Remote behaviour change service for inactive adults with lung disease: A non-randomised controlled study

Abstract

Objectives

To determine whether a remote behaviour change service (Active Steps), delivered by a UK charity, can increase physical activity over a 12-month period in adults with lung disease.

Methods

Active Steps (n=166) consisted of 1:1 telephone health coaching for 12 months alongside printed (e.g. activity diary and chart, information and goal setting booklet) and digital behaviour change strategies (e.g. exercise videos, motivational newsletters, pedometer). A concurrent control group was recruited (n=80) for comparison. Self-reported physical activity (Short Active Lives Survey) and self-reported health (EQ-VAS) were collected at baseline, 3, 6, and 12 months. Logistic regression models (adjusted for confounders) were used to analyse the odds of becoming physically active or average changes in physical activity time and EQ-VAS.

Results

Primary outcome: The number of participants defined as physically active at 12 months was greater with Active Steps (46%) compared to controls (23%) which was statistically significant in an unadjusted (odds ratio (OR) [95% confidence intervals (CI)]:2.8 [1.3-6.1], p=0.011) but not adjusted model (OR: 2.2 [0.9-5.4], p=0.097). Secondary outcomes: Compared to controls, Active Steps increased the number of participants physically active at 3 and 6 months, increased the weekly minutes of moderate-vigorous physical activity and improved self-reported health (EQ-VAS) across all timepoints.

Discussion

Our findings suggest that charity-led remote behaviour change support is effective in the short-to-medium-term in adults with lung disease, but high participant attrition necessitates caution in interpreting sustained effect estimates at 12 months. Further long-term evaluations are needed to address adherence and sustainability of such services.

Keywords

behaviour change physical activity inactivity lung disease respiratory COPD

Background

Adults diagnosed with chronic lung disease such as asthma, bronchiectasis, chronic obstructive pulmonary disease (COPD), and interstitial lung disease have reduced levels of daily physical activity when compared to healthy counterparts of a similar age, with many not meeting the recommended physical activity guidelines.^1–5

Increasing physical activity in people with chronic lung disease is associated with improved symptoms, physical function and quality of life whereas inactivity is associated with adverse clinical outcomes, including increased hospitalisation and all-cause mortality.^6–8 Structured physical activity programmes, such as pulmonary rehabilitation are one of the key treatments for chronic lung disease. A six-to-eight week course of pulmonary rehabilitation is known to have well-established benefits including clinically important improvements in exercise capacity, quality of life and respiratory symptoms such as dyspneoa.⁹ However, there are barriers preventing people with chronic lung disease attending pulmonary rehabilitation including accessibility, under-referral and long waiting times.¹⁰ Even when a pulmonary rehabilitation program is completed, this does not always translate to changes in long-term daily physical activity for people with chronic lung disease.¹¹

More strategies are needed to increase physical activity across a wider range of settings for people with chronic lung disease including catering for when structured physical activity programmes cannot be accessed or before starting such programmes. Asthma + Lung UK (formerly British Lung Foundation) received funding to develop, deliver and test a new remote health coaching behaviour change service called Active Steps. The co-design and development of this service have been previously reported.¹² The primary aim of this study was to determine whether the remote behaviour change service could increase physical activity in adults with lung disease who are physically inactive. The secondary aim was to assess the impact of the behaviour change service on self-reported health and disease-specific quality of life.

Methods

Design

This was a non-randomised controlled study conducted between February 2019 and October 2021.

Participants

Eligibility criteria

The target population were physically inactive adults living with a chronic lung disease (asthma, bronchiectasis, COPD, interstitial lung disease, lung cancer, mesothelioma). Physical inactivity was defined - in accordance with public health services in England^13,14 - as undertaking less than 30 minutes in total of moderate-vigorous intensity physical activity per week. People unable or unwilling to provide their consent for the study were excluded.

Interventions

Intervention (Active Steps) : The full detail of the development of the remote behaviour change service has been reported previously.¹² The Behaviour Change Wheel¹⁵ was used as an intervention development framework. The overarching delivery model of the Active Steps service is shown in Figure 1. An outline of the components of Active Service including which domains of the COM-B model of behaviour (the hub of the behaviour change wheel) were targeted is provided in the supplementary material (Table S1). All service users received 12 months of 1:1 telephone health coaching, regular email newsletters and an information pack (Keep Active booklet and an Active Steps plan including behaviour change tasks, activity diary and wall chart), Stay active, stay well activity DVD a letter to give to their GP, and optionally a pedometer. Prior to service launch, the anticipated phasing of the telephone calls was fortnightly in the first 3 months, monthly calls from month 4-6 and then bimonthly calls thereafter. However, the telephone health coaching was person-centred, considering individual needs and therefore frequency and intensity of the intervention varied.

Figure 1.

Active Steps delivery model.

Control (Comparator) : All participants in the control group received National Health Service standard care for their lung disease with no active intervention from the research team. Participants were informed that their clinical outcome data would be used to evaluate services of the British Lung Foundation but were not told any specific information about the Active Steps service.

Outcomes

Primary

The primary outcome was a between-group comparison of the number of participants completing 30 minutes or more of moderate-to-vigorous physical activity per week at 12 months. Physical activity was measured by the Short Active Lives Survey (SALS). The Active Steps service was funded as part of a larger programme by Sport England to impact on physical activity measured by SALS. Since January 2018 SALS has been Sport England’s recommended physical activity measurement tool for its funded projects. SALS is a self-reported tool to measure adult physical activity levels. SALS consists of three questions, each containing three sub-questions, which establish activity levels in relation to walking, cycling and sport/fitness activities. The tool asks the number of days the activity was undertaken, the duration, and whether the effort was enough to raise their breathing rate. SALS has previously been validated against other recognised self-reported and device-based measures of physical activity.¹⁶

Secondary

Between-group comparisons of the number of participants completing 30 minutes or more of moderate-to-vigorous physical activity per week at 3 and 6 months were secondary outcomes. Between-group changes in total self-reported moderate-to-vigorous physical activity were also assessed by SALS at 3, 6 and 12 months. Between-group changes in self-reported health at 3, 6 and 12 months were measured using the visual analogue scale (EQ VAS) and descriptive system of the EQ-5D-5L quesstionnaire.¹⁷ The intervention participants completed a short-form chronic respiratory disease questionnaire (CRQ)^18–20 as a measure of respiratory-specific quality of life at baseline, 3 months, 6 months and 12 months.

Recruitment

Behaviour change service (Active Steps) : The Active Steps service was delivered by the British Lung Foundation. The service was funded for delivery from January 2019 to October 2020. During this time, people meeting the eligibility criteria were recruited into the service either via the charity helpline, or self-referral (enquiry via the British Lung Foundation website). Service users consented for collection and use of their data for evaluation purposes upon entry into the service. All procedures were approved by the University of Lincoln Human Ethics committee (2019-Jan-0378).

Control group : To account for changes over time (independent of Active Steps) in the collection of study outcomes, an independent control group was recruited over a period of time that matched the service delivery of Active Steps. The recruitment of this independent control group was publicly registered (NCT04080583) and had received Health Research Authority and NHS Westminster Research Ethics Committee approval (19/LO/1046). The independent control group was recruited through the following recruitment routes: mailshot to eligible participants (adults with a chronic lung disease) from Lincolnshire East Clinical Commissioning Group (CCG), Lincolnshire West CCG, South Lincolnshire CCG, and Countywide Respiratory Services of Lincolnshire Community Health Services NHS Trust; opportunistic participant identification by nurses, GPs, or respiratory physiotherapists at the above participant identification sites; previous research participants on a study of physical activity following pulmonary rehabilitation (NCT03660644) who were happy to be contacted for future research; and support groups for people living with chronic lung disease in Lincolnshire, United Kingdom. The information sheet advised potential participants for the control group that the aim of their participation was to measure and understand self-reported physical activity levels in people with chronic lung disease and to help inform evaluation of British Lung Foundation services. All control participants provided written consent to the study.

Data collection and follow-up

The British Lung Foundation collected baseline, 3, 6 and 12-month data for intervention participants over the telephone with data entry to purpose-built web-based surveys (Qualtrics, Utah, United States). Data were collected by a member of the Active Steps team who was not involved in delivering intervention calls to the specific participant, with data submitted directly to an independent research team. In September 2019, data collection procedures were expanded to include the option of intervention participants being emailed a link to the web-based surveys to self-complete, if staff were unable to schedule a telephone follow-up. Baseline and outcome data for the control group were collected at baseline, 3, 6 and 12 months by the research team using the same web-based surveys (Qualtrics) as the Active Steps service users.

Sample size

This was a non-randomised study where a concurrent control group was sought during the same study period as a group receiving a new service implemented by a national charity. The evaluation was designed to detect an absolute difference of 10% in the proportion of participants being physically active at 12 months. Based on 80% statistical power and a 5% level of significance, 58 participants per group were required. British Lung Foundation had originally estimated capacity to enrol 200 participants into Active Steps over the service period. However, we conservatively estimated that at least 80 participants would be recruited for both the intervention and control groups in the study period. This would allow for a 25% dropout rate over the 12-month follow-up period.

Data analysis

All analyses were performed using SPSS v27 (IBM Armonk, New York). The primary approach for analysis was a “per protocol” analysis. This referred to the analysis of only those participants who strictly adhered to the project (protocol) including receipt of the intervention for 12 months and/or provided data at follow-up.

The primary outcome was assessed using binary logistic regression. Given the study design, we first identified potential confounders for the outcome of physical activity. A directed acyclic graph (DAG)²¹ was constructed (Supplementary Material, Figure S1) to identify the minimal sufficient set of potential confounders for adjusting the effect of the intervention on physical activity. Based on prior literature,^22,23 baseline differences in demographic or clinical factors between groups in the study, and inspection of the DAG, we identified age, gender, breathlessness severity, diagnosis of diabetes, heart condition or mental health condition, and social deprivation as potential confounders of the effect of the intervention on physical activity.

Binary regression models with and without adjustment for these potential confounders were used to estimate the effect (interpreted as an odds ratio) of the intervention on number of participants completing 30 minutes or more of moderate-vigorous physical activity per week at 3, 6 and 12 months. Changes from baseline at 3, 6 or 12 months in the secondary outcomes of total minutes of physical activity and self-reported health (EQ-VAS) were assessed using logistic regression with or without adjustment for the potential confounders. The potential confounders for the outcome of self-reported health (EQ-VAS) were considered similar to that of physical activity. Analysis of the effect of intervention on self-reported health (EQ-VAS) was performed with or without adjustment for the potential confounders. Descriptive statistics of the health profiles from the EQ-5D questionnaire were reported in accordance with published methods (see supplementary material).²⁴ Changes from baseline to 12 months in CRQ domain scores were assessed using paired t-tests. The short to medium changes (at 3 and 6 months) in CRQ domain scores were assessed using one-way repeated measures analysis of variance followed by post-hoc paired t-tests.

Given the primary focus of the intervention was to increase the number of people physically active at 12 months, we also undertook a sensitivity analysis using non-responder imputation. All participants lost to follow-up at 12 months were classified as inactive to provide a “worst-case” conservative estimate alongside the per-protocol analysis. Baseline characteristics of participants who completed the 12-month follow-up or were lost to follow-up were compared within both groups. The DAG informed confounders used in the primary analysis were also applied in the non-responder imputation. To support interpretation of results across the study duration, non-responder imputation was also conducted with physical activity at 3 months and 6 months.

Results

Completeness of follow-up

Baseline data were collected on 166 intervention participants and a concurrent control group of 80 participants. Intervention participants who provided baseline data by October 2019 (12 months before end of service funding period) were included in the 12-months follow-up including primary outcome (n= 124), whereas impact on outcomes at 3 and 6 months included all intervention participants recruited up to March 2020 (n=166). The baseline characteristics are provided in Table 1. Of the 124 intervention participants recruited up until October 2019, 68 (55%) and 49% participants provided data for the primary outcome and secondary outcomes respectively at 12 months. At 3 and 6 months (n=166, recruited up to March 2020), 51% of all intervention participants provided data. The mean (± standard deviation) number of health coaching calls across intervention participants were 6 (±4) calls each. The completeness of follow-up for the control group was 93%, 88% and 75% at 3 months, 6 months and 12 months respectively. The completeness of follow-up of intervention and control group is presented in Figure 2.

Table 1.

Baseline characteristics of participants in Intervention and Control groups.

	Intervention [recruited up to October 2019]	Intervention [recruited up to March 2020]	Control
Gender, n (%)
Female	92 (74)	118 (71)	43 (54)
Male	32 (26)	48 (29)	37(46)
Age, mean ± SD	64 ± 12	65 ± 11	70 ± 10
Ethnicity, n (%)
White	118 (95)	157 (95)	78 (98)
Asian or Asian British	2 (2)	3 (2)	1 (1)
Black or Black British	2 (2)	3 (2)	1 (1)
Mixed	2 (2)	2 (1)	0 (0)
Index of Multiple Deprivation Decile, n (%)
1 (most deprived)	8 (7)	14 (8)	11(14)
2	12 (10)	17 (10)	8 (10)
3	14 (11)	15 (9)	10 (13)
4	19 (15)	21(13)	7 (9)
5	7 (6)	15 (9)	4 (5)
6	9 (7)	12 (7)	11 (14)
7	12 (10)	18 (11)	10 (13)
8	15 (12)	18 (11)	4 (5)
9	16 (14)	20 (12)	10 (13)
10 (least deprived)	5 (4)	9 (6)	5 (6)
Lung disease, n (%)
COPD	79 (64)	111 (67)	60 (75)
Asthma	26 (21)	39 (24)	25 (31)
Bronchiectasis	22 (18)	24 (15)	12(15)
Interstitial lung disease	16 (13)	22(14)	6 (8)
Lung cancer	1 (1)	1 (1)	1 (1)
Other	4 (3)	6 (4)	1 (1)
MRC Dyspnoea Scale, mean ± SD	3 ± 1	3 ± 1	3 ± 1
Oxygen Use, n (%)
None	107 (86)	143 (86)	74 (93)
Long-term oxygen	9 (7)	12 (7)	4 (5)
Ambulatory oxygen	8 (7)	11 (7)	2 (3)
Comorbidities, n (%)
Arthritis or ongoing back/joints problem	50 (40)	69(41)	38 (48)
Other long-term condition or disability	45 (36)	61 (37)	20 (25)
Mental health condition	40 (32)	53 (32)	15 (19)
High blood pressure	29 (23)	46 (28)	31(39)
Heart condition	21 (17)	26 (16)	26(33)
Diabetes	11 (9)	19 (11)	18(23)
Cancer diagnosis/treatment (last 5 years)	8 (7)	10 (6)	6 (8)
Kidney or liver disease	7 (6)	10 (6)	6 (8)
Blindness or partial sight	3 (2)	5 (3)	7 (9)
Neurological condition	3 (2)	6 (4)	1 (1)
Learning disability	2 (2)	2 (1)	1 (1)
Stroke which affects day-to-day life	2 (2)	4 (2)	2 (3)
Dementia	1 (1)	1 (1)	2 (3)

Figure 2.

Completeness of follow up of Intervention and Control participants.

Primary outcome

Number completing 30 minutes or more of moderate-vigorous physical activity at 12 months: The intervention increased the number of adults physically active at 12 months relative to control, but this effect was only statistically significant in the unadjusted model (Table 2). The analysis was repeated with an additional nine participants who withdrew from intervention but provided the primary outcome at 12 months. This had minimal impact on the point estimate of effect of the intervention and the pattern between unadjusted and adjusted estimates (Unadjusted OR [95% CI]: 2.4 [1.1-5.3], p=0.023; Adjusted OR: 2.1[0.9-5.3], p= 0.105).

Table 2.

Number completing 30 minutes or more of moderate intensity physical activity per week in Intervention and Control groups at each follow-up of the evaluation.

	Physical activity classification		Model effect estimates
	Active n (%)	Inactive n (%)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)
3 months
Intervention (n=85)	46 (54)	39 (46)
Control (n=74)	9 (12)	65 (88)	8.3 (3.7 – 18.8)	9.1 (3.7 – 22.5)
6 months
Intervention (n= 84)	47 (56)	37 (44)
Control (n = 70)	6 (9)	64 (91)	13.5 (5.3 – 35.0)	17.3 (5.8 – 51.4)
12 months
Intervention (n= 59)	27 (46)	32 (54)
Control (n = 60)	14 (23)	46 (77)	2.8 (1.3 – 6.1)	2.2 (0.9 – 5.4)

Secondary outcomes

Physical activity: The interventions increased the number of adults physically active at both 3 months and 6 months, which was statistically significant when unadjusted and adjusted for potential confounders (Table 2). The intervention increased the total minutes of moderate-vigorous physical activity per week at 12 months, but this effect was only statistically significant in the unadjusted model (Unadjusted Mean difference (MD) [95% CI]): 52[2-102], p =0.041; Adjusted MD: 48[-9-105], p= 0.096) (Table 3). The intervention increased the total minutes of moderate-vigorous physical activity per week at 3 months (Unadjusted MD: 49[14-83], p=0.005; Adjusted MD: 50[13-86], p= 0.008) and 6 months (Unadjusted MD: 65[27-103], p= 0.001; Adjusted MD: 50[13-86], p= 0.008), all of which were statistically significant in the unadjusted and adjusted models (Table 3).

Table 3.

Change in absolute values of moderate-vigorous physical activity (minutes per week) at each follow-up in Intervention and Control group.

	Intervention mean (95% CI)	Control mean (95% CI)
Change at 3 months (Intervention n =85, Control n = 74)
Moderate-vigorous physical activity (minutes per week)	75 (50 - 100)	26 (4- 48)
Change at 6 months (Intervention n =84, Control n = 70)
Moderate-vigorous physical activity (minutes per week)	100 (69 – 132)	25 (-3 – 55)
Change at 12 months (Intervention n = 59, Control n = 60)
Moderate-vigorous physical activity (minutes per week)	91 (45 – 137)	34 (13 – 56)

Generic health: The intervention improved self-reported health (EQ-VAS) at 12 months, but this effect was only statistically significant in the unadjusted model (Unadjusted MD: 13 [3-23], p=0.011; Adjusted MD: 7 [-5-18], p= 0.272) (Table 4). The intervention had statistically significant effects on reported health across all models at 3 months when unadjusted and adjusted for potential confounders (Unadjusted MD: 10 [2 – 17], p = 0.011; Adjusted MD: 9 [1-17], p=0.026) (Table 4) but not at 6 months (Unadjusted MD: 5 [-2–12], p=0.164; Adjusted MD: 6 [2-14],p=0.026) (Table 4). Descriptive summaries of responses to the EQ-5D by dimension and level as well as changes in health over time for the intervention and control groups are presented in the supplementary material (Table S3-S5).

Table 4.

Changes in EQ-VAS at each follow-up in Intervention and Control group.

	Intervention			Control
	Change at 3 months (n = 85)	Change at 6 months (n = 84)	Change at 12 months (n = 49)	Change at 3 months (n = 74)	Change at 6 months (n = 70)	Change at 12 months (n = 60)
EQ-VAS	11.3 (6.2-16.2)	11.2 (6.9–15.5)	15.2 (7.3–23.1)	1.7 (-3.7–7.2)	6.2 (0.3–12.1)	2.0 (-4.7–8.6)

CRQ: At 12 months, the mean ratings (95% CI) improved by 0.5 (0.2-1.0, p=0.008), 0.6 (0.2–1.0. p=0.005), 0.8 (0.4–1.2, p<0.001) and 0.5 (0.1–1.0, p=0.026) for dyspnoea, fatigue, emotional function and mastery domains respectively, all of which were statistically significant and of a magnitude considered to be clinically meaningful. Repeated measures ANOVA revealed a main effect of time on dyspnoea (p< 0.001), fatigue (p< 0.001), emotional function (p<0.001) and mastery (p=0.006) at 3 and 6 months. At 3 months, the mean ratings (95%CI) improved by 0.5 (0.1–0.8, p=0.014), 0.6 (0.6, 0.2–0.9, p=0.002), 0.7 (0.2–1.2, p=0.005) and 0.6 (0.1–1.0, p=0.007) for dyspnoea, fatigue, emotional function and mastery domains respectively, all of which were statistically significant and of a magnitude considered to be clinically meaningful. At 6 months, the mean ratings (95%CI) improved by 0.8 (0.8, 0.3–1.2, p<0.001), 0.6 (0.3–1.0, p<0.001), 0.9 (0.6–1.3, p<0.001) and 0.4, (0.1–1.9, p=0.147) for dyspnoea, fatigue, emotional function and mastery domains respectively, all of which were statistically significant and of a magnitude considered to be clinically meaningful except for the mastery domain.

Sensitivity analyses

Under the conservative non-responder imputation analysis, the estimate of effect of Active Steps on the primary outcome was reduced (Adjusted OR 1.2, [0.5–2.70]) (Table 5). A comparison of baseline characteristics between study completers and those lost to follow-up at 12-months did not reveal any differences (Supplementary Table S2). The widening gap between the per-protocol and non-responder imputation estimates coincided with an increase in the cumulative attrition rate (Table 5). Sensitivity analysis using non-responder imputation (NRI) at 3 months and 6 months demonstrated that the effect of Active Steps remained statistically significant even under worst-case assumptions (Table 5).

Table 5.

Comparison of per-protocol efficacy and non-responder imputation (worst-case) effectiveness of intervention on physical activity (number defined as physically active) across all follow-up timepoints.

Timepoint	Per-protocol adjusted OR (95% CI)	Non-responder imputation adjusted OR (95% CI)	Attrition rate (%)
3 months	9.1 (3.7 - 22.5)	3.4 (1.5 - 7.8)	35%
6 months	17.3 (5.8 – 51.4)	5.4 (2.1 - 14.0)	37%
12 months	2.2 (0.9 – 5.4)	1.2 (0.5-2.7)	42%

Discussion

Summary of main findings

Per protocol analysis of adults who engaged with a remote behaviour change service (Active Steps) were twice as likely to be physically active at 12 months than a control group. However, this intervention effect was attenuated (to 1.2 times) in a conservative sensitivity analysis assuming all participants lost to follow-up remained inactive, reflecting the impact of high participant attrition on statistical power and certainty of the long-term effects of the service. Conversely, per protocol analysis revealed Active Steps service users were 9 times and 17 times more likely to be physically active at 3 months and 6 months. Sensitivity analysis confirmed these short-to-medium term changes were robust (3 and 5 times more likely to be physically active at 3 and 6 months) even under worst-case assumptions. Changes in physical activity in the per-protocol analysis were accompanied by clinically meaningful improvements in self-reported health and disease-specific quality of life.

Existing literature

Systematic reviews of remote coaching interventions delivered through internet and telephone have shown potential to support behaviour change and promote physical activity in people with chronic lung disease,²⁵ but they are not readily available in the UK. Similar to randomised controlled trials,^26–28 which have assessed the effects of health-coaching interventions, we observed benefits in the short-term and long term. The current study targeted the intervention at adults who were defined as physically inactive, with this at the time being the UK Chief Medical Officers/Sport England definition of less than 30 minutes of moderate-vigorous physical activity per week.^13,14,29 The current study indicates the potential for health coaching interventions to support the most physically inactive people in lung disease populations. Although the absolute between group differences in weekly physical activity were ∼ 60 minutes or less, such changes with the intervention were accompanied with clinically meaningful improvements in self-reported health and quality of life. Health coaching interventions, with a focus on physical activity behaviours have been shown in randomised controlled trials to increase quality of life in COPD³⁰ but less evidence is available for other chronic lung diseases.

Strengths and limitations of the study

To our knowledge, this study is the first to report the analysis of a service delivered remotely by a UK charity to modify physical activity in inactive adults with lung disease. We acknowledge that our outcome measures were all self-reported, which may be prone to recall and social desirability bias. However, outcomes were selected based on physical activity measures (SALS) used at population level in England¹⁴ and tools (EQ5D, CRQ) that are frequently used to assess the health impact of behaviour change interventions in people living with lung disease.^31,32 Methods were also undertaken to limit risk of bias through keeping data collection and intervention delivery in the service as independent as possible and the data submitted immediately to the research team via a web-based system upon collection. The specific nature of the study design (i.e. lack of randomisation) does not eliminate the greater risk of bias with observational studies. For example, participants in the intervention and control groups relied on different recruitment sources albeit the use of a control group allowed for better scrutiny of longitudinal changes in physical activity and health status compared to other possible designs of service implementation in the ‘real world’ (e.g. single arm before-and-after study designs). It is worthy to note that usual care in the control group was not strictly controlled, where some participants had unexpected increases in physical activity. The CRQ was only assessed within Active Steps participants, hence we cannot be certain that the clinically meaningful improvements in CRQ domains at all timepoints were due to the intervention alone.

The eligibility criteria for this study were designed to be as broad as possible to replicate those likely to seek support from the British Lung Foundation. Although the behaviour service engaged with people with a range of lung diseases from various socioeconomic backgrounds, the service largely recruited females. Baseline imbalances between the two study groups were also observed in other characteristics such as age, mental health, heart disease and diabetes. The analyses on physical activity and self-reported health were performed with adjustments for these baseline imbalances. While these adjustments helped to reduce the risk of confounding bias it impacted statistical power in the per protocol analysis, as seen with the shift away from statistically significant effects in the primary outcome between the unadjusted and adjusted model. The current study, both in intervention delivery and data collection also met the difficult conditions of the COVID-19 pandemic. Although the study was aided by remote service delivery and data collection, the impact of COVID-19 on the charity’s services and delivery (including financial position and staffing) meant the total number of participants enrolled in the service over the study period was less than expected and those recruited between late 2019 and March 2020 were unable to be followed-up for the full 12-month period.

A key limitation of this study was the participant attrition rate, particularly important when interpreting true effectiveness of the intervention in the ‘real world’. The primary approach to this study was a per protocol analysis to evaluate the efficacy of the service, in those who remain engaged throughout the 12-month follow-up. To support collection of the primary outcome, the British Lung Foundation sent a letter to all service users prior to the follow-up at 12 months to encourage completion. Service users were given the option of only providing data for the primary outcome. This was partly successful in having service users who had withdrawn from the service come forward to provide data at 12 months. To address the loss of data in the analysis, we employed a non-responder imputation as a conservative sensitivity analysis, where all participants lost to follow-up were classified as physically inactive. It is worthy to note that his approach is recognised in the literature to be overly conservative, particularly in behavioural interventions where missingness may not strictly equate to treatment failure.³³ As this sensitivity analysis is considered to give the most conservative estimate of minimum treatment effect, it may underestimate the true clinical benefit.³⁴ Our comparison of participants who completed the 12-month follow-up to those who were lost to follow-up demonstrated a higher proportion of males in the control group were lost to follow-up at 12 months than completed (65% vs. 35%), but gender was not a significant predictor of activity in the final adjusted models. Hence, while attrition was high, it was largely non-systematic and unlikely to have introduced significant selection bias on our physical outcomes. The possibility of survivor bias, however, cannot be entirely excluded.

Implications for future research and practice

The current study had relatively specific target populations and outcomes. The primary objective of this study dichotomised physical activity outcomes into active or inactive. Active Steps was a component of the ‘Movement for All’ programme, a strategic partnership between multiple health and social care charities in the UK and Sport England. This initiative (including the ‘We Are Undefeatable’ campaign) was designed to support adults with chronic disease to become physically active and rooted in UK national policies stating unquestionable health benefit or greatest return on public investment would be achieved by supporting the most inactive members of the community.¹⁴ ²⁹Hence, the current study targeted reduction of the number of people who would be defined as being physically inactive in the UK (completing less than 30 minutes of moderate-vigorous activity per week). Therefore, the success of the intervention in the current study may not be generalisable to all countries, which may use different definitions for physical inactivity. Challenges in the long-term follow-up of populations with lung disease, including those assessing physical activity, are common.³⁵ The interpretation of the long-term effects on physical activity in the present study was sensitive to participant attrition. The uncertainty of the effect at 12 month is likely attributable to a combination of a significant loss of statistical power and a decline in the intensity of behaviour change over time. The true effect of the behaviour change support provided in this study likely sits somewhere between our per protocol analysis and conservative sensitivity analysis. This highlights a dual challenge for any future remote behaviour change services; the potential need for additional strategies (between 6 and 12 months) to maintain long-term behaviour, and refined retention strategies to maintain the sample size necessary for long-term evaluation. This study demonstrates that charity-led remote support can be an effective strategy to initiate change in physical activity behaviour in adults with lung disease with robust effectiveness sustained for at least six months. Future services will need to focus on identifying and supporting participants at high risk of disengagement to bridge the gap between initial adoption and long-term maintenance.

Conclusions

Active Steps was an effective charity-led remote service to initiate changes in physical activity behaviour in the short-to-medium-term in adults with lung disease, but high participant attrition necessitates caution in interpreting longer term effects at 12 months. Changes in physical activity in people who engaged with the service coincided with clinically meaningful improvements in self-reported health and respiratory specific quality of life.

Supplemental material

Supplemental material - Remote behaviour change service for inactive adults with lung disease: A non-randomised controlled study

Supplemental material for Remote behaviour change service for inactive adults with lung disease: A non-randomised controlled study by Jones AW, Robinson H, Saunders M, Brewster A, Ridyard C, Smith M, Peel J, Shopland J, Wilson C, Dean A, Hinchliffe M, Siriwardena AN, Law G in Chronic Respiratory Disease

Footnotes

Acknowledgements

We would like to thank and acknowledge: all of the participants who took part in this study; Susan Frizzell, British Lung Foundation Active Steps and data teams for supporting data collection; Lincolnshire Community Health Services NHS Trust, Lindum Medical Practice, Nettleham Medical Practice, Spilsby Surgery, Glebe Practice and National Institute for Health Research Clinical Research Network East Midlands for supporting recruitment of the control group.

ORCID iDs

Jones AW

Smith M

Funding

The delivery of the behaviour change service and an independent evaluation was funded by Sport England and the British Lung Foundation.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental material

Supplemental material for this article is available online.

References

Bradley

Wilson

Hayes

, et al. Sedentary behaviour and physical activity in bronchiectasis: a cross-sectional study. BMC Pulm Med 2015; 15: 61. https://doi.org/10.1186/s12890-015-0046-7

Camcioglu

Karadalli

Guclu

, et al. Respiratory and peripheral muscle strength, functional exercise capacity, dyspnea, fatigue and physical activity in patients with interstitial lung disease. Eur Respir J 2014; 44: P4283. https://doi.org/10.1183/13993003/erj.44.suppl_58.p4283

Cordova-Rivera

Gibson

Gardiner

, et al. Physical Activity and Exercise Capacity in Severe Asthma: Key Clinical Associations. J Allergy Clin Immunol Pract 2018; 6: 814–822. https://doi.org/10.1016/j.jaip.2017.09.022

Vorrink

Kort

Troosters

, et al. Level of daily physical activity in individuals with COPD compared with healthy controls. Respir Res 2011; 12: 33. https://doi.org/10.1186/1465-9921-12-33

Watz

Pitta

Rochester

, et al. An official European Respiratory Society statement on physical activity in COPD. Eur Respir J 2014; 44: 1521–1537. https://doi.org/10.1183/09031936.00046814

Benzo

Chang

Farrell

, et al. Physical activity, health status and risk of hospitalization in patients with severe chronic obstructive pulmonary disease. Respir Int Rev Thorac Dis 2010; 80: 10–18.

Brumpton

Langhammer

Henriksen

, et al. Physical activity and lung function decline in adults with asthma: The HUNT Study. Respirology 2017; 22: 278–283. https://doi.org/10.1111/resp.12884

Waschki

Kirsten

Holz

, et al. Physical activity is the strongest predictor of all-cause mortality in patients with COPD: a prospective cohort study. Chest 2011; 140: 331–342. https://doi.org/10.1378/chest.10-2521

Man

Chaplin

Daynes

, et al. British Thoracic Society Clinical Statement on pulmonary rehabilitation. Thorax 2023; 78(Suppl 4): s2–s15. https://doi.org/10.1136/thorax-2023-220439

10.

Steiner

Holzhauer‐Barrie

Lowe

, et al.

Pulmonary Rehabilitation: Steps to breathe better

National Chronic Obstructive Pulmonary Disease (COPD) Audit Programme: Resources and organisation of pulmonary rehabilitation services in England and Wales 2015: Royal College of Physicians, 2016.

11.

Spruit

Singh

Garvey

, et al. An official American Thoracic Society/European Respiratory Society statement: key concepts and advances in pulmonary rehabilitation. Am J Respir Crit Care Med 2013; 188: 13–64. https://doi.org/10.1164/rccm.201309-1634ST

12.

Peel

Webb

Jones

. A remote behaviour change service for increasing physical activity in people with chronic lung conditions: intervention development using the Behaviour Change Wheel. Perspect Public Health 2020; 140(1): 16–21. https://doi.org/10.1177/1757913919880928

13.

Department of Health . Start Active Stay Active: A Report on Physical Activity from the Four Home Countries’ Chief Medical Officers: Department of Health, 2011.

14.

Sport England . Towards an active nation strategy 2016-2021: DCMS, 2016.

15.

Michie

Atkins

West

. The behaviour change wheel: A guide to designing interventions. : Silverback Publishing, 2014.

16.

Milton

Engeli

Townsend

, et al. The selection of a project level measure of physical activity. Sport England, 2017. Available from. https://evaluationframework.sportengland.org/media/1348/short-active-lives-research-report.pdf

17.

Janssen

Pickard

Golicki

, et al. Measurement properties of the EQ-5D-5L compared to the EQ-5D-3L across eight patient groups: a multi-country study. Qual Life Res 2013; 22(7): 1717–1727. https://doi.org/10.1007/s11136-012-0322-4

18.

Charalambous

Molassiotis

. Preliminary validation and reliability of the Short Form Chronic Respiratory Disease Questionnaire in a lung cancer population. Eur J Cancer Care (Engl) 2017; 26(1): e12418. https://doi.org/10.1111/ecc.12418

19.

Moitra

Adan

Akgün

, et al. Less social deprivation is associated with better health-related quality of life in asthma and is mediated by less anxiety and better sleep quality. J Allergy Clin Immunol Pract 2023; 11(7): 2115–2124.e7. https://doi.org/10.1016/j.jaip.2023.03.052

20.

Tsai

C-L

Hodder

Page

, et al. The short-form Chronic Respiratory Disease Questionnaire was a valid, reliable, and responsive quality-of-life instrument in acute exacerbations of chronic obstructive pulmonary disease. J Clin Epidemiol 2008; 61(5): 489–497. https://doi.org/10.1016/j.jclinepi.2007.07.003

21.

Williamson

Aitken

Lawrie

, et al. Introduction to causal diagrams for confounder selection. Respirology 2014; 19(3): 303–311. https://doi.org/10.1111/resp.12238

22.

Gimeno-Santos

Frei

Steurer-Stey

, et al. Determinants and outcomes of physical activity in patients with COPD: a systematic review. Thorax 2014; 69: 731–739. https://doi.org/10.1136/thoraxjnl-2013-204763

23.

Cordova-Rivera

Gibson

Gardiner

, et al. A Systematic Review of Associations of Physical Activity and Sedentary Time with Asthma Outcomes. J Allergy Clin Immunol Pract 2018; 6(6): 1968–1981. https://doi.org/10.1016/j.jaip.2018.02.027

24.

Devlin

Parkin

Janssen

. Methods for Analysing and Reporting EQ-5D Data: Springer, 2020.

25.

Mantoani

Rubio

McKinstry

, et al. Interventions to modify physical activity in patients with COPD: a systematic review. Eur Respir J 2016; 48(1): 69–81. https://doi.org/10.1183/13993003.01744-2015

26.

Altenburg

ten Hacken

Bossenbroek

, et al. Short- and long-term effects of a physical activity counselling programme in COPD: a randomized controlled trial. Respir Med 2015; 109: 112–121. https://doi.org/10.1016/j.rmed.2014.10.020

27.

Coultas

Jackson

Russo

, et al. Home-based physical activity coaching, physical activity, and health care utilization in chronic obstructive pulmonary disease. Chronic Obstructive Pulmonary Disease Self-Management Activation Research Trial secondary outcomes. Ann Am Thorac Soc 2018; 15: 470–478. https://doi.org/10.1513/AnnalsATS.201704-308OC

28.

Jolly

Sidhu

Hewitt

, et al. Self management of patients with mild COPD in primary care: randomised controlled trial. BMJ 2018; 361: k2241. https://doi.org/10.1136/bmj.k2241

29.

Department of Health and Social Care . Physical Activity Guidelines: UK Chief Medical Officers’ Report. : Crown Copyright, 2019. Available at. https://www.gov.uk/government/publications/physical-activity-guidelines-uk-chief-medical-officers-report

30.

Benzo

Hoult

McEvoy

, et al. Promoting Chronic Obstructive Pulmonary Disease Wellness through Remote Monitoring and Health Coaching: A Clinical Trial. Ann Am Thorac Soc 2022; 19(11): 1808–1817. https://doi.org/10.1513/AnnalsATS.202203-214OC

31.

Nolan

Longworth

Lord

, et al. The EQ-5D-5L health status questionnaire in COPD: validity, responsiveness and minimum important difference. Thorax 2016; 71(6): 493–500. https://doi.org/10.1136/thoraxjnl-2015-207782

32.

Cazzola

Hanania

MacNee

, et al. A review of the most common patient-reported outcomes in COPD--revisiting current knowledge and estimating future challenges. Int J Chron Obstruct Pulmon Dis 2015; 10: 725–738. https://doi.org/10.2147/COPD.S77368

33.

Tomson

Björnström

Gilljam

, et al.

Are non-responders in a quitline evaluation more likely to be smokers?

BMC Public Health 2005; 5: 52. https://doi.org/10.1186/1471-2458-5-52

34.

Papp

Fonjallaz

Casset-Semanaz

, et al. Analytical approaches to reporting long-term clinical trial data. Curr Med Res Opin 2008; 24(7): 2001–2008. https://doi.org/10.1185/03007990802215315

35.

Burge

Cox

Abramson

, et al. Interventions for promoting physical activity in people with chronic obstructive pulmonary disease (COPD). Cochrane Database Syst Rev 2020; 4(4): CD012626. https://doi.org/10.1002/14651858.CD012626.pub2

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.37 MB