Machine learning approaches for early DRG classification and resource allocation

Daniel Gartner, Rainer Kolisch, Daniel Neill, Rema Padman

Research output: Contribution to journalArticle

Abstract

Recent research has highlighted the need for upstream planning in healthcare service delivery systems, patient scheduling, and resource allocation in the hospital inpatient setting. This study examines the value of upstream planning within hospital-wide resource allocation decisions based on machine learning (ML) and mixed-integer programming (MIP), focusing on prediction of diagnosis-related groups (DRGs) and the use of these predictions for allocating scarce hospital resources. DRGs are a payment scheme employed at patients' discharge, where the DRG and length of stay determine the revenue that the hospital obtains. We show that early and accurate DRG classification using ML methods, incorporated into an MIP-based resource allocation model, can increase the hospital's contribution margin, the number of admitted patients, and the utilization of resources such as operating rooms and beds. We test these methods on hospital data containing more than 16,000 inpatient records and demonstrate improved DRG classification accuracy as compared to the hospital's current approach. The largest improvements were observed at and before admission, when information such as procedures and diagnoses is typically incomplete, but performance was improved even after a substantial portion of the patient's length of stay, and under multiple scenarios making different assumptions about the available information. Using the improved DRG predictions within our resource allocation model improves contribution margin by 2.9% and the utilization of scarce resources such as operating rooms and beds from 66.3% to 67.3% and from 70.7% to 71.7%, respectively. This enables 9.0% more nonurgent elective patients to be admitted as compared to the baseline.

Original languageEnglish (US)
Pages (from-to)718-734
Number of pages17
JournalINFORMS Journal on Computing
Volume27
Issue number4
DOIs
StatePublished - Jan 1 2015

Fingerprint

Resource allocation
Learning systems
Hospital beds
Operating rooms
Integer programming
Planning
Machine learning
Diagnosis-related groups
Scheduling
Resources
Prediction

Keywords

  • Attribute selection
  • Classification
  • Diagnosis-related groups
  • Machine learning
  • Mathematical programming

ASJC Scopus subject areas

  • Software
  • Information Systems
  • Computer Science Applications
  • Management Science and Operations Research

Cite this

Machine learning approaches for early DRG classification and resource allocation. / Gartner, Daniel; Kolisch, Rainer; Neill, Daniel; Padman, Rema.

In: INFORMS Journal on Computing, Vol. 27, No. 4, 01.01.2015, p. 718-734.

Research output: Contribution to journalArticle

Gartner, Daniel ; Kolisch, Rainer ; Neill, Daniel ; Padman, Rema. / Machine learning approaches for early DRG classification and resource allocation. In: INFORMS Journal on Computing. 2015 ; Vol. 27, No. 4. pp. 718-734.
@article{f3587dcaf5e64979b19c801abc0faea0,
title = "Machine learning approaches for early DRG classification and resource allocation",
abstract = "Recent research has highlighted the need for upstream planning in healthcare service delivery systems, patient scheduling, and resource allocation in the hospital inpatient setting. This study examines the value of upstream planning within hospital-wide resource allocation decisions based on machine learning (ML) and mixed-integer programming (MIP), focusing on prediction of diagnosis-related groups (DRGs) and the use of these predictions for allocating scarce hospital resources. DRGs are a payment scheme employed at patients' discharge, where the DRG and length of stay determine the revenue that the hospital obtains. We show that early and accurate DRG classification using ML methods, incorporated into an MIP-based resource allocation model, can increase the hospital's contribution margin, the number of admitted patients, and the utilization of resources such as operating rooms and beds. We test these methods on hospital data containing more than 16,000 inpatient records and demonstrate improved DRG classification accuracy as compared to the hospital's current approach. The largest improvements were observed at and before admission, when information such as procedures and diagnoses is typically incomplete, but performance was improved even after a substantial portion of the patient's length of stay, and under multiple scenarios making different assumptions about the available information. Using the improved DRG predictions within our resource allocation model improves contribution margin by 2.9{\%} and the utilization of scarce resources such as operating rooms and beds from 66.3{\%} to 67.3{\%} and from 70.7{\%} to 71.7{\%}, respectively. This enables 9.0{\%} more nonurgent elective patients to be admitted as compared to the baseline.",
keywords = "Attribute selection, Classification, Diagnosis-related groups, Machine learning, Mathematical programming",
author = "Daniel Gartner and Rainer Kolisch and Daniel Neill and Rema Padman",
year = "2015",
month = "1",
day = "1",
doi = "10.1287/ijoc.2015.0655",
language = "English (US)",
volume = "27",
pages = "718--734",
journal = "INFORMS Journal on Computing",
issn = "1091-9856",
publisher = "INFORMS Inst.for Operations Res.and the Management Sciences",
number = "4",

}

TY - JOUR

T1 - Machine learning approaches for early DRG classification and resource allocation

AU - Gartner, Daniel

AU - Kolisch, Rainer

AU - Neill, Daniel

AU - Padman, Rema

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Recent research has highlighted the need for upstream planning in healthcare service delivery systems, patient scheduling, and resource allocation in the hospital inpatient setting. This study examines the value of upstream planning within hospital-wide resource allocation decisions based on machine learning (ML) and mixed-integer programming (MIP), focusing on prediction of diagnosis-related groups (DRGs) and the use of these predictions for allocating scarce hospital resources. DRGs are a payment scheme employed at patients' discharge, where the DRG and length of stay determine the revenue that the hospital obtains. We show that early and accurate DRG classification using ML methods, incorporated into an MIP-based resource allocation model, can increase the hospital's contribution margin, the number of admitted patients, and the utilization of resources such as operating rooms and beds. We test these methods on hospital data containing more than 16,000 inpatient records and demonstrate improved DRG classification accuracy as compared to the hospital's current approach. The largest improvements were observed at and before admission, when information such as procedures and diagnoses is typically incomplete, but performance was improved even after a substantial portion of the patient's length of stay, and under multiple scenarios making different assumptions about the available information. Using the improved DRG predictions within our resource allocation model improves contribution margin by 2.9% and the utilization of scarce resources such as operating rooms and beds from 66.3% to 67.3% and from 70.7% to 71.7%, respectively. This enables 9.0% more nonurgent elective patients to be admitted as compared to the baseline.

AB - Recent research has highlighted the need for upstream planning in healthcare service delivery systems, patient scheduling, and resource allocation in the hospital inpatient setting. This study examines the value of upstream planning within hospital-wide resource allocation decisions based on machine learning (ML) and mixed-integer programming (MIP), focusing on prediction of diagnosis-related groups (DRGs) and the use of these predictions for allocating scarce hospital resources. DRGs are a payment scheme employed at patients' discharge, where the DRG and length of stay determine the revenue that the hospital obtains. We show that early and accurate DRG classification using ML methods, incorporated into an MIP-based resource allocation model, can increase the hospital's contribution margin, the number of admitted patients, and the utilization of resources such as operating rooms and beds. We test these methods on hospital data containing more than 16,000 inpatient records and demonstrate improved DRG classification accuracy as compared to the hospital's current approach. The largest improvements were observed at and before admission, when information such as procedures and diagnoses is typically incomplete, but performance was improved even after a substantial portion of the patient's length of stay, and under multiple scenarios making different assumptions about the available information. Using the improved DRG predictions within our resource allocation model improves contribution margin by 2.9% and the utilization of scarce resources such as operating rooms and beds from 66.3% to 67.3% and from 70.7% to 71.7%, respectively. This enables 9.0% more nonurgent elective patients to be admitted as compared to the baseline.

KW - Attribute selection

KW - Classification

KW - Diagnosis-related groups

KW - Machine learning

KW - Mathematical programming

UR - http://www.scopus.com/inward/record.url?scp=84948694241&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84948694241&partnerID=8YFLogxK

U2 - 10.1287/ijoc.2015.0655

DO - 10.1287/ijoc.2015.0655

M3 - Article

AN - SCOPUS:84948694241

VL - 27

SP - 718

EP - 734

JO - INFORMS Journal on Computing

JF - INFORMS Journal on Computing

SN - 1091-9856

IS - 4

ER -