4.7 Article

Thirty-day hospital readmission prediction model based on common data model with weather and air quality data

Journal

SCIENTIFIC REPORTS
Volume 11, Issue 1, Pages -

Publisher

NATURE PORTFOLIO
DOI: 10.1038/s41598-021-02395-9

Keywords

-

Funding

  1. Ministry of Trade, Industry & Energy, South Korea [20004927]
  2. Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI) - Ministry of Health & Welfare, Republic of Korea [HI19C0378]

Ask authors/readers for more resources

This study developed a model to predict unplanned hospital readmission within 30 days, considering weather and air quality factors, and demonstrated high performance in various machine learning methods, especially with gradient boosting machine (GBM). The model based on weather and air quality factors showed transportability and adaptability to other hospital systems.
Although several studies have attempted to develop a model for predicting 30-day re-hospitalization, few attempts have been made for sufficient verification and multi-center expansion for clinical use. In this study, we developed a model that predicts unplanned hospital readmission within 30 days of discharge; the model is based on a common data model and considers weather and air quality factors, and can be easily extended to multiple hospitals. We developed and compared four tree-based machine learning methods: decision tree, random forest, AdaBoost, and gradient boosting machine (GBM). Above all, GBM showed the highest AUC performance of 75.1 in the clinical model, while the clinical and W-score model showed the best performance of 73.9 for musculoskeletal diseases. Further, PM10, rainfall, and maximum temperature were the weather and air quality variables that most impacted the model. In addition, external validation has confirmed that the model based on weather and air quality factors has transportability to adapt to other hospital systems.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available