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.

A deep learning system for heart failure mortality prediction

Heart failure (HF) is the final stage of the various heart diseases developing. The mortality rates of prognosis HF patients are highly variable, ranging from 5% to 75%. Evaluating the all-cause mortality of HF patients is an important means to avoid death and positively affect the health of patients. But in fact, machine learning models are difficult to gain good results on missing values, high dimensions, and imbalances HF data. Therefore, a deep learning system is proposed. In this system, we propose an indicator vector to indicate whether the value is true or be padded, which fast solves the missing values and helps expand data dimensions. Then, we use a convolutional neural network with different kernel sizes to obtain the features information. And a multi-head self-attention mechanism is applied to gain whole channel information, which is essential for the system to improve performance. Besides, the focal loss function is introduced to deal with the imbalanced problem better. The experimental data of the system are from the public database MIMIC-III, containing valid data for 10311 patients. The proposed system effectively and fast predicts four death types: death within 30 days, death within 180 days, death within 365 days and death after 365 days. Our study uses Deep SHAP to interpret the deep learning model and obtains the top 15 characteristics. These characteristics further confirm the effectiveness and rationality of the system and help provide a better medical service.

The EPICTER score: a bedside and easy tool to predict mortality at 6 months in acute heart failure

Aims

Estimating the prognosis in heart failure (HF) is important to decide when to refer to palliative care (PC). Our objective was to develop a tool to identify the probability of death within 6 months in patients admitted with acute HF.

Methods and results

A total of 2848 patients admitted with HF in 74 Spanish hospitals were prospectively included and followed for 6 months. Each factor independently associated with death in the derivation cohort (60% of the sample) was assigned a prognostic weight, and a risk score was calculated. The accuracy of the score was verified in the validation cohort. The characteristics of the population were as follows: advanced age (mean 78 years), equal representation of men and women, significant comorbidity, and predominance of HF with preserved ejection fraction. During follow‐up, 753 patients (26%) died. Seven independent predictors of mortality were identified: age, chronic obstructive pulmonary disease, cognitive impairment, New York Heart Association class III–IV, chronic kidney disease, estimated survival of the patient less than 6 months, and acceptance of a palliative approach by the family or the patient. The area under the ROC curve for 6 month death was 0.74 for the derivation and 0.68 for the validation cohort. The model showed good calibration (Hosmer and Lemeshow test, P value 0.11). The 6 month death rates in the score groups ranged from 6% (low risk) to 54% (very high risk).

Conclusions

The EPICTER score, developed from a prospective and unselected cohort, is a bedside and easy‐to‐use tool that could help to identify high‐risk patients requiring PC.

Association Between Albumin-Bilirubin Score at Admission and In-Hospital Mortality in Patients with Acute Heart Failure

Liver dysfunction is one of the most recognized complications in patients with acute heart failure (HF) and therefore a liver function score may be useful for risk-stratification in those patients. Recently, the albumin-bilirubin (ALBI) score was developed as a new model to assess liver function in liver disease. We explored the association between the ALBI score at admission and in-hospital mortality in patients with acute HF.We enrolled 262 patients (median age, 86 years, 137 males) who were admitted to our hospital for treatment of acute HF. The following data were recorded: vital signs, laboratory data including B-type natriuretic peptide (BNP) level, echocardiographic data at admission, demographic and clinical characteristics, and treatment and prognostic information. The Get With the Guidelines-Heart Failure (GWTG-HF) risk score was calculated as an established risk model for each patient. The primary outcome was all-cause in-hospital mortality.During hospitalization, 37 patients (14.1%) died. The in-hospital mortality rate was significantly higher in patients with ALBI scores > -2.25 compared with patients with ALBI scores ≤ -2.25 (21.1% versus 4.5%, respectively; P = 0.0001). Multivariate analysis revealed that the GWTG-HF score (odds ratio [OR] 1.16, 95% confidence interval [CI] 1.08-1.25, P < 0.0001), BNP level (OR 1.0007, 95% CI 1.0003-1.001, P = 0.0003) and ALBI score (OR 6.0, 95% CI 1.8-19.6, P = 0.0017) were independently associated with in-hospital mortality.Our results indicated that the ALBI score was independently associated with in-hospital mortality in patients hospitalized for acute HF.