresponded to a 3.02% (95%CI: 0.72%, 5.38%), 5.03% (95%CI: 1.84%, 8.32%) and 2.18% (95%CI: 0.40%, 4.00%) increase in daily MDs hospitalizations, respectively. These effects were modified by sex, season and cause-specific MDs. The C-R curves of SO2 and NO2 with MDs indicated nonlinearity and the slops were steeper at lower concentrations. Overall, using current standards as reference concentrations, 0.27% (95%CI: 0.07%, 0.48%) and 0.06% (95%CI: 0.02%, 0.10%) of MDs hospitalizations could be attributable to extra SO2 and NO2 exposures, and the corresponding economic costs accounted for 0.34% (95%CI: 0.08%, 0.60%) and 0.07% (95%CI: 0.03%, 0.11%) of hospitalization expenses, respectively. Moreover, using threshold values detected from C-R curves as reference concentrations, the above mentioned morbidity and economic burdens increased a lot. These findings suggest more strict emission control regulations are needed to protect mental health from gaseous pollutants.

Limited Access to Healthcare among Hispanics in the US-Mexico Border Region.

OBJECTIVE: Using large national databases, we investigated how living in the US-Mexico border region further limited access to healthcare among the non-elderly Hispanic adult population after controlling individual and county-level characteristics. METHODS: The 2008-2012 individual-level data of non-elderly Hispanic adults from the Behavioral Risk Factor Surveillance System (BRFSS) were merged with county-level data from Area Health Resources File (AHRF). Multivariate logistic analyses were performed to predict insurance status and access to doctors using residency in the US-Mexico border region as the key predictor, adjusting individual and county-level factors. RESULTS: Controlling only individual characteristic, Hispanics living in the US-Mexico border region had significantly lower odds of having health insurance (adjusted odds ratio [AOR] = 0.51; 95% confidence interval [CI], 0.49-0.54) and access to doctors (AOR = 0.69; 95% CI, 0.66-0.72). After including county-level measurements of healthcare system capacity and other local characteristics, the border region continued to be associated with lower likelihood of healthcare access. CONCLUSION: Hispanic residents in the U.S.-Mexico border had less access to healthcare than their inland counterparts. The findings highlight unique features in this region and support policies and initiatives to improve minority healthcare access, particularly among disadvantaged populations in this region.