Surveillance Beyond the ICU

Prevent serious health events and improve clinician experience through smarter, data-driven workflows and interventions

By John Zaleski, PhD, NREMT

COVID-19 has raised the general public’s awareness of healthcare functioning amid a pandemic. With more than 2.2 million cases confirmed in the U.S., approximately 120,000 deaths, and nearly 700,000 recovered, the intense focus on caring for and managing these patients has drawn much attention worldwide (John Hopkins University & Medicine Coronavirus Resource Center, n.d.). The evolving state of the virus, together with our understanding from the medical perspective, has caused changes in treatment and management of these patients. Whereas intense mechanical ventilation was the end state early on in treatment for a small fraction of unfortunate patients, the prevailing objective is now to keep patients from being intubated and placed on mechanical ventilation for as long as possible. The focus of much of the acute treatment has been in ICUs, although many health systems had evolved other units into makeshift ICUs, in preparation for an expected overwhelming patient intake. Reality has proven that the great overwhelm, in general and countrywide, has not come to pass, although there are hot spots of intense focus—principally in the major cities on the seaboards with dense populations.

While much of the focus has been on the ICU and the many monitoring and therapeutic patient care devices used to maintain and manage patients’ airway, breathing, and circulation (ABC), there is also a growing need outside of these high-acuity settings for monitoring patients. What the general public—and even hospital leadership—might not be aware of is that outside of critical care, there is a patient safety concern that could be growing, in part, due to clinical staffing pressures associated with COVID-19. In these non-critical units, patients with lower-acuity COVID-19 or other conditions may be at greater risk of suffering adverse events simply by being overlooked. These assertions are expressed in multiple ways, including reports of the need for emergency intubation and mechanical ventilation in more than 40,000 patients per year in the U.S. alone, and with interventions for opioid-induced respiratory depression estimated at $2 billion per year. Furthermore, estimates of complications associated with mechanically ventilated patients are estimated to exceed $1 billion annually (Wong, 2018).

Additional staffing in these units would help prevent these events, but it is not always feasible considering that hospitals face significant financial challenges, which have been exacerbated under COVID-19. Rather, clinicians need support in preventing adverse events through an enhanced clinical surveillance strategy that involves new data-driven workflows. Enhanced real-time clinical surveillance goes beyond monitoring physiologic parameters to detect emergent clinically actionable events and improve operational efficiency, enabling clinicians to safely care for larger numbers of patients.

More devices, less actionable information

Outside of the ICU, monitoring of vital signs can be episodic, roughly every four to six hours, depending on hospital or unit policies. Episodic data collection can result in gaps that do not capture key vital sign activity. If tachycardia or bradycardia occur, even for a sustained duration of 15 to 20 seconds, they may not be captured if discrete heart rate measurements are taken at one-minute or five-minute intervals.

Hence, when data collection is not continuous, events such as these are unknown to the clinician unless observed directly. Continuous data collection is important, particularly in lower-acuity settings, because the occurrence of “in-hospital respiratory failure [has been] associated with a mortality of nearly 40%” (Morris et al., 2017). This necessitates close surveillance of patients to catch respiratory decline, which in turn necessitates continuous monitoring.

Continuous capture of medical device data alone, however, will not solve this challenge, even when the devices are integrated through a single platform. A single clinician responsible for numerous patients could find it impossible to coherently interpret a stream of unfiltered data, especially with the multiple devices that many patients require. Such an unmanaged data flow would also contribute to nuisance alerts that feed into the alarm fatigue reported as a patient safety concern, as well as cognitive overload, which is another cause of clinician burnout.

With COVID-19, such alarms create unnecessary virus exposure risk for clinicians. Responding in person also requires clinicians to don personal protective equipment (PPE) before entering the patient’s room. Considering that PPE is still in short supply—and that demand will only increase now that hospitals are resuming elective procedures—wasting such gear on false alarms should be avoided.

Streamlining clinical surveillance

Instead of responding to every alert or concerning device output in person, hospitals need to be more strategic with their surveillance workflows. That requires clinicians in the ICU and in lower-acuity departments to have remotely available analytic tools to make sense of the stream of real-time patient data coming from devices. The data flow can be compared to historical EHR information to identify clinically relevant trends, sustained conditions, reoccurrences, and combinatorial indications to help clinicians predict health decompensation before it occurs, which can prevent harm and improve outcomes.

Along with analytics, clinicians should also be able to configure alerts so that they are relevant and clinically actionable. For example, in patients who receive postoperative opioid analgesia, remote alerts can be configured by taking a combination of vital signs into account and notifying clinicians only after a specific period of time has passed that indicates a worsening trend. A smarter alert such as this can reduce the alarm frequency without increasing risk to the patient. Clinicians can also respond to such alerts more fully informed about the patient’s status and prepared to intervene.

For hospitals using mobile secure messaging applications, these alerts should be integrated and bolstered with additional context, such as patient ID and location, trended data, device settings, and other measurements at the time the alert occurred. Mobile communication is crucial when surveilling patients with COVID-19 to minimize potential virus exposure while ensuring a rapid and prepared response to a critically ill patient.

Early detection saves lives

Whether for COVID-19 or any other condition, early detection of patient deterioration saves lives. Such early detection, however, is especially challenging outside of the ICU where the ratios of patients to clinicians are typically higher. Key advancements in real-time healthcare and advanced data analysis are driving continuous clinical surveillance and present a significant opportunity for hospitals and health systems to push patient safety initiatives into the realm of proactive healthcare, both during and after the pandemic.

John Zaleski is head of clinical informatics for Capsule Technologies.


John Hopkins University & Medicine Coronavirus Resource Center. (n.d.). COVID-19 dashboard by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU). Retrieved June 18, 2020, 6:32 a.m., from

Morris, T. A., Gay, P. C., MacIntyre, N. R., Hess, D. R., Hanneman, S. K., Lamberti, J. P., Doherty, D. E., Chang, L., & Seckel, M. A. (2017). Respiratory compromise as a new paradigm for the care of vulnerable hospitalized patients. Respiratory Care, 62(4), 497–512.

Wong, M. (2018, March 18). The case for continuous surveillance monitoring and real-time analytics. Healthcare Business Today.