Rationale and Design of Healthy at Home for COPD: an Integrated Remote Patient Monitoring and Virtual Pulmonary Rehabilitation Pilot Study

Abstract Chronic Obstructive Pulmonary Disease (COPD) is a common, costly, and morbid condition. Pulmonary rehabilitation, close monitoring, and early intervention during acute exacerbations of symptoms represent a comprehensive approach to improve outcomes, but the optimal means of delivering these services is uncertain. Logistical, financial, and social barriers to providing healthcare through face-to-face encounters, paired with recent developments in technology, have stimulated interest in exploring alternative models of care. The Healthy at Home study seeks to determine the feasibility of a multimodal, digitally enhanced intervention provided to participants with COPD longitudinally over six months. This paper details the recruitment, methods, and analysis plan for the study, which is recruiting 100 participants in its pilot phase. Participants were provided with several integrated services including a smartwatch to track physiological data, a study app to track symptoms and study instruments, access to a mobile integrated health program for acute clinical needs, and a virtual comprehensive pulmonary support service. Participants shared physiologic, demographic, and symptom reports, electronic health records, and claims data with the study team, facilitating a better understanding of their symptoms and potential care needs longitudinally. The Healthy at Home study seeks to develop a comprehensive digital phenotype of COPD by tracking and responding to multiple indices of disease behavior and facilitating early and nuanced responses to changes in participants’ health status. This study is registered at Clinicaltrials.gov (NCT06000696).


INTRODUCTION
2][3] This condition, characterized by chronic symptoms such as shortness of breath, coughing, and mucus production, leads to frequent episodes of acute exacerbation. 4[12] Early detection and treatment of acute COPD exacerbation have been shown to improve patient quality of life and decrease the need for emergency services utilization and hospital admission. 10Furthermore, interventions upstream of an index admission for COPD are more effective than interventions during hospital admission or post-discharge, to prevent further admissions. 9Put another way, the best way to prevent COPD readmission is to prevent index admission. 9However, uncertainty remains about the optimal means of predicting and intervening upon impending COPD exacerbations.
4][15] However, the burden of daily symptoms and vital sign collection is onerous, and real-world evidence has demonstrated limited adherence to active vital sign collection that requires patient participation.Alternatively, passive collection of biometric data such as oxygen saturation and activity levels minimally adds to patient burden and has been shown to have prognostic value for acute care needs. 16Further validation of the utility of biometric data signals as clinically actionable alerts for patients and their providers in real-world practice is needed.Additionally, early antibiotic and steroid therapy have been shown to improve both outcomes and patient quality of life but to initiate treatment, clinicians must be alerted to clinical deterioration and be empowered to intervene promptly. 10,17Therefore, in tandem with the optimization of monitoring strategies, a systematic approach to review and respond to alerts based on consolidated data streams is needed to prevent added burden for the patients and their medical team.The present study has been designed to leverage ongoing efforts to con gure a multimodal virtual care intervention for COPD patients and evaluate the feasibility of its implementation.Empowering patients with tools to help them expedite the recognition of worsening symptoms and removing barriers to care access are two critical components of improving clinical outcomes and patient quality of life.
The objective of the current study is to conduct a feasibility trial of the "Healthy at Home" program, a multimodal intervention aimed at preventing emergency services utilization and hospitalization among patients living with COPD.The program features a digital platform linked to a wearable health-monitoring device, virtual pulmonary support to facilitate complex care management, and access to an around-theclock mobile integrated health (MIH) service that provides on-demand, in-home visits for evaluation and treatment of acute symptoms.

Setting and Participants
This study was performed at an urban, academic tertiary care center.Inclusion criteria for participation included receiving healthcare through the hospital system of interest, being 18 years of age or older at the time of recruitment, and carrying a diagnosis of COPD.Subjects were also required to have access to a smartphone (iPhone or Android) with internet access to download and use the study apps and to live within the regional geographic area served by the system's MIH program.Those who lacked the capacity to consent, did not understand English, did not have internet access, were enrolled in another investigational clinical trial at the time of enrollment, or were (or had previously been) enrolled in any Wellinks pulmonary support program were excluded from the study.The study was approved by the WIRB-Copernicus Group Institutional Review Board and is registered at Clinicaltrials.gov(NCT06000696).

Recruitment
Screening for participant eligibility was performed via a query of the system's electronic health record (EHR) for patients with a diagnosis of COPD.The number of ED visits and hospitalizations, along with the number of COPD-related medication changes in the two years before study enrollment were considered proxy variables for COPD severity and treated as a count variable.We re ned our recruitment to those patients that were within the second through fourth quintile of the count variable such that our cohort was comprised of patients with moderate-severity COPD.On a per-case basis, exceptions to this strategy were accommodated when patients were directly referred to the research team by the clinical team.
All participants were initially invited via email.Non-response to email solicitation was followed by two text messages within two weeks of the email invitation and a paper mailer sent to prospective participants' homes.Patients who were identi ed as prospective participants and who were hospitalized or had upcoming ambulatory pulmonary clinic appointments were approached in person by the study team at the medical center.Finally, yers advertising the study were placed in high-visibility places such as clinic waiting rooms.The recruitment strategy and number of participants approached are summarized in Fig. 1.Invitations for recruitment were conducted in waves ranging from 500-3,000 patients based on the composition of the study cohort after each cycle to achieve a balanced representation of risk categories and sociodemographic factors.

Procedure
Participants who expressed interest in participating were given instructions to download the study app, a custom interface created through the MyDataHelps platform (Supplemental File 1).MyDataHelps is a mobile application created by CareEvolution to give a platform for researchers and participants to engage in clinical research.Through this study app, participants completed eligibility questions to ensure that they met all inclusion criteria for the study.If they were determined to be eligible, participants were subsequently prompted to digitally review and sign an informed consent form through the study app.Because eligibility screening and enrollment could be performed through the app, participants could enroll entirely remotely.
Once participants completed consenting procedures, a welcome kit was shipped to their residence containing all necessary study-related materials including a Fitbit Charge 5 smartwatch and additional literature about the study and the a liate MIH program.Subjects were then prompted to schedule an onboarding call with a member of the study team where they were guided through setting up and using the smartwatch, educated on the use of the study app, and provided any additional support needed to initiate participation in the study.Once onboarded, participants were asked to partake in study procedures for six months.A summary of all components of the study and their interconnectedness is depicted in

Participant Reported Outcomes
Study participants were asked to complete a series of instruments throughout the six-month study period through the MyDataHelps app.Participants were noti ed of outstanding surveys and prompted to complete them through app push noti cations according to the study schedule.In addition to demographic questions asked at baseline, participants were prompted to complete surveys including the NIH-PROMIS COPD questionnaires, Patient Activation Measure, Modi ed Medical Research Council Dyspnea Scale, and a patient satisfaction measure at enrollment, 3 months, and 6 months (Table 2). 18- 21Participants were also prompted to complete the COPD Assessment Test monthly and a single-item wellness measure weekly. 22Momentary assessments for participants were assessed using CLEAR-Sx, Ex, and Rx tools; these were triggered based on participants reporting not feeling well on the single-item wellness measure or by wearable device data (Table 3). 16,23

Remote patient activity monitoring
Participants were asked to wear the smartwatch daily, including at night, to collect data including daily steps, heart rate, oxygen saturation, and sleep patterns.This information was visible to study coordinators and investigators, as well as MIH paramedics, through the MyDataHelps platform.The data were used to develop automated prompts for participants to provide momentary assessments using CLEAR surveys.

Virtual Comprehensive Pulmonary Support Service
All participants were presented with the option to enroll in a virtual pulmonary care program that provides multimodal support for COPD patients, through a company called Wellinks.The option to enroll in this was made available through the MyDataHelps during the rst 90 days of study participation.The program includes coaching to support participant education, treatment adherence, and goal setting, and a virtual pulmonary rehabilitation program that provides a home-based exercise plan including safety instructions and COPD-speci c breathing techniques.Additionally, Wellinks participants received a second equipment kit containing a spirometer, pulse oximeter, ten-foot tape measure, exercise band, and an exercise guide.The virtual pulmonary support program team coordinated introductory calls with participants to download their app and con gure the spirometer and pulse oximeter devices via Bluetooth which are used to collect additional biometric data.After a comprehensive personalized assessment, a schedule for regular sessions with a health coach was created for participants.All coaching was scheduled and joined through the program's app, with board-certi ed health and wellness coaches or nurse practitioners.
In the event of an acute clinical concern, the Wellinks team's protocol was to instruct participants to contact the MIH clinician or their healthcare provider.The Wellinks team was also empowered to contact the on-duty MIH clinician directly.Participants who enrolled in the Wellinks program were classi ed as the "full intervention" group and those who enrolled in the study but opted not to participate in the Wellinks program were considered the "partial intervention" group.

MIH Integration
To on-demand, eld-based clinical support to participants, all enrolled participants were invited to utilize the institution's a liate MIH program for the duration of their participation. 24MIH programs are new models of care that leverage mobile resources, including specially trained paramedic-level clinicians, called community paramedics, to evaluate and care for patients in an out-of-hospital environment in coordination with health systems. 24,25By providing eld-based care, such programs are intended to eliminate barriers that patients encounter in accessing timely management of acute illness.They complement digital and remote clinical interventions by providing exible, face-to-face evaluation and treatment to patients.During episodes of acute clinical symptoms (such as worsening shortness of breath), participants, their caregivers, and the Wellinks team were empowered to request an MIH visit.The community paramedics are available 24 hours a day, 7 days a week, and present to participants' homes within 2 hours of a request.Upon arrival, they evaluate and treat participants, aided by mobile diagnostic tools and medications as well as telehealth support from an on-call supervising physician, then develop a plan of care with the participant, which usually entails treatment in the home and intensive care coordination with the participant's ambulatory providers.The MIH program is speci cally equipped to initiate treatment for COPD exacerbation with inhaled bronchodilators and parenteral antibiotics and steroids.If a participant is too acutely ill to remain at home, the participant is diverted into the emergency services system.To streamline care, the community paramedic team and their supervising physicians had access to participants' study dashboards so that they could review their aggregated clinical data and recent changes in biometric patterns.
In addition to the education and literature provided during study onboarding, the participant-facing MyDataHelps app contained a "Call MIH" button which enabled the participant to easily contact MIH for an in-home assessment if they felt unwell by directly connecting them with the MIH request line.The MyDataHelps app was set to issue a prompt to the participant to consider calling MIH when pre-set metrics are consistent with a potential COPD exacerbation (Table 3).Finally, research coordinators were instructed to refer participants to the MIH hotline if any clinical concerns were raised while on a phone call with participants.The MIH program did not proactively provide care because of alerts or ags observed from participant-reported data or biometric signals in this pilot phase.Clinical care provided by the MIH paramedics for the study's participants was identical in practice to the care provided to patients outside the study.

ANALYSIS PLAN
Participants are followed longitudinally for six months in the pilot study, during which measures of feasibility, clinical e cacy, and operational e cacy will be examined.The feasibility measures observed in the two intervention arms include the response rate for study the adoption rate for virtual pulmonary support, adherence for survey and sensor data, adherence to coaching activities, ability to aggregate sensor data, and ability to measure clinical outcomes through participant-provided data.Figure 3 outlines all data collected during the study.
Clinical and operational e cacy will be compared across all three groups of COPD participants: fullintervention, partial-intervention, and a control group (no intervention) identi ed from the EMR using a novel synthetic control study strategy that has been endorsed by the FDA for generating real world evidence. 26,27The clinical e cacy measures include medication requirements, healthcare utilization related to acute or urgent care, total medical expenditure, and, in the intervention arms, participantreported outcomes (COPD severity, COPD exacerbations, patient activation measure, patient satisfaction, modi ed Medical Research Council Dyspnea Scale).Lastly, operational e cacy include time spent during MIH visits, MIH visit escalation to acute care rate, and care management time.
Descriptive analysis will be used to characterize the feasibility outcomes.Continuous measures will be described using mean or median depending on the underlying distribution, and categorical variables will be described as proportions.Associations with sociodemographic characteristics and GOLD classi cation of COPD severity will be assessed for feasibility outcomes.Count models (Poisson, negative binomial, or zero-in ated negative binomial) will be used to compare healthcare utilization over the course of the study.Continuous models will be used to compare medical expenditure.GEE models with logit or linear link will be used to assess changes over time for COPD severity measurements.

DISCUSSION
The Healthy at Home study addresses a need to remove barriers to preventative and rehabilitative care for COPD patients and recognize potential indicators of COPD exacerbations.The current mainstay of COPD care is centered around acute disease management and services in post-acute settings.However, there is an unmet need for providing proactive care to avert or diffuse an acute exacerbation before hospitalization occurs.Prior studies have focused on remote patient monitoring to detect evidence of acute exacerbation, but there is a dearth of knowledge about the operational impact and implementation of these interventions.There is also limited data on whether such monitoring initiatives are costeffective. 15This study aims to demonstrate the feasibility of conducting a larger-scale study using the novel synthetic control design to generate real-world evidence for clinical, operational, and cost outcomes related to the provision of the "Healthy at Home" program.
By employing the use of a Fitbit device, on-demand acute in-home healthcare visits, and, for full intervention participants, a pulse oximeter, spirometer, and virtual health coaching, participants will have a greater ability to view and act on their personal health metrics.In addition to monitoring their own health and being encouraged to contact the MIH program for health-related concerns, participants' health and survey data will be used to understand the most important factors predictive of COPD-related exacerbations.The use of integrated data steams combining participant-reported information, biometric signals, and EHR patterns will generate a rich multifaceted digital phenotype of the COPD disease process which can be used to predict future clinical outcomes and facilitate proactive intervention when deterioration is predicted.
We also expect that this important addition to the knowledge base of COPD care will come with a limited burden to participants, as the majority of the data collected is passive.It is expected that participants will spend about 10 minutes per week on study activities, with an additional hour per week of pulmonary coaching and exercises for those enrolled in the full intervention portion of the study.We predict that this consciously low physical and cognitive burden to participants will result in excellent implementation outcomes such as high retention, delity to the study protocol, and adoption of the interventions including use of the MIH program.
The current study is one of many programs implemented in recent years attempting to reduce COPDrelated readmissions.The Centers for Medicaid and Medicare Services has listed COPD as a target condition for the Hospital Readmission Reduction Program (HRRP).This program imposes penalties on health systems in relation to their readmission rates in an effort to promote innovation and reduce 30-day readmission.9][30][31][32] Existing literature has focused on the root causes of readmissions, identifying areas in which improvement is needed. 3Evidence surrounding COPD readmissions shows that multimodal interventions that include frequent assessment, care management, and pulmonary coaching or rehabilitation have the highest e cacy for preventing readmission. 9 is expected that this study will have implications for the standard of care provided to patients at risk of admission for COPD exacerbations.If successful, increased implementation of at-home monitoring and virtual pulmonary support services may be established for patients who would bene t from this level of care.Additionally, this information can be shared with healthcare providers, such as community paramedics to assist them in providing a more proactive and robust level of care for their patients.It is also a goal of the study that these activities will help to decrease the cost acquired for patients who accumulate high levels of healthcare spending due to repeated hospitalizations and acute care needs.As a feasibility study, the current project aims to establish the impacts of study activities and expand both the population and the research partners in future analysis of this population.
By enrolling all-comers, agnostic to insurance type, residing in the towns serviced by the a liate MIH program, it is expected that we will enroll a diverse population in reference to race, ethnicity, and socioeconomic status.However, this study is limited to those who have established health care within the system of interest and those who reside in the fairly urban areas surrounding these hospitals.
Additionally, those who do not speak English are excluded from this study.Future studies should focus on a more diverse population based on geographical location and language, as there are barriers to care in those populations that will not be seen in the study.Additionally, participants are required to have a smartphone with the ability to access the internet at home to enroll on the study app, potentially excluding a subset of the population who do not have these resources.
The Healthy at Home program's feasibility study aims to create a foundation around which future, more expansive studies can be constructed.Independent or ancillary studies could be designed, possibly in coordination with pre-existing studies in digital medicine, to allow this novel multimodal platform to be harnessed by those aiming to better understand other disease states and populations.In the case of studying other disease states, the structure of Healthy at Home could easily be modi ed to work with other healthcare vendors who may be specialized to care for a speci c population's needs.Past simple expansion of study protocols, the information gathered from studies built on the foundation of Healthy at Home may be used in the long-term to grow patient care using this digital, multimodal approach.It is the hope to enhance both study protocols for future investigations and medical care for patients in a multiphasic approach, by which optimizations are made over a period of time, building upon each other.Using biometric data gathered from participants, prognostic digital biomarkers may be able to be used to help patients understand when they may be at a heightened risk of needing acute care.This preventative care may also be used in coordination with a virtual, digital coach, enabling patients to be better noti ed of when they may be at risk or when their vital signs are outside a speci ed range and to seek interventive care early in the course of a threatened exacerbation.These optimizations, as well as others yet-to-bearticulated, may eventually aid in the elevation of the standard of care provided in both research and practice.

Conclusions
The Healthy at Home study's long-term goals are to develop new tools and modes of care to better treat and understand the needs of patients with COPD.This is done by determining the feasibility of this type of digital study, as well as examining the e cacy of the multimodal interventions, across both the partialand full-intervention groups.Furthermore, we hope to use the data gathered from this current study to inform an improved digital dashboard for use by the MIH clinical team to provide them with a more comprehensive, longitudinal history of their patients' health.This multimodal approach may aid patients to better manage their COPD symptoms by more closely monitoring their health and providing them with multiple avenues to prevent acute care needs, including through the ability to call community-based medical providers and through regular use of the virtual pulmonary support coaching.

Figure 2 SPIRIT
Figure 2

Table 1 .
A Standard Protocol Items: Recommendations for Intervention Trials (SPIRIT) Figure for the study is depicted in Fig. 2.

Table 1 Healthy
At Home Intervention Components: This multimodal intervention included a study app, smartwatch, access to acute care through a mobile integrated health program, and virtual pulmonary support services.

Table 2
Participant-Facing Assessments: Participants were prompted to complete all assessments through the Care Evolution Healthy At Home study app.The app prompted participants to complete each assessment as they were due and sent reminders for delinquent assessments.

Table 3 Clinical
Alert Triggers.The clinical alert metrics triggered app noti cations to prompt participants to request a medical evaluation if they were feeling unwell.The app enabled participants to quickly call the mobile integrated team for an on-demand acute home visit with a "one-touch" option to dial the MIH hotline phone number.