key: cord-1043163-9vgrvuv7
authors: Michard, Frederic; Saugel, Bernd; Vallet, Benoit
title: Rethinking the post-COVID-19 pandemic hospital: more ICU beds or smart monitoring on the wards?
date: 2020-07-01
journal: Intensive Care Med
DOI: 10.1007/s00134-020-06163-7
sha: d97b0f8fe5164d14f7540a510f5684f83f3274e4
doc_id: 1043163
cord_uid: 9vgrvuv7

nan

equipped their wards with wireless pulse oximeters to cope with that surge.

On hospital wards, the nurse/patient ratio is much lower than in ICUs and clinicians are not used to deal with life-threatening clinical situations. Therefore, we cannot envision to simply acquire existing ICU monitoring systems to install them on the wards. We have to implement systems specifically designed for the wards. Such systems are currently under development or evaluation. They usually include wearable sensors which are easy to set up (e.g., adhesive patch on the skin) and enable wireless monitoring, which, in the context of epidemics, has the advantage to decrease the number of physical interactions and the risk of contamination. Artifacts and false alarms, common in non-sedated patients, need to be filtered, and machine learning algorithms may help to achieve that goal [5] . Medical-grade connectivity protocols are necessary to prevent connectivity disruptions, common with classical Bluetooth connections. Smart algorithms are also needed to aggregate and fuse vital signs into a single early warning score or deterioration index so that nurses can only respond to "smart" alerts received at a central station or directly on their pager [4] . By doing so, one may envision the possibility to ask ward nurses to look after a large number of patients without increasing their workload [2] . Indeed, focusing on patients who are deteriorating, as soon as they do, should improve work efficiency. However, this remains to be confirmed by large clinical studies done in real-life conditions. Finally, to be useful, such remote monitoring systems need to direct the alert to the right person. For instance, if ward nurses could be informed first in case of mild deterioration, rapid response teams would need to *Correspondence: frederic.michard@bluewin.ch 1 MiCo (michardconsulting.com), Denens, Switzerland Full author information is available at the end of the article be alerted directly in case of life-threatening clinical condition requiring immediate intervention.

Therefore, in the hospital of the future, ICU beds could be reserved to patients requiring organ support such as mechanical ventilation and renal replacement therapy. Patients who would simply need close monitoring could safely stay on the wards and benefit from smart and continuous monitoring (Fig. 1) . This strategy may help to improve patient safety and quality of hospital care without dramatically increasing the number of ICU beds and associated costs. 

Conflicts of interest FM is the founder and managing director of MiCo (micha rdcon sulti ng.com), a Swiss consulting firm that provides advisory and research services to medtech companies, some of them developing mobile monitoring solutions. MiCo does not sell any medical products and FM does not own shares from any medtech company. BS has received honoraria for consulting, honoraria for giving lectures, refunds of travel expenses or institutional restricted research grants from Edwards Lifesciences (Irvine, CA, USA), Pulsion Medical Systems (Feldkirchen, Germany), CNSystems (Graz, Austria), Retia Medical (Valhalla, NY, USA), Philips (Böblingen, Germany), and Tensys Medical (San Diego, CA, USA). BV has no conflicts of interest to declare. Fig. 1 On hospital wards, continuous monitoring of vital signs with wearable and wireless sensors may enable the early detection of clinical deterioration, timely intervention, and the prevention of severe adverse events such as cardiac arrest, ICU transfer, and death

ICU beds: less is more? No

The rise of ward monitoring: opportunities and challenges for critical care specialists

Beyond "failure to rescue": the time has come for continuous ward monitoring

Effect of an automated notification system for deteriorating ward patients on clinical outcomes

Using supervised machine learning to classify real alerts and artifact in online multisignal vital sign monitoring data

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Accepted: 22 June 2020