Risk stratification in heart failure decompensation in the community: HEFESTOS score

Drug Burden Index Is a Modifiable Predictor of 30-Day Hospitalization in Community-Dwelling Older Adults With Complex Care Needs: Machine Learning Analysis of InterRAI Data

BACKGROUND

Older adults (≥65 years) account for a disproportionately high proportion of hospitalization and in-hospital mortality, some of which may be avoidable. Although machine learning (ML) models have already been built and validated for predicting hospitalization and mortality, there remains a significant need to optimize ML models further. Accurately predicting hospitalization may tremendously affect the clinical care of older adults as preventative measures can be implemented to improve clinical outcomes for the patient.

METHODS

In this retrospective cohort study, a data set of 14 198 community-dwelling older adults (≥65 years) with complex care needs from the International Resident Assessment Instrument-Home Care database was used to develop and optimize 3 ML models to predict 30-day hospitalization. The models developed and optimized were Random Forest (RF), XGBoost (XGB), and Logistic Regression (LR). Variable importance plots were generated for all 3 models to identify key predictors of 30-day hospitalization.

RESULTS

The area under the receiver-operating characteristics curve for the RF, XGB, and LR models were 0.97, 0.90, and 0.72, respectively. Variable importance plots identified the Drug Burden Index and alcohol consumption as important, immediately potentially modifiable variables in predicting 30-day hospitalization.

CONCLUSIONS

Identifying immediately potentially modifiable risk factors such as the Drug Burden Index and alcohol consumption is of high clinical relevance. If clinicians can influence these variables, they could proactively lower the risk of 30-day hospitalization. ML holds promise to improve the clinical care of older adults. It is crucial that these models undergo extensive validation through large-scale clinical studies before being utilized in the clinical setting.

Prognostic Significance of Phenylalanine in Heart Failure: Clinical Insights and Inter-Organ Crosstalk Snapshot

Background: Heart failure (HF) remains a leading cause of morbidity and mortality globally, necessitating the identification of reliable prognostic biomarkers to guide therapeutic interventions. Recent clinical observations have underscored phenylalanine (PHE) as a prognostic marker in HF, although the mechanisms involving inter-organ crosstalk remain understood. Methods: This study adopted a dull approach, with a retrospective analysis of 550 HF patients to establish the prognostic value of pre-discharge PHE levels and a study on the inter-organ crosstalk of PHE among 24 patients. We analyzed the correlations between PHE concentrations and clinical outcomes, alongside a comprehensive examination of PHE metabolism across the skeletal muscle, liver, heart, kidney, and lung. Results: In the clinical prognostic analysis of 550 patients hospitalized for acute decompensated HF, elevated PHE levels (≥65.6 μM) were significantly and independently associated with increased all-cause mortality during a median follow-up of 4.5 years (log rank = 36.7, p < 0.001), underscoring its value as a prognostic marker in HF. The inter-organic crosstalk study elucidated the mechanism associated with PHE elevation in patients with HF, characterized by an increase in PHE output in skeletal muscle and a decrease in hepatic and cardiac PHE uptakes. Notably, PHE concentration gradients across these organs were correlated with HF severity, such as the NYHA functional class, B-type natriuretic peptide levels, and the presence of acute HF. Conclusions: Our findings confirm the prognostic significance of PHE in patients with HF and unveil the complex metabolic interplay among key organs that contribute to PHE dysregulation. These insights not only reinforce the importance of metabolic monitoring in HF management but also open avenues for therapeutic targets.

Application of machine learning approaches in predicting clinical outcomes in older adults - a systematic review and meta-analysis

BACKGROUND

Machine learning-based prediction models have the potential to have a considerable positive impact on geriatric care.

DESIGN

Systematic review and meta-analyses.

PARTICIPANTS

Older adults (≥ 65 years) in any setting.

INTERVENTION

Machine learning models for predicting clinical outcomes in older adults were evaluated. A random-effects meta-analysis was conducted in two grouped cohorts, where the predictive models were compared based on their performance in predicting mortality i) under and including 6 months ii) over 6 months.

OUTCOME MEASURES

Studies were grouped into two groups by the clinical outcome, and the models were compared based on the area under the receiver operating characteristic curve metric.

RESULTS

Thirty-seven studies that satisfied the systematic review criteria were appraised, and eight studies predicting a mortality outcome were included in the meta-analyses. We could only pool studies by mortality as there were inconsistent definitions and sparse data to pool studies for other clinical outcomes. The area under the receiver operating characteristic curve from the meta-analysis yielded a summary estimate of 0.80 (95% CI: 0.76 - 0.84) for mortality within 6 months and 0.81 (95% CI: 0.76 - 0.86) for mortality over 6 months, signifying good discriminatory power.

CONCLUSION

The meta-analysis indicates that machine learning models display good discriminatory power in predicting mortality. However, more large-scale validation studies are necessary. As electronic healthcare databases grow larger and more comprehensive, the available computational power increases and machine learning models become more sophisticated; there should be an effort to integrate these models into a larger research setting to predict various clinical outcomes.

Evaluation of the HEFESTOS scale to predict outcomes in emergency department acute heart failure patients

The HEFESTOS scale was developed in 14 Spanish primary care centres and validated in 9 primary care centres of other European countries. It showed good performance to predict death/hospitalisation during the first 30 days after an episode of acute heart failure (AHF), with c-statistics of 0.807/0.730 in the derivation/validation cohorts. We evaluated this scale in the emergency department (ED) setting, comparing it to the EHMRG and MEESSI scales in the ED and the EFFECT and GWTG scales in hospitalised patients, to predict 30-day outcomes, including death and hospitalisation. Consecutive AHF patients were enrolled in 34 Spanish EDs in January-February 2016, 2018, and 2019 with variables needed to calculate outcome scores. Thirty-day hospitalisation/death (together and separately) and post-discharge combined adverse event (ED revisit or hospitalisation for AHF or all-cause death) were determined for patients discharged home after ED care. Predictive capacity was assessed by c-statistic with 95% confidence intervals. Of 10,869 patients, 4,044 were included (median age: 83 years, 54% women). The performance of HEFESTOS was modest for 30-day hospitalisation/death, c-statistic=0.656 (0.637-0.675), hospitalisation, 0.650 (0.631-0.669), and death, 0.610 (0.576-0.644). Of 1,034 patients with scores for the 5 scales, HEFESTOS had the numerically highest c-statistic for hospitalisation/death at 30 days, 0.666 (0.627-0.704), vs. MEESSI= 0.650 (0.612-0.687, p=0.51), EFFECT=0.633 (0.595-0.672, p=0.21), GWTG=0.618 (0.578-0.657, p=0.06) and EHMRG=0.617 (0.577-0.704, p=0.07). Similar modest performances were observed for predicting hospitalisation [ranging from HEFESTOS=0.656 (0.618-0.695) to GWTG=0.603 (0.564-0.643)]. Conversely, prediction of 30-day death was good with the MEESSI=0.787 (0.728-845), EFFECT=0.754 (0.691-0.818) and GWTG=0.749 (0.689-0.809) scales, and modest with EHMRG=0.649 (0.581-0.717) and HEFESTOS=0.610 (0.538-0.683). Although the HEFESTOS scale was numerically better for predicting 30-day hospitalisation/death in ED AHF patients, its modest performance precludes routine use. Only 30-day mortality was adequately predicted by some scales, with the MEESSI achieving the best results.