Home-based Pulmonary Rehabilitation and Health Coaching in Fibrotic Interstitial Lung Disease: Implementation and Qualitative Assessment of a Pilot Telehealth Program

Abstract

PURPOSE:

Pulmonary rehabilitation (PR) is a behavioral modification intervention shown to improve exercise tolerance and patient-reported quality of life in patients with fibrotic interstitial lung disease (f-ILD). Home-based rehabilitation may provide easier access for those who struggle to complete center-based rehabilitation programs due to increased symptom burden or frailty.

METHODS:

We present the quantitative and qualitative findings of a pilot study of 21 patients with f-ILD who participated in a 12-wk home-based PR program with activity monitoring and health coaching.

RESULTS:

Pre- and post-intervention patient-reported outcome questionnaires suggested improvements in dyspnea and respiratory-related quality of life but were underpowered to meet statistical significance. Half had increases in mean daily step counts while a quarter declined due to disease progression. Qualitative analysis of semi-structured participant interviews suggested a significant baseline disease burden with related secondary impacts, including anxiety regarding disease progression and prognosis. Many who participated had no specific program expectations or self-determined goals but still found the program impactful, particularly on their abilities to adapt and cope with the disease.

CONCLUSION:

Our study suggests feasibility in a diverse set of patients with varying severity and diagnostic subtypes. We also provide quantitative and qualitative aspects of the program’s impact on patient well-being and highlight the complex interaction between measured physical and self-reported outcomes and disease experience.

CONDENSED ABSTRACT:

Pulmonary rehabilitation is an intervention shown to improve patient-reported quality of life (QoL) in fibrotic interstitial lung disease patients (f-ILD). Findings of a study of f-ILD participants in a home-based program suggests improvements in dyspnea and respiratory-related QoL. Our study suggests feasibility in a diverse set of symptomatic patients.

The fibrotic interstitial lung diseases (f-ILD) are characterized by radiologic fibrosis and restrictive physiology, often presenting with progressive dyspnea, cough, and fatigue contributing to significant disease burden and loss of quality of life (QoL)(1, 2). While current available therapies include immunosuppressants and anti-fibrotics, patients often continue to have functional decline despite directed treatment(3). Pulmonary rehabilitation (PR) is a behavioral modification intervention shown to improve exercise tolerance and patient-reported QoL in those with chronic lung disease(4, 5). Unfortunately, there is limited availability of center-based programs in both rural and urban settings and a shortage of experienced exercise therapists(6). Distance and travel complicate by the need to transport cumbersome oxygen delivery devices for care(7, 8). Home-based PR programs have been used in f-ILD and idiopathic pulmonary fibrosis (IPF) patients, demonstrating feasibility and similar adherence to center-based programs(9-13).

We implemented a home-based PR program utilizing health coaching and activity monitoring in patients with f-ILD with targeted endpoints of improved patient-reported QoL and physical activity. Using structured qualitative interviews, we assessed patient experience of disease burden, physical activity, program feasibility, and effect size.

METHODS

This was a single-site, pre-/post-evaluation of patients with f-ILD seen at Mayo Clinic (Rochester, MN) between September 2020 to August 2021. Eligible patients had radiologic fibrosis of > 10% on chest computed tomography and a modified Medical Research Council (mMRC) dyspnea score of >1. Patients completing center-based PR within 3 mo, experiencing an acute exacerbation within 4 wk, or enrolling in hospice were excluded. Additional exclusions included cognitive impairment with inability to follow directions and orthopedic/cardiac limitations precluding safe participation. There were no exclusions for baseline lung function, oxygen use, or directed medical therapy.

HOME-BASED PULMONARY REHABILITATION PROGRAM AND ACTIVITY MONITORING AND HEALTH COACHING

The Mayo Clinic Mindful Breathing Lab developed the home-based PR program through grants R44 HL114162 and tested in R01 HL140486 (R Benzo PI).

Remote physical activity (PA) monitoring and software application:

The home-based PR monitoring system (Figure 1) consists of three commercial devices and two software applications. The devices are a Garmin Vívofit 4^™, a Nonin 3150 WristOx2^® Pulse Oximeter (PO), and a 7" Android^™ tablet with WIFI or cellular service (Verizon). The Garmin Vívofit 4^™ and pulse oximeter wirelessly communicate with the tablet via Bluetooth. During the upper extremity exercise routine and walks, the PO measures the participants' heart rate (HR) and oxygen saturation (SpO₂). Near the end of the day, the tablets periodically sound a musical alarm and display short health status questions asking participants to rate their wellbeing, ease of breathing, and energy level (options are excellent-very good, good, fair, or poor). At night, the tablets securely transmit the device data and wellbeing questionnaire answers to a remote Web server, storing the data and generating reports. In addition, as part of the weekly health coaching phone calls, when the participants and the coaches will discuss their progress in PR and wellbeing.

Figure 1:

Home based Program (created in R44 HL114162 and tested in R01 HL140486)

Participant application (figure 2):

Figure 2:

Tablet Main Screen

When the participant turns on the Android^™ tablet, the application logs in, and the participant accesses the main screen. The user selects the corresponding action from the "To-Do" column to initiate one of these activities. Selecting "Exercise" from the "To-Do" column activates a video guiding the user through the PR exercises. Similarly, choosing "Walk" from the "To-Do" column starts a 6-mi countdown once the application detects that the participant has initiated walking. The application provides instructions before and between each activity. Patients are asked to follow the videos in all PR sessions to maintain a minimal and equal "dose" of PR for all participants. The PO data during exercise is used to measure adherence and exercise time. The "my journey" screen plots historical data, allowing participants to see trends (figure 3). A Message section allows to send and receive messages to the coach anytime.

Figure 3:

Weekly view in the participant’s tablet (coaches can see from anywhere they connect). The weekly encounter will collaboratively discuss the lifestyle shown in the figure daily steps, daily exercise (simple standing or sitting yoga focusing of the breath) and the daily check in questions: well-being (How do you feel today?), breathlessness (how is your breathing today), fatigue (how is your energy today? and the perception of progression to the goal of daily activity.

Initial encounter

Participants meet by phone with their coach to introduce the 12-wk program, discuss the patient's motivation and confidence with technology, propose daily exercise, assess expectations, and schedule subsequent health coaching calls.

Subsequent encounters (weeks 2-12)

The assigned coach was consistent for a given participant for the entirety of the intervention to promote engagement and therapeutic alliance. Each encounter focused on discussing the lifestyle of the participant based on the weekly report and the progress of the last goal. The health coaching encounters followed the four processes of motivational interviewing (MI): engagement of the participant (avoiding cheerleading), focusing on what to work on the following week (chosen by the participant), the evocation of the participant's reasons to work on that topic and planning (goal setting). In addition, the coach helped the patient articulate their question to their providers to make medical visits more effective.

Home Rehabilitation

(SDC 1): Participants were expected to engage in the home-based exercise practice ≥ 6 d/wk. The PR routine began with a slow breathing awareness exercise, "pursed lips breathing," followed by upper-extremity exercises, then two walks (at the home or outside) for 6 min and finishing with another slow and mindful breathing exercise. The exercise protocol lasted about 20 min. The Garmin Vívofit 4^™ activity monitor was to be always worn to capture steps/d and the metabolic equivalent of tasks (METs)/d.

We monitored the unsupervised proposed in-home practice > 6000 times with no falls or adverse events.

Health coaching

An assigned health coach called all participants weekly to review their program experience, facilitate patient-directed goals, and address barriers to participation as part of an implementation assessment.

The coach was trained in MI through printed material, interactive classes (6) and role-playing, and mindfulness (undivided attention, silence, deep listening training) as reported.(15, 18) In addition, health coaches used specific MI techniques like open-ended questions, affirmations, reflections, and summaries to guide the conversation and promote patients' expression.(19) Two randomized clinical trials to date confirm the feasibility and effectiveness of this program in COPD improving all outcomes including daily PA. (15, 16).

PATIENT-REPORTED QUALITY OF LIFE, PHYSICAL ACTIVITY OUTCOMES, AND QUALITATIVE INTERVIEWS

Baseline demographic data, disease type, lung function (percent predicted forced vital capacity (FVC%), and diffusion capacity for carbon monoxide (DLCO%) were collected.

Measured endpoints reflecting changes in patient-reported QoL included pre/post-intervention mMRC and Chronic Respiratory Questionnaire-Self Administered Survey (CRQ-SAS) as obtained by self-administered paper questionnaires.

The mMRC is an established dyspnea scale ranging from 0-4, with higher scores representing greater dyspnea(21). The minimal clinical important difference is a one-point increase suggesting worsening dyspnea. The CRQ-SAS is an established and validated respiratory-related QoL questionnaire of 20 items assessing four clinical domains (Dyspnea, Fatigue, Emotional, and Mastery), measured on a scale of 1-7 with higher scores suggesting better/positive findings(22, 23). The minimal clinical important difference is 0.5 for any of the CRQ domains, with a change of >1.25 considered significant.

PA was defined by the mean steps/d as measured by the a wgt3x-bt monitor (Actigraph) worn for 5 d prior to initiation and 1wk after completion. A separate activity Garmin Vivofit 4 monitor was worn throughout the intervention and uploaded into the software.

Participants who completed the program were invited to participate in a qualitative interview. Telephonic interviews were conducted by a member of the study team trained in qualitative interview methods (JDD) using a semi-structured interview guide which were recorded, transcribed for analysis, and reviewed by two study members for accuracy.

STATISTICAL ANALYSIS

Categorical data are presented as n (%) and continuous data as mean ± SD or median and (IQR). Post-intervention combined those with stable or declining symptoms/function versus those with positive improvement. mMRC and CRQ domain improvements were defined by changes greater than the published minimal clinical important difference with improved PA defined as > 600 or more steps/d as measured by Actigraph at the end of intervention(24, 25). Continuous data were compared to baseline using paired t-test.

Data analysis followed an inductive approach(26). Two members of the study team independently reviewed interview transcripts for familiarization, then read them line-by-line and annotated topics that represented specific content. Annotated data was reviewed to identify major (>75% of interviews) and minor themes (>30% of interviews). Final patterns with representative quotations, a description of the analytic process, and quantitatively outcomes were presented to the study team of clinical and social science investigators.

RESULTS

Twenty-four patients were enrolled in the implementation study, with twenty-one completing the full 12-wk program (one patient declined participation after enrollment and two died from underlying ILD progression) and twenty having comparison data (one patient did not complete post-intervention data). The final cohort was comprised of older patients (median 70 yr) with a slight female predominance (57%), moderate FVC% severity (median 53% [IQR 43-71]), and severely decreased DLCO% (median 39% [IQR 30-51]) (Table 1). Most were on oxygen (76%) at rest or with exertion at the time of enrollment. Idiopathic pulmonary fibrosis and fibrotic hypersensitivity pneumonitis comprised a third each of the cohort (66% total). Over half had stable FVC% change (57%) and 28% (N= 6) were considered worsening prior to study enrollment (defined as a drop in FVC% > 5% in the prior 6 months). The median number of health coach calls was 6 (25-75 IQR of 4-7)/participant over the 12-wk period.

Table 1).

Pilot cohort demographic and clinical data

Age, yr	70 (60-76)
Sex, M/F	9/12 (43/57)
FVC%	53 (43-71)
DLCO%	39 (30-51)
Oxygen use	16 (76)
ILD dx distribution
-IPF	2 (9)
-Probable/Indeterminate UIP	5 (24)
-f-HP	7 (33)
-CTD-ILD	4 (19)
-unclassifiable/mixed/other	3 (14)
FVC trend prior to intervention (relative FVC% change ≥ 5% in last 12 mo), N (%)
stable	12 (57)
improved	2 (10)
worsening	7 (33)
Number of health coach calls, median (IQR) (out of 12 possible)	6 (4-7)

Data are presented median (IQR) or n (%)

Abbreviations: CTD-ILD: connective tissues disease interstitial lung disease DLCO: diffusion capacity for carbon monoxide, f-HP: fibrotic hypersensitivity pneumonitis, FVC: forced vital capacity, ILD: interstitial lung disease, IPF: idiopathic pulmonary fibrosis, UIP: usual interstitial pneumonia.

PATIENT-REPORTED OUTCOMES AND PHYSICAL ACTIVITY

Pre- and post-intervention PA and PRO changes are presented in Table 2). Mean steps/d as measured by Actigraph varied widely at baseline, with slight decrease overall at the end of intervention (6944 ± 2489 vs 6462 ± 1809; P = .134). The CRQ total and physical summary score (the combination of the dyspnea and fatigue domain) and emotional summary score (the combination of the emotions and mastery domains) trended towards improvement along with dyspnea (mMRC), though none met statistical significance due to our smaller sample size.

Table 2).

Pre- and post-intervention PAand patient-reported outcome changes

Characteristic	Baseline	End of intervention	P value¶
Steps/d (Actigraph)	6944 ± 2489 (3106-12557)	6462 ± 1809 (3206-10794) (N= 20)	.134
CRQ Physical domain	3.56 ± 1 (1.7-5.1)	3.74 ± 1.2 (1-5.6)	.36
CRQ Emotional subscale,	4.04 ± 1.2 (1.3-6.2)	4.37 ± 1.1 (1.1-5.8)	.24
mMRC	2.8 ± 0.6 (2-4)	2.5 ± 0.8 (1- 4)	.14

^¶Paired t-test

Data are presented as mean ±SD (range).

Abbreviations: CRQ-SAS: Chronic Respiratory Questionnaire-Self Administered Survey, mMRC: modified Medical Research Council

Categorical counts for positive change in individual patients (>600 steps/d, ≥0.5 change in CRQ domains, or one-point change in mMRC) are presented in Table 3). Ten patients (50%) had improvement in mean steps/d from baseline, and one-third reported clinically meaningful improvement in dyspnea based on mMRC. Five patients had a decline in mean steps/d (range loss of 1763 to 4364 steps/d from baseline), which resulted in lower mean daily steps for the whole cohort on paired analysis. Only three of those with a decline in daily steps had evidence of lung function decline in the 6 mo prior to study participation. Improved CRQ findings were seen in 50% of patients for total CRQ and emotional domain scores.

Table 3).

Number of patients (n=20) with positive changes in PA and patient-reported outcomes after intervention

Characteristic	Improved	Stable/worsened
Mean steps/d (Actigraph)a	10 (50)	10 (50)
CRQ Physical domain b	6 (30)	14 (70)
CRQ Emotional subscale b	10 (50)	10 (50)
mMRC c	6 (30)	14 (70)

Data presented as n (%).

^aIncrease in 600 steps/d by Actigraphy

^bIncrease of 0.5 point for CRQ domain subscales

^cDecrease of 1 point as MCID

Abbreviations: CRQ-SAS: Chronic Respiratory Questionnaire-Self Administered Survey, mMRC: modified Medical Research Council

QUALITATIVE OUTCOMES

Twelve participants completed qualitative interviews. Major and minor findings are presented below and in SDC 2, organized by the main assessment topics: 1) baseline disease experience, 2) program experience, and 3) program impact. Additional exemplar quotations are provided in SDC 3-5.

Baseline disease experience:

Patients were assessed for their understanding of their lung disease prior to study participation and most described significant impact from symptom burden and related secondary impacts of anxiety or distress with future progression or morbidity.

Morbidity:

Most patients described being incompletely informed about their exact diagnoses and outcomes, expressing significant anxiety about the accuracy and timeliness of their treatments and the diminishment of their well-being and long-term survival. In the absence of concrete information from clinicians, many gathered information from the internet, often with conflicting or misleading information.

‘When I first heard that I had IPF, the doctor came in and said, "Oh no, oh no, I am so sorry". I was thinking, how long do I have to live the way she was carrying on? She said, "You've got pulmonary fibrosis". You know, I didn't know what pulmonary fibrosis was, and well Google said I had three to five years to live.’

Significant disease burden:

Daily symptoms, often progressed from the time of diagnosis, was endorsed by a majority. Many alluded to the use of oxygen as a personal and social marker of disease extent and severity with time. Some patients avoided oxygen therapy for this reason, potentially creating increased risk for future problems related to chronic hypoxia.

‘I have changed my lifestyle drastically, due to my lung disease. My wife and I are retired. We had a fairly active lifestyle. As I've aged, that's changed a little bit. But really, my lung disease put the biggest, absolutely the biggest change in my life.’

Hope:

Hope was often placed in ‘medical cures or treatments,’ with few mentioning the utility PR programs as a management approach. Transplantation was also mentioned by several patients with a few being denied from listing, shattering their view of a future life being disease-free.

‘I always thought, ‘I'm going to beat this. They're gonna fix me, they're gonna find the magic pill, the magic thing, something is gonna change in this visit, and they're gonna say "oh my gosh, we don't need to see [you] anymore!’… Then I had that epiphany moment… I'm in denial about my lung disease. That was a pivotal moment for me. It changed how I looked at it.’

Program experience:

Direct patient experience with the program involved a review of pre-intervention expectations, positives and negatives of participation, and any changes in behavior or activity to complete the program.

Limited program expectations:

Most entered the program with an ‘open-mind,’ having few specific goals or expectations prior to participating. The health coach was seen as a general positive and appreciated for program support and adherence. Goals were equally solicited or evoked by the health coach, such as completing daily exercises or targeting a particular number of steps/d, which often changed each wk.

Program Feasibility:

Flexibility in terms of the program’s home-based/telehealth approach was noted, and all felt safe doing the exercises in their homes. All participants described being more readily able to participate in a home-based program than center-based programs, with only minor setbacks from technical or logistical issues (syncing activity monitors or connecting to internet as examples).

Time management or setting up routines:

A key element was participation consistency with patients reorganizing their day-to-day to adhere to the program’s expectations.

‘When I was starting to slack, I’d say, ‘no you committed to do this,’ so that helped me. I always felt like this meant more than just me. That helped me stay focused. But then once I got into it, it was easier to not miss a day.’

Program Impact:

Positive impact on PA and symptoms:

Most participants had increased awareness of their symptoms and physical limitations after program completion with improved confidence in adaptability to daily challenges. A change in perspective of what constituted ‘exercise’ or ‘PA’ was also described with a majority expressing improved exercise or activity tolerance and dyspnea. A majority still felt similarly about their disease in terms of progressive symptom burden.

‘What I learned from this was that you can move at a slower pace and still reap benefits. It may not be as much as the person who's running on the treadmill at the gym. Because of my current situation, something is better than nothing.’

Recommended the program to others:

All felt the program was still beneficial and would recommend it to others with similar disease or symptom burden.

‘No. I mean I really enjoyed it. I think any lung patient would benefit from it, just because for a lot of us, we feel that exercises are impossible, but this has shown it is possible.’

Continuation of the program:

Few admitted to continuing with the breathing exercises or using a daily step tracker weeks after completing the program, with online links to the videos and a step tracker being provided as part of study participation.

DISCUSSION

This pilot study involved lesser-used modalities for PR in patients with ILD. Our results indicate feasibility, patient acceptance, and provide an effect size for future well-powered studies. A 2021 Cochrane review of current PR practices in ILD described most programs as occurring in the outpatient setting with center-based programs(27). Most programs were 8 - 12 wk in duration with an average of 2-3 sessions/wk. Our program involved more commitment, and yet > 70% of participants were adherent to the program, demonstrating the feasibility of this approach among patients with f-ILD. Prior home-based programs report similar compliance and feasibility, with some achieving primary activity or QoL outcomes in patients with f-ILD and IPF(10-12, 28).

Most PR programs focus on improving exercise or PA and often use 6-min walk test or other measures of exercise tolerance as targeted endpoints(29, 30). While total mean steps/d did not improve in our study, 50% had improvement in PA. Several patients with initially higher steps/d who declined over the study period contributed to a lower post-intervention steps/d average for the cohort. These findings are confirmed by the qualitative data describing patient illness trajectories despite program participation. Perez-Bogerd et al. described a minimal increase in mean step counts or time spent in moderate-intensity PA yet long-term improvement in strength and walk distance(31). Reported improvements in traditional and home-based PR programs as measured by 6-min walk test distance range anywhere from 295 −551m (32), while PA as measured by our program was defined by mean steps/d over a minimum of 5-d, pre- and post-intervention.

Health coaching and motivational interviewing, utilized in our study, have been reported in the management of chronic lung diseases(20),(33, 34). Benzo et al. reported a reduction in COPD-related rehospitalizations in patients receiving a written exacerbation plan and motivational interviewing-based coaching versus usual care(35). Disease-specific QoL and treatment adherence also improved in the health coaching group compared with control at 6 and 12 mo baed on CRQ (P < .05)(35),(36). Generally viewed as a positive influence, coaching calls were specifically mentioned by participants in the qualitative interviews as a source for program accountability. Careful alignment of patient and program goals earlier in the program participation may be helpful. Awareness and acceptance of disease progression may also impact self-efficacy during participation(37).

Another goal of PR is improvement in patient-reported health-related QoL (HRQoL) and symptom burden(29). Improvements in CRQ and St. George’s Respiratory Questionnaire have been reported in patients with ILD, as highlighted by the IPF subgroup of a pooled Cochrane analysis(27). Our study demonstrated improvements in HRQoL and dyspnea as measured by CRQ and mMRC, respectively but did not meet statistical significance due to lack of statistical power. A post-hoc analysis also did not suggest baseline demographic or functional findings in our smaller cohort correlated with CRQ or mMRC (data not shown) outcomes, though caution should be taken considering the sample size.

Additional insights from the study’s qualitative analysis highlight that many patients with f-ILD expressed expectations of a poor outcome with little hope of improvement unless novel medical therapies were discovered and became widely available. Pulmonary rehabilitation was often seen as adjunctive to medical treatments with participation hampered by the expectation of little benefit. Understanding motivation, which is often tied to different expectations during times of disease, may play an important role in supporting behavior change(1).

Lastly, reframing interpretations of “physical well-being’ may be important for gauging program impact. While mean steps/d served as a surrogate for PA in our study, additional program impacts that were not explicitly described to patients include better engagement with daily activities, self-awareness of breathing patterns, and mindfulness of attitude with physical limitations. Explicitly highlighting these important experiences as relevant outcomes for program participation may improve program development, engagement, and long-term intervention adherence.

Our study has several limitations. A pilot cohort of only 21 participants may limit the generalizability of our findings, including the ability to meet statistical significance of the endpoints and lack of rigor in outcome assessment. The intent of the pilot was to review process delivery for patients with f-ILD and assess effectiveness in real-world settings, with an appraisal of program impact on patient experience. A larger study is underway utilizing qualitative and process insights gained from this early analysis. Another study limitation is a less limited, but more heterogeneous cohort of multiple f-ILD subtypes with varied disease severity that may confound intervention findings. Our choice of measured endpoints (change in selected PROs and steps/d) may not have captured unanticipated benefits from study participation, hence the addition of a qualitative review and analysis. While the choice of PROs was based on the considerations previously noted, other PROs, such as COPD Assessment Test, have growing literature for their use in the ILD population and may be more financially accessible. Additionally, given the home-based nature of the program occurring in multiple states, no established testing like a 6-min walk test or shuttle walk test was incorporated. Patients were also recruited from a large tertiary care center and may represent patients who are more well-resourced and willing to participate but whose motivation to do so may also more directly impact the outcomes of behavior-based interventions. Similarly, only people who completed the program participated in interviews which may have skewed results; future work may explore experiences of completers and non-completers. Finally, variations in process delivery, including challenges with technology (damaged battery, wifi connection issues, etc) or participants not attending health coach calls may influence program delivery and confound study results.

CONCLUSION

Home-based PR with telemonitoring and health coaching is likely a feasible approach for many patients unable to participate in center-based programs. Improved exercise tolerance and patient-reported QoL in patients with f-ILD was seen with a pilot assessment of a home-based program though a larger study is needed to determine scope and effect size. Qualitative analysis provided additional insights for improved implementation and optimizing intervention impact.

KEY PERSPECTIVE.

• What is novel? The study utilized relatively novel home-based pulmonary rehabilitation program for patients with interstitial lung disease who may struggle to complete center-based rehabilitation programs due to increased symptom burden or frailty.

• What are the clinical and/or research implications? Our results suggest improvements in dyspnea and respiratory-related quality of life with qualitative analysis of semi-structured participant interviews suggested a significant baseline disease burden with related secondary impacts, highlight the feasibility of this home program in a diverse set of patients with varying severity and diagnostic subtypes of fibrotic lung disease.