(295) ICU Telemedicine, Severity of Illness, and 30-Day Mortality Risk: A Retrospective Cohort Analysis

Monday, January 22, 2024

2:30 PM – 3:30 PM MT

Location: Connections Central - RST 13

First Author(s)

CU

Chiedozie I. Udeh, MBA, MHEcon, MBBS

Staff Intensivist, Medical Director, ICU Operations, Medical Director, EHospital
Cleveland Clinic Foundation
Cleveland, Ohio

Co-Author(s)

CC

Christina Canfield, CCRN-E, CNS, MSN, RN

Clinical Program Manager
Cleveland Clinic Foundation, Ohio, United States
AB

Abigail Brown, DNP, MBA, ACNP

Manager, Critical Care Transport
Cleveland Clinic, Ohio, United States
JC

Jonathan Castro, MD, MS

ICU Medical Director
Marymount Hospital, United States
SH

Steven Hata, MD, MSc

Staff Intensivist (rtd)
Cleveland Clinic, United States

Introduction: ICU telemedicine (ICU-TM) is an expanding model of care. Studies show variable impact on outcomes and lack randomized trials. Some observational studies have found that severity of illness may influence ICU-TM’s outcomes. This study models ICU-TM exposure for differences in 30-day mortality incident rate ratios (IRR), across ascending quartiles of APACHE IV scores, adjusted for competing risks.

Methods: We analyzed consecutive adult ICU patients from 9 hospitals by ICU-TM exposure from 2010-2019 (n=151,780), stratified by APACHE IV quartiles (Q1-Q4). Mortality risk (unadjusted) was assessed with chi-square and survival analyses. A Poisson multivariate regression model for IRR adjusted for 14 variables and 3 interaction terms. We targeted interaction between ICU-TM and Q1-Q4.

Results: Mortality risk ratios of ICU-TM/no ICU-TM were different among Q1(0.83, 95%CI 0.71,0.97), Q2(0.98,95%CI 0.88,1.10), Q3(0.63, 95%CI 0.57,0.69), and Q4(0.77, 95% CI 0.74,0.82) (p < 0.001). Unadjusted survival analyses showed ICU-TM exposure with reduced IRR in Q3 [0.60, 95%CI 0.54,0.66], and Q4 [0.72, 95%CI 0.68,0.77].
The Poisson model showed reduced IRR with ICU-TM (IRR 0.82, 95%CI 0.70,0.97). APACHE quartiles showed increased risk for Q2 (IRR 2.15, 95%CI 1.83,2.52), Q3 (IRR 3.93, 95%CI 3.39,4.56), and Q4 (IRR 9.30., 95%CI 8.10,10.67). IRR was higher in males (IRR 1.20,95%CI 1.15,1.26), sepsis (IRR 1.22, 95%CI 1.16,1.29), emergency admits (IRR 1.17, 95%CI 1.12,1.22) and weekend admits (IRR 1.24, 95%CI 1.14,1.34), non-white race (IRR 1.17 95%CI 1.11,1.22), cardiac arrest (IRR 1.96, 95%CI 1.80,2.14), and multi-ICU admits (IRR 1.55, 95%CI 1.47,1.63). IRR was lower with surgery (IRR 0.71, 95%CI 0.66,0.75) and neurologic care (IRR 0.85, 95%CI 0.76,0.95).
Importantly, significant interaction was noted between ICU-TM and APACHE Q3 (IRR 0.80, 95%CI 0.67,0.96) and Q2(IRR 1.24, 95%CI 1.02,1.50). There was no significant interaction in APACHE Q4 (IRR 1.00, 95%CI 0.85,1.19). Interaction also occurred with ICU-TM and admits at night (IRR 0.88, 95%CI 0.80,0.96) or weekends (IRR 0.85, 95%CI 0.77,0.94).

Conclusions: In summary, ICU telemedicine’s association with reduced mortality risk appears to be modified by APACHE IV severity of illness, particularly in Q3 and Q2. Our model supports that other factors are also influential.