European Journal of Cardiovascular Nursing

http://cnu.sagepub.com/

Effects of community based cardiac rehabilitation: Comparison with a hospital-based programme

Sultan M Mosleh, Christine M Bond, Amanda J Lee, Alice Kiger and Neil C Campbell
Eur J Cardiovasc Nurs published online 8 January 2014
DOI: 10.1177/1474515113519362

The online version of this article can be found at:

http://cnu.sagepub.com/content/early/2014/01/08/1474515113519362

Published by:

http://www.sagepublications.com

On behalf of:

European Society of Cardiology

Additional services and information for European Journal of Cardiovascular Nursing can be found at:

Email Alerts: http://cnu.sagepub.com/cgi/alerts

Subscriptions: http://cnu.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - Jan 8, 2014

What is This?

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
519362
research-article2014
CNU0010.1177/1474515113519362European Journal of Cardiovascular NursingMosleh et al.

EUROPEAN
SOCIETY OF
Original Article CARDIOLOGY ®

European Journal of Cardiovascular Nursing

Effects of community based cardiac

201X, Vol. XX(X) 1­–9
© The European Society of Cardiology 2014
Reprints and permissions:
rehabilitation: Comparison with a sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1474515113519362

hospital-based programme cnu.sagepub.com

Sultan M Mosleh1, Christine M Bond2, Amanda J Lee2, Alice Kiger3

and Neil C Campbell3

Abstract
Background: With typically fewer than 35% of eligible patients attending outpatient cardiac rehabilitation (CR),
more accessible provision is required. Community-based cardiac rehabilitation is one option but its effects need to be
compared with those of hospital-based CR.
Aims: The purpose of this study was to compare changes in health-related quality of life (HRQOL), anxiety and
depression, and exercise and smoking rates, between attendees at community-based and hospital-based CR programmes.
Method: A prospective comparative cohort design was used. Consecutive patients admitted to Aberdeen Royal
Infirmary and eligible for CR were recruited and followed up by self-report questionnaire. Outcomes were health status
(RAND-36), Hospital Anxiety and Depression Scale (HADS), Godin Leisure-Time Exercise and smoking status.
Results: There were 136 of 179 (75%) attenders at community-based CR, compared to 169 of 209 (80%) at hospital-
based CR (p=0.242). In univariate analysis, there were no significant differences between the two groups in health status,
HADS, and frequency or intensity of exercise immediately after the CR programme or six months later. Adjusting for
other significant factors, patients who attended community CR reported higher RAND-36 energy scores at six months
compared with attenders at hospital CR (p=0.020), but were less likely to undertake frequent exercise (p=0.041).
Conclusions: Community-based CR appears to achieve similar attendance rates and effects on health status and health
behaviour as hospital-based CR. This option might help overcome the poor attendance of patients with long travelling
times to hospital-based CR.

Keywords
Community-based cardiac rehabilitation, hospital-based cardiac rehabilitation, health status, health behaviour

Date received: 10 October 2013; revised 12 December 2013; accepted 15 December 2013

Introduction
Despite accumulating evidence for the benefits of car-
diac rehabilitation (CR), the attendance rate worldwide
and in the UK is poor with less than 35% of eligible
patients participating.1–3 Many barriers to CR attendance
are reported.4 Organisational barriers include work con-
flicts, transport difficulties and long distances to a reha-
bilitation centre. In a quantitative review of 32 studies,
patients were found to be more likely to attend CR when
Community and home-based programmes are ways to
provide CR locally. A Cochrane review found that home-
based CR had similar effects and costs to hospital-based
1Department of Fundamentals and Adult Nursing, University of Mutah,
Jordan
2Centre of Academic Primary Care, University of Aberdeen, Scotland
3Medical Statistics Team, University of Aberdeen, Scotland

the programmes were easily accessible.5 Bittner et al.

Corresponding Author:
found that referral to CR was more likely if the patient Sultan Mosleh, Department of Fundamentals and Adult Nursing, Faculty
lived in area in which the rehabilitation facility was of Nursing, University of Mutah, PO Box 7, Karak 61710, Jordan.
located.6 Email: s.mosleh@mutah.edu.jo

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
2 European Journal of Cardiovascular Nursing XX(X)
programmes.7 Some patients, however, value the opportu-
nity for peer support and this may improve their uptake and
adherence.8 Community-based group programmes offer
this opportunity and early evaluations have reported
improvements in exercise levels and reductions in angina
similar to those reported by hospital-based programmes.9
Since then, however, there have been changes in both CR
programme content and participant demography. Early pro-
grammes consisted of exercise training, whereas now a
comprehensive approach is recommended, including
assessment of risk factors, psychological and educational
interventions, risk factor correction, stress management
and relaxation training, and delivered by a multidiscipli-
nary group according to national standards.4,10,11 Patients
now enrolled in cardiac rehabilitation are older and more
likely to have co-morbidities such as diabetes.12
As community-based programmes may have less access
to specialist staff and resources compared to their hospital
counterparts, it is important to investigate whether, in
practice, they have similar benefits or not. The aim of this
study was to compare changes in health related quality of
life (HRQOL), anxiety, depression, exercise and smoking
rates, between attendees at community-based versus hos-
pital-based CR programmes.
Methods
Design
This was a prospective comparative cohort study.
Setting
The study was set in northeast Scotland. The locations
included one hospital-based programme and four commu-
nity programmes. Recruitment and baseline data collec-
tion (time zero: T0) were between January–December
2007. First follow-up (T1) commenced in June 2007 and
ceased in June 2008. Second follow-up (T2) commenced
in February 2008 and ceased in November 2008.
Cardiac rehabilitation
The eight-week outpatient CR programme in all five loca-
tions was developed by the same team to comply with the
British Association of Cardiovascular Prevention and
Rehabilitation (BACPR) and Scottish Intercollegiate
Guideline Network (SIGN) guidelines which are evidence-
based.10,11 They normally commence six weeks after hos-
pital discharge, although in some cases they begin later
depending on the waiting list. There are two sessions of
exercise training and relaxation, and one session of educa-
tion per week over eight weeks. The hospital-based pro-
gramme is provided at the University Hospital, a teaching
hospital. It is run by a multidisciplinary team (a CR nurse,
a physiotherapist and a dietician) and a psychologist and
Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
physician are available if needed. The community-based
programmes are delivered from sports centres in
Aberdeenshire at Banchory, Banff, Peterhead and Inverurie
by a nurse and a physiotherapist
Participants
All patients enrolled in CR were eligible to take part in the
study. Patients eligible for the programme were those
admitted to the University Hospital with acute myocardial
infarction or undergoing coronary artery bypass surgery or
coronary angioplasty and referred to either the hospital-
based CR in Aberdeen city or one of the four community-
based CR programmes in Aberdeenshire based upon where
they lived. Patients were excluded from the CR programme
and this study if they had terminal illness, arrhythmia,
alcohol or drug abuse, or mental or physical disability.
Outcome measures
Five health outcome measures were used. The first three
are suggested in both the National Service Framework
(NSF) for coronary heart disease and the national guide-
lines for CR as valid, reliable instruments for audit of CR
programmes.13
HRQOL
The RAND-36 (RAND: Research and Development,
http://www.rand.org/) was used to evaluate patients’
HRQOL. The 36 items on the questionnaire are divided
into eight scales: physical functioning (10 items), social
functioning (two items), role limitation due to physical
problem (five items), role limitation due to emotional
problem (three items), mental health (five items), energy/
fatigue (four items), bodily pain (two items), and general
health (five items).14 The RAND-36 includes the same
items as the SF-36 (SF: Short-Form Health Survey), as
developed in the Medical Outcomes Study. The SF-36 has
been found to be a sensitive, reliable generic HRQOL
instrument for cardiac patients.15
Anxiety and depression
The Hospital Anxiety and Depression Scale (HADS) is a
well-validated screening tool that was included to describe
the patient’s mental health status. It has a sensitivity and
specificity greater than 80% in cardiac patients and good
internal consistency with a mean Cronbach’s alpha 0.83
for anxiety and 0.82 for depression.16
Physical exercise
The Godin Leisure-Time Exercise questionnaire was used to
measure the time spent performing exercise during leisure.17
It is a self-administration seven-day recall questionnaire and
Mosleh et al. 3
has three levels of exercise, strenuous; moderate; and mild
exercise. The participants were asked how often per week
they engaged in these three levels exercise for 15 min or
more during their free time. It also enquires into the weekly
frequency of sweat-inducing activity. Both the NSF for coro-
nary heart disease and the national guidelines for CR have
selected the Godin Leisure-Time Exercise questionnaire as a
valid, reliable instrument to evaluate patients who attend a
CR programme and to develop a dataset for audit of the ben-
efit of CR.13 The internal reliability for questions is satisfac-
tory, with the mean of Cronbach’s α ranging from
0.62–0.81.18
Exercise intention
The Exercise Intention and Planning questionnaire was
used to provide information on patients’ intention to
change their behaviour and practice physical exercise in
future. The questionnaire was developed by Luszczynska
and Schwarzer19 and it is composed of three subscales:
intention to maintain exercise (six items); action planning
(five items), which is about when, where, and how exer-
cise is taken; and coping planning (five items), which
assesses respondents’ plans to maintain physical exercise
in difficult situations (e.g. having little time or experienc-
ing a setback). Reported Cronbach’s alpha values are
0.82–0.88 for intentions, 0.92–0.95 for action planning
and 0.90–0.91 for coping planning.19
Smoking status
Smoking status was measured with items developed
from previous studies.20,21 Participants selected the state-
ment that best described their smoking status from the
following options: ‘I smoke daily’; ‘I smoke occasion-
ally, but not every day’; ‘I used to smoke daily, but do
not smoke at all now’; ‘I used to smoke occasionally, but
do not smoke at all now’ and ‘I have never smoked’.
Patients who used to smoke were then asked to indicate
the date they stopped or the number of years since they
had stopped smoking.
Additional variables collected
All of the above measures were combined into a simple
questionnaire with additional items on demography (age,
sex, co-habitation, employment status). Patient’s depriva-
tion level was described using Carstairs scores for Scottish
postcode sectors from the 2001 Census. Carstairs decile
scores were assigned to each participant based on their
address postcode of residence.22 These scores, were col-
lapsed into three categories. The first category represent-
ing the least deprived 30% of the Scottish population
(1=least deprived), the third category representing the
most deprived 30% (3=most deprived) and the second
Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
level representing the remaining 40% of Scottish popula-
tion (2= moderate deprivation). Data were also collected
on key medical co-morbidities: diabetes, stroke, cancer,
myocardial infarction, cardiac surgery, percutaneous trans-
luminal coronary angioplasty, hypertension, respiratory
diseases (including asthma, emphysema, chronic bronchi-
tis), and joint diseases (including rheumatism, arthritis,
chronic back pain).
Data collection and method of follow-up
Smoking status, HADS and Godin Leisure-Time Exercise
exercise were measured pre-hospital discharge (T0) by
self-completion questionnaire. All measures except smok-
ing status were applied at T1 (immediately after the eight-
week course of CR had been completed) by postal
questionnaire and all measures were then repeated at T2
(six months later) again by postal questionnaire. Two
reminder questionnaires were sent to non-responders at
two-weekly intervals. Participants who did not respond to
the first follow-up (T1) were still eligible for second fol-
low-up (T2) questionnaire unless they had withdrawn or
died.
Sample size calculation
With a sample of 212 participants at baseline, the study
had 80% power at the two-sided 5% significance level to
detect a 10-point (moderate) difference in the mean
RAND-36 scores between those patients who had attended
hospital versus those who had attended community based
CR.23,24
Statistical methods
Statistical analyses were conducted using SPSS 18.
Missing data were treated as missing at random and
excluded from statistical analysis. The two-sample t-test
was used to compare continuous health outcomes (exer-
cise intention score, HADS) between attenders at the
hospital and community based programmes. Scores in
the RAND-36 and the Godin Leisure-Time Exercise
questionnaire were not normally distributed, so the
Mann-Whitney rank test was used. Between groups dif-
ferences in smoking status was investigated using the
Chi-square test.
In order to identify potential confounders, appropriate
univariate tests were used to identify associations
between outcome variables and the additional demo-
graphic, socio-economic and morbidity variables col-
lected at baseline. The Pearson correlation coefficient
was used to investigate the strength of the linear associa-
tion between age, which was normally distributed, and
the T2 scores of the RAND-36, the Godin Leisure-Time
Exercise questionnaire and the HADS. The Spearman’s
4 European Journal of Cardiovascular Nursing XX(X)

115 declined
Invited to take part in the study 49 excluded:
(n=551) 29 medical reasons
4 died
16 not referred to CR

Before hospital discharge (T0) Agreed to participate

Questionnaire (Godin Leisure Time
n=387
Exercise, HADS and Smoking)

Invited to Invited to
hospital CR community CR
n=208 n=179

Attended Attended
hospital CR community CR
n=169 n=134

Completion of CR programme (T1) T1 T1

Questionnaire (RAND-36, the Godin
Leisure Time Exercise, Exercise
142 (84%) 115 (86%)
Intention and Planning and HADS) respondents respondents

Six months after T1 (T2)

Questionnaire (RAND-36, the Godin T2 T2
Leisure Time Exercise, Exercise 128 (76%) 109 (81%)
Intention and Planning, HADS and
respondents respondents
Smoking)

Figure 1.  Patient recruitment and follow-up process.

HADS: Hospital Anxiety and Depression Scale.

rank correlation coefficient was used to investigate the Ethical approval

strength of the linear association between the scores of
the RAND-36 and the Godin Leisure-Time Exercise
questionnaire and HADS.
Factors that emerged from univariate analysis with a
significance level of ≤0.05 were included in multivariate
models as potential confounders. Forward stepwise linear
regression modelling was used to identify whether com-
munity or hospital-based CR was independently associ-
ated with the continuous health outcomes (HRQOL,
HADS and physical exercise) at T2 after adjustment for
baseline factors. Forward stepwise logistic regression was
similarly used to identify whether site of CR was indepen-
dently associated with smoking status at T2 after adjust-
ment for baseline confounders.
Full ethical approval was sought from the Grampian
Research Ethics Committee and was granted on 17
November 2006 (ref. no.: 06/S0802/119).
Results
Of the 551 patients eligible to take part in the study, 115
(21%) declined and 49 (9%) became ineligible for CR
(29 became unwell, 16 were not referred as the referral
sheets were lost, and four died prior to data collection).
The remaining 387 (68%) patients agreed to take part in
the study and completed the baseline questionnaire
(Figure 1).

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
Mosleh et al. 5

Table 1.  Baseline comparison between attenders and non-attenders at cardiac rehabilitation (CR).

Attenders Non-attenders p-
(n=303) (n=80)a value
Age 61 (11) 67 (10) <0.001
Gender (male) 219 (72) 47 (59) 0.028
Married or living as married 230 (76) 53 (66) 0.108
Living alone 54 (18) 20 (25) 0.198
Medical history  
High blood pressure 116 (38) 42 (52) 0.030
Diabetes 44 (14) 14 (17) 0.627
Respiratory disease 38 (13) 9 (11) 0.903
Joint disease 46 (15) 18 (22) 0.164
Cancer 8 (3) 1 (1) 0.692
Stroke 6 (2) 4 (5) 0.134
Myocardial infarction 215 (71) 61 (76) 0.425
Cardiac surgery 121 (40) 27 (34) 0.378
Angioplasty 93 (30) 27 (34) 0.697
Working status  
Employed 131 (44) 21 (26)  
Unemployed 15 (5) 3 (4) 0.014
Sick 15 (5) 8 (10)  
Retired 134 (46) 48 (60)  
Smoking status  
Current smoker 46 (15) 16 (20)  
Ex-smoker 166 (55) 41 (51) 0.585
Never smoked 90 (30) 23 (29)  
Anxiety level at admission  
Not anxious (score <8) 170 (60) 43 (58) 0.700
Borderline anxious 65 (23) 20 (27)  
Anxious (score >10) 50 (17) 11 (15)  
Depression level at admission  
Not depressed (score <8) 230 (81) 57 (77) 0.770
Borderline depressed 41 (14) 13 (18)  
Depressed (score >10) 14 (5) 4(5)  
Deprivation level  
Level 1 (low deprivation) 194 (65) 55 (71) 0.145
Level 2 77 (26) 12 (16)  
Level 3(high deprivation) 28 (9) 10 (13)  
aFour of non-attenders had died, so they were excluded from analysis.

Values are n (%) or mean (standard deviation (SD)).

One hundred and seventy-nine patients were invited to
community-based CR, of whom 136 (75%) attended.
Two hundred and eight patients were invited to hospital-
based CR of whom 169 (80%) attended. This difference
in attendance rate by site of CR was not statistically sig-
nificant (p=0.242). At baseline, univariate analyses
showed that attenders tended to be male (p=0.028),
younger (p<0.001), employed (p=0.014) and less likely
to have hypertension (p=0.030) (Table 1), compared to
non-attenders.
Comparing baseline (T0) demographic, medical history
and risk factor data from attenders at hospital and commu-
nity programmes, the only significant difference was in
deprivation level. Attenders at the hospital programme
were more likely to be from socially-deprived areas com-
pared with attenders at the community-based programmes
(p=0.001, Table 2).
At both follow-up points (T1 and T2), there were no
significant differences between the two CR site groups in
either overall health status or in any of the RAND-36 sub-
scales (Table 3). There were also no significant differences
in exercise levels, exercise intentions and planning, nor
HADS scores. At T2 there were no significant CR site dif-
ferences in smoking status(Table 4).
Adjusting for potential baseline confounders using lin-
ear regression modelling, CR programme location was a
significant independent predictor for the energy subscale
of health status (RAND-36) and Godin Leisure-Time

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
6 European Journal of Cardiovascular Nursing XX(X)

Table 2.  Baseline comparison of attenders at hospital and community cardiac rehabilitation (CR).

Hospital-CR Community-CR p-value

(n=169) (n=136)
Age 61.7 (11.6) 60.5 (12.1) 0.383
Gender (male) 118 (70) 101 (75) 0.346
Married or living as married 121 (72) 109 (81) 0.067
Cohabiting 135 (80) 114 (85) 0.307
Medical history  
Hypertension 66 (39) 51 (38) 0.954
Diabetes 25 (15) 18 (13) 0.926
Respiratory disease 149 (88) 116 (87) 0.808
Joint disease 139 (82) 118 (88) 0.215
Cancer 3 (2) 5 (4) 0.488
Stroke 5 (3) 1 (0.7) 0.388
Myocardial infarction 116 (67) 99 (74) 0.384
Cardiac surgery 73 (43) 49 (37) 0.294
Angioplasty 51 (30) 42 (31) 0.926
Working status  
Employed 74 (44) 63 (47) 0.674
Unemployed 7 (4) 8 (6)  
Sick 10 (6) 5 (4)  
Retired 78 (46) 58 (43)  
Smoking status  
Current smoker 28 (17) 18(13) 0.450
Ex-smoker 87(52) 79(59)  
Never smoked 53 (31) 37(28)  
Anxiety level at admission  
Not anxious (score <8) 92 (58) 78 (61)  
Borderline anxious 35 (22) 30 (24) 0.394
Anxious (score >10) 31 (20) 19 (15)  
Depression level at  
admission
Not depressed (score <8) 127 (80) 102 (81)  
Borderline depressed 26 (16) 16 (13) 0.814
Depressed (score >10) 6 (4) 8 (6)  
Deprivation level  
Level 1 (low deprivation) 100 (60) 94 (72) 0.001
Level 2 41 (24) 36 (27)  
Level 3(high deprivation) 27 (16) 1 (1)  

Values are n (%) or mean (standard deviation (SD)).

Exercise levels at T2 (Table 5). Patients who attended a
community programme reported higher energy scores
compared with attenders at a hospital programme, but
reported taking less exercise. There were no other signifi-
cant differences between attenders and hospital or commu-
nity-based programmes.
Discussion
Our study showed that there were no significant differ-
ences in HRQOL, anxiety and depression mood, physical
exercise, exercise intention and smoking status, between
patients who attended a community-based programme and
those who attended a hospital-based programme. This
suggests that both types of programmes are similarly ben-
eficial. In addition, the attendance rates at both pro-
grammes were similar.
This study had both strengths and limitations. The com-
munity and hospital CR programmes had the same objectives
and management strategies, with the only difference being
that the community CR is run by fewer, less-specialised staff.
A limitation of the study is that CR site was not randomised
between patients since this was a pragmatic real-life compari-
son. Random allocation would be the best choice to assess
differences in health outcomes between community and hos-
pital-based programme CR but, in practice, such a project
would be difficult to deliver as it would involve setting up
community and hospital programmes co-located to each

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
Mosleh et al. 7

Table 3.  Health status, exercise, exercise intentions and planning, anxiety and depression comparing attenders at hospital and
community cardiac rehabilitation (CR) at first and second follow-up.

First follow-up (T1) Second follow-up (T2)

p-value
  Hospital CR Community CR p-value Hospital CR Community CR

  n Median (IQR) n Median n Median (IQR) n Median (IQR)

(IQR)
Physical functioning 142 85 (60–95) 115 85 (65–95) 0.518 128 85 (60–95) 109 85 (70–95) 0.882
Role limitation due to 141 75 (25–100) 114 87 (18–100) 0.907 126 100 (25–100) 107 75 (25–100) 0.607
physical problems
Role limitation due to 140 100 (66–100) 114 100 (66–100) 0.698 126 100 (66–100) 107 100 (100–100) 0.120
emotional problems
Energy/ fatigue 142 65 (45–80) 113 65 (50–80) 0.744 126 65 (44–80) 108 70 (55–80) 0.053
Emotional well-being 142 80 (64–92) 114 84 (64–92) 0.842 126 84 (64–92) 108 88 (76–92) 0.110
Social functioning 142 100 (72–100) 114 100 (62–100) 0.745 126 100 (72–100) 108 100 (87–100) 0.248
Body pain 142 90 (67–100) 114 90 (67–100) 0.752 126 89 (67–100) 108 100 (77–100) 0.231
General health 142 70 (50–85) 115 75 (50–85) 0.583 128 70 (50–83) 109 75(50–85) 0.513
Godin Leisure-Time 142 37 (21–56) 114 41 (21–56) 0.513 126 34 (19–56) 109 31 (19–53) 0.397
Exercise score
  Mean (SD) Mean (SD) Mean (SD) Mean (SD)  
Intention to maintain 140 3.11 (0.54) 113 3.22 (0.53) 0.103 126 3.05 (0.58) 107 2.99 (0.53) 0.062
exercise
Action planning for 136 2.62 (0.74) 109 2.67 (0.71) 0.589 124 2.54 (0.76) 104 2.41 (0.59) 0.740
exercise
Coping planning for 135 2.39 (0.65) 109 2.49 (0.63) 0.265 123 2.32 (0.70) 104 2.41 (0.59) 0.277
exercise
Anxiety (HADS) 142 12.62 (2.20) 114 12.69 (2.17) 0.810 128 12.64 (2.22) 109 13.02 (1.86) 0.150
Depression (HADS) 142 8.70 (1.24) 114 8.62 (1.33) 0.633 128 8.63 (1.31) 109 8.83 (1.37) 0.260

HADS: Hospital Anxiety and Depression Scale; IQR: interquartile range; SD: standard deviation.

Table 4.  Smoking status of attenders at hospital and community cardiac rehabilitation (CR) at baseline and second follow-up.

Smoking status Baseline (T0) Second follow-up (T2)

Hospital Community p-value Hospital Community p-value

CR CR CR CR
Smoker 28 (17) 18 (13) 16 (12) 10 (9)
Ex-smoker 87 (52) 79 (59) 0.450 68 (54) 63 (58) 0.661
Never smoked 53 (31) 37 (28) 43 (34) 36 (33)

Values are n (%).

other. Only in this way would it be possible to avoid the con-
founding impact of a participant’s place of residence. The
main difference between site groups was in deprivation, with
more deprived participants attending the hospital-based pro-
gramme. Deprivation can be associated with poorer health
status and health behaviour (so may have disguised poorer
outcomes among community-based programme participants),
but we adjusted for it and related variables (employment and
morbidity) in our multivariate analysis.
The study was powered to detect a significant differ-
ence in the mean change between the groups, however
loss of follow-up is a potential source of limitation related
to the longitudinal design of this study. In a study which
follows people over time, it is inevitable that some indi-
viduals are no longer available for study. The percentage
of loss at first and second follow-up was 19% and 27.2%
respectively. Compared with those who responded, the
lost individuals were significantly more likely to be
younger and current smokers, and less likely to have
hypertension. The fact that the non-respondents were
younger and had less experience of hypertension contra-
dicts previous research which found patients with old age
and more co-morbidities were less likely to respond to a
postal questionnaire.25,26 It could support the hypothesis

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
8 European Journal of Cardiovascular Nursing XX(X)

Table 5.  Differences in health status, anxiety, depression, exercise and smoking at T2 between attenders at community-based
cardiac rehabilitation (CR) and attenders at hospital-based CR, adjusted for confounders identified from univariate analyses.

Model Adjusted difference (attenders 95% CI p-value

at community CR versus
attenders at hospital CR)
Model 1a Physical functioning 2.35 –3.60–8.31 0.437
Model 2b Role limitation due to physical problems –1.07 –11.25–9.11 0.836
Model 3c Role limitation due to emotional problems 6.56 –1.83–14.96 0.125
Model 4d Energy/ fatigue 6.45 1.04–11.86 0.020
Model 5e Emotional well-being 3.77 –0.97–8.52 0.118
Model 6f Social function 3.89 –1.41–9.20 0.150
Model 7g Body pain 2.46 –4.80–9.73 0.506
Model 8h General health 3.23 –2.05–8.52 0.229
Model 9i Depression 0.19 –0.15–0.53 0.274
Model 10j Anxiety 0.26 –0.25–0.77 0.314
Model 11k Godin Leisure-Time Exercise score –10.06 –19.69– –0.42 0.041
Model 12l Smoking (odds ratio) 0.90 0.23–3.52 0.880

CI: confidence interval.

aAdjusted for age, depression, gender, marital status, working status, smoking status, joint diseases badjusted for age, depression, working status,

joint disease, smoking status; cadjusted for age, depression, working status, joint disease, smoking status; dadjusted for depression, anxiety, working
status, joint disease; eadjusted for depression, anxiety, joint disease and history of hypertension; fadjusted for depression, anxiety, working status,
joint disease; gadjusted for working status, joint disease; hadjusted for depression, anxiety, working status, smoking status, joint disease, respiratory
diseases; iadjusted for depression at baseline; jadjusted for marital status, joint disease, hypertension, anxiety at baseline and follow-up; kadjusted for
age, working status, joint disease, smoking status; llogistic regression adjusted for deprivation score and marital status.

that young patients are more likely to be healthy and less
likely to seek secondary prevention after their hospitali-
sation, or that they are less interested as they return to
work. It is difficult to know what effect the exclusion of
those individuals had on the study’s findings.
Overall, this study suggests that patients who attend
community-based CR can expect similar benefits to
their health status (as measured by RAND-36 and
HADS) and exercise levels to patients who attend hos-
pital-based CR. However, there were two significant
differences that we identified. Six months after they
completed CR, attenders at community-based CR had
more energy (felt less fatigued), but undertook less
exercise than attenders at hospital-based CR. These dif-
ferences were not apparent immediately after CR, but
developed subsequently. If they had been due to differ-
ences in the CR programmes, they might have been
expected immediately after CR. Perhaps community-
based programmes were less effective at equipping par-
ticipants to sustain physical activity levels. There were,
however, no significant differences in exercise inten-
tions or planning, the behavioural mechanisms used to
sustain exercise. It is possible that facilities for sustain-
ing exercise are less available in rural areas. The Godin
Leisure-Time Exercise questionnaire focuses mainly
on formal exercise, so we may have not fully captured
other forms of exercise undertaken as part of employed
work which may be higher in the community-based
attendees. None of this explains the higher energy lev-
els reported by attenders at community-based CR, so
this apparently paradoxical finding may be due to
chance.
There are few studies comparing the effects of commu-
nity-based CR with traditional hospital-based programmes.
In one recent randomised controlled trial in Leicester, 100
low-risk myocardial infarction patients were assigned to
either Phase 3 hospital-based CR (n=54) or a fast-tracked
group in a community Phase 4 scheme led by an exercise
instructor (n=46).27 The researchers reported similar ben-
efits between groups in HRQOL. Another qualitative study
explored the perceptions of patients who attended hospital
and community CR programmes in Leeds. The authors
concluded that hospital-based and community-based CR
programmes were viewed by patients and staff as a valua-
ble additional service for cardiac patients.28 The findings
of these small studies, set in different parts of the United
Kingdom, are in line with our pragmatic comparison, sup-
porting their validity and suggesting that they may be more
widely generalisable.
Although we did not compare community-based CR
with home-based programmes, previous research has dem-
onstrated that home-based programmes also have similar
effects to traditional hospital-based CR. 7 The implications
for patients are that the three methods of receiving CR –
hospital, home and community-based programmes –
appear to offer similar benefits, at least in terms of health
behaviour and quality of life. Providing patients with
choice might help to improve the poor attendance rates
reported for CR, especially for patients with long travel-
ling times to centralised programmes.

Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
Mosleh et al. 9
12. Evans J, Bethell H, Turner S, et al. Characteristics of patients
Implications for practice
•• Community-based programmes have similar
effects to traditional hospital-based CR.
•• Community programmes are a suitable alternative
to traditional hospital-based CR programmes.
•• Community-based CR might help overcome the
poor attendance of patients with long travelling
times to hospital-based CR
Conflicts of interest
The authors declare that there are no conflicts of interest.
Funding
This research received no specific grant from any funding agency
in the public, commercial, or not-for-profit sectors.
References
1. The National Audit for Cardiac Rehabilitation (NACR).
Annual reports, www.cardiacrehabilitation.org.uk/nacr/
reports.htm (2007–2013, accessed 10 August 2013).
2. Department of Health. National service framework for coronary
heart disease, www.dh.gov.uk/en/Publicationsandstatistics/
Publications/PublicationsPolicyAndGuidance/DH_4094275
(2000, accessed 15 June 2010).
3. Bethell H, Lewin R, Evans J, et al. Outpatient cardiac reha-
bilitation attendance in England: Variability by region and
clinical characteristics. J Cardiopulm Rehabil Prev 2008;
28: 386–391.
4. Bethell H, Lewin R and Dalal H. Cardiac rehabilitation in
the United Kingdom. Heart 2009; 95(4): 271–275.
5. Jackson L, Leclerc J, Erskine Y, et al. Getting the most out
of cardiac rehabilitation: A review of referral and adherence
predictors. Heart 2005; 91: 10–14.
6. Bittner V, Sanderson B, Breland J, et al. Referral patterns
to a university-based cardiac rehabilitation program. Am J
Cardiol 1999; 83: 252–255.
7. Dalal HM, Zawada A, Jolly K, et al. Home based versus
centre based cardiac rehabilitation: Cochrane systematic
review and meta-analysis. Brit Med J 2010; 340: b5631.
8. Clark AM, Munday C, McLaughlin D, et al. Peer support to
promote physical activity after completion of centre-based
cardiac rehabilitation: Evaluation of access and effects. Eur
J Cardiovasc Nurs 2012; 11: 388–395.
9. Bethell HJ and Mullee MA. A controlled trial of community
based coronary rehabilitation. Br Heart J 1990; 64: 370–375.
10. Scottish Intercollegiate Guideline Network (SIGN). Cardiac
rehabilitation. SIGN Guideline No. 57. Edinburgh: SIGN,
2002.
11. British Association for Cardiovascular Prevention and
Rehabilitation. The BACPR standards and core components
for cardiovascular disease prevention and rehabilitation 2012,
http://www.bacpr.com/resources/8BZ_BACPR_Standards_
and_Core_Components_2012.pdf (accessed 11 May 2013).
Downloaded from cnu.sagepub.com at MICHIGAN STATE UNIV LIBRARIES on January 15, 2014
entering cardiac rehabilitation in the United Kingdom 1993–
2006: Implications for the future. J Cardiopulm Rehabil
Prev 2011; 31: 181–187.
13. Lewin B, Roebuck A and Thompson D. Coronary heart
disease national service framework: Cardiac reha-
bilitation, meeting the information needs. Northern &
Yorkshire Public health Observatory Occasional Paper
No.4, http://www.cardiacrehabilitation.org.uk/nacr/docs/
PHO_Article.pdf (2003, accessed 15 January 2009).
14. Hays RD, Sherbourne CD and Mazel RM. The RAND
36-Item health survey 1.0. Health Econ 1993; 2: 217–227.
15. Dempster M and Donnelly M. Measuring the health related
quality of life of people with ischaemic heart disease. Heart
2000; 83: 641–644.
16. Bjelland I, Dahl AA, Haug TT, et al. The validity of the
Hospital Anxiety and Depression Scale: An updated review.
J Psychiat Res 2002; 52: 69–77.
17. Godin G and Shephard RJ. A simple method to assess exer-
cise behaviour in the community. J Applied Sport Sciences
1985; 10: 141–146.
18. Godin G and Shephard RJ. (1997) Godin Leisure-Time
Exercise questionnaire. Med Sci Sports Exerc 1997; 29:
S36–S38.
19. Luszczynska A and Schwarzer R. Planning and self-efficacy
in the adoption and maintenance of breast self-examination:
A longitudinal study on self-regulatory cognitions. Psychol
Health 2003; 18: 93–108.
20. Lim LL, Johnson NA, O’Connell RL, et al. Quality of life
and later adverse health outcomes in patients with sus-
pected heart attack. Aust N Z J Public Health 1998; 22:
540–546.
21. Bengtsson I, Hagman M, Währborg P, et al. Lasting impact
on health-related quality of life after a first myocardial
infarction. Int J Cardiol 2004; 97: 509–516.
22. Morgan O and Baker A. Measuring deprivation in England
and Wales using 2001 Carstairs scores. Health Stat Q 2006;
31: 28–33.
23. Shi H, Lee H, Chiu C, et al. (2008) Responsiveness and
minimal clinically important differences after cholecystec-
tomy: GIQLI versus SF-36. J Gastrointest Surg 2008; 12:
1275–1282.
24. Petrie A and Samib C. Medical statistics at a glance. 2nd ed.
Oxford: Blackwell Publishing, 2005.
25. Une H, Miyazaki M and Momose Y. Comparison of mor-
tality between respondents and non-respondents in a mail
survey. J Epidemiol 2000; 10: 136–139.
26. Criqui MH, Barrett-Connor E and Austin M. Differences
between respondents and non-respondents in a population-
based cardiovascular disease study. Am J Epidemiol 1978;
108: 367–372.
27. Robinson HJ, Samani NJ and Singh SJ. Can low risk cardiac
patients be ‘fast tracked’ to Phase IV community exercise
schemes for cardiac rehabilitation? A randomised con-
trolled trial. Int J Cardiol 2011; 146: 159–163.
28. Blake E, Tsakirides C and Ingle L. Hospital versus commu-
nity-based phase III cardiac rehabilitation. Br J Nurs 2009;
18: 116–122.