What is the clinical evidence from this study?

Study design: Observational. Population: Hospital readmission (n=25000). Intervention: Machine learning predictive models. Primary outcome: Prediction of hospital readmission.

What does this research mean for the field?

Machine learning models predict hospital readmission using routine clinical and utilization variables, with Random Forest achieving the highest F1-score (0.5428) and Logistic Regression achieving the highest ROC-AUC (0.6434). Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To evaluate hospital readmission rates and identify high-risk patients using supervised machine learning algorithms to improve healthcare delivery.

October 5, 2026EPH - International Journal of Medical and Health ScienceOpen Access

Analysis of Hospital Readmission Rates Using Predictive Analytics for Improved Healthcare Delivery

View Full Paper

Ask AI

Bookmark

Key result

Random Forest model predicts hospital readmission with an F1-score of ~0.54 using clinical utilization metrics.

n=25,000

Why the study?

Hospital readmission is a key measure of healthcare quality and safety, and predictive analytics may help identify patients at greater risk of readmission to improve healthcare delivery.

Population

25,000 patients from a structured hospital readmission dataset

Comparison

Predictive analytics using Logistic Regression, Decision Tree, and Random Forest models

Design

Observational study using machine learning models

Authors

SKSandeep KulkarniJain University

Discussion

Member takes

Implication

Moderate ML readmission prediction warrants clinical caution; leaves open need for prospective validation and refinement.

Key Points

To evaluate hospital readmission rates and identify high-risk patients using supervised machine learning algorithms to improve healthcare delivery.
Analyzed a structured hospital dataset of N=25,000 patients evaluating 17 clinical and utilization variables without missing data or duplicate records.
Trained and evaluated three supervised machine learning algorithms—Logistic Regression, Decision Tree, and Random Forest—using accuracy, precision, recall, F1-score, and ROC-AUC metrics.
A total of 11,754 patients (47.02%) were readmitted, while 13,246 patients (52.98%) were not readmitted.
Random Forest demonstrated the highest F1-score (0.5428) with an accuracy of 0.5991, precision of 0.5851, recall of 0.5062, and ROC-AUC of 0.6283; Logistic Regression achieved the highest accuracy (0.6060) and ROC-AUC (0.6434).
Feature importance analysis identified the number of lab procedures (importance score 0.1743), number of medications (0.1447), and time in hospital (0.0983) as the top three predictors of readmission.

Study Design

Type

Observational (n=25,000)

Structured PICO

Population

25,000 hospital patient records were analyzed to predict the risk of readmission using demographic, clinical, and healthcare utilization variables.

Exposure

Predictive analytics using machine learning models (Logistic Regression, Decision Tree, Random Forest)

Outcome

Hospital readmission

Machine learning models, particularly Random Forest, can predict hospital readmission using clinical and utilization variables, though performance is moderate and requires further refinement.

Limitations

Did not consider socio-economic factors such as income, education, occupation, family support, housing and follow-up care.
Dataset did not include hospital characteristics such as hospital size, staffing, discharge policies, treatment cost, insurance and access to follow-up care.
Dataset was cross-sectional and did not include longitudinal data.
Models had moderate rather than extremely high performance scores.
Did not consider socio-economic factors such as income, education, occupation, family support, housing and follow-up care
Dataset did not include hospital characteristics such as hospital size, staffing, discharge policies, treatment cost, insurance and access to follow-up care
Dataset was cross-sectional and did not include longitudinal data
Models had moderate rather than extremely high performance scores

Cite This Study

Sandeep Kulkarni (2025) conducted an observational in Hospital readmission (n=25,000). Machine learning predictive models was evaluated on Prediction of hospital readmission. A Random Forest machine learning model predicted hospital readmission with an F1-score of 0.5428, identifying the number of laboratory procedures, medications, and length of stay as the top predictors.

synapsesocial.com/papers/6ac2fd087180a622ddd4df3b https://doi.org/10.69980/g5f4jg42

View Full Paper

Ask AI

Bookmark

Analysis of Hospital Readmission Rates Using Predictive Analytics for Improved Healthcare Delivery

Discussion

Member takes

Implication

Key Points

Study Design

Structured PICO

Limitations

Cite This Study

Analysis of Hospital Readmission Rates Using Predictive Analytics for Improved Healthcare Delivery

Discussion

Member takes

Implication

Key Points

Study Design

Structured PICO

Limitations

Cite This Study

Also Consider

Also Consider