View in EDS HTML Full Text PDF Full Text

Patient Outcome Predictions Improve Operations at Hartford HealthCare.

Saved in:
Bibliographic Details
Title: Patient Outcome Predictions Improve Operations at Hartford HealthCare.
Authors: Na, Liangyuan (AUTHOR), Villalobos Carballo, Kimberly (AUTHOR), Pauphilet, Jean (AUTHOR), Haddad-Sisakht, Ali (AUTHOR), Kombert, Daniel (AUTHOR), Boisjoli-Langlois, Melissa (AUTHOR), Castiglione, Andrew (AUTHOR), Khalifa, Maram (AUTHOR), Hebbal, Pooja (AUTHOR), Stein, Barry (AUTHOR), Bertsimas, Dimitris (AUTHOR)
Source: INFORMS Journal on Applied Analytics. May/Jun2026, Vol. 56 Issue 3, p224-242. 19p.
Subjects: Discharge planning, Prognostic models, Medical care costs, Hospital administration, Death forecasting, Artificial intelligence in medicine, Prognosis, Length of stay in hospitals
Geographic Terms: Hartford (Conn.)
Abstract: We build and deploy machine learning models that accurately predict short- and medium-term inpatient outcomes for Hartford HealthCare, including 24–48 hour discharge, ICU transfer, mortality, and discharge disposition (AUC 76%–93%). More than 200 clinicians currently use these predictions in daily rounds, leading to earlier discharge planning, shorter length of stay (0.63 days per patient), and substantial financial benefits (between $52 and $67 million annually) for the healthcare system. Access to accurate predictions of patients' outcomes can enhance decision making within healthcare institutions. Hartford HealthCare has been collaborating with academics and consultants to predict short- and medium-term outcomes for all inpatients across their seven hospitals. We develop machine learning models that predict the probabilities of next 24-hour/48-hour discharge and intensive care unit transfers, end-of-stay mortality, and discharge dispositions. All models achieve high out-of-sample area under the receiver operating curve (75.7% – 92.5%) and are well calibrated. In addition, combining 48-hour discharge predictions with doctors' predictions simultaneously enables more patient discharges (10%–28.7%) and fewer 7-day/30-day readmissions (p < 0.001). We implement an automated pipeline that extracts data and updates predictions every morning, as well as user-friendly software and a color-coded alert system to communicate these patient-level predictions to clinical teams. Since its deployment, more than 200 doctors, nurses, and case managers across seven hospitals have been using the tool in their daily patient review process. With our tool, we find that doctors start the administrative discharge process earlier, leading to a significant reduction in the average length of stay (0.63 days per patient). We anticipate substantial financial benefits (between $52 and $67 million annually) for the healthcare system. History: This paper was refereed. Supplemental Material: The online appendix is available at https://doi.org/10.1287/inte.2024.0170. [ABSTRACT FROM AUTHOR]
Copyright of INFORMS Journal on Applied Analytics is the property of INFORMS: Institute for Operations Research & the Management Sciences and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Psychology and Behavioral Sciences Collection
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: pbh
DbLabel: Psychology and Behavioral Sciences Collection
An: 194057700
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Patient Outcome Predictions Improve Operations at Hartford HealthCare.
– Name: Author
  Label: Authors
  Group: Au
  Data: &lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Na%2C+Liangyuan%22&quot;&gt;Na, Liangyuan&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Villalobos+Carballo%2C+Kimberly%22&quot;&gt;Villalobos Carballo, Kimberly&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Pauphilet%2C+Jean%22&quot;&gt;Pauphilet, Jean&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Haddad-Sisakht%2C+Ali%22&quot;&gt;Haddad-Sisakht, Ali&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Kombert%2C+Daniel%22&quot;&gt;Kombert, Daniel&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Boisjoli-Langlois%2C+Melissa%22&quot;&gt;Boisjoli-Langlois, Melissa&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Castiglione%2C+Andrew%22&quot;&gt;Castiglione, Andrew&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Khalifa%2C+Maram%22&quot;&gt;Khalifa, Maram&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Hebbal%2C+Pooja%22&quot;&gt;Hebbal, Pooja&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Stein%2C+Barry%22&quot;&gt;Stein, Barry&lt;/searchLink&gt; (AUTHOR)&lt;br /&gt;&lt;searchLink fieldCode=&quot;AR&quot; term=&quot;%22Bertsimas%2C+Dimitris%22&quot;&gt;Bertsimas, Dimitris&lt;/searchLink&gt; (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: &lt;searchLink fieldCode=&quot;JN&quot; term=&quot;%22INFORMS+Journal+on+Applied+Analytics%22&quot;&gt;INFORMS Journal on Applied Analytics&lt;/searchLink&gt;. May/Jun2026, Vol. 56 Issue 3, p224-242. 19p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: &lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Discharge+planning%22&quot;&gt;Discharge planning&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Prognostic+models%22&quot;&gt;Prognostic models&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Medical+care+costs%22&quot;&gt;Medical care costs&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Hospital+administration%22&quot;&gt;Hospital administration&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Death+forecasting%22&quot;&gt;Death forecasting&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Artificial+intelligence+in+medicine%22&quot;&gt;Artificial intelligence in medicine&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Prognosis%22&quot;&gt;Prognosis&lt;/searchLink&gt;&lt;br /&gt;&lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Length+of+stay+in+hospitals%22&quot;&gt;Length of stay in hospitals&lt;/searchLink&gt;
– Name: SubjectGeographic
  Label: Geographic Terms
  Group: Su
  Data: &lt;searchLink fieldCode=&quot;DE&quot; term=&quot;%22Hartford+%28Conn%2E%29%22&quot;&gt;Hartford (Conn.)&lt;/searchLink&gt;
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: We build and deploy machine learning models that accurately predict short- and medium-term inpatient outcomes for Hartford HealthCare, including 24–48 hour discharge, ICU transfer, mortality, and discharge disposition (AUC 76%–93%). More than 200 clinicians currently use these predictions in daily rounds, leading to earlier discharge planning, shorter length of stay (0.63 days per patient), and substantial financial benefits (between $52 and $67 million annually) for the healthcare system. Access to accurate predictions of patients&#39; outcomes can enhance decision making within healthcare institutions. Hartford HealthCare has been collaborating with academics and consultants to predict short- and medium-term outcomes for all inpatients across their seven hospitals. We develop machine learning models that predict the probabilities of next 24-hour/48-hour discharge and intensive care unit transfers, end-of-stay mortality, and discharge dispositions. All models achieve high out-of-sample area under the receiver operating curve (75.7% – 92.5%) and are well calibrated. In addition, combining 48-hour discharge predictions with doctors&#39; predictions simultaneously enables more patient discharges (10%–28.7%) and fewer 7-day/30-day readmissions (p &lt; 0.001). We implement an automated pipeline that extracts data and updates predictions every morning, as well as user-friendly software and a color-coded alert system to communicate these patient-level predictions to clinical teams. Since its deployment, more than 200 doctors, nurses, and case managers across seven hospitals have been using the tool in their daily patient review process. With our tool, we find that doctors start the administrative discharge process earlier, leading to a significant reduction in the average length of stay (0.63 days per patient). We anticipate substantial financial benefits (between $52 and $67 million annually) for the healthcare system. History: This paper was refereed. Supplemental Material: The online appendix is available at https://doi.org/10.1287/inte.2024.0170. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: &lt;i&gt;Copyright of INFORMS Journal on Applied Analytics is the property of INFORMS: Institute for Operations Research &amp; the Management Sciences and its content may not be copied or emailed to multiple sites without the copyright holder&#39;s express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.&lt;/i&gt; (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=pbh&AN=194057700
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1287/inte.2024.0170
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 19
        StartPage: 224
    Subjects:
      – SubjectFull: Discharge planning
        Type: general
      – SubjectFull: Prognostic models
        Type: general
      – SubjectFull: Medical care costs
        Type: general
      – SubjectFull: Hospital administration
        Type: general
      – SubjectFull: Death forecasting
        Type: general
      – SubjectFull: Artificial intelligence in medicine
        Type: general
      – SubjectFull: Prognosis
        Type: general
      – SubjectFull: Length of stay in hospitals
        Type: general
      – SubjectFull: Hartford (Conn.)
        Type: general
    Titles:
      – TitleFull: Patient Outcome Predictions Improve Operations at Hartford HealthCare.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Na, Liangyuan
      – PersonEntity:
          Name:
            NameFull: Villalobos Carballo, Kimberly
      – PersonEntity:
          Name:
            NameFull: Pauphilet, Jean
      – PersonEntity:
          Name:
            NameFull: Haddad-Sisakht, Ali
      – PersonEntity:
          Name:
            NameFull: Kombert, Daniel
      – PersonEntity:
          Name:
            NameFull: Boisjoli-Langlois, Melissa
      – PersonEntity:
          Name:
            NameFull: Castiglione, Andrew
      – PersonEntity:
          Name:
            NameFull: Khalifa, Maram
      – PersonEntity:
          Name:
            NameFull: Hebbal, Pooja
      – PersonEntity:
          Name:
            NameFull: Stein, Barry
      – PersonEntity:
          Name:
            NameFull: Bertsimas, Dimitris
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 05
              Text: May/Jun2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 26440865
          Numbering:
            – Type: volume
              Value: 56
            – Type: issue
              Value: 3
          Titles:
            – TitleFull: INFORMS Journal on Applied Analytics
              Type: main
ResultId 1