Impacts of the COVID-19 Pandemic on Cardiac Rehabilitation Delivery around the World

Background: We investigated impacts of COVID-19 on cardiac rehabilitation (CR) delivery around the globe, including virtual delivery, as well as effects on providers and patients. Methods: In this cross-sectional study, a piloted survey was administered to CR programs globally via REDCap from April to June 2020. The 50 members of the International Council of Cardiovascular Prevention and Rehabilitation (ICCPR) and personal contacts facilitated program identification. Results: Overall, 1062 (18.3% program response rate) responses were received from 70/111 (63.1% country response rate) countries in the world with existent CR programs. Of these, 367 (49.1%) programs reported they had stopped CR delivery, and 203 (27.1%) stopped temporarily (mean = 8.3 ± 2.8 weeks). Alternative models were delivered in 322 (39.7%) programs, primarily through low-tech modes (n = 226,19.3%). Furthermore, 353 (30.2%) respondents were re-deployed, and 276 (37.3%) felt the need to work due to fear of losing their job, despite the perceived risk of contracting COVID-19 (mean = 30.0% ± 27.4/100). Also, 266 (22.5%) reported anxiety, 241(20.4%) were concerned about exposing their family, 113 (9.7%) reported increased workload to transition to remote delivery, and 105 (9.0%) were juggling caregiving responsibilities during business hours. Patients were often contacting staff regarding grocery shopping for heart-healthy foods (n = 333, 28.4%), how to use technology to interact with the program (n = 329, 27.9%), having to stop their exercise because they have no place to exercise (n = 303, 25.7%), and their risk of death from COVID-19 due to pre-existing cardiovascular disease (n = 249, 21.2%). Respondents perceived staff (n = 488, 41.3%) and patient (n = 453, 38.6%) personal protective equipment, as well as COVID-19 screening (n = 414, 35.2%), and testing (n = 411, 35.0%) as paramount to in-person service resumption. Conclusion: Given the estimated number of CR programs globally, these results suggest approximately 4400 CR programs globally have ceased or temporarily stopped service delivery. Those that remain open are implementing new technologies to ensure their patients receive CR safely, despite the challenges. Highlights: – COVID-19 has impacted cardiac rehabilitation (CR) delivery around the globe. – In this cross-sectional study, a survey was completed by 1062 (18.3%) CR programs from 70 (63.1%) countries. – The pandemic has resulted in at least temporary cessation of ~75% of CR programs, with others ceasing initiation of new patients, reducing components delivered, and/or changing of mode delivery with little opportunity for planning and training. – There is also significant psychosocial and economic impact on CR providers. – Alternative CR model (e.g., home-based, virtual) reimbursement advocacy is needed, to ensure safe, accessible secondary prevention delivery.


Introduction
The coronavirus disease 2019 (COVID-19) is an infectious disease with high transmissibility caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), [1] which was declared a pandemic on March 11, 2020 by the World Health Organization (WHO).
COVID-19 disproportionately affects older adults, and there is a higher case fatality rate in those with underlying conditions such as cardiovascular diseases (CVD) [2,3]. In addition to respiratory effects, COVID-19 has negative cardiovascular effects as well [4,5]. COVID-19 has placed tremendous pressures on health care organizations around the world.
CVDs are among the most prevalent non-communicable diseases worldwide [6]. Cardiac rehabilitation (CR) is recommended as the standard of care for secondary prevention of CVD [7,8] as it significantly reduces morbidity and mortality [9]. Although models vary, CR is generally delivered in clinical settings [10], involving semi-weekly visits for exercise and education/counselling sessions over several months, for a median of 24 sessions globally [11].
These in-person visits therefore, if continued during COVID-19, carry a high risk of exposure.

Design and Procedure
This research was cross-sectional in design. This work was undertaken by the International Council of Cardiovascular Prevention and Rehabilitation (ICCPR; globalcardiacrehab.com); York University's Office of Research Ethics (Toronto, Canada) approved secondary use of the data for publication purposes. Participation in the survey constituted voluntary consent to participate. Responses were confidential.
All 40 ICCPR member associations and 10 "friend" members were contacted, and requested to circulate the survey to their members and any other programs of which they were aware in their regions. Contacts of the investigators were also approached to circulate and complete the survey. Potential respondents were also recruited from the attendance list of the ICCPR webinar on COVID-19 and CR [26], which was attended by more than 300 participants.
Data collection occurred from March to June 2020 via online survey administered through REDCap (English). The original survey was not available within China. Therefore, the survey was translated (Simplified Chinese) by co-author XL, a bilingual nurse with CR expertise, and disseminated through Sojump, a professional online survey platform.
in the global audit only a random subsample of programs was recruited in the United States, but all programs were emailed for this study) [28]. Information from key informants and respondents were used to update these numbers where available.

Measures
A 33-item questionnaire was developed by the investigators to examine the objectives of this study (Online Appendix 1). Non-COVID-related questionnaire items were based on ICCPR's previous CR program survey [18]. The questionnaire was divided into three sections: (1) CR program characteristics and impacts of COVID-19, (2) barriers and facilitators to delivering CR programs virtually, and (3) COVID-related impacts on staff and patients. The items had forced-choice (some were select all that apply) and open-ended response options, and skip-logic was used to get more detail where applicable. Input from CR providers was solicited, with minor changes made to questions and wording prior to launch of the survey.
Respondents were asked to report their country and the date. To contextualize responses, information on COVID-19 cases at the time of survey completion in each country was extracted from the website "Our World in Data" [29] and confirmed with data available on each country's government website. The Government Response Stringency Index -information on 9 common policy responses that governments have taken to respond to the pandemic such as workplace closures and stay-at-home requirements (rescaled to a value from 0 to 100, with 100 = strictest responses) -was also extracted from this website for each country, with information corresponding to the mean date of survey completion by respondents in each country. number of responses for each question varied due to missing data (e.g., respondent did not answer a question due to inapplicability, skip logic, or decided to not answer for other reasons); for descriptive analyses, percentages were computed using the number of valid responses for the specific item as the denominator.
Descriptive statistics (e.g., frequency with percentage) were applied for all closed-ended items in the survey. All open-ended responses were coded.

Respondents
Overall, 1062 responses were received, of which 1044 (98.3%) identified their country. Table 1, responses were received from 70 (34.5%) of the approximately 203 countries in the world (covering all 6 WHO regions; response rate by region ranged from 14.1-100.0%; median number of responding programs per country=98.0; Q25-75=36.0-108.0), or 63.1% of the 111 countries identified to have any CR programs in ICCPR's global audit [28].

As shown in
Given the estimated number of programs globally 28 and updates gleaned through this process (no additional countries were determined to have CR, but 13 countries had increases in the number of programs since ICCPR's 2016/17 audit, with a total of 60 additional programs globally; Table   1), responses were received from 18.3% of the estimated 5,813 programs worldwide.
In regards to location, 392 (37.4%) CR programs were located in a community hospital, 364 (34.7%) in a referral center, quaternary or tertiary facility, and/or academic center, 46 (4.4%) in a rehabilitation hospital or residential facility, and 70 (6.7%) in other locations, including private clinics, universities, and fitness centers. With regards to reimbursement, 690 (66.4%) programs have their services covered by the government, 557 (52.8%) programs' services were All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; paid directly by patients, and 484 (47.5%) were paid by health insurance (many programs had multiple sources).

Impact of COVID-19 on CR Program Delivery
Overall, 106 (14.4%) programs had suspected or positive COVID-19 patients (see Table   1 for COVID-19 cases by region). At the time of survey completion, 367 (49.1%) programs reported they had completely stopped CR delivery for the COVID-19 pandemic, and 203 (27.1%) stopped for a period but had already resumed services (mean 8.3±2.8 [standard deviation] weeks); 178 (23.8%) programs did not stop (see Table 1 for this information by country). As shown, Nigeria has the poorest CR density, and all programs closed. Of those that stopped for any amount of time, 363 (30.0%) made no other arrangements to provide patient care, and the 202 (16.7%) that did describe the following: home-based CR/ telehealth, online consultations (with respondents reporting WeChat, MS Teams, and Zoom), phone or email consultations, and education offered more often via phone, online or via postal mail to patients. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; During COVID-19, 202 (42.3%) programs were only treating existing patients, while all others were still accepting new patients; programs reported having the capacity to serve a median of 20.0 (Q25-27 =10.0-40.0) patients / month pre-COVID, and 3.0 (Q25-75=0.0-15.0) during. The most common program adaptations were: reducing the number of program elements offered (n=228, 19.5%), deferring graduation until post-program assessments could be completed (n=120; 10.3%), shortening the program duration (n=95, 8.2%), and graduating patients more quickly (n=85, 7.3%). On the other hand, 177 (15.1%) respondents reported adapting all program elements to retain service levels (see Table 2 for adaptations by WHO region). Table 3 displays the components that are delivered by CR programs, and continue to be provided during the COVID-19 pandemic. As shown, most impacted was supervised exercise training, resistance training, inclusion of family / informal caregivers, end of program reassessment, and functional capacity testing. Indeed, of the programs providing services or without COVID-19 restrictions, 119 (14.5%) had changed the type of functional capacity test they use. Of these, 30 (26.3%) were performing functional capacity testing remotely, most commonly: a walk test (n=19; 1.6%), step test (n=6; 0.5%); 15 (1.3%) cited other tests, such as the Duke Activity Status Index and timed sit-to-stand.
Of those that continued providing services, the COVID-19 pandemic had impacted the way exercise is prescribed for 306 (37.5%) programs (e.g., virtual consultations with no structured supervised exercise program), and monitored for 226 (27.8%; e.g., focus on rating of perceived exertion rather than target heart rate, all done virtually). Where patients were encouraged to exercise at home or in their community, the following changes were made to manage patient symptoms, reporting or concerns: education about symptoms enhanced, regular All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; virtual (phone and online) consultations with CR staff were added, and patients were encouraged to directly contact their physician if symptoms were experienced.
Furthermore, of programs continuing to provide services, in 111 (13.6%) programs, the type of provider delivering CR had changed, and in 122 (15.0%) patients have had to interact with a different provider (e.g., mainly one professional interacting with patients -mostly nurses, as access to allied health care providers ceased due to COVID-19).
Online Appendix 2 provides details of patient education delivery modes and providers, in light of the COVID-19 pandemic. As shown, there was a drastic reduction in exposure to all disciplines. Modes of delivery were chiefly phone, email, online resources, and mail; live video was only exploited in approximately 20% of programs. For those that offered it, education session frequency before COVID-19 was 8.7±8.4 / month / patient (mean of 50.1±47.5 minutes / session), dropping to 4.1±7.5 (mean of 20.2±26.0 minutes / session) during the pandemic.
Forms of communication used in remote models are also shown in Table 4. Overall, 182 (62.8%) programs reported barriers to using these communication tools, most commonly: lack of patient access (i.e., patients not having computer with email; n=146, 12.5%), difficulties for the patients (e.g., lack of technical knowledge; n=111, 9.6%), logistical problems such as connectivity issues (n=107, 9.2%), and difficulties for the clinical staff (e.g., being too busy, lack of staff; n=53, 4.6%).
Barriers to delivering CR remotely were: patients did not have the technology to connect with program staff remotely (n=195, 16.8%), lack of equipment /program resources for secure and private remote delivery (n=148, 12.7%), insufficient funding (n=123, 10.5%), insufficient staff (n=118, 10.1%), patients' risk is too high for unsupervised exercise/safety concerns (n=111, 9.5%), too inefficient (i.e., cannot see as many patients because they do not offer remote program elements in a group, but only individually; n=90, 7.7%), staff need training (n=90, 7.7%), and no policy (n=69, 5.9%); other barriers (n=24, 2.1%) included language proficiency, remote delivery not reimbursed, not enough referrals and hours.
Finally, to increase their capacity to deliver home-based/remote CR services to patients (open-ended item), respondents perceived they would need: time to research and develop the model (including appropriate services for illiterate patients); secure /private means for staff to communicate with patients electronically; equipment to communicate remotely with patients (e.g., multiple laptops with built-in cameras and microphones, headsets); facilities / space (e.g., videoconferencing in private room); home equipment to loan patients including tablets, wearables/activity trackers for exercise monitoring and assessment of all risk factors (e.g., sphygmomanometers/cuffs); cheap and reliable wireless technology to monitor for adverse events remotely (e.g., heart rate and rhythm); reliable and low-cost high-speed internet access for staff and patients (including those living in rural and remote areas); a dedicated multidisciplinary team (including more nurses and exercise physiologists); staff training (by those with experience in remote models), physician champions; administrative staff to facilitate scheduling of virtual sessions; a structured, evidence-based home-based CR program software platform or smartphone All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; app (including that it has capacity to reach groups of patients, that family / informal caregivers can also access it, that it is comprehensive [e.g., includes structured exercise livestreaming, education], and includes tracking diaries with two-way communication; or funding for technology upgrades), with associated assessment tools (including functional capacity) and policies to implement; technology support staff (with time) to train patients (potentially in-person when they are in acute care) and staff to use the remote technology (e.g., zoom) and equipment, and also for database management support; ability to have at least one safe in-person session with each patient to ensure safety and education; financial resources as well as reimbursement of remote model; and patient as well as provider awareness of availability of the remote model (including referrals).

Impact on CR staff
CR team communication was also impacted. Overall, 295 (25.0%) reported having only ad-hoc meetings as needed, 274 (23.0%) reported they were having routine staff meetings via remote means (e.g., videoconferencing), 137 (11.8%) reported not being able to communicate as much as needed (e.g., staff re-deployed, staff lack home technology), and 88 (7.6%) reported they could not really communicate except via email.
Occupational impacts on staff are displayed in Table 5 by WHO region. Moreover, 114 (15.3%) staff reported feeling stigmatized outside of work for being a healthcare worker (i.e., greater risk of COVID-19 exposure), 275 (36.9%) felt valued for providing care, and 325 (43.6%) felt no stigma (others responded 'not applicable'). Negative psychosocial impacts that do or could impact their work are shown in Figure 1 (respondents were directed to check all that apply; n=98, 8.4% responded 'not applicable' as they have experienced no COVID-19 impacts; All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; n=267, 22.8% of respondents reported no personal suffering due to COVID-19; and n=233, 20.0% have been using the opportunity to catch up on things and learn).
When asked whether they had the equipment (e.g., video camera, remote access to hospital servers) and space / privacy to work from home efficiently during business hours, 198 (26.6%) responded affirmatively, 141 (19.0%) responded only for work that does not involve patient care, and 272 (36.6%) reported no (for reasons provided including: no access to work materials and hospital server remotely). Table 5 also displays whether programs have treated suspected or COVID-19 positive patients by WHO region, and associated concerns. Figure 2 displays perceived risk in contacting COVID-19 through their CR work by country (global mean=30.0%±27.4). Generalized linear mixed models, accounting for country as a higher-order variable, revealed degree of perceived risk was associated with country stringency index (p=0.01), but not number of cases (p=0.80).

Impact on Patients
CR staff reported patients contacting them and expressing the following reservations and concerns regarding COVID-19: whether they need to change their exercise prescriptions (n=337, 28.7%), adherence to a heart-healthy diet while concerned about going to grocery stores (n=333, 28.4%), about their mental well-being (n=331; 28.2%), using technology to interact with the program (n=329, 27.9%), questions about medications (n=318, 26.9%), stopping exercise due to a lack of space to exercise in their homes or inability to exercise outside without potential COVID-19 exposure (n=303, 25.7%), questions about how to safely receive CR care without COVID-19 exposure (n=255, 21.4%), their risk of death from COVID-19 due to pre-existing CVD (n=249, 21.2%), and many patients contacted programs about reservations regarding receiving CR care and staff had to let them know they closed down the program temporarily All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; (n=230, 19.6%). On the other hand, 142 (12.2%) reported patients had not really been contacting the program more often due to COVID-19, and 62 (5.3%) reported patients seemed to be avoiding contacting their programs at all, under the assumption that services were suspended.

Resumption of CR services
Overall, 273 (36.8%) reported that their institution had a policy regarding the circumstances under which regular services could resume (for n=167, 22.5% this was not applicable).

Discussion
This first-ever global survey on the impacts of COVID-19 on outpatient rehabilitation has confirmed that the impact of this infectious disease goes well beyond those suffering from it, affecting availability, structure, delivery format, and components of chronic disease care, as well as the mental health of CR providers and patients alike. Given that three-quarters of responding programs reported complete or temporary cessation of services, we estimate COVID-19 has been responsible for the closure of approximately 4400 CR programs worldwide; it is unknown when or if these programs will resume care delivery. Capacity was incredibly low in the programs that remained open (i.e., 3 patients/month, with no new patients) and key risk-reducing components were no longer delivered in more than 60% of on-going programs, including supervised exercise training. This substantial service loss will likely result in greater CVD morbidity and mortality in the coming months and years, further straining already-taxed healthcare systems.
All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; To balance both infection risk and risk of an adverse event, some in-person contact at the beginning of the CR program was advocated, following screening and with PPE (indeed onequarter of respondents were concerned about delivery due to lack of PPE), perhaps at reduced volume to enable physical distancing. Programs have transitioned to using different functional All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; capacity testing methods to inform exercise prescription and different means to monitor exercise.
There was great interest in using technology to remotely monitor risk factors, with the goal of The social restriction, anxiety and occupational impacts resulting from coronaviruses adversely affects the psychological well-being of healthcare workers, and this impact can be substantial and long-lasting [36][37][38]. Although about a fifth of respondents in this study reported they have been coping "alright" with COVID-19, many are experiencing anxiety, fear of exposing family, loneliness, difficulty sleeping, and stress due to higher workloads (including answering many questions from patients about their concerns). Over a third felt the need to work despite perceived risk, due to fear of losing their job or pay (and one in five had no sick pay).
Approximately a third of CR program respondents had been redeployed, rendering it impossible to deliver CR services without staff. Almost 10% have needed to work double-duty, serving as caregivers during business hours. The inequities uncovered by this outbreak [39] are also evidenced in the data from this study, with variation in occupational / economic and All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; psychosocial impacts by WHO region (although inferential tests were not performed due to some small cell sizes).
Caution is warranted in interpreting the findings of this study. First, this was a convenience sample; results may be biased, representing mainly programs that still had some staff working and available to respond to the survey. However, there was good representation from programs in community and academic hospitals alike. We did assume that each response was received from a unique program, however it is possible different staff members from the same program completed a survey, given most respondents were working remotely. Response rate was lower in the Americas and Europe, so generalizability to those regions is more questionable; country representation was high, however. Second, the reliability and validity of the survey is unknown; it was pilot-tested, but responses in a random subsample of programs were not verified against actual delivery. Programs may have responded in a socially-desirable manner, although the survey was confidential. Third, because of small cell sizes in some case, inferential comparisons by WHO region could not be performed. Finally, since the survey was mainly completed in English, some concepts could have been misunderstood by respondents for whom English was not their first language.

Conclusion
The COVID-19 pandemic has impacted CR programs worldwide, including cessation of services or of initiation of new patients, a decrease in CR components delivered, change of mode delivery without much opportunity for planning and training, as well as psychosocial and economic impact on healthcare providers. Technology is seen as a safe means to ensure cardiac patients receive the care they need during this difficult time, within the context of screening, testing and sufficient PPE. Remote delivery is not often reimbursed, and therefore advocacy is All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; needed if we are to fulfill our mission in supporting patients in secondary prevention, which will ultimately reduce the burden on a strained healthcare system. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Percy E, Luc
The copyright holder for this this version posted November 12, 2020. ; https://doi.org/10.1101/2020.11.11.20230045 doi: medRxiv preprint *number of respondents surpassed the number of CR programs previously identified. †received updated information regarding number of programs in country since ICCPR's global audit. §represents number of CR spots available per year per incident ischemic heart disease patient (higher numbers represent worse density; i.e., 1 CR "spot" per n patients). ‡16 respondents did not identify which country they are from.

Tables
°average date of survey completion.     Note: respondents instructed to check all that apply.
All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  I  h  a  v  e  b  e  e  n  v  e  r  y  w  o  r  r  i  e  d  a  b  o  u  t  e  x  p  o  s  i  n  g  m  y  f  a  m  i  l  y  t  o  C  O  V  I  D  -1  9   Y  e  s  ,  I  a  m  f  e  e  l  i  n  g  a  n  x  i  o  u  s  a  t  t i m e s All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020. ; https://doi.org/10.1101/2020.11.11.20230045 doi: medRxiv preprint All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  O  u  r  i  n  s  t  i  t  u  t  i  o  n  h  a  s  p  o  l  i  c  i  e  s  i  n  p  l  a  c  e  t  o  p  r  o  t  e  c  t  p  a  t  i  e  n  t  s  (  e  .  g  .  ,  m  a  s  k  s  ,  s  u  f  f  i  c  i  e  n  t  s  p  a  c  e  f  o  r  p  h  y  s  i  c  a  l  d  i  s  t  a  n  c  i  n  g  a  t  o  u  r  c  e  n  t  e  r  ) S t  a  f  f  h  a  v  e  s  u  f  f  i  c  i  e  n  t  p  e  r  s  o  n  a  l  p  r  o  t  e  c  t  i  v  e  e  q  u  i  p  m  e  n  t  a  v  a  i  l  a  b  l  e   P  e  r  c  e  n  t  a  g  e  o  f  v  a  l  i  d  r  e  s  p  o  n  s  e  s All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this this version posted November 12, 2020.