Article

The Promise of Remote Patient Monitoring:

Lessons Learned During the COVID-19 Surge
in New York City
Paul N. Casale, MD, MPH1,2, Medha Vyavahare, MPH3,4, Shauna Coyne, BSc5,
Ian Kronish, MD, MPH6,7, Peter Greenwald, MD, MS8,9, Siqin Ye, MD7, Emme Deland, MBA3,
and Peter M. Fleischut, MD3

Abstract
The coronavirus pandemic catalyzed a digital health transformation, placing renewed focus on using remote monitoring
technologies to care for patients outside of hospitals. At NewYork-Presbyterian, the authors expanded remote monitoring
infrastructure and developed a COVID-19 Hypoxia Monitoring program—a critical means through which discharged
COVID-19 patients were followed and assessed, enabling the organization to maximize inpatient capacity at a time of
acute bed shortage. The pandemic tested existing remote monitoring efforts, revealing numerous operating challenges
including device management, centralized escalation protocols, and health equity concerns. The continuation of these
programs required addressing these concerns while expanding monitoring efforts in ambulatory and transitions of care
settings. Building on these experiences, this article offers insights and strategies for implementing remote monitoring
programs at scale and improving the sustainability of these efforts. As virtual care becomes a patient expectation, the
authors hope hospitals recognize the promise that remote monitoring holds in reenvisioning health care delivery.
Keywords
digital health, remote monitoring, COVID-19, care delivery

COVID-19 has rapidly reshaped how hospitals inter- equal footing with in-person patient interactions.
act with patients, making digital and telehealth inter- Although consumer demand and provision of tele-
actions a vital part of the delivery model.1 The current health services have grown exponentially, recent
crisis has accelerated the transition to a hybrid care analysis by McKinsey and Company suggests that the
delivery environment, in which virtual care exists on opportunity for virtualization remains vast. That
research estimates that up to 24% of all outpatient
1
Clinical Medicine and Population Health Sciences, Weill encounters can be virtualized, leveraging digital
Cornell Medical College, Adjunct Professor Columbia health tools to reimagine the office visit and redesign
University, NewYork-Presbyterian/Weill Cornell Medical Center,
home health services.2
New York, NY
2
Population Health, NewYork-Presbyterian Healthcare System NewYork-Presbyterian (NYP) is a comprehensive
Inc., New York, NY integrated academic health care system with 10 hos-
3
NewYork-Presbyterian Healthcare System Inc., Digital Health, pital campuses across the Greater New York area
New York, NY including Manhattan, Queens, Brooklyn, Westchester,
4
Harvard Medical School, Boston, MA
5
and Putnam counties. NYP is affiliated with 2 aca-
IS & Telehealth, NewYork-Presbyterian Heatlhcare System,
Inc., New York, NY demic medical centers, Weill Cornell Medical Center,
6
Center for Behavioral Cardiovascular Health. Columbia and the Columbia University Vagelos College of
University Irving Medical Center. New York, NY Physicians and Surgeons, and has >200 primary and
7
Columbia University Irving Medical Center, New York, NY specialty care clinics and medical groups providing
8
Weill Cornell Medicine. New York, NY
9
an array of telemedicine services.
Clinical Emergency Medicine, Weill Cornell Medicine. New
York, NY At NYP, virtual visits represented 53% of all out-
patient interactions at the peak of the COVID-19 cri-
Corresponding Author: sis, up from 2% before the pandemic (unpublished
Paul N. Casale, MD, MPH, Clinical Medicine and Population
data; NYP internal weekly digital health volume
Health Sciences, Weill Cornell Medical College, Adjunct
Professor Columbia University. NewYork-Presbyterian/Weill reports from July 2016 to April 2020). Remote moni-
Cornell Medical Center. New York, NY 10065-4870. toring, in particular, became a core component of the
Email: [email protected] COVID-19 patient management strategy. Remote
American Journal of Medical Quality 2021, Vol. 36(3) 139–144
© The Authors 2021
patient monitoring (RPM) is a mode of health care
DOI: 10.1097/01.JMQ.0000741968.61211.2b delivery that gathers and integrates patient data
140 American Journal of Medical Quality 36(3)

outside of traditional health care settings, allowing toward integration of medical records and building a
providers to track, assess, and engage patients regard- centralized escalation model in which a designated
less of location.3 Remote monitoring programs were team of providers managed system-wide alerts and
a critical means through which care teams could escalations across campuses. In the heart failure RPM
closely follow discharged COVID-19 patients and program, there was a renewed need to manage acute
assess changes in health status, enabling the organiza- exacerbations outside of the hospital during the pan-
tion to maximize inpatient capacity at a time of acute demic. Enterprise leadership met early on to stan-
bed shortage. Emergently, expanding remote moni- dardize clinical escalation protocols for all providers
toring has focused on equipping patients with new seeing these patients. Cross-EMR permissions were
tools in their homes while providing infrastructure established and an “escalation physician” was desig-
for care teams to respond to the information gath- nated to streamline the reporting infrastructure and
ered. As we look to the future, robust remote moni- manage all nurses and physician assistants communi-
toring programs will require equipping health systems cating with patients remotely. Standardization and
with an expanded capacity to manage care remotely leadership buy-in allowed for a more efficient and
and integrate remote monitoring within the broader scalable program that was able to continue offering
continuum of care. NYP significantly expanded services during the significant disruption of the
remote monitoring programs during the peak of the pandemic.
COVID-19 crisis, and this article offers insights and Device management was another challenge across
strategies for implementing remote monitoring pro- RPM programs. Although NYP partnered with device
grams at scale. vendors to manage inventory and deliver devices,
there was considerable progress to be made in inte-
grating the device ordering and inventory within the
Maturing Existing RPM Programs at NYP
organization’s own EMR. In scaling these programs
Before COVID-19, NYP successfully built remote during the pandemic, device visibility and availability
monitoring into digitally oriented care delivery mod- were paramount concerns. To better address this
els for heart failure and hypertension. Both programs issue, an order was created in the Epic health record
leveraged a mix of in-home devices and asynchro- to directly link device provision on discharge, simpli-
nous communication with nurses and physicians to fying the workflow for providers and increasing the
manage exacerbations. ability to track and manage device inventory
Development of these RPM programs hinged on in-house.
the ability to optimize device management, standard- Finally, RPM programs also faced inherent health
ize escalations protocols, and develop a coordinated equity challenges, exacerbated by the isolation of
escalations infrastructure. Inventory management some patient communities during the pandemic. The
was an early-stage priority, determining the appropri- requirement for a reliable and secure internet connec-
ate types of devices needed, creating device support tion to operate devices and share information can be
resources, and monitoring device security and con- a major barrier for some patients. Variable familiarity
nectivity to ensure proper use of these technologies. with these devices can make RPM implementation
As RPM programs expanded across sites, the second- more challenging in particular patient populations.
stage efforts focused on developing standardized To that end, patient education, translation of materi-
escalations protocols to ensure appropriate clinical als into multiple languages, and the provision of
criteria were met and all teams understood how and Bluetooth-enabled devices that did not require an
when escalations should take place. Finally, maturing internet connection to share information became part
RPM at NYP required investing in centrally housed of the broader strategy to make RPM more accessible
teams and the infrastructure to manage high escala- to patients, regardless of their health literacy and
tions volume across the system. circumstances.
Although COVID-19 accelerated the pace of
Addressing RPM Challenges During the change in RPM, the goal remained to develop an
infrastructure for real-time monitoring while creating
COVID-19 Pandemic
a seamless care environment spanning home and hos-
RPM programs across the enterprise faced numerous pital for all patients. Learnings were taken from the
challenges at the outset of the pandemic. The system expansion of existing RPM programs—improving
was in the midst of transitioning to a single electronic centralized escalation, creating seamless device
medical record (EMR) across all 10 campuses at the deployment orders, and creating solutions for patients
start of the pandemic, hindering initial progress with limited access to internet and other devices—to
Casale et al 141

design the COVID-19 Hypoxia Monitoring Program, monitored through this program (Figure 3). Early
a novel approach to COVID-19 patient management data from the largest volume inpatient and outpatient
utilizing the RPM infrastructure. sites are promising—of the patients monitored in the
inpatient and outpatient programs, 21 were sent to
the ED for monitoring and 33 patients were readmit-
Designing and Deploying the COVID-19
ted after monitoring.
Hypoxia Monitoring Program
During the COVID-19 crisis, the potential benefits of Impact of the COVID-19 Hypoxia
remote monitoring quickly became clear. As NYP
Monitoring Program
hospitals began responding to high volumes of
COVID-19 patients, a significant concern was the Rather than simply converting an in-person encoun-
potential for rapid deterioration of seemingly stable ter into a virtual encounter, the COVID-19 Hypoxia
patients. However, maintaining and expanding hospi- Monitoring Program and other RPM efforts repre-
tal capacity to respond to the influx of COVID-19 sent the second generation of digital heath initiatives,
patients also meant ensuring appropriate and timely providing patients with longitudinal care beyond the
discharge from inpatient units and the emergency visit-oriented delivery model. Creating these pro-
department (ED). To preserve patient safety, RPM grams during a time of crisis when many providers
efforts were refocused to respond to the needs of dis- were idle because of the cessation of elective services
charged COVID-19 patients by quickly prototyping allowed for new members of the clinical team to take
and implementing a COVID-19 Hypoxia Monitoring on a critical role in management while conserving
Program at scale. capacity and improving the targeting and delivery of
Specifically, COVID-19 patients discharged from scarce care resources. It brought together inpatient,
the ED and inpatient wards, as well as those from outpatient, and ED teams in a coordinated effort to
outpatient clinics who did not require immediate ED develop standardized workflows and algorithms for
referral, were carefully assessed for clinical acuity and remote monitoring, examining data, and responding
potential risk of decompensation. Depending on in real time. Ultimately, improving the integration of
symptomatology and degree of hypoxia, patients these services within clinical workflows was critical
were discharged with pulse oximeters and, if clini- to their scale and success, allowing conservation of
cally appropriate, oxygen concentrators to facilitate personal protective equipment, improved patient
recovery at home in a monitored setting (Table). In safety, and increased capacity of key services without
the midst of the crisis, a diverse team, including nurses compromising patient care.
and nurse practitioners, physician assistants, care However, as these programs grow, the organiza-
managers, and volunteer medical and nursing stu- tion recognizes the ongoing need to address usabil-
dents, was organized to monitor patient oxygen read- ity concerns and ensure that devices are relaying
ings remotely and assess any worsening symptoms. accurate information to providers while also
Physician assistants and medical students conducted addressing the challenge of building the infrastruc-
daily follow-up calls to enrolled patients for a 14-day ture needed to deliver data to physicians in an
period, monitoring COVID-19 symptoms, pulse, tem- actionable and centralized manner across all remote
perature, and oxygen saturation. Designated escala- monitoring programs. Ultimately, progress is being
tion physicians were alerted of any concerning made toward a future in which remote monitoring
readings, and directed patients to additional assess- programs are integrated within digitally oriented
ments via telemedicine or, if needed, readmission to care models, expanding the scope of the patient–
the ED. A detailed workflow and clinical assessment provider relationship.
criteria can be found in Figures 1 and 2, respectively.
During the initial surge, >2000 patients from the ED Lessons Learned
and >3650 inpatients and outpatients were
In scaling and building programs during the peak of
the epidemic, the approach to RPM was able to be
Table. Clinical Guidelines for Receiving Pulse Oximeter and
Oxygen Concentrator.
tested and modified, leveraging the strengths of both
clinical providers and automated systems to deliver
Oxygen saturation (on room air) Devices given at discharge
optimal patient care. Based on this experience, the fol-
≥95% No devices lowing recommendations are offered for health sys-
92%–94% Pulse oximeter only
90%–92% Pulse oximeter and oxygen concentrator
tems seeking to build on the virtual visit and expand
remote monitoring beyond the current pandemic.
142 American Journal of Medical Quality 36(3)

Figure 1. COVID-19 hypoxia monitoring program workflow.

Figure 2. COVID-19 hypoxia monitoring program clinical escalation criteria.

Effective Integration Allows Technology integral part of the care continuum, informing and
to Do the Work connecting patients and providers across care settings.
Early efforts of NYP’s heart failure and other home
Remote monitoring services should be embedded monitoring programs demonstrated the challenges of
within the care model. Rather than approaching managing these programs separately; an overreliance
remote monitoring as a separate modality with a dis- on in-person follow-up visits with nurse practitioners
parate management structure, it can become an and disjointed enrollment and escalation processes
Casale et al 143

Centralize Device Management and

Support
The maturity of remote monitoring programs hinges
on the development of a centralized device manage-
ment plan. To appropriately and efficiently deliver,
monitor, and recoup devices from patients enrolled in
these programs, hospitals will need to develop proto-
cols for enrolling patients, tracking devices, and
ensuring privacy and security of patient information.
Figure 3. Patient volume in COVID-19 hypoxia monitoring
program (April 17, 2020–present). NYP has recognized that inventory management
requires strong vendor partnerships, and a clear del-
egation of responsibility for device recovery. Device
made coordination across programs challenging.
management also will require working internally and
RPM programs must be designed thoughtfully to pro-
with vendors to provide technical support, servicing,
vide relevant streams of data, work with a leaner in-
and education to patients to ensure data accuracy
person workforce, and provide timely information to
and ease of use.
providers in the medical record. The benefits of
remote monitoring can be fully realized when RPM
data do not simply replace an in-person touch point, Identifying Populations Who Benefit
but when these data are used as part of the treatment Most Requires Equity-Centered
plan and data trends can inform patient care. Design
Integration will require building in postdischarge
workflows, automated escalations infrastructure, and As remote monitoring programs scale, identifying
device delivery protocols that are standardized and populations for whom RPM will deliver value is criti-
centrally managed. As the EMR increasingly becomes cal to its success. Ideally, remote monitoring pro-
a shared tool between patients and providers, RPM grams can build on existing devices in patients’
data will not only deliver insights to providers about homes; however, these programs run the risk of exac-
how patients’ conditions change over time but it also erbating existing disparities if equity is not consid-
will empower patients to use this information to ered in program design. Delivering devices to patients
coproduce health. most in need and addressing logistical issues like
access to Wi-Fi can help ensure these programs
address barriers to care access. Digital literacy should
Coordination Across Providers not be a determinant of RPM success—rather, pro-
Boosts Engagement vider organizations must recognize and meet the
Remote monitoring is a powerful tool for diverse care needs of patients who require additional support. To
settings, including transitions of care and outpatient ensure equity and value of these programs, reaching
chronic disease management. Making coordination out to all stakeholders, applying user-centered design
across providers a priority—informing and involving to understand barriers, and clearly communicating
primary care teams in escalations management and fol- program motivations can foster uptake by physicians
low-up alongside specialists—improves the value of and patients. Given the cost of device management,
these services. Engaging social work teams, care manag- effective targeting and a clear understanding of how
ers, and community health workers in follow-up signifi- RPM data can influence clinical decision-making will
cantly broadened the reach and ensured that the care maximize the impact of these resources on patient
team was familiar with local resources while signifi- outcomes.
cantly reducing care fragmentation. Coordinating pro-
viders was achieved, in part, by shared accountability Sustainability Depends on a
for training, education, and program management. Supportive Policy Environment
Rather than creating implementation silos, the most
successful remote monitoring programs ensured that all The sustainability and scale of remote monitoring
stakeholders were engaged and trained to maximize the programs is partly contingent on a reliable reimburse-
utility of the data. As the RPM programs achieve scale, ment infrastructure. The Centers for Medicare &
NYP seeks to further centralize staffing to improve the Medicaid Services recently expanded the physician
efficiency and sustainability of these programs. fee schedule to include billable codes for remote
144 American Journal of Medical Quality 36(3)

monitoring services, spurring considerable growth in Conflicts of Interest

the delivery of RPM services.4 However, variability in The authors have no conflicts of interest to disclose.
RPM reimbursement in the private sector persists and
considerable uncertainty remains about qualifying Author Contributions
parameters, devices, and types of data. Disparate cov-
Coyne, Drs Casale, and Fleischut contributed to the con-
erage policies can complicate the administration of
ception and design of the manuscript. Vyavahare and Dr
these programs and impact patient communication Casale were involved in drafting the manuscript, gathering
and experience. As RPM becomes a mainstay of vir- and interpreting data, and making revisions. Deland and
tual care, reimbursement should be informed by fur- Drs Kronish, Greenwald, and Ye provided critical revi-
ther characterization of how RPM delivers value and sions and guidance on the development of the piece.
impacts preventable admissions, ED visits, and
chronic care management.5 Appropriate reimburse- Funding
ment policy that accommodates the diversity and
This research received no specific grant from any funding
evolution of remote monitoring programs will be agency in the public, commercial, or not-for-profit sectors.
critical to future growth.
References
Conclusion 1. Hollander JE, Carr BG. Virtually perfect? Telemedicine
When the dust settles from COVID-19, there will be for Covid-19. N Engl J Med. 2020;382:1679–1681.
2. Bestsennyy O, Gilbert G, Harris A, Rost J. Telehealth:
an enormous opportunity to identify which remote
a quarter-trillion-dollar post-COVID-19 reality?
monitoring services and technologies deliver the most
McKinsey and Company. 2020. Accessed October 6,
value to patients and providers. Even when the capac- 2020. https://www.mckinsey.com/industries/health-
ity burden is lifted, hospitals should recognize that care-systems-and-services/our-insights/telehealth-a-
the ground beneath them has shifted. Many predict quarter-trillion-dollar-post-covid-19-reality.
that virtual care enabled by RPM tools will become 3. Center for Connected Health Policy. Remote Patient
an expectation as patients become more familiar with Monitoring (RPM). Accessed September 25, 2020.
the full range of telemedicine services that can be https://www.cchpca.org/about/about-telehealth/
offered. Hospitals should build on the learnings of remote-patient-monitoring-rpm.
this crisis to evolve and recognize the promise that 4. Centers for Medicare & Medicaid Services, Department
remote monitoring programs hold in reenvisioning of Health and Human Services. Medicare telemedicine
health care provider fact sheet. Accessed January 6,
health care delivery.
2021. https://www.cms.gov/newsroom/fact-sheets/
medicare-telemedicine-health-care-provider-fact-sheet.
Acknowledgments 5. Noah B, Keller MS, Mosadeghi S, et al. Impact of
The authors would like to acknowledge Amelia Shapiro, remote patient monitoring on clinical outcomes: an
MBA, Director, Strategy, NewYork-Presbyterian updated meta-analysis of randomized controlled trials.
Healthcare System, for contributions to this article. NPJ Digital Med. 2018;1:20172.