Advanced Care Planning for Hospitalized Patients Following Clinician Notification of Patient Mortality by a Machine Learning Algorithm - Original Investigation - Health Informatics

Q: What was the main objective of this cohort study?

To examine goals of care discussions in community hospitals for patients identified as high mortality risk by a machine learning mortality prediction algorithm.

Q: How did the intervention work in the intervention hospital?

Physicians were notified when a patient had high risk of 30-day mortality, and they were encouraged to arrange goals of care discussions.

Q: What were the key findings regarding goals of care discussions before discharge?

Patients in the intervention group were associated with about five times higher likelihood of documented goals of care discussions by discharge, and GOCDs occurred earlier compared with matched controls.

Aug 5, 2026

Upload by Finn

Goal-concordant care remains difficult in hospital settings, especially when clinicians must identify patients at high risk of death within 30 days to prompt serious illness conversations and documentation of goals of care. This cohort study evaluated goals of care discussions in community hospitals. Adult inpatients with high ML-predicted 30-day mortality were included from January 2 to July 15, 2021. Physicians received notification and encouraged GOCDs in the intervention hospital, compared with matched controls.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Abstract

Importance

Objective

Design, Setting, and Participants

Intervention

Main Outcomes and Measures

Results

Conclusions and Relevance

Key Points

Washington University School of Medicine Digital Commons@Becker | 2020-Current year OA Pubs | Open Access Publications | | --- | --- | 4-3-2023 Advanced care planning for hospitalized patients following clinician notification of patient mortality by a machine learning algorithm Stephen Chi Seunghwan Kim Matthew Reuter Katharine Ponzillo Debra Parker Oliver See next page for additional authors Follow this and additional works at: [https://digitalcommons.wustl.edu/oa_4](https://digitalcommons.wustl.edu/oa_4) Part of the Medicine and Health Sciences Commons Please let us know how this document benefits you. Authors Stephen Chi, Seunghwan Kim, Matthew Reuter, Katharine Ponzillo, Debra Parker Oliver, Randi Foraker, Kevin Heard, Jingxia Liu, Kyle Pitzer, Patrick White, and Nathan Moore Original Investigation | Health Informatics Advanced Care Planning for Hospitalized Patients Following Clinician Notification of Patient Mortality by a Machine Learning Algorithm Stephen Chi, MD; Seunghwan Kim, MS; Matthew Reuter, MD; Katharine Ponzillo, MD; Debra Parker Oliver, PhD, MSW; Randi Foraker, PhD; Kevin Heard, BS; Jingxia Liu, PhD; Kyle Pitzer, PhD; Patrick White, PhD, MD; Nathan Moore, MD | Abstract
IMPORTANCE Goal-concordant care is an ongoing challenge in hospital settings. Identification of high mortality risk within 30 days may call attention to the need to have serious illness conversations, including the documentation of patient goals of care.

OBJECTIVE To examine goals of care discussions (GOCDs) in a community hospital setting with patients identified as having a high risk of mortality by a machine learning mortality prediction algorithm.
DESIGN, SETTING, AND PARTICIPANTS This cohort study took place at community hospitals within 1 health care system. Participants included adult patients with a high risk of 30-day mortality who were admitted to 1 of 4 hospitals between January 2 and July 15, 2021 . Patient encounters of inpatients in the intervention hospital where physicians were notified of the computed high risk mortality score were compared with patient encounters of inpatients in 3 community hospitals without the intervention (ie, matched control) .
INTERVENTION Physicians of patients with a high risk of mortality within 30 days received notification and were encouraged to arrange for GOCDs.

MAIN OUTCOMESAND MEASURES The primary outcome was the percentage change of documented GOCDs prior to discharge. Propensity-score matching was completed on apreintervention and postintervention period using age, sex, race, COVID-19 status, and machine learning-predicted mortality risk scores. A difference-in-difference analysis validated the results.

RESULTS Overall, 537 patients were included in this study with 201 in the preintervention period (94 in the intervention group; 104 in the control group) and 336 patients in the postintervention period. The intervention and control groups included 168 patients per group and were well-balanced in age (mean [SD], 79.3 [9.60] vs 79.6 [9.21] years; standardized mean difference [SMD], 0.03), sex (female, 85 [51%] vs 85 [51%]; SMD, 0), race (White patients, 145 [86%] vs 144 [86%]; SMD 0.006), and Charlson comorbidities (median [range], 8 .00 [2 .00-15.0] vs 9.00 [2 .00 to 19.0]; SMD, 0.34) . Patients in the intervention group from preintervention to postintervention period were associated with being 5 times more likely to have documented GOCDs (OR, 5 .11 [95% CI, 1 .93 to 13 .42]; P = .001) by discharge compared with matched controls, and GOCD occurred significantly earlier in the hospitalization in the intervention patients as compared with matched controls (median, 4 [95% CI, 3 to 6] days vs 16 [95% CI, 15 to not applicable] days; P < .001) . Similar findings were observed for Black patient and White patient subgroups.
CONCLUSIONSAND RELEVANCE In this cohort study, patients whose physicians had knowledge of high-risk predictions fro

FAQs

What was the main objective of this cohort study?

To examine goals of care discussions in community hospitals for patients identified as high mortality risk by a machine learning mortality prediction algorithm.

How did the intervention work in the intervention hospital?

Physicians were notified when a patient had high risk of 30-day mortality, and they were encouraged to arrange goals of care discussions.

What were the key findings regarding goals of care discussions before discharge?

Patients in the intervention group were associated with about five times higher likelihood of documented goals of care discussions by discharge, and GOCDs occurred earlier compared with matched controls.

Aug 5, 2026

Upload by Finn