SCCM Resource Library-16-ICU Telemedicine and Clinical Factors Related to 30-Day Mortality: A Retrospective Cohort Study

Video Transcription

Good afternoon. My name is Dr. Chieh Dozieh-Ude. I am the Enterprise Medical Director of ICU Operations for the Cleveland Clinic Health System. I also serve as Medical Director of  eHospital. eHospital is our Cleveland Clinic telemedicine program for the ICUs.  It is my pleasure to be presenting today some aspects of our work with ICU telemedicine.  The topic of my talk is ICU telemedicine and clinical risks associated with 30-day mortality, a retrospective cohort study. First, I would like to recognize my co-investigators,  as shown here on the slide. And on their behalf, I also want to acknowledge our colleagues who  work with us in the ICU telemedicine program. They have achieved the outcomes that I will be discussing today. None of the investigators have any conflicts of interest to declare.  This work was supported solely by institutional funds.  Now, I'm getting right to it. As you are probably aware, 24-7 intensivist presence in every ICU is  simply not feasible at the present time. In fact, some have argued that 24-7 coverage is not  necessary. In any case, this is where ICU telemedicine is playing a growing role. Yet, after  Yet, after almost 20 years of experience, there is still some controversy about the benefits and  risks of ICU telemedicine. Most of the current evidence we have comes from observational studies.  And as you can imagine, there are significant practical difficulties in conducting a randomized controlled trial in this arena.  Nonetheless, continued research is still needed to better define when and how ICU telemedicine  is beneficial. It was against this backdrop that the Cleveland Clinic Health System established  its ICU telemedicine program. It was deployed from 2014 to 2016. And since inception, it has  run only as a nighttime program from 7 p.m. to 7 a.m. It covers and supports ICUs across the health  system, both in Ohio and in Florida. And the operating base is in Cleveland. The program is  staffed with a mix of intensivists, acute care nurse practitioners, and critical care nurses.  Now, after about five to six years of continuous operation, we begin to report some of our clinical  outcomes. So the primary goal of this study was to evaluate differences in 30-day mortality  amongst patients based on their exposure to ICU telemedicine within a health system.  The secondary goals were also to evaluate the impact of exposure or not  to ICU telemedicine on hospital and ICU length of stay.  The study was approved by our enterprise IRB with a waiver of informed consent as a retrospective  cohort study. As such, it was conducted in accordance with the stroke guidelines for retrospective studies. The patient cohort was identified and relevant data abstracted from our  enterprise ICU data mart and the Apache registry. The cohort included all adult ICU patients The cohort included all adult ICU patients admitted to ICUs across a health system that  used the Apache severity of illness model. Patients admitted to our main campus ICUs were excluded.  This was because all our main campus ICUs have 24-7 in-house intensivist coverage, and so there is no significant exposure to ICU telemedicine at our main campus.  The study was powered to 90% to detect a 5% difference in mortality between the two groups,  assuming a baseline estimated mortality of 20%. Standard descriptive statistics were used to  summarize the data, while multivariable logistic regression modeling and survival curves were used for the statistical analysis. All the tests were done using Stata version 16.  This flowchart shows patient selections for the study.  Out of 642,000 patients in the data mart for the study period,  there were about 154,000 that met the inclusion criteria.  Approximately 108,000 of these patients had received ICU telemedicine coverage, while just over 45,000 had no ICU telemedicine coverage during this day.  This slide summarizes the demographics and relevant characteristics of the patients in the  study. In summary, patients exposed to ICU telemedicine were statistically more likely to be  older and of non-white race. Also, they had more medical and pulmonary diagnosis, as well as more post-cadet care rest admissions, emergency admissions, and weekend admissions.  In the non-ICU telemedicine group, the day one Apache score was marginally higher, 60.5 versus  60.1. There was no difference in gender distribution between both groups. So, what did we find?  This slide shows our non-adjusted results first.  30-day mortality was significantly lower in the ICU telemedicine group,  compared to the non-ICU telemedicine group, by about 20%. This was also similar when we looked  at 30-day mortality rate per 1,000 patient days. In the Kaplan-Meyer survival analysis,  the ICU telemedicine group is represented by the red line, and the non-ICU telemedicine group  by the blue line. The log-rank test showed that patients exposed to ICU telemedicine were statistically more likely to be alive at 30 days, compared to those who were not exposed  to ICU telemedicine. As I mentioned, we had also performed a multivariable logistic regression analysis to assess clinical and operational factors that may be associated with 30-day  mortality. This forest plot summarizes the results of that analysis. As you can see, there were 12 covariates and two interaction times between them. At the top of the slide,  you can see that ICU telemedicine was associated with an odds ratio of 0.82 for 30-day mortality,  and the confidence interval was quite narrow at 0.1. There were five variables that were found  to be associated with increased 30-day mortality, non-white race, emergency admissions, cardiac  arrest, sepsis, and weekend admissions. Notably, though, when we analyzed the interaction between  ICU telemedicine and weekend admissions, we found a positive modification of that risk. The final odds ratio for mortality in that setting was 0.8 in the ICU telemedicine group.  We assessed the overall discriminative performance of the model using a receiver operating curve as shown on this slide. As you can see, the area under the curve was 0.77,  indicating that our model had quite good discriminative ability.  Secondary analysis also had similar findings. As shown on this slide, we can see that ICU length of stay and hospital length of stay were significantly lower in the ICU telemedicine  exposed group of patients. When we put all this together, our results support a statistically  significant association between ICU telemedicine exposure and 30-day mortality within a large  patient population. This risk reduction was seen in both adjusted and non-adjusted analysis.  Our results are directionally concordant with similar studies that have been published from  other centers, and it adds to the growing body of evidence that supports the benefits of ICU  telemedicine. However, there are limitations and cautions to keep in mind. The design was a retrospective pre-post design, and this can affect the validity of the results.  In particular, there might have been a measure of co-founders that could have affected the results.  Nonetheless, though, we think that this provides further reassurance about the value of ICU  telemedicine, particularly in the light of our collective experience in 2020 with the COVID-19  pandemic, during which period telemedicine in general has assumed greater permanence and utilization. That said, the mechanisms that underlie this observation of decreased 30-day  mortality does require further study, and we're looking into this. In conclusion,  in our large patient cohort from a single health system, exposure to ICU telemedicine appears to  be one of several clinical and operational factors that are associated with reduced 30-day mortality.  Thank you very much for your time and audience. I also reiterate my thanks to my co-investigators  and to our colleagues in the ICU telemedicine service, and I also want to thank our hospital administration for their support in setting up the program and continued support of the  program through these years. Thank you, and enjoy the rest of the conference.

Video Summary

Dr. Chieh Dozieh-Ude, Enterprise Medical Director of ICU Operations for the Cleveland Clinic, presented findings from a retrospective study on ICU telemedicine and its association with 30-day mortality. The study evaluated the outcomes of patients exposed to ICU telemedicine versus those who were not. The results showed that patients exposed to ICU telemedicine had a significantly lower 30-day mortality rate. Additionally, ICU length of stay and hospital length of stay were also lower in the telemedicine group. These findings support the benefits of ICU telemedicine in reducing mortality and improving patient outcomes. Further research is still needed to understand the mechanisms behind these outcomes.

Asset Subtitle

Administration, Research, 2021

Asset Caption

The Society of Critical Care Medicine's Critical Care Congress features internationally renowned faculty and content sessions highlighting the most up-to-date, evidence-based developments in critical care medicine. This is a presentation from the 2021 Critical Care Congress held virtually from January 31-February 12, 2021.

Chiedozie I. Udeh

Introduction/Hypothesis: ICU telemedicine (ICU-TM) has expanded in the U.S addressing critical care coverage demands. Evidence of its impact on clinical outcomes is growing. This study evaluated the association of ICU telemedicine exposure or non-exposure and relevant clinical factors with 30-day mortality.

Methods: This was a retrospective, cohort study of 153,987 consecutive ICU patients drawn from an institutional, ICU dataset for 9 hospitals with APACHE IV scoring, from 2014 to 2020. Analyses included summary statistics for demographics, 30-day mortality, multivariate logistic regression modeling, and survival analysis.

Results: For the entire cohort the unadjusted 30-day mortality proportion was significantly different between patients with ICU telemedicine (5.5%) or without ICU-TM (6.9%) with a risk ratio of 0.80 (95% CI 0.77, 0.84) (p <0.0001). Mortality rate for ICU-TM and no ICU-TM was 2.45/1000 versus 3.18/1000 patient-days respectively (p < 0.0001). Multivariate logistic regression modeling showed that ICU-TM exposure was associated with reduced 30-day mortality (OR 0.82, 95% CI 0.77, 0.87). In the final model, increased risk was seen with admission after cardiac arrest (1.42, 95% CI 1.26, 1.59), weekend admission (OR 1.29, 95% 1.18, 1.41), emergency admission (1.18, 95% CI 1.12, 1.24), race (non-white) (OR 1.11, 95% CI 1.05, 1.17), sepsis (OR 1.06, 95% CI 1.00, 1.12), day 1 APACHE score (OR 1.03, 95% CI 1.03, 1.03), and ICU LOS, (OR 1.01, 95% CI 1.01, 1.02). Risk reduction occurred with hospital LOS (OR 0.95, 95% CI 0.95, 0.96), surgical admission (OR 0.67, 95% CI 0.63, 0.72), coma (OR 0.48, 95% CI 0.36, 0.64) and 2 interaction terms (weekend admission with ICU telemedicine (OR 0.80, 95% CI 0.72, 0.90) and afterhours admission with ICU telemedicine (OR 0.78, 95% CI 0.73, 0.82)). The model has a c-statistic of 0.77. Secondary analyses showed that ICU and hospital length of stays were significantly reduced in the ICU telemedicine group (-1.6 days, 95% CI -1.5, -1.7) and -2.1 days (95% CI -1.9, -2.4), respectively.

Conclusions: In this large cohort, ICU telemedicine appears to be one of several factors associated with reduced 30-day mortality of ICU patients.

Meta Tag

Content Type Presentation

Knowledge Area Administration

Knowledge Area Research

Knowledge Level Intermediate

Knowledge Level Advanced

Membership Level Select

Tag Telemedicine eICU

Tag Outcomes Research

Year 2021

Keywords

ICU telemedicine

30-day mortality

patient outcomes

ICU length of stay

hospital length of stay

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.