A latent variable approach to potential outcomes for emergency department admission decisions

Amy L. Cochran, Paul Rathouz, Keith E. Kocher, Gabriel Zayas-Cabán

Research output: Contribution to journalArticle

Abstract

In emergency departments (EDs), care providers continuously weigh admissions against continued monitoring and treatment often without knowing their condition and health needs. To understand the decision process and its causal effect on outcomes, an observational study must contend with unobserved/missing information and a lack of exchangeability between admitted and discharged patients. Our goal was to provide a general framework to evaluate admission decisions from electronic healthcare records (EHRs). We describe admission decisions as a decision-making process in which the patient's health needs is a binary latent variable. We estimate latent health needs from EHR with only partial knowledge of the decision process (ie, initial evaluation, admission decision, length of stay). Estimated latent health needs are then used to understand the admission decision and the decision's causal impact on outcomes. For the latter, we assume potential outcomes are stochastically independent from the admission decision conditional on latent health needs. As a case study, we apply our approach to over 150 000 patient encounters with the ED from the University of Michigan Health System collected from August 2012 through July 2015. We estimate that while admitting a patient with higher latent needs reduces the 30-day risk of revisiting the ED or later being admitted through the ED by over 79%, admitting a patient with lower latent needs actually increases these 30-day risks by 3.0% and 7.6%, respectively.

Original languageEnglish (US)
Pages (from-to)3911-3935
Number of pages25
JournalStatistics in Medicine
Volume38
Issue number20
DOIs
StatePublished - Jan 1 2019

Fingerprint

Potential Outcomes
Latent Variables
Emergency
Hospital Emergency Service
Health
Healthcare
Delivery of Health Care
Electronics
Emergency Medical Services
Exchangeability
Causal Effect
Observational Study
Binary Variables
Observational Studies
Length of Stay
Decision Making
Outcome Assessment (Health Care)
Estimate
Monitoring
Partial

Keywords

  • causal inference
  • emergency department admission decisions
  • latent variables
  • potential outcomes

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

A latent variable approach to potential outcomes for emergency department admission decisions. / Cochran, Amy L.; Rathouz, Paul; Kocher, Keith E.; Zayas-Cabán, Gabriel.

In: Statistics in Medicine, Vol. 38, No. 20, 01.01.2019, p. 3911-3935.

Research output: Contribution to journalArticle

Cochran, Amy L. ; Rathouz, Paul ; Kocher, Keith E. ; Zayas-Cabán, Gabriel. / A latent variable approach to potential outcomes for emergency department admission decisions. In: Statistics in Medicine. 2019 ; Vol. 38, No. 20. pp. 3911-3935.
@article{689a65dc3f6e4546ad70f2a919fe6262,
title = "A latent variable approach to potential outcomes for emergency department admission decisions",
abstract = "In emergency departments (EDs), care providers continuously weigh admissions against continued monitoring and treatment often without knowing their condition and health needs. To understand the decision process and its causal effect on outcomes, an observational study must contend with unobserved/missing information and a lack of exchangeability between admitted and discharged patients. Our goal was to provide a general framework to evaluate admission decisions from electronic healthcare records (EHRs). We describe admission decisions as a decision-making process in which the patient's health needs is a binary latent variable. We estimate latent health needs from EHR with only partial knowledge of the decision process (ie, initial evaluation, admission decision, length of stay). Estimated latent health needs are then used to understand the admission decision and the decision's causal impact on outcomes. For the latter, we assume potential outcomes are stochastically independent from the admission decision conditional on latent health needs. As a case study, we apply our approach to over 150 000 patient encounters with the ED from the University of Michigan Health System collected from August 2012 through July 2015. We estimate that while admitting a patient with higher latent needs reduces the 30-day risk of revisiting the ED or later being admitted through the ED by over 79{\%}, admitting a patient with lower latent needs actually increases these 30-day risks by 3.0{\%} and 7.6{\%}, respectively.",
keywords = "causal inference, emergency department admission decisions, latent variables, potential outcomes",
author = "Cochran, {Amy L.} and Paul Rathouz and Kocher, {Keith E.} and Gabriel Zayas-Cab{\'a}n",
year = "2019",
month = "1",
day = "1",
doi = "10.1002/sim.8210",
language = "English (US)",
volume = "38",
pages = "3911--3935",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "20",

}

TY - JOUR

T1 - A latent variable approach to potential outcomes for emergency department admission decisions

AU - Cochran, Amy L.

AU - Rathouz, Paul

AU - Kocher, Keith E.

AU - Zayas-Cabán, Gabriel

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In emergency departments (EDs), care providers continuously weigh admissions against continued monitoring and treatment often without knowing their condition and health needs. To understand the decision process and its causal effect on outcomes, an observational study must contend with unobserved/missing information and a lack of exchangeability between admitted and discharged patients. Our goal was to provide a general framework to evaluate admission decisions from electronic healthcare records (EHRs). We describe admission decisions as a decision-making process in which the patient's health needs is a binary latent variable. We estimate latent health needs from EHR with only partial knowledge of the decision process (ie, initial evaluation, admission decision, length of stay). Estimated latent health needs are then used to understand the admission decision and the decision's causal impact on outcomes. For the latter, we assume potential outcomes are stochastically independent from the admission decision conditional on latent health needs. As a case study, we apply our approach to over 150 000 patient encounters with the ED from the University of Michigan Health System collected from August 2012 through July 2015. We estimate that while admitting a patient with higher latent needs reduces the 30-day risk of revisiting the ED or later being admitted through the ED by over 79%, admitting a patient with lower latent needs actually increases these 30-day risks by 3.0% and 7.6%, respectively.

AB - In emergency departments (EDs), care providers continuously weigh admissions against continued monitoring and treatment often without knowing their condition and health needs. To understand the decision process and its causal effect on outcomes, an observational study must contend with unobserved/missing information and a lack of exchangeability between admitted and discharged patients. Our goal was to provide a general framework to evaluate admission decisions from electronic healthcare records (EHRs). We describe admission decisions as a decision-making process in which the patient's health needs is a binary latent variable. We estimate latent health needs from EHR with only partial knowledge of the decision process (ie, initial evaluation, admission decision, length of stay). Estimated latent health needs are then used to understand the admission decision and the decision's causal impact on outcomes. For the latter, we assume potential outcomes are stochastically independent from the admission decision conditional on latent health needs. As a case study, we apply our approach to over 150 000 patient encounters with the ED from the University of Michigan Health System collected from August 2012 through July 2015. We estimate that while admitting a patient with higher latent needs reduces the 30-day risk of revisiting the ED or later being admitted through the ED by over 79%, admitting a patient with lower latent needs actually increases these 30-day risks by 3.0% and 7.6%, respectively.

KW - causal inference

KW - emergency department admission decisions

KW - latent variables

KW - potential outcomes

UR - http://www.scopus.com/inward/record.url?scp=85067662772&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067662772&partnerID=8YFLogxK

U2 - 10.1002/sim.8210

DO - 10.1002/sim.8210

M3 - Article

VL - 38

SP - 3911

EP - 3935

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 20

ER -