— Closed ICU model demonstrates benefits longer-term after implementation
October 21, 2026 • 4 min read
Hospitals using remote intensivist staffing for their intensive care unit (ICU) saw shorter stays overall without negative safety signals, a longer-term follow-up study showed.
Overall mean hospital length of stay dropped by 1.7 days in the 12 months after the 6-month implementation and ramp-up period, about a 19% change compared with the 22 months prior (7.1 vs 8.8 days, P=0.032).
ICU length of stay remained stable at an average of 5 days, and ventilator utilization was unchanged as well, reported Chidinma Chima-Melton, MD, MBA, cardiopulmonary medical director at Pipeline Health in Los Angeles and founder of Tele-ICU, at the CHESTopens in a new tab or window meeting hosted by the American College of Chest Physicians.
The program used a closed ICU model, in which remote intensivists provided direct care and wrote orders with around-the-clock coverage, including structured night rounds, in conjunction with other physician specialties on site and the nursing team. Chima-Melton said she aimed to replicate "high-quality, high-touch critical care that you would get in person in a tertiary care center. I was thinking, 'How can we divorce that from location and actually bring it in through a different modality into a hospital that does not have an intensivist?'"
And that model might be why the results were positive, suggested Alice Gallo De Moraes, MD, a pulmonologist and critical care expert at the Mayo Clinic in Rochester, Minnesota, who was not involved in the study.
Tele-ICU has been tried under a variety of models, such as hospitals only calling in ad-hoc assistance as needed, just using it for monitoring alerts and equipment, or having only nurses and physician assistants in the ICU. Results have been mixed.
In the large TELESCOPEopens in a new tab or window trial of remote daily multidisciplinary rounds led by a board-certified intensivist through telemedicine and provision of evidence-based clinical protocols, ICU length of stay and all secondary endpoints remained unchanged after implementation across 30 ICUs.
Reducing length of stay not only translates into big dollar savings, but even showing no increase in risk is a positive as well, as it helps keep patients safely at their home hospital, Gallo De Moraes said.
"That's important to families. That's financially important too, because sometimes with transfer to another hospital, the patients have to pay for the ambulance ride, for example," she said. "I work at a tertiary health center and sometimes the patients are transferred to us and families stop visiting because they don't have the means to come, either financial or physical, like they don't have a vehicle."
The study included four safety-net hospitals in Los Angeles with a total of 62 ICU beds and 93.2% of admissions publicly insured. None of the hospitals used in-house intensivists before launch of the tele-ICU program in September 2024.
The tele-ICU had contiguous scheduling and structured bedside nursing presentations. Family meetings were held within 72 hours. The program had protocolized orders and provided nursing education.
Across the study period from November 2022 through February 2026, 6,386 ICU admissions occurred and were included in the retrospective cohort study. Analyses accounted for baseline mortality trends and facility differences.
Mortality in the ICU had been declining before implementation, trended upward during the tele-ICU implementation, and then was relatively flat during the post-ramp period.
The mortality odds ratios indicated no credible increase in risk, Chima-Melton noted. "That is reassuring, but it does not prove safety or establish a mortality benefit."
Still, she expressed optimism that mortality trends could improve as the program continues to mature at the hospitals, pointing to a Veterans Affairs studyopens in a new tab or window that did not show an impact on mortality until 5 years after implementation.
"Health systems are like cruise ships, and it takes a while to turn and actually start having them go in the right direction," she said.
The study was limited by its observational design without a concurrent control group. Residual confounding was possible, and uncertainty limits causal and safety claims, she added.
The study also couldn't determine what components of the program were more or less effective, although Chima-Melton said she speculated that one factor in particular was responsible for the reduction in overall length of stay without changing ICU duration: discharge planning.
"From the time the patient lands in the ICU, we're actually thinking about disposition plan, we're actually thinking about discharge plan for them," she said. "After they leave the ICU, they're not hanging around for days on the floor waiting to be discharged to the right facility, because we have that connection with the discharge planners, like case managers, really from the get-go."