EUVIMEDEUVIMED entdecken
Crossref

Identifying inpatient mortality in MarketScan claims data using machine learning

Fenglong Xie iD, Timothy Beukelman iD, Dongmei Sun, Huifeng Yun, Jeffrey R Curtis

Pharmacoepidemiology and Drug Safety · 2023

DOI: 10.1002/pds.5658 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Pharmacoepidemiology and Drug Safety, 2023. Verfasst von Fenglong Xie und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medicine, Division of Clinical Immunology and Rheumatology, University of Alabama at Birmingham, Birmingham, Alabama, USA. Hauptthemen laut MEDLINE: Inpatients, Machine Learning. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 12 Literaturangaben. Gefördert von: National Institute of Arthritis and Musculoskeletal and Skin Diseases.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Inpatient mortality is an important variable in epidemiology studies using claims data. In 2016, MarketScan data began obscuring specific hospital discharge status types for patient privacy, including inpatient deaths, by setting the values to missing. We used a machine learning approach to correctly identify hospitalizations that resulted in inpatient death using data prior to 2016.

Methods

All hospitalizations from 2011 to 2015 with discharge status of missing, died, or one of the other subsequently obscured values were identified and divided into a training set and two test sets. Predictor variables included age, sex, elapsed time from hospital discharge until last observed claim and until healthcare plan disenrollment, and absence of any discharge diagnoses. Four machine learning methods were used to train statistical models and assess sensitivity and positive predictive value (PPV) for inpatient mortality.

Results

Overall 1 307 917 hospitalizations were included. All four machine learning approaches performed well in all datasets. Random forest performed best with 88% PPV and 93% sensitivity for the training set and both test sets. The two factors with the highest relative importance for identifying inpatient mortality were having no observed claims for the patient on days 2–91 following hospital discharge and patient disenrollment from the healthcare plan within 60 days following hospital discharge.

Conclusion

We successfully developed machine learning algorithms to identify inpatient mortality. This approach can be applied to obscured data to accurately identify inpatient mortality among hospitalizations with missing discharge status.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

InpatientsMachine LearningHumansAlgorithmsHospitalizationPatient DischargeRetrospective Studies

Stichwörter der Autor:innen

claims datamachine learningmortality

Autor:innen und Institutionen

  • Department of Medicine, Division of Clinical Immunology and Rheumatology, University of Alabama at Birmingham, Birmingham, Alabama, USA.
  • Foundation for Science, Technology, Education, and Research (FASTER), Birmingham, Alabama, USA.

ORCID-Profile: Fenglong Xie, Timothy Beukelman

Förderung

  • National Institute of Arthritis and Musculoskeletal and Skin Diseases

Zitieren

APA

Xie, F., Beukelman, T., Sun, D., Yun, H., & Curtis, J. R. (2023). Identifying inpatient mortality in MarketScan claims data using machine learning. Pharmacoepidemiology and Drug Safety. https://doi.org/10.1002/pds.5658

Vancouver

Xie F, Beukelman T, Sun D, Yun H, Curtis JR. Identifying inpatient mortality in MarketScan claims data using machine learning. Pharmacoepidemiology and Drug Safety. 2023. doi:10.1002/pds.5658

BibTeX
@article{Xie2023,
  author = {Xie, Fenglong and Beukelman, Timothy and Sun, Dongmei and Yun, Huifeng and Curtis, Jeffrey R},
  title = {Identifying inpatient mortality in MarketScan claims data using machine learning},
  journal = {Pharmacoepidemiology and Drug Safety},
  year = {2023},
  doi = {10.1002/pds.5658}
}
Kostenlos starten →