How AI is fueling the new Tele-ICU

February 21, 2022
Article written by Equum Medical

Critical Care Quality and Outcomes Metrics have benefitted from Tele-ICU programs, but now AI is adding greater efficiency from scheduling to proactive intervention.

In the COVID-19 era, Intensive Care Unit (ICU) patient capacities have ranged from 52% up to 289% during surge conditions1.  ICU patients represent the most critically ill acute care population with mortality rates as high as 41% in the COVID-19 era2.  Overall, medical care in The United States accounts for ~17% of gross domestic product with ICU care accounting for up to 4.1% of all healthcare spending3.  Devising methods to improve patient outcomes and maximize efficiency in care delivery is of paramount importance.  Tele-ICU has been used to reduce mortality and improve patient safety outcomes4-6.  

Two prevailing models for Tele-ICU delivery are “The Centralized Monitoring Model” (CCM) and the “Virtual Consultant Model” (VCM)7.  The continuous CCM is considered an active monitoring modality whereby a coordinated team of nurses, doctors and clerical staff are fed a continuous stream of fully integrated data.  The episodic VCM episodic form where patient care support is on demand via an audiovisual device aimed at tele-consultation.  Once integrated, the Tele-ICU can be compared to that of a central nervous system where sensory inputs (ie nurse/physician/pharmacy patient review, alert reporting, vital signs, lab values) are streamed to the Tele-ICU team where that data can be analyzed in real-time.  The Tele-ICU team can act on the data monitoring to then act on patient needs as they arise.  The Tele-ICU team, assisted by technology, can detect an issue with a patient and allow the healthcare provider to treat that patient in real-time.  

Artificial intelligence and specifically machine learning applications have been used in medicine as clinical prediction tools and severity of illness scores for decades but are cumbersome without computer automation8,9.  Classic modelling techniques such as logistic and linear regression are being surpassed in accuracy by models employing tree-based or neural-network methodologies10.  Tele-ICU systems have the potential to incorporate machine learning tools to predict severity of illness in real time as well as create benchmarking systems to find ICU population-wide deficiencies11.  Identification of problems which can be fixed within the ICU can lead to the most pronounced improvement in outcomes5,6.  It can be difficult to know what deficiencies might be present in an ICU without looking specifically.  Evidence-based and actionable determinants of successful patient outcomes should be identified as targets for improvement.  Machine learning classifiers and regressors can be programmed to automatically compare your ICU performance to that of peers and find the most clinically meaningful deficiencies.  Once identified, sophisticated Tele-ICU programs can institute best practices to address those deficiencies and improve overall ICU quality.

The future of medicine is going to be largely driven by use of technology to improve patient care outcomes and to maximize efficiency of healthcare delivery.  As healthcare costs continue to grow, methods for improving healthcare efficiency will be more and more important.  Tele-ICU is one technology which if properly implemented can help to improve both patient care outcomes and care delivery.  Artificial intelligence can be an important tool for creating approaches to identify and improve upon actionable outcomes which lead to improved patient care overall.

Dr. Mario Fusaro, MD serves as the Chief Innovation Officer at Equum Medical and helps to bridge the gap between clinical excellence and technologic innovation. Dr. Fusaro received a degree in Chemistry with honors from Pennsylvania State University before attending Temple University School of Medicine for his medical degree.  His training in Internal Medicine was completed at NYU School of Medicine and then at The University of Maryland School of Medicine for Pulmonary Diseases and Critical Care Medicine.  He holds a master’s degree in Applied Health Economics and Outcomes Research from Thomas Jefferson University and is the author of over 35 peer-reviewed publications, book chapters, and podcasts.


  1. Douin DJ, Ward MJ, Lindsell CJ, et al. ICU bed utilization during the coronavirus disease 2019 pandemic in a multistate analysis-march to june 2020. Crit Care Explor. 2021;3(3):e0361. doi: 10.1097/CCE.0000000000000361 [doi].
  2. Armstrong RA, Kane AD, Cook TM. Outcomes from intensive care in patients with COVID-19: A systematic review and meta-analysis of observational studies. Anaesthesia. 2020;75(10):1340-1349. doi: 10.1111/anae.15201 [doi].
  3. Society for Critical Care Medicine. Critical care statistics . 
  4. Becker CD, Fusaro MV, Al Aseri Z, Millerman K, Scurlock C. Effects of telemedicine ICU intervention on care standardization and patient outcomes: An observational study. Crit Care Explor. 2020;2(7):e0165. doi: 10.1097/CCE.0000000000000165 [doi].
  5. Fusaro MV, Becker C, Miller D, Hassan IF, Scurlock C. ICU telemedicine implementation and risk-adjusted mortality differences between daytime and nighttime coverage. Chest. 2021;159(4):1445-1451. doi: S0012-3692(20)35108-4 [pii].
  6. Fusaro MV, Becker C, Scurlock C. Evaluating tele-ICU implementation based on observed and predicted ICU mortality: A systematic review and meta-analysis. Crit Care Med. 2019;47(4):501-507. doi: 10.1097/CCM.0000000000003627 [doi].
  7. Ramnath VR, Ho L, Maggio LA, Khazeni N. Centralized monitoring and virtual consultant models of tele-ICU care: A systematic review. Telemed J E Health. 2014;20(10):936-961. doi: 10.1089/tmj.2013.0352 [doi].
  8. Zimmerman JE, Kramer AA, McNair DS, Malila FM. Acute physiology and chronic health evaluation (APACHE) IV: Hospital mortality assessment for today's critically ill patients. Crit Care Med. 2006;34(5):1297-1310. doi: 10.1097/01.CCM.0000215112.84523.F0 [doi].
  9. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: A severity of disease classification system. Crit Care Med. 1985;13(10):818-829.
  10. Syed M, Syed S, Sexton K, et al. Application of machine learning in intensive care unit (ICU) settings using MIMIC dataset: Systematic review. Informatics (MDPI). 2021;8(1):10.3390/informatics8010016. Epub 2021 Mar 3. doi: 16 [pii].
  11. Badawi O, Liu X, Hassan E, Amelung PJ, Swami S. Evaluation of ICU risk models adapted for use as continuous markers of severity of illness throughout the ICU stay. Crit Care Med. 2018;46(3):361-367. doi: 10.1097/CCM.0000000000002904 [doi].

Connect with Equum

Lead Form 2 Equum Medical

Related Posts