The Society of Critical Care Anesthesiologists

COVID-19: The Rise of the Machines

Artificial intelligence (AI) is a revolutionary inclusion to current medical practice, particularly in the field of Critical Care medicine. The actualization of AI is empowered with the intellectual acumen to analyze and generate vastly complex data, as well as integrate from experience, based on recognition of formed patterns. This is achieved through machine learning and perception, compounded by natural language processing, and automated simple repetitive tasks or exposure. These are attributes or skills born from the innovative integration of human intelligence and computer systems. Thus, allowing for a multitude of decision-making and task-executing functions. Critical Care widely encompasses an intricate and detail-oriented multidisciplinary approach in diagnosing diseases and analyzing their progression with simultaneous treatment interventions. While the established conventional approach to medical management has proven to be undoubtedly reliable, it is also unremarkably challenging and time-consuming to medical professionals in the face of disease progression and overwhelming workload. The field of Critical Care establishes multiple landscapes and opportunities in which AI is a transformative asset to researchers and Critical Care physicians alike.

As standards of medical care have been continually refined, the clinical outcomes of critically ill patients have improved exponentially. However, the traditional Critical Care approach still maintains limitations in terms of comprehending the complexities of acuity, dealing with significant individual heterogeneity, forecasting deterioration, and delivering early treatment techniques prior to decompensation (Yoon et al., 2022). Artificial intelligence, along with its conceptual role in Critical Care, has the capacity to, not only guide clinical decisions, but to accurately diagnose detected diseases and evaluate their predicted outcomes. While the concept of merging such systems in improving patient care is in emergence, several studies on its application have been published, further examining its validity, with the viewpoint to allow for rapid intervention for such patients. This early assisted detection is crucial to the management of critically ill patients in the ICU, as deterioration of their clinical status can either be insidious, rapid, or even a combination of both.  Through pattern recognition from complex data, AI can, not only detect diseases but recognize the many phenotypes they can manifest as well. Not to mention, the use of electronic health records conjugated with AI to predict mortality of certain diseases. An example highlighted in the 2022 study by Yoon et al., shows that excess alveolar fluid cannot be presumed to be the cause of pulmonary infiltrates, as the underlying etiology could be cardiac, infectious, or even trauma related. However, timely management tends to be delayed, in the face of limited clinical context and ongoing investigations. The role of AI in this clinical scenario can assist in prompt focused diagnosis, through advanced text and image processing capabilities. This collaborative approach toward patient care, allows for medical doctors to manage patients more efficiently in the Critical Care unit.

Another aspect of the use of technology in Critical Care is Telehealth and Telemedicine. Telehealth refers to the delivery of healthcare service via remote use of telecommunications technology. Telemedicine, a subset of telehealth, is the remote diagnosis and treatment of patients by means of telecommunications technology. It aims to permit real-time, two-way interaction between physician and patient from a distance. With its inception dating as far back as the Ancient Egyptian era, this prominent concept was first employed in scrolls disseminated to apprise others of various health-related events such as diseases (Kichloo et al., 2022). Along with its distinguished past, telemedicine has made profound strides in modern-day medicine as well, with its most recent evolution occurring amidst the COVID-19 pandemic. Implemented extensively in what is known as ‘forward triage,’ telehealth and telemedicine have accommodated the screening of patients while in a state of quarantine, both containing the spread of a prospective virus, and preserving the health of the physicians and other health-care providers. Such instances can also be conducted via remote monitoring, which is devoid of real-time interaction. Supplementary to remote monitoring and interactive patient care, telemedicine also provides the opportunity to ‘store-and-forward.’ This is the process of collecting clinical information and sending it to another location for evaluation. Demographic data, medical history, documents such as lab reports, images, video, and/or sound files are all common types of information utilizing this process. Furthermore, the principle of telemedicine is intimately intertwined with that of an AI relationship that has also come into fruition amidst the COVID-19 pandemic. In addition to remote screening, the collusion of telehealth and AI has paved the path for improved protein structure prediction, therapy monitoring, awareness, social control, and digital health (El-Sherif et al., 2022).  Lastly, AI in telemedicine offers expanded healthcare access, less exposure to diseased patients, and the preservation of supplies.

Intensive care unit telemedicine (Tele-ICU) refers to a technology-based apparatus that employs artificial intelligence to deliver effective Critical Care from remote locations, in response to addressing the increasing patient load and shortage of intensivists. Combined with highly qualified and experienced Critical Care staff, tele-ICU provides remarkable improvements to patient survival and quality care. Risk prediction algorithms, smart alarm systems and machine learning tools augment conventional coverage and can potentially improve the quality of care as Critical Care physicians are able to access patient data, off-site and efficiently implement clinical decisions through the tele-ICU (Khurrum et al., 2021). Despite said advantages, acceptance by healthcare professionals, as well as markedly expensive costs to maintain and operate systems have both proven to be limiting factors to the seamless integration of Tele-ICU into global healthcare.

The COVID-19 pandemic created a perilous environment for the healthcare sector, affecting patients from the general population as well as medical professionals and available resources. Hence, it motivated the undeniable need for the delivery of digital medical services in the form of diagnostic, therapeutic, and prognostic models. This provided a unique opportunity for further integration and advancement of AI use in Critical Care management. The burden for Critical Care services has significantly increased compared to patients requiring ICU admissions and mechanical ventilation before the pandemic (Jansson et al., 2020). This inevitably propelled the extensive revision of the quality and extent of health care readily offered, allowing for the inculcation of AI into practice. AI assisted diagnostic interventions for a variety of clinical presentations has been integrated to facilitate COVID-19 detection through radiological and biochemical techniques, combined with advanced contact tracing. Lesions on chest x-rays and CT scans are better identified via methods assisted by AI. Artificial intelligence algorithms have also been devised to predict COVID-19-related mortality, along with need for ICU admission, and to determine the most suitable drug or treatment option for patients based on their unique clinical profiles. According to Jansson et al. (2020), AI accommodating ventilator adjustments allowed for the improvement of time-sensitive decisions and negated the risk of COVID-19 exposure in Critical Care professionals, as well as reduced the costs of personal protective equipment and labor. Not to mention, the application of deep learning systems can accommodate the rapid detection of COVID-19 via RT-PCR, thus reducing the time to diagnose COVID-19, aiding in the management of insufficient isolation-bed resources, and adequately accommodating critically ill patients (Lee et al., 2022). 

It is evident that AI in medicine, a still rudimentary yet promising concept, has the potential to elicit revolutionary strides in the medical world. To not only summarize the aforementioned, but also emphasize the profound impact that AI has started to have on medicine, a brief synopsis of its benefits. Firstly, and arguably most importantly, the standardization of AI in Critical Care could prove to be highly instrumental in identifying diseases, predicting the evolution of diseases, categorizing diseases into their corresponding phenotypes, and guiding clinical decisions (Yoon et al., 2022). Secondly, both telehealth and tele-ICU could also be revamped, with more thorough and efficient data analysis and collaboration, providing more accurate diagnoses, protecting physicians from burn-out, and enhanced convenience in monitoring patients remotely. Lastly, AI has shown to be propitious in addressing the most recently established pillar of modern medicine: battling the COVID-19 pandemic. Artificial systems aid in the quick and effortless diagnosing of COVID-19 patients as well as focusing on drugs that could combat this illness through AI screening and predicting future variants of the virus.

Even though AI promises outstanding refinements in modern medical practices and procedures, it has impediments that, if not catered to, may hinder the seamless integration of this “other-worldly” concept into current practices. One such fear is that AI systems are overly complex, therefore reducing interpretability. This would require training of medical personnel which may be cumbersome. Additionally, AI technology is still rather new and rife with unanswered questions, which can only be clarified by further trials and experimentation. Supplementary to the above, ethical concerns have also risen, AI automation could slowly replace human-oriented jobs, resulting in the obsolescing of educated medical professionals. Given that these encumbrances are slowly but surely surmounted, we feel artificial intelligence will be ubiquitous in modern medicine. Above all, fulfilling the emotional needs of the patients, the expectation of human touch and “treating the patient, not the disease” will continue to be important in the field of medicine.

Artificial intelligence is far from perfect, but its reliance on the fundamental building blocks of human intelligence, namely learning, reasoning, problem solving, perception and communicative skills, may be programmed in a way that is refreshingly free from the behavioral detritus of human judgment. While the conventional approach to medical management is well-grounded and dependable, it remains with its known limitations. To not only surmount the current plateau of medical advancements but to go far beyond it, the integration of artificial intelligence in modern medicine, particularly Critical Care, is pivotal.


El-Sherif, D. M., Abouzid, M., Elzarif, M. T., Ahmed, A. A., Albakri, A., & Alshehri, M. M. (2022, February 18). Telehealth and artificial intelligence insights into healthcare during the COVID-19 pandemic. Healthcare (Basel, Switzerland). Retrieved April 30, 2022, from

Jansson, M., Rubio, J., Gavaldà, R., & Rello, J. (2020, December). Artificial Intelligence for clinical decision support in critical care, required and accelerated by covid-19. Anaesthesia, critical care & pain medicine. Retrieved April 30, 2022, from

Khurrum, M., Asmar, S., & Joseph, B. (2021). Telemedicine in the ICU: Innovation in the Critical Care Process. Journal of intensive care medicine, 36(12), 1377–1384.

Kichloo, A., Albosta, M., Dettloff, K., Wani, F., El-Amir, Z., Singh, J., Aljadah, M., Chakinala, R. C., Kanugula, A. K., Solanki, S., & Chugh, S. (2020, August). Telemedicine, the current COVID-19 pandemic and the future: A narrative review and perspectives moving forward in the USA. Family medicine and community health. Retrieved April 30, 2022, from

Lee, Y., Kim, Y.-S., Lee, D.-in, Jeong, S., Kang, G.-H., Jang, Y. S., Kim, W., Choi, H. Y., Kim, J. G., & Choi, S.-hoon. (2022, January 24). The application of a deep learning system developed to reduce the time for RT-PCR in COVID-19 detection. Nature News. Retrieved April 30, 2022, from

Yoon, J. H., Pinsky, M. R., & Clermont, G. (2022, March 22). Artificial Intelligence in critical care medicine - critical care. BioMed Central. Retrieved April 30, 2022, from


Vinodkumar Singh, MD
University of Alabama Medicine
Birmingham, Alabama
Ayesha Bryant, MD, MSPH
University of Alabama Medicine
Birmingham, Alabama
Meghana Muppuri, MBBS
Cornwall Regional Hospital, Jamaica
Montego Bay, Jamaica
Arnav Muppuri
Cornwall College, Jamaica
Montego Bay, Jamaica