Article Text

Original research
Machine-learning-based hospital discharge predictions can support multidisciplinary rounds and decrease hospital length-of-stay
  1. Scott Levin1,2,
  2. Sean Barnes2,3,
  3. Matthew Toerper1,2,
  4. Arnaud Debraine2,
  5. Anthony DeAngelo2,
  6. Eric Hamrock2,
  7. Jeremiah Hinson1,2,
  8. Erik Hoyer4,
  9. Trushar Dungarani5,
  10. Eric Howell6
  1. 1Emergency Medicine, Johns Hopkins University, Baltimore, Maryland, USA
  2. 2StoCastic, Baltimore, Maryland, USA
  3. 3Decision, Operations, and Information Technologies, University of Maryland at College Park, College Park, Maryland, USA
  4. 4Physical Medicine and Rehabilitation, Johns Hopkins University, Baltimore, Maryland, USA
  5. 5Community Physicians, Johns Hopkins Medicine, Baltimore, Maryland, USA
  6. 6Society of Hospital Medicine, Philadelphia, Pennsylvania, USA
  1. Correspondence to Dr Scott Levin, Emergency Medicine, Johns Hopkins University, Baltimore, MD 21218, USA; slevin33{at}jhmi.edu

Abstract

Background Patient flow directly affects quality of care, access and financial performance for hospitals. Multidisciplinary discharge-focused rounds have proven to minimise avoidable delays experienced by patients near discharge. The study objective was to support discharge-focused rounds by implementing a machine-learning-based discharge prediction model using real-time electronic health record (EHR) data. We aimed to evaluate model predictive performance and impact on hospital length-of-stay.

Methods Discharge prediction models were developed from hospitalised patients on four inpatient units between April 2016 and September 2018. Unit-specific models were implemented to make individual patient predictions viewable with the EHR patient track board. Predictive performance was measured prospectively for 12 470 patients (120 780 patient-predictions) across all units. A pre/poststudy design applying interrupted time series methods was used to assess the impact of the discharge prediction model on hospital length-of-stay.

Results Prospective discharge prediction performance ranged in area under the receiver operating characteristic curve from 0.70 to 0.80 for same-day and next-day predictions; sensitivity was between 0.63 and 0.83 and specificity between 0.48 and 0.80. Elapsed length-of-stay, counts of labs and medications, mobility assessments and measures of acute kidney injury were model features providing the most predictive value. Implementing the discharge predictions resulted in a reduction in hospital length-of-stay of over 12 hours on a medicine unit (p<0.001) and telemetry unit (p=0.002), while no changes were observed for the surgery unit (p=0.190) and second medicine unit (p<0.555).

Conclusions Incorporating automated patient discharge predictions into multidisciplinary rounds can support decreases in hospital length-of-stay. Variation in execution and impact across inpatient units existed.

  • assistive technology
  • delivery
  • medical apps
http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

View Full Text

Statistics from Altmetric.com

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Contributors SL, SB and JH planned the study. MT, ArD, AnD, SB and SL developed the technical components of the technology and performing data analyses. EHa led technology implementation and implementation-focused writing. JH, EHoy, TD, EHow performed clinical interpretation for model development, results evaluation and critically revised the manuscript. EHow also performed a non-conflicted review of the data and manuscript.

  • Funding The technology development and assessment was supported by the United States National Science Foundation (NSF) Award #0927207, SBIR 1621899 and SBIR 1738440 to StoCastic LLC, with a subaward to The Johns Hopkins University School of Medicine. SL, SB, ArD, AnD and EHa received funding from this award.

  • Competing interests The technology described in this manuscript was developed and implemented by StoCastic LLC. Under a license agreement between StoCastic and the Johns Hopkins University, Dr Levin and the University are entitled to royalty distributions related to technology described in this publication. SL is a founder of StoCastic and he, JH, EHa, AnD, MT and the University own equity in the company. This arrangement has been reviewed and approved by the Johns Hopkins University in accordance with its conflict of interest policies. A patent (US9311449B2) for the discharge prediction tool has been issued by Johns Hopkins University. Rights for commercial development of this intellectual property (IP) have been exclusively licensed by StoCastic, LLC.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement No data are available. Patient-level hospital data are prohibited for sharing under both federal privacy (HIPAA) and contractual constraints. However, authors are happy to share additional methodological detail and code if requested. Please email slevin33@jhmi.edu

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.