Feasibility of home-based cardiac rehabilitation in frail older patients: a clinical perspective

ABSTRACT

(a) Background

Home-based cardiac rehabilitation (CR) is an attractive alternative for frail older patients who are unable to participate in hospital-based CR. Yet, the feasibility of home-based CR provided by primary care physiotherapists (PTs) to these patients remains uncertain.

(b) Objective

To investigate physiotherapists’ (PTs) clinical experience with a guideline-centered, home-based CR protocol for frail older patients.

(c) Methods

A qualitative study examined the home-based CR protocol of a randomized controlled trial. Observations and interviews of the CR-trained primary care PTs providing home-based CR were conducted until data saturation. Two researchers separately coded the findings according to the theoretical framework of Gurses.

(d) Results

The enrolled PTs (n = 8) had a median age of 45 years (IQR 27–57), and a median work experience of 20 years (IQR 5–33). Three principal themes were identified that influence protocol-adherence by PTs and the feasibility of protocol-implementation: 1) feasibility of exercise testing and the exercise program; 2) patients’ motivation and PTs’ motivational techniques; and 3) interdisciplinary collaboration with other healthcare providers in monitoring patients’ risks.

(e) Conclusion

Home-based CR for frail patients seems feasible for PTs. Recommendations on the optimal intensity, use of home-based exercise tests and measurement tools, and interventions to optimize self-regulation are needed to facilitate home-based CR.

KEYWORDS:

• Cardiovascular diseases
• cardiac rehabilitation
• home-based
• frail older patients
• comorbidity

Introduction

Comprehensive cardiac rehabilitation (CR) is recommended for older patients who experienced a hospital-admission for cardiovascular diseases (CVD) (Achttien et al., 2013, 2015; Rohrbach et al., 2017). Likewise, the benefits of CR in frail older populations have been researched and documented (Afilalo, 2019; Doll et al., 2015; Shields et al., 2018; Taylor et al., 2015). Frailty may be defined as a syndrome of physiological decline characterized by marked vulnerability to adverse health outcomes such as hospital readmission and mortality (Buurman et al., 2012; Krumholz et al., 2013). Consequently, providing CR for the frail older patient with CVD is met with distinctive challenges (Doll et al., 2015; Shields et al., 2018), like low participation rates (i.e. 20–30%) in hospital-based CR because of a lack of transportation facilities, the patient’s perception about the worth of the rehabilitation program, and their apprehensions regarding the risks of exercising (Ruano-Ravina et al., 2016).

To mitigate these reservations, home-based CR is gradually becoming a good alternative to hospital-based CR to effectively improve the physical functioning of low-to-moderate risk (non-frail) patients (Taylor et al., 2015; Verweij et al., 2019). Yet, the feasibility of a home-based CR program for frail older patients from a primary care physiotherapist’s perspective is uncertain. The current CR guidelines for physiotherapists (PTs) are largely based on research involving non-frail patients, not adjusted to the home situation, and there is a lack of specific recommendations for modifying CR in the presence of frailty and comorbidities (Achttien et al., 2013, 2015; Dekker, Buurman, and van der Leeden M, 2019; Piepoli et al., 2014; van Weel and Schellevis, 2006; Vigorito et al., 2017). Consequently, physiotherapists may be reluctant to prescribe exercise therapy, resulting in less adequate treatment. Therefore, the current study aimed to investigate physiotherapists’ clinical experiences in light of the present guidelines for CR and the home-based CR protocols that are followed for frail older patients.

Methods

Study design

The conduct of this qualitative study involved observing primary care PTs who performed home-based CR for frail older patients, followed by semi-structured personal face-to-face interviews. The home-based CR component was a part of a larger randomized trial, the “Cardiac Care Bridge (CCB)” (Jepma et al., 2021b; Verweij et al., 2018) in collaboration with one teaching hospital and five other regional hospitals in and around Amsterdam. The study was registered with Netherlands Trial Register 6316, https://www.trialregister.nl/trial/6169. The Medical Ethical Review Board of the Amsterdam University Medical Center approved the study protocol for the CCB (MEC2016_024). All participants provided written informed consent before participating. This manuscript follows the COREQ checklist for reporting qualitative research (Tong, Sainsbury, and Craig, 2007).

Home-based CR in the CCB intervention

The CCB trial, conducted from June 2017 to March 2020, evaluated the effects of a transitional care program to prevent hospital readmission and mortality in frail, older patients with CVD (Figure 1). The detailed study design and outcomes of the CCB study have been described in a previously published report (Jepma et al., 2021b; Verweij et al., 2018). Briefly, the CCB program is an integrated transitional-care program from the hospital to the patients’ home and is composed of three components: 1) case management; 2) disease management; and 3) home-based PT-led CR. Component 1 consisted of an in-patient comprehensive geriatric assessment-based integrated care plan implemented during the hospital stay and followed-up at home. In component 2, cardiac and community nurses coordinated with an affiliated pharmacist to carry out medication-checks, monitor medical parameters, and administer lifestyle coaching. In the last component, the PTs focused on facilitating participation in activities of daily living, performing regular exercises, monitoring weight, heart rate, and blood pressure, besides educating the patient striving for an active lifestyle.

Figure 1. Overview of the CCB program including home-based cardiac rehabilitation.

Blue depicts the nurse coordinated care, red describes the exercise components of home-based cardiac rehabilitation by physiotherapists. In addition, physiotherapists coached on active lifestyle and monitored complications; DSMS = Dutch Safety Management System, BORG = Borg scale of perceived exertion

The CCB-integrated-home-based-CR protocol was based on the Dutch cardiac rehabilitation guidelines for PTs (Achttien et al., 2011, 2013, 2015) and was adapted for the CCB by two skilled PTs with more than 5 years of experience in the field of cardiopulmonary rehabilitation. The principal modifications done to the CR-guidelines include a substitution of regular exercise field tests, namely the shuttle walk test (Singh et al., 1992) or six-minute walk test (Bellet, Adams, and Morris, 2012; Butland et al., 1982) and the 6–10 repetition maximum strength test, which require sufficient floor space or strength testing equipment, with a home-executable version as a two-minute step test (Bohannon and Crouch, 2019; Rikli and Jones, 1999) and the 30-second chair-stand test (Jones, Rikli, and Beam, 1999), respectively. The two-minute step test requires patients to march in place, raising their knees to a height halfway between the patients’ iliac crest and patella, and has proven to have good validity and reliability for the assessment of exercise capacity in older patients (Bohannon and Crouch, 2019; Rikli and Jones, 1999). For the chair-stand test, the times that a patient comes to a full stand within 30 seconds are counted, when rising from a chair with their arms crossed in front of their chest. Research has shown good reliability and validity for the assessment of strength in older patients (Jones, Rikli, and Beam, 1999). Further modifications entailed elaborate directions for increasing exercise intensity and volume (Appendix A) and physical activity in old patients with CVD; specific instructions on when and how to communicate and collaborate with the community nurse; and how to act in the case of an emergency in the home setup. The CCB-CR protocol also consisted of recommendations for goal setting, exercise intensity, and exercise forms tailored for these patients (Figure 1). Following discharge from the hospital, the first 3 weeks of home-based CR had 2 sessions/week which focused on familiarization with exercise at Borg RPE (Borg Rating of Perceived Exertion) of 10–11 points (Noble et al., 1983), and gain in muscle strength and increasing participation in activities of daily living (ADL) as indicated by an increase from 250 metabolic equivalents (METs) in the first week to 350 METs by week-3. In the following weeks, week 4–6, the prescribed sessions were once-a-week, ADL performance was further increased by 50 METs/week to achieve 500 METs by week-6, and exercise intensity was raised to Borg-RPE 12–13 indicating improved functional capacity, and if possible, the aerobic capacity.

Patient population

Patients aging ≥ 70 years falling in the high-risk category of functional loss measured by the Dutch Safety Management System (DSMS) screening instrument (Appendix B) and admitted for more than 48 hours to the cardiology or cardiac surgery departments of the participating hospitals, were eligible for participation in the CCB study. The DSMS screens for four geriatric conditions: 1) limitation in Activities of Daily Living (ADL); 2) falls; 3) malnutrition; and 4) the risk of delirium. Patients are considered at high risk of functional decline if they are 70–79 years old with ≥ 2 geriatric conditions, or aged ≥ 80 years with ≥ 1 geriatric condition (Heim et al., 2015). The exclusion criteria followed were: failure to provide informed consent and follow instructions owing to severe cognitive impairment (Mini-Mental State Examination, MMSE < 15); known case of congenital heart disease; a life expectancy of ≤ 3 months; those transferring to another hospital or nursing home; and unable to communicate in the Dutch language. The CCB-research nurses collected baseline data for patients during their hospital admission (Verweij et al., 2018).

Physiotherapists

PTs enrolled in the CCB study were either members of a local PT network ‘LoRNA, affiliated with homecare organizations or were experienced in the field of rehabilitation for CVDs along with the relevant educational qualification necessary to provide CR. Additionally, they had at least one year of experience with home-based care and treatment for old patients with comorbidities. These PTs received five additional training sessions on cardiac disease and frailty, and on the CR protocol followed in the CCB study. The training sessions covered the following items: 1) pathology of congestive heart failure; 2) adaptations of exercise frequency, intensity, time, and type (FITT-factors) according to comorbidity and frailty; polypharmacy; motivational interviewing; when and how to collaborate with CCB community nurses; and practical application of the study protocol (Verweij et al., 2018). The PTs in the CCB study were then approached for the current qualitative study after they had completed the provision of the home-based CR program for at least one patient.

Data collection and measurements

The process of data collection consisted of observations and interviews performed by MT, a PT researcher with experience in CR and trained in performing qualitative research. The observations and interviews were structured according to the Gurses et al. (2010) framework (Appendix C) on compliance to evidence-based guidelines.

The observations regarding PTs’ adherence/non-adherence to the CCB-CR protocol and identifying the potential barriers and facilitators were executed in weeks 2–5 of the treatment period. MT recorded patient’s characteristics (e.g. present comorbidity, motivation), provider’s characteristics (e.g. displayed habits), and the system characteristics (e.g. serviceability of the patient registry, used exercise materials, and measurement instruments) on the observation form (Appendix D).

Following an observation, MT interviewed the PT about their experiences regarding the feasibility of the home-based CCB-CR protocol and perceived barriers and facilitators. Similar to the observation-phase, demographic characteristics (i.e. age, gender, work experience, and education) for all PTs were recorded. The interview topics (Appendix E) based on Gurses’ framework were: 1) provider characteristics (e.g. knowledge of the content of the CCB-CR protocol); 2) guideline characteristics (e.g. compatibility of the CCB-CR protocol with daily practice, experienced complexity of the protocol); 3) system characteristics (e.g. experienced workload and work environment); and 4) implementation characteristics (e.g. involvement of their organization’s management) (Gurses et al., 2010). Each interview lasted approximately 45 minutes and all proceedings were audio-recorded and transcribed ad verbum. Our sample size was based on when data saturation (i.e. when the same information is repeated) was reached (Malterud, Siersma, and Guassora, 2016).

Data analysis

First, two researchers (MT and LV) independently coded (coding category based on Gurses’ framework, Appendix C) the extended field notes of the observations and “ad verbum” transcriptions of the interviews in Max-QDA 12 (Gurses et al., 2010). Subsequently, axial and selective coding was used to compare categories and identify the central themes related to the feasibility of the CCB-CR protocol (Corbin and Strauss, 2008). The examiners added newly-identified categories to the observation and interview forms after each analysis, wherever applicable. Disagreements between the two researchers regarding the coding categories were settled by discussion. Descriptive statistics (mean and standard deviation SD, median and interquartile range IQR or percentages and range) for the baseline participant and patient characteristics were computed and tabulated.

Results

Study population

Data saturation was reached after the sixth and confirmed in the seventh and eighth observation and interview. The eight observed and interviewed PTs (median age 45 years, IQR 27–57) had a median work experience of 20 years (IQR 5–33) (Table 1). Four PTs were skilled in CR and seven had practiced geriatric care. All eight PTs had more than one year of experience in home-based treatment, and seven had previously worked in multidisciplinary teams. All PTs had treated at least one patient in the CCB before observations and interviews took place. The cumulative number of patients treated by the PTs at the time of observations was 21.

Table 1. Characteristics physiotherapists

Therapist (n = 8)	A	B	C	D	E	F	G	H	Total (Median)
Age (years)	23	64	25	56	54	58	57	28	45 (IQR 27–57)
Gender (m/f)	f	m	f	f	f	f	f	m	Male: 25%
Work experience (years)	2	38	2	32	10	35	29	6	20 (IQR 5–33)
Specific work experience (More than one year)
Cardiac Rehabilitation (CR)	-	-	-	-	+	+	+	+	n = 4 (50%)
Geriatrics	+	+	-	+	+	+	+	+	n = 7 (87.5%)
Home-based treatment	+	+	+	+	+	+	+	+	n = 8 (100%)
Working in multidisciplinary team	+	+	-	+	+	+	+	+	n = 7 (87.5%)
Education on CR	-	-	-	-	+	MSc	+	-	n = 3 (37.5%)
Education Geriatrics	-	+	-	-	MSc	MSc	-	-	n = 3 (37.5%)
Other relevant education	COPD	Home care	-	Home care	-	COPD PAD MI	COPD	-

CR = cardiac rehabilitation; COPD = Chronic Obstructive Pulmonary Disease; PAD = Peripheral Artery Disease; MI = motivational interviewing; MSc = Master of Science level; IQR = Interquartile range; n = number; – = no; + = yes

The average age of patients in the observed treatment sessions was 81 ± 8.2 years, of which five were males (Table 2). The majority of the patients had at least two geriatric conditions out of: 1) limitations in Activities of Daily Living (ADL); 2) falls; 3) malnutrition, and 4) risk of delirium according to the DSMS screening. All patients were diagnosed with congestive heart failure and the cause for hospital admissions were decompensated congestive heart failure (n = 5), endocarditis (n = 1), angina pectoris (n = 1), and pacemaker implantation (n = 1). The following comorbidities were reported – diabetes mellitus (n = 4), peripheral arterial disease (n = 2), chronic obstructive pulmonary disease (n = 1), knee osteoarthritis (n = 1), stroke (n = 1), and renal failure (n = 2).

Table 2. Baseline characteristics of the patients (n = 8) in the observed treatment sessions

Characteristic	Mean (SD) or n (%)
Age (years)	82.8 (SD 7.8)
Gender (male)	5 (62.5%)
Nationality
Dutch	6 (75%)
Suriname	1 (12.5%)
German	1 (12.5%)
Highest Education
Primary school	1 (12.5%)
Secondary school	2 (25%)
Lower professional education	3 (37.5%)
Higher professional education or university	2 (25%)
Screening instrument for frail older adults of the Dutch Safety Management System (DSMS)
Score 1	1 (12.5%)
Score 2	5 (62.5%)
Score 3–4	2 (25%)
Mini-Mental State Examination (MMSE)	25.1 (SD 2.9)
Reason hospital admission
Heart failure	5 (62.5%)
Atrial Fibrillation	1 (12.5%)
Endocarditis	1 (12.5%)
Unstable angina	1 (12.5%)
Number of medicines	10.3 (SD 1.5)
Charlson comorbidity index
0 comorbidities	2 (25%)
1–2 comorbidity	3 (37.5%)
3–4 comorbidities	3 (37.5%)

SD, standard deviation; n, number; DSMS: 1 point is scored for each present geriatric condition (max 4): limitations in Activities of Daily Living (ADL), falls, malnutrition and risk of delirium; MMSE: score < 24 indicates cognitive impairment

Adherence and feasibility themes

Table 3 summarizes the reasons reported by PTs for adherence or non-adherence to the CCB-CR protocol (Gurses et al., 2010). Mean intercoder agreement in the analyses of the combined observations and interviews was 94.5% (range 90.3–100%). Our analyses revealed three sets of themes that influenced the practicability of the protocol. The subsequent paragraphs elaborate on each theme structured according to the system, provider, guideline, implementation, and patient characteristics (Gurses et al., 2010). Briefly, the first theme encompasses the feasibility of exercise testing and the exercise program. The second theme covers the motivational aspects of home-based CR, consisting of both patient motivation or self-regulation and the use of motivational techniques by the PT. The third theme elucidates on the interdisciplinary-collaboration and task-division between the community nurse and other healthcare providers in monitoring the patients’ health status and risks of hospital readmission.

Table 3. Reasons for physiotherapists’ adherence or non-adherence to the CCB Cardiac Rehabilitation protocol

Characteristics Gursus	Adherence	Non-adherence
System characteristics	Positive culture for change Preexisting network for organized teamwork (e.g. community care organization) Availability of measurement instruments (e.g. for blood pressure, oxygen saturation)	Task: high workload, time restrictions Tools: Manageability of paper care file No functioning communication transfer system from hospital to physiotherapist (e.g. secure e-mail system)
Provider characteristics	Positive attitude and outcome expectancies intervention Motivation to participate in research Normative beliefs that measurements should be performed properly Familiarity with tools (BORG, METs) Knowledge of content CCB exercise protocol	Usual habits of practice with frail elderly Preexistent high level of expertise with frail elderly Low outcome expectancies of parts of the program (e.g., application of METs or aerobic exercise)
Guideline characteristics (provided exercise program)	Low complexity of exercise program Observability of patient improvements	Limited compatibility with patients Limited time for try-out (only nine sessions of 30 minutes)
Implementation characteristics	Intervention supported by direct leadership Financial support for the intervention	Limited financial support for long term implementation More time needed for treatment and traveling than was financed
Patient characteristics	High level of motivation to be independent Higher activity level before hospital admission	Comorbidities: diabetes, wounds, gout, COPD, anemia, knee osteoarthritis, kidney failure, stroke, psychiatric condition Low level of motivation Sedentary lifestyle before hospital admission

BORG = Borg scale of perceived exertion ranging from 6–20, MET = Metabolic Equivalents Table, COPD = Chronic Obstructive Pulmonary Disease

Theme 1: feasibility of exercise testing and the exercise program according the CCB-CR protocol

System characteristics

Most PTs reported that the most common barrier was the administration of exercise tests and the time consumed in taking repeated blood pressure measurements. It was further observed that availability of sufficient time facilitated the application of the Borg-RPE scale and METs table, supplemented with the availability of practical tools like dumbbells and mobile sitting bicycle and the PT’s experience in using the equipment, encouraged the PT to prescribe higher intensity exercises. As a facilitator to adhere to the CCB-CR protocol, some PTs reported a positive culture for change in the organization, (quote 1 (Q1)).

(Q1) “Their (the management’s) opinion is that you should improve yourself every time by following education or by any way possible. So, they create space to do this (e.g. for participation in the CCB intervention).” (PT C)

Provider characteristics

As facilitators for adherence to the CCB-CR protocol, most PTs reported an increase in the patients’ physical activity levels (Q2) resulted in a positive attitude and expectations of better outcomes from the home-based CR. All PTs reported a positive attitude about participating in research and displayed sufficient knowledge of the CCB-CR protocol during the observations.

(Q2) “Nine out of ten patients sit too much and that won’t help them. So, I am convinced that the intervention can prevent hospital readmissions.” (PT C)

All PTs adhered to the suggested exercise tests (i.e. two-minute step test and 30-second chair stand test) and described them as feasible for evaluating patients’ endurance and strength (Bohannon and Crouch, 2019; Jones, Rikli, and Beam, 1999). Most PTs did not frequently use the table with metabolic equivalents of tasks (METs), which was intended to provide insight into patients’ daily activities energy expenditure (Ainsworth et al., 2000). Additionally, most PTs reported a lack of knowledge on how to calculate METs and their non-applicability to frail patients and preferred the “patient-specific functioning scale” or the activity log (Cleland, Fritz, Whitman, and Palmer, 2006). Likewise, most PTs reported difficulty in applying the Borg-RPE scale of perceived exertion to estimate exercise intensity due to the patients’ limited ability to discriminate between the scale’s levels (Q3) (Noble et al., 1983).

(Q3) “With this lady, I didn’t even try to explain the BORG properly, because it just costs too much time and energy in contrast to what it yields. That is something you judge within two times when they indicate the same score.” (PT E)

It was observed that all PTs prescribed functional exercises (e.g. walking, climbing stairs) as recommended in the CCB-CR protocol. All PTs reported that their objective of home-based CR was not an improvement in cardiovascular fitness alone, but encouraging the patients to improve their physical functioning and self-management.

Guideline characteristics

All PTs reported that adherence to the CCB-CR protocol was facilitated by the simplicity of the protocol (Q4), except for the description of the METs. Some PTs also reported a need for more examples of home exercises.

(Q4) “Well, I think the protocol, which can be found in the CCB file online, is really quite clear. I mean considering its complexity. I don’t think it is very complex.” (PT D)

Patient characteristics

A reported barrier for exercise therapy was the presence of a psychiatric condition. To adapt the exercise therapy to the patient’s characteristics, most PTs indicated that they adjusted the following FITT-factors: 1) training intensity because congestive heart failure limited the patient’s exercise capacity; 2) exercise type and intensity because comorbidities limited the exercise possibilities (e.g. gout); and 3) timing, because exercise had to be postponed if kidney failure was present. For patients with higher activity goals PTs were able to prescribe aerobic training (e.g. walking or cycling > 5 minutes) in the last two weeks of the treatment period. For patients with lower activity goals, PTs deemed that aerobic training was not feasible. Exercise intensity could vary between training sessions and no PTs reported any adverse events in patients during or after the exercise sessions. The researcher did not observe and neither did any PT report any implementation characteristics related to this theme.

Theme 2: motivational aspects

Provider characteristics

All PTs spent time on motivating patients to increase daily activities and the patients’ self-regulation, as instructed in the CCB-CR protocol. PTs used goal setting to match the exercise needs of the patients, like exercising to walk independently to the mailbox. However, some PTs found it difficult to motivate sedentary patients, while it was easier if the patient was self-motivated to be independent in daily activities (Q5). All PTs had to balance the time spent on motivational interviewing and exercise. One PT prioritized patient-coaching over exercise therapy, while some PTs indicated a need for further training in motivational techniques that may be used in older patients with CVD.

(Q5) “No, because he is committed, because he really wants to. That is to say: his youngest son is handicapped. And he really wants to be able to independently take a cab to go to his son, and then be able to walk with his walking aid together with his son and go to a restaurant for a cup of coffee. And that is just not possible at this moment.” (PT D)

Further, PTs inspired self-management in patients by spreading CR-sessions over a longer period than the suggested six weeks, asking open-ended questions, and letting the patient formulate their own goals and concrete actions, such as restrictions of fluid intake in congestive heart failure (Q6).

(Q6) “How should we approach this? Do you want to do it yourself? Or with a dietician? Or do you want to use a fluid diary?” (PT F)

Patient characteristics

Some PTs indicated that it was harder to motivate patients for exercise or physical activity if the patient had a sedentary lifestyle or anxiety (Q7). They also reported that two patients had dropped out from physiotherapy because they found the number of care workers visiting them overwhelming.

(Q7) “No. I already did it this morning. I’m finished. I’m afraid to end up horizontally.”

No relevant system, guideline, or implementation characteristics for this theme were observed or reported during the PT interviews.

Theme 3: interdisciplinary collaboration in monitoring health status and risks of readmission

System characteristics

All therapists monitored blood pressure, heart rate, and weight. Some PTs reported that occasionally, the home-care nurses measured and reported the patients’ weight, allowing the PT or CCB-community nurse to focus on the interpretation of the measurement outcomes.

All PTs indicated that communication with other caregivers was facilitated through personal interactions (like from the training sessions or joint intake), and having collaboration-agreements with home care organizations. An often reported barrier for inter-professional communication was the absence of an institutional, secured communication system. All PTs preferred to communicate by telephone or e-mail if a situation was urgent, or through the documented patient-log, if less urgent.

Provider characteristics

All PTs said they were more aware of their responsibility in the interdisciplinary team to monitor patients’ health status by measuring blood pressure, heart rate, and weight, and reported three potential risk situations for hospital readmission, namely hypotension after medication changes, weight loss (i.e. risk of sarcopenia), and weight gain (i.e. potential decompensated congestive heart failure) (Q8). Most PTs said they were satisfied with their collaboration with the community and hospital nurses (Q9).

(Q8) “I had measured the blood pressure, which was quite low. And then it turned out that she had stopped with a medicine. She didn’t really know which one. Luckily, someone from homecare came along and found out which medicine it was.” (PT C)

(Q9) “Yes, I really appreciate that (meaning close collaboration with the nurse from the hospital). Because last week I called, because he had a low blood pressure, around 100 (systolic), so that is fine, but he was worried about it. I said: “Well, if it worries you. We discussed dizziness and other signs to monitor and how to respond when this happens ….” “Yes, but I do worry about it”. … I said: “Well, let’s call.” So, I called the hospital. I got her on the phone right away (meaning the cardiac nurse). She said the same as I did. And that, yes, that. He felt understood. And that was really satisfying.” (PT F)

Neither did the researcher observe, nor did the PTs report any guideline, implementation, or patient characteristics relevant to this Theme 3.

Discussion

The present study efficiently demonstrates that PTs find home-based cardiac rehabilitation (CR) to be feasible in frail older patients with CVD. Accordingly, no adverse events were reported. We identified three main issues that may be targeted for further improvements in home-based CR in these patients: 1) room for personalization of exercise testing and exercise intensity according to patients’ capabilities; 2) accounting for patients’ motivation and level of self-regulation when formulating activity goals; and 3) facilitating interdisciplinary communication in primary care. Through well-executed interviews, PTs reported experienced leadership as a facilitator and timing issues as an important barrier for the implementation of home-based CR. Our study is a first step toward offering frail older patients with CVD an alternative to traditional hospital-based CR, and the results may aid in developing a personalized, home-based CR program for this population.

We observed sufficient PT-reported CR-content acceptance and satisfactory levels of PT protocol adherence, ascertaining the feasibility of our home-based CR program. Based on our results and supported by the current literature, the guidelines for exercise prescription should, ideally, provide sufficient room for customization of exercise intensity and volume according to patients’ performance capabilities (O’Neill and Forman, 2019). This recommendation translates into CR-exercise protocols that suggest a range of exercises from low to high intensity, beginning with improving the patients’ muscle strength through functional training and practicing ADLs in the first three weeks of CR (e.g. chair rising for one frail patient and stair climbing for another) (Khadanga, Savage, and Ades, 2019; O’Neill and Forman, 2019; Tamuleviciute-Prasciene et al., 2018). For patients who can perform more than 5 minutes of exercise at > 3 METs, CR-exercise protocols can incorporate endurance-building exercises according to the patient’s activity goals. Although the 6MWT and incremental shuttle walk test are valid and reliable in clinically stable patients (Bellet, Adams, and Morris, 2012; Pepera, McAllister, and Sandercock, 2010) we chose to evaluate physical capacity within the patient’s home setup with the 2-minute step test, and the 30-second chair stand test, which were practical alternatives for the regular hospital-based cardiac exercise testing protocol (Jones, Rikli, and Beam, 1999; Rikli and Jones, 1999). Validity and reliability of these tests were demonstrated in various older patient populations, but our study is the first to demonstrate the feasibility in frail cardiac patients (Bohannon and Crouch, 2019). Minimal clinical important difference of these tests in frail patients should be further investigated. In addition, the present study’s findings suggest that a physical-activity log or METs table was less feasible in older patients with CVD. A more attractive alternative for the home situation can be the use of an activity tracker, which provides the patient an insight into their recovery process while encouraging them to become more active and engaged (Bostrom, Sweeney, Whiteson, and Dodson, 2020; Forman et al., 2014; Pol et al., 2019).

Our findings reaffirm the concept that the individual’s level of motivation for physical activity and self-regulation in undertaking and sustaining such activities should be considered when formulating activity goals for frail older patients (Nied and Franklin, 2002). Additional recommendations to communicate these goals to the target geriatric population, such as: tailored information about the risks and benefits of physical activity in this population (Bell and Saraf, 2014); motivational interviewing techniques (Schertz et al., 2019); and tools that support shared-decision-making (Backman, Levine, Wenger, and Harold, 2020), can potentially help in personalizing CR-protocols for frail older patients.

Finally, secured transmural interdisciplinary communication systems may aid primary care PTs in monitoring and exchanging health and risk information. Current communication systems and electronic patient records vary with organizations, forcing health professionals to fall back to less secure communication channels such as writing, faxing, e-mail, or telephone.

Our results are corroborated by systematic reviews and the American Heart Association’s guidelines, in which they reported that home-based CR was as effective and safe as center-based CR in patients with low-to-moderate risk (Doll et al., 2015; Taylor et al., 2015; Verweij et al., 2019). In addition, home-based CR has shown to be cost-effective, and improves QALY’s in patients with heart failure (Bakhshayesh et al., 2020). O’Neill and Forman (2019) also reaffirmed the safety and effectiveness of CR in older patients. The results of our study ascertain that home-based CR is also feasible in very old (above 80 years) and frail populations and it is possible to organize home-based CR in primary-care PT practices in the Netherlands. Patients’ adherence and satisfaction to the CCB-intervention have been evaluated and published separately (Jepma et al., 2021a; Verweij et al., 2021).

Strengths and limitations

There are several strengths of the study. We used a qualitative design that combined observations with consecutive interviews, thereby obtaining a unique in-depth insight into how contextual variables and characteristics of frail older patients influence a PT’s professional practice (Jensen, Gwyer, and Shepard, 2000). This way we were able to evaluate the observed-adherence of PTs to the suggested CCB-CR protocol and identify points of improvement for improving the feasibility of home-based CR in frail older patients. Second, by separately coding all transcripts we reduced the risk of confirmation bias. The different professional backgrounds of the coders (nurse and PT) resulted in an inter-professional assessment of the relevant themes that influenced feasibility. After a consultation, the researchers attained an excellent intercoder agreement which supports the robustness of our study outcomes. Finally, the PT-sample varied in age, experience, and education, giving a realistic representation of Dutch primary care PTs. The review of Taylor et al. shows that PTs in other countries like Australia, Canada, and the United Kingdom were able to perform home-based CR in non-frail patients (Taylor et al., 2015). Our findings add to the body of knowledge by suggesting that trained primary care PTs can also utilize home-based CR in frail older patients.

Some aspects of our study warrant consideration. First, the sample size in our study was small, potentially limiting the scope of our findings. However, our observations and interviews with the last two PTs did not reveal new codes or themes. Combining data from observations and interviews may have accelerated saturation. Second, not all included patients in the CCB-study were willing to participate in home-based CR. Also, PTs reported that two patients had dropped out of the home-based CR component of the CCB-study because they found the many healthcare professionals visiting them to be overwhelming. Our result may therefore not apply to highly unmotivated frail patients. Finally, our study was conducted in the Netherlands. Studies in other countries with their specific healthcare systems may provide additional insights on the local challenges for implementation.

Recommendations for further research and implementation

Findings in our pilot study indicate that home-based CR is feasible, but with adjustments to meet the needs of frail older patients, such as: 1) replacing the usual exercise tests like the 6-minute walk test with more functional (i.e. 30-second chair stand test) and feasible tests (e.g. 2-minute step test) (Rikli and Jones, 1999); 2) using different tools for measuring activity levels (e.g. activity tracker instead of METs) (Theou, Jakobi, Vandervoort, and Jones, 2012); and 3) adapting exercise type and intensity according to the patient’s comorbidities and capability (Dekker, de Rooij, and van der Leeden M, 2016). Further research is needed to confirm the applicability and responsiveness of exercise tests and to select the best tool for measuring daily activity levels. It is unclear which exercise components lead to the best results regarding physical functioning, hospital readmission, and mortality. Sandercock, Cardoso, Almodhy, and Pepera (2013) found a limited effect on mortality and morbidity for low volumes of exercise prescription, however for this frail population aiming on physical independence seems more important. The optimal amount of exercise and physical activity remains unclear. Since there are no norms for frail older cardiac patients (Tudor-Locke et al., 2011) exercise programs should be tailored based on patients capacity, activity levels, and personal goals. To objectively assess the effects of physical therapy in frail older patients, activity levels (e.g. activity monitor) are more practical than the VO₂-max measures (Brubaker et al., 2009; Witham, Daykin, and McMurdo, 2008). Home-based cardiac rehabilitation including tele monitoring showed similar outcome effects and higher satisfaction rates in the low- to moderate risk patients, and may also be feasible for coaching frail older cardiac patients to increase physical activity (Brouwers et al., 2020). Future research should investigate whether this approach will lead to improved activity in this older population, whilst PTs experience difficulty in motivating these patients face to face. Considering the heterogeneity of frail older patients with CVD, more research analyzing subgroups within this population is needed to identify groups for whom home-based CR is feasible, safe and effective (Vaduganathan et al., 2015).

Conclusion

This study suggests that primary care PTs found the CCB-home-based CR protocol to be feasible, and that it provided practical guidance in the treatment of a sample of frail older patients. Important challenges to further improve home-based CR are the identification of an optimal level of CR intensity, selection and identification of exercise tests and measures that are suited for home-based CR, and the integration of interventions to optimize the patients’ self-regulation.

Acknowledgments

We are grateful to the participants in the pilot of the CCB transitional care program and the contribution of all participating physiotherapists and nurses. We especially thank Ferdinand de Haan for his work on the CCB-CR protocol.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Netherlands Organisation for Health Research and Development [520002002]; nederlandse organisatie voor wetenschappelijk onderzoek [0.23.008.024,023.013.017].

Appendices Appendix A. Directions for increasing exercise intensity and physical activity in old patients with cardiovascular disease

Table

General information and instructions It is well known that shortly after hospital discharge it is important to improve activities of daily living (ADL). Patients’ independence and physical health can deteriorate. Therefore, it is important to monitor their medical status, e.g. cardiac decompensation, recurrent angina, arrhythmias. Knowledge of these conditions (Dutch cardiac rehabilitation guideline: https://www.kngf2.nl/binaries/content/assets/kennisplatform/onbeveiligd/guidelines/cardiac_rehabilitation_practice_guidelines_2011.pdf) (Achttien et al., 2011) is considered a prerequisite. If in doubt about the cardiovascular status of the patient, please consult the cardiac team in the hospital. For other questions or serious adverse effects, contact the general practitioner. The structure of the exercise program is primarily based on the expertise of the healthcare professional. There is no fixed training schedule because the situation can differ too much per patient. Therefore, per week the structure is outlined. The following principles do apply: an important goal is to strive for the exercise standard of 30 minutes of moderate physical activity per day; better aerobic endurance reduces the risk of one year mortality; given the age, it is very important to match the exercise program with the goals and wishes at the limitation and participation level of the patient. Before each training session, always: check the medical status of the patient and whether he/she has recovered from the previous training session; check the resting heart rate and, if indicated, the blood pressure. Start slow and build up slowly. Start with muscular endurance first, followed by muscular strength and only later in the course build up aerobic endurance. When the exercise capacity is very low, you can train in the phosphate system (training stimulus < 15 seconds (sec), recovery 30 sec. or more if necessary). Train functionally as much as possible, e.g. climbing stairs, carrying shopping bags.
Week and Objective	Frequency	Intensity	Type and Time
Week 1–2 Including intake: Medical status Check vital functions Set activity goals. 2MST 30 sec chair stand test Setting up functional exercises (ADL) Accustom to physical exertion Removing uncertainties Advising what is permissible Promote healthy lifestyle Support self-management	2 x per week 30 minutes	11–12 RPE on the Borg scale Achieve 250 MET in the first week Achieve 300 MET in the second week	- Muscle endurance 10 min - Muscle strength 5 min - Aerobic endurance 5 min Integrate muscle strength and muscular endurance exercises into functional activities as much as possible If possible, strive for 2 minutes of stepping on a 20 cm high step in the 2^nd week Step at patients own pace Stepping up a bench is a powerful cardiovascular training stimulus
Week 3–4 Check how the patient responds to exertion after the habituation phase If the patient’s capacity allows it, try to increase the exercise intensity Interim evaluation: repeat 2MST and 30SCST to evaluate and show progress Identify risk situations and barriers to healthy exercise behaviors – Build new skills & routines	2 x per week And 1 x per week	11–13 RPE on the Borg scale Achieve 350 MET in the third week Achieve 400 MET in the fourth week	Muscle endurance 10 min Muscle strength 7,5 min Aerobic endurance 7,5 min Integrate muscle strength and muscular endurance exercises into functional activities as much as possible If possible, build up to 3 minutes of stepping up 20 cm high step (stair step) Train ADL activities, e.g.: walking; climbing stairs; lifting; vacuuming; mopping; dancing
Week 5–6 Build up to the improvement of the aerobic endurance Final evaluation ninth session and adjust program: repeat 2MST; repeat 30SCST	1 x per week	12–13 RPE on the Borg scale Strive for an intensity of 13 RPE on the Borg scale during exercise, depending on the patient’s load capacity. Achieve 450 MET in the fifth week Achieve 500 MET in the sixth week Maintain 450–500 MET per week	Muscle endurance 5 min Aerobic endurance 20 min Integrate muscle strength and muscular endurance exercises into functional activities as much as possible If possible, in the sixth week see if 4 minutes or longer of stepping on a 20 cm high step at own pace is feasible Train ADL activities, e.g.: walking; climbing stairs; lifting; vacuuming; mopping; dancing

ADL = Activities of Daily Living; Sec = Seconds; Min = Minutes; 2MST = Two Minute Step Test; 30SCST = 30-Second Chair Stand Test; RPE = Rate of Perceived Exertion; MET = Metabolic Equivalent of Task

Appendix B. Screening tool for frail elder of the Dutch Safety Management System (DSMS)

Table

Risk domain	Instrument	Questions	Cut-off	Score^a
Fall risk	Single question	Did you fall in the last six months?	yes	1
Malnutrition	SNAQ [2]	Assessing whether the patient: 1) lost weight unintentionally in the last month (>3 kg) or last six months (>6 kg) and/or 2) experiences a decreased appetite and 3) used supplemental drinks or tube feeding	Question 1 = yes or Question 2 + 3 = yes	1
Delirium	Single questions	Assessing whether: 1) the patient has cognitive impairment; 2) the patient needed help with self-care in the last 24 hours; 3) the patient has previously undergone a delirium	≥ 1 point	1
ADL-limitations	KATZ-6 [3]	Assessing whether the patient needs help with: 1) bathing, 2) dressing, 3) toileting, 4) transferring from bed to a chair, 5) eating, and 6) whether the patient uses incontinence material	≥ 2 points	1
Total score				0–4

^aPatients are at high risk of functional decline if aged 70–79 years and score ≥ 2 or aged ≥ 80 years and score ≥ 1; ADL-limitations = Limitations in Activities of Daily Living; KATZ-6 = Modified KATZ-6 index; kg = kilogram; SNAQ = Short Nutritional Assessment Questionnaire

Appendix C. Theoretical framework of Gurses

Figure C1. Conceptual interdisciplinary framework of clinicians’ compliance with evidence-based guidelines, by Gurses 2010.

The framework of Gurses (Figure 1) shows the expected interrelationships among four major categories of factors (System, Provider, Guideline, and Implementation characteristics) that influence guideline compliance. System (e.g. a checklist or communication system) and guideline characteristics (e.g. relative advantage, complexity, compatibility) can positively and negatively influence provider characteristics (e.g. awareness, agreement with the guidelines, self-efficacy). Pre-existing characteristics influence compliance with the guideline through their impact on implementation characteristics. These implementation characteristics (e.g. tension for change) function as mediators and moderators of the pre-existing characteristics toward the behavior of clinicians’ compliance or noncompliance to guidelines. For example, low self-efficacy (provider characteristic) may diminish implementation quality and thus lead to low compliance (mediation). Or, as a moderator, adequate implementation can diminish the impact of low self-efficacy. Clinicians’ compliance to guidelines influence patient outcomes, but patient outcomes are also influenced by patient characteristics. For example, present comorbidity may influence guideline adherence, e.g. PTs may prescribe too low exercise intensity because they don’t understand complex interactions between diseases and as a consequence decide to underload the patient just to be safe. Finally, next to unintentional errors (e.g. forgetting to check glucose levels before and after exercise), careful and deliberate clinical decision-making may lead PTs to intentionally deviate from the guidelines (e.g. postponing aerobic exercise due to a COPD exacerbation).

Appendix D. Observation form adherence/non-adherence to study protocol provided in CCB training

Table

Factors	Barriers	Possible actions
Caregiver
Knowledge about the protocol/guideline Does the caregiver know how to follow the protocol/guideline?
Attitude toward the protocol/guideline Does the caregiver believe that following the protocol/guideline is of value?
Usual habits in daily practice What does the caregiver normally do in comparable situations?
Work environment
Task Who is responsible for following the protocol/guideline?
Means and technical tools Which materials are available and are being used? (e.g. training material: e.g. dumbbells)
Decision support How often are tools available and how often are they in use?
Physical environment In what way does the environment influence adherence or non-adherence to the protocol/guideline?
Organizational structure In what way does the organizational structure influence adherence or non-adherence of the protocol/guideline?
Administrative support In what way does administrative support structure influence adherence or non-adherence of the protocol/guideline?
Monitoring/feedback of performance? How does the caregiver know that they adhere to the protocol/guideline?
Culture in the organization In what way does the culture in the organization influence adherence or non-adherence of the protocol/guideline?
The CR protocol/guideline
Applicability to patients in the CCB study
Difficulty/Ease of following the protocol/guideline In what way does the protocol/guideline influence the workload of the caregiver?

Appendix E. Interview form adherence/non-adherence to study protocol provided in CCB training

Table

Factors	Exemplary questions
Caregiver
Knowledge of the protocol/guideline Does the caregiver know how to follow the protocol/guideline?	Can you explain in what way you apply the protocol/guideline in daily practice? Can you provide an example? How new is the information in the protocol for you? How do you experience the extensiveness of the protocol? How do you experience the complexity of the protocol? What is your opinion about the applicability of the BORG-scale, MET-scale? What is your opinion about monitoring vital parameters such as: weight, medication, blood pressure?
Attitude toward the protocol/guideline Does the caregiver believe that following the protocol/guideline is of value?	Can you explain how important it is for you to adhere to the protocol? Can you give an example of what will or will not happen if you do not follow the protocol? Which parts of the protocol do you disagree with? To what extent do you feel capable of following the protocol? How hard is it for you to follow the protocol? What is your opinion about the prescribed intensity in the protocol? To what extent did you experience a need to implement this protocol in daily practice? What is your experience with interprofessional collaboration within this study? Did you have contact with another caregiver? In what way? What was your contact about? What went well/wrong in interprofessional collaboration? What is your opinion about performing measurements in this study? (BORG/MET/Exercise tests: two-minute step test)
Usual habits in daily practice What does the caregiver normally do in comparable situations?	Can you explain the usual procedure prior to this protocol? How satisfied are you with the usual procedure? To what extent is it possible to change “old” habits in accordance with the study protocol?
Self-experienced compliance with the study protocol How often does the care provider do everything according to the study protocol?	Can you tell me to what extent you are currently following the study protocol? E.g. 80%/90%. What do you encounter when you do or do not follow the protocol? To what extent does the therapist need advice in the protocol regarding abnormalities, e.g. due to comorbidity?
Work environment
Task Who is responsible for following the study protocol?	Can you tell who in your institution is/are responsible for following the study protocol? What is the general workload? How often do you have a lot to do? How much time do you have to carry out the study protocol? How tired were you last week? How tense did you feel last week? How satisfied are you with the quality of your care? How satisfied are you with your work? Social norms To what extent did you feel pressure to follow the study protocol? To what extent were you rewarded for following the study protocol?
Means and technique What stocks and materials are available or in use?	What materials are currently available to carry out the care according to the protocol? Think of checklist, training schedule, logbook, workbook, point software, digital map. How up to date is this material? Also think about blood pressure meter/saturation meter etc. What materials are actually used?
Decide support How often are tools available and used?	What tools are available to follow the study protocol? (e.g. overview map, workbook)
Physical environment In what way does the environment influence whether or not the study protocol is followed?	Can you tell how the environment influences whether you can follow the study protocol? For example, how busy or noisy is the environment?
Organizational structure In what way does the organizational structure (e.g. staffing) influence whether or not the study protocol is followed?	Can you tell me how the organizational structure affects whether or not you can follow the study protocol? How do you normally experience the effectiveness of your organization? To what extent do you experience support from your organization for following the study protocol?
Administrative support In what way does the current administrative support influence whether the study protocol is followed?	Can you tell me how the administrative support affects whether or not you can follow the study protocol? Is information available on time?
Monitoring/feedback on performance How does the care provider know that they are/are not following the study protocol?	Can you tell me how you get feedback on following the study protocol? If applicable: Do you need that feedback? Can you tell how this is monitored? How do you experience the financial support for the implementation according to the study protocol?
Culture How does (working) culture influence whether you follow the study protocol?	Can you tell me in what way the (working) culture influences whether or not you are able to follow the study protocol? How is the culture (e.g. open, about deviations from prescribed actions may be communicated)?
The study protocol
Applicability to patients in the CCB Is the study protocol applicable to the specific patients in the cardiological care bridge? Think of: age, type and severity of comorbidity, patient motivation, degree of independence.	Can you tell me to what extent the study protocol is applicable to the patients you have seen within the cardiological care bridge? To what extent do you feel that you have been able to give cardiac rehabilitation instead of geriatric rehabilitation? To what extent does comorbidity influence the following of the protocol? To what extent do you have knowledge about the comorbidity the patient had? In what way does the motivation of the patient(s) influence following of the protocol? Do you lack knowledge in motivating the patient? To what extent did the patient’s fear influence the following of the protocol?
Ease of following the study protocol How does the study protocol influence the workload of the care provider?	Can you tell how following the study protocol affects your workload? What makes the time pressure higher when performing cardiac rehabilitation in this specific patient group? What is the benefit or additional benefit of the study protocol? To what extent do you experience room to try out the study protocol through trial and error? What do you think of the level of evidence that supports the study protocol?