Predicting 120-day hospital readmission using medical administrative patient data

Research output: Contribution to conferencePaperScientificpeer-review

Abstract

Hospitals and health-care insurers routinely use models to predict patient readmission, extrapolating from historical data. Subsequently, the predicted quantities can be used for contracting and pricing negotiations between these hospitals and healthcare insurers. The Dutch healthcare system uses unique standardized Care Trajectories (so-called DBCs) for administration and billing of care. Here, we compared supervised machine learning methods on predicting 120-day readmission as an operationally significant metric. We used administrative patient data from 21 common Care Trajectories, in combination with demographic information. A lightGBM model using undersampling to tackle class imbalance yielded an AUROC score of 0.86 and provided the highest recall score (73.8%).
Original languageEnglish
Publication statusPublished - 2019
Event28th Belgian Dutch Conference on Machine Learning - Brussels, Belgium
Duration: 6 Nov 20198 Nov 2019
Conference number: 28
https://bnaic19.brussels/benelearn-call-for-papers/

Conference

Conference28th Belgian Dutch Conference on Machine Learning
Abbreviated titleBenelearn
CountryBelgium
CityBrussels
Period6/11/198/11/19
Internet address

    Fingerprint

Keywords

  • Administrative patient data
  • Care trajectory
  • Health analytics
  • Hospital readmission

Cite this

Van Wingerden, M., De Boer, J., & Postma, E. (2019). Predicting 120-day hospital readmission using medical administrative patient data. Paper presented at 28th Belgian Dutch Conference on Machine Learning, Brussels, Belgium.