Evaluating the Angina Plan in Patients Admitted to Hospital with Angina: A Randomized Controlled Trial


Dr. Stella Zetta, Thettalon 19, Karditsa, 43100, Greece.
Tel.: 00302441029380;
E-mail: skzetta@aim.com


The aim of this trial was to evaluate the Angina Plan (AP), a cognitive-behavioral nurse-facilitated self-help intervention against standard care (SC). A randomized controlled trial of 218 patients hospitalized with angina assessed participants predischarge and 6 months later. Data were collected during a structured interview using validated questionnaires, self-report, and physiological measurement to assess between group changes in mood, knowledge and misconceptions, cardiovascular risk, symptoms, quality of life, and health service utilization. The intention-to-treat (ITT) analysis found no reliable effects on anxiety and depression at 6 months. AP participants reported increased knowledge, less misconceptions, reduced body mass index (BMI), an increase in self-reported exercise, less functional limitation, and improvements in general health perceptions and social and leisure activities compared to those receiving SC. Sensitivity analysis excluding participants with high baseline depression revealed a statistical significant reduction in depression levels in AP compared to the SC participants. Analysis excluding participants receiving cardiac surgery or angioplasty removed the ITT effects on physical limitation, self-reported exercise and general health perceptions and the improvements seen in social and leisure activities, while adaptive effects on knowledge, misconceptions and BMI remained and between-group changes in depression approached significance. Initiating the AP in a secondary care setting for patients with new and existing angina produces similar benefits to those reported in newly diagnosed primary care patients. Further evaluation is required to examine the extent of observed effects in the longer term.


International statistics show 9.8 million people in the USA and 1.98 million in the UK suffer from angina pectoris and experience considerable mortality and morbidity [1,2] from this common clinical manifestation of coronary heart disease (CHD). It is estimated a further 96,000 and 785,000 new cases of stable angina are identified in the UK and USA, respectively [1,2]. This results in an annual increase in hospital admissions associated with angina; however, this group is not always included in traditional rehabilitation and recovery programs [3].

Many patients with angina report psychological distress (anxiety and depression), which can influence the development and prognosis of CHD [4–6]. Symptoms including chest pain, fatigue, impaired physical function, and quality of life may affect recovery and may result in disability and increased hospital readmissions [4,7–11]. Therefore, when evaluating the effectiveness of healthcare interventions that focus on angina, outcomes such as anxiety and depression, symptoms, physical function, quality of life, and service uptake should be assessed.

Evaluation of cardiac rehabilitation (CR) and secondary prevention programs targeting lifestyle modification following a cardiac event have shown improved knowledge and risk factor profiles, increased prescription and uptake of secondary prevention treatments, improved mood, clinical and physical status, all-cause mortality and cardiac mortality, quality of life, and a reduction in hospital admissions [12–14]. Traditionally, however, angina patients who are likely to gain similar benefit from such interventions have often been excluded from CR programs [3]. One trial, which evaluated the health outcomes of patients with angina 6 months after discharge from a chest pain service, showed that 58% of patients reported ongoing symptoms 42% reported anxiety and 23% moderate depression [10] and limited change in risk factors postdischarge has been reported [10,15].

Previous research has highlighted the need to identify and correct misconceptions or maladaptive beliefs related to CHD which are associated with increased levels of anxiety, increased physical limitations, and more admissions to hospital [9,16]. Correcting misconceptions relating to angina is thought to give patients the confidence to engage in lifestyle changes and to enable more effective self-management of their long-term condition [16]. A randomized controlled trial (RCT) in selected primary care patients has demonstrated that a brief cognitive-behavioral nurse-facilitated self-help intervention, the Angina Plan (AP), can reduce anxiety and depression, self-report of angina and physical disability when added to a routine secondary prevention clinic in community patients [9]. However, such intervention is currently not widely available to angina patients in the UK. There is a need, therefore, to offer structured support to angina patients to target their educational needs, correct any misconceptions, reduce anxiety and depression, optimize symptom management, and encourage lifestyle modification [11].

The aim of this current trial was to establish the effectiveness of the AP in an unselected group of the CHD population. This unselected population would reflect angina patients who are being admitted every day into an acute admission or cardiology area. This may include patients with a past history of cardiac disease or this may be their first presentation with CHD. Since patients may be more willing to change behavior at this time, there is an opportunity to initiate this established intervention in a hospital setting following an acute event and to continue the follow-up in the community postdischarge. Anxiety as determined by the Hospital Anxiety and Depression Scale (HADS) [17] was the primary outcome measure in this trial following the work of Lewin et al. (2002) [9]. An abstract of this trial has been previously presented [18].


This pragmatic RCT with broad inclusion criteria targeted patients who are commonly admitted to a medical admissions or coronary care unit in daily practice in two hospitals incorporating a teaching hospital and smaller district general hospital. This allowed the AP intervention to be evaluated within the context of current clinical practice over a 22-month period. Data were collected by a structured interview predischarge and at 6-month follow-up using a series of validated questionnaires, self-report and physiological measurement. Ethical approval for this trial was obtained from the local Committee on Medical Research Ethics. This trial was powered (98%) to be able to detect a difference of 1.5 units (SD = 3) on the HADS anxiety scale with a sample size of 260 participants at the 6-month evaluation. This difference was comparable to that shown by Lewin et al. (2002). Following screening for entry (see inclusion and exclusion criteria, Figure 1) written informed consent was obtained from 233 participants who had been admitted to either the acute medical admissions wards or Coronary Care Units of either of the two hospitals (see Figure 2). Baseline data were collected and participants were randomized to either:

Figure 1.

Inclusion, exclusion, and withdrawal criteria.

Figure 2.

Flowchart of participating patients.

  • 1Standard Care (SC): (N = 117) a minimal intervention by nurses during their admission which identified patient's risk factors, provided advice on their condition and risk factor reduction where possible depending on staff workload and skill mix.
  • 2Angina Plan (AP): (N = 116) participants received additional support in the form of a cognitive-behavioral nurse-facilitated self-help intervention over the following 12 weeks.

Random allocations were computer generated, allocated to permuted fixed blocks of 20 and stratified for site [19]. Following recruitment, the research officer in an off site university research department initiated the randomization. The research officer contacted the AP nurse who then informed participants of their group allocation. The researcher (SZ) was blinded to group allocation throughout the trial.

The Angina Plan Intervention

The AP is a systematically developed intervention based on theoretical principles of Cognitive Behavioral Therapy (CBT). The aim of CBT is to educate patients about the reciprocal relationship between thoughts, feelings and behaviors and to increase awareness of the automatic thoughts that occur in response to situations, events, and interactions. The aim of CBT is to allow the person to take control of his or her own problems and for him or her to manage his/her thoughts, feelings, and behaviors in such a way that future problems are dealt with in an adaptive way [20]. The AP involves specific and structured training for the nurses who provide the intervention and could potentially be replicated elsewhere. As it is short and does not require extensive resources it could potentially be incorporated to existing rehabilitation services. During a 45-minute in-hospital consultation the AP nurse completed an assessment and initiated the AP intervention, which was then facilitated over the next 12 weeks. The patient's cardiac misconceptions were identified using the brief questionnaire within the AP pack at the start of the consultation to allow the nurse to proactively target and correct these misconceptions. Individual cardiovascular risk was assessed and advice on risk factor modification given. Participants received the AP, which included a patient-held “work-book” and an audiotaped relaxation and information program. The workbook provided information on angina and its management, cardiovascular risk, relaxation, exercise and goal setting and pacing techniques. Over the following 12 weeks a method of “goal setting and pacing” based on the principles of CBT was used by the AP facilitator to introduce lifestyle changes and support recovery during telephone follow-up at weeks 1, 4, 8, and 12 for all participants in the AP group [21].


Data, including demographic and clinical information, cardiac risk factors, motivation to change, measures of knowledge and misconceptions, mood, cardiovascular symptoms, quality of life, health service utilization, and satisfaction with care, were collected during a structured interview by the researcher at baseline (predischarge) and 6-month postdischarge in the research clinic. For participants unable to attend the follow-up clinic, these data were collected in the participants’ home.

Mood was assessed using the Hospital Anxiety and Depression Scale (HADS) a widely used and well-validated 14-item tool with 2, seven item subscales to measure anxiety and depression within a nonpsychiatric population [17]. A score from 0 to 3 for each item generated a total score (range 0 to 21) for each subscale. Scores between 8 and 10 indicate borderline presence of anxiety or depression and those above 10 suggest that these states may be present. Anxiety was the primary outcome measure in this trial.

Knowledge and misconceptions were assessed using the 14-item York Angina Beliefs Questionnaire [22]. This uses a Likert scale response format ranging from “strongly agree” to “strongly disagree” and has been reported to be reliable and valid in this client group. Items targeted the cause, physiology and coping with angina. Summation and transformation of the item scores generated a scale total ranging from 0–56 with higher numbers indicating more misconceptions.

Cardiovascular symptoms were measured using 2 questionnaires. The first, Seattle Angina Questionnaire (SAQ) [23] is a disease-specific health related quality of life measure comprised of a 19 item questionnaire measuring five dimensions of coronary artery disease: physical limitation, angina stability, anginal frequency, treatment satisfaction and disease perception. Each dimension is scored separately on a 0–100 scale with higher scores indicating better functioning. The Cardiovascular Limitations and Symptoms Profile (CLASP) [24] measures nine physical and functional dimensions, including four symptom subscales (angina, shortness of breath, tiredness, ankle swelling) and five subscales focusing on functional limitations (mobility, social life and leisure activities, activities within the home, concerns and worries, sexual activity). Each of the nine subscales is scored separately to calculate a specific measure of impairment. It is widely used in clinical practice and research studies and its psychometric validation has been published [24,25].

Quality of life (QoL) was measured using two instruments. The Short Form-36 Health Survey (SF-36) is a 36-item questionnaire assessing general health and QoL, which has been used extensively with health and medical populations and has demonstrated good validity, reliability and sensitivity to change over time. The eight dimensions of the SF-36 (physical functioning, role limitations caused by physical problems, bodily pain, social functioning, mental health, role limitations caused by emotional problems, vitality-energy/fatigue and general health perception) generate scores on each dimension between 0 and 100, with higher scores representing better health status [26]. The second the Schedule for the Evaluation of Individual Quality of Life – Direct Weighting (SEIQoL – DW) is an interview-based tool specifically designed for the assessment of individual quality of life [27]. Using the SEIQoL – DW participants define five areas that comprise their individual ‘quality’ of life. These items are rated in terms of level of importance. An overall score ranging from 0–100 is then calculated with higher scores reflecting better quality of life. The SEIQoL-DW is totally subjective and patient-centered and provides a relatively unique measure of quality of life.

Individuals’ motivation to change behavior associated with cardiovascular risk was assessed using an adaptation of the stages of change model [28]. Individuals were categorized into active or nonactive stages reflecting their readiness to change.

Cardiovascular risk assessment included blood pressure (BP) measurement, following British Hypertension Society Guidelines [29], plasma cholesterol levels and body mass index (kg/m2). Current smoking, dietary and exercise habits were assessed by self-report questionnaires as part of the structured interview. Smoking behavior was verified using a calibrated Bedford EC50-piCO Smokerlyser breath CO monitor.

Health service utilization was assessed by recording the frequency and reasons for contact with primary and secondary care during the 6-month follow-up period. This was assessed by self-report and verified by screening GP and hospital case notes.

To control for order effects, questionnaires were administered sequentially to the participants in three different agreed and preallocated orders. HADS was administered first in each sequence as responses to the HADS may be influenced by preceding questionnaires [30]. SEIQoL-DW was the last in each sequence as it was the most time consuming measure to complete. Data collection took between 45 and 90 minutes per participant on average.

Data Analysis

Data were entered into the SPSS 11.5 (SPSS Inc., Chicago, IL), cleaned and screened for normality of distribution, presence of outliers and missing data. Characteristics of participants in each arm were compared descriptively. Two variables (HADS anxiety and depression) were transformed using square root transformation and one variable (CLASP concerns) was transformed using logarithmic transformation (log+1). Assessment of the effect of the intervention was conducted using two-way analysis of variance (ANOVA) using two fixed factors, that is, the group (SC or AP) and sequence of questionnaires (sequence 1, 2, and 3) for the YABQ, CLASP, SAQ, SF-36, and HEQ. Difference variables were created by subtracting follow-up values from baseline. HADS, SEIQoL-DW, body mass index (BMI), and BP were analysed using an independent samples t test because group was the only fixed factor. All tests were conducted using a 2-tailed significance level of P < 0.05. A nonparametric approach was adopted for variables (total cholesterol, HEQ-A, SAQ angina treatment satisfaction and anginal stability, SF-36 role emotional, role physical, social functioning and mental health and CLASP Home score) that were not normally distributed (see Table 3). The main intention-to-treat (ITT) analysis included all participants in the group to which they were randomly assigned, regardless of their adherence to the entry criteria, treatment assignment and receipt, withdrawal from the program or deviation from the protocol. The missing values strategy of replacing the mean of the other group for the patient group was adopted in order to minimize the type 1 error since the dropout rate was less than 20%[31]. Two sensitivity analyses were performed. The first excluded participants who during the 6 months time being on the trial, underwent either Coronary Artery Bypass Grafting (CABG) or Percutaneus Coronary Intervention (PCI). The second excluded participants who scored more than 11 on HADS depression at baseline (see Figure 3).

Table 4.  Distribution of participants in BMI categories at two time points
BMI categoriesStandard care N (%)Angina Plan N (%)
T1 BaselineT2 Follow-upT1 BaselineT2 Follow-up
Underweight1 (0.9)1 (0.9)1 (0.9)1 (0.9)
Normal29 (26.6)24 (22.0)25 (22.9)30 (27.5)
Preobese43 (39.4)44 (40.4)50 (45.9)43 (39.4)
Obese class I, II &III36 (33.1)40 (36.7)33 (30.3)35 (32.1)
Figure 3.

Data analysis process—sensitivity analyses.



Between August 2003 and October 2004, 566 eligible patients were admitted to the two hospital sites. Four hundred and sixteen patients were available and approached for participation in the trial. A total of 233 patients (56% of those available) agreed to participate in the trial (Figure 2), which ended in May 2005. During the first 4 months of the recruitment period, patients were given 24 hours or at least overnight to consider participation. However, this strategy resulted in approximate 60 eligible patients being missed due to being discharged within 24 hours (see Figure 2). The recruitment process was then changed with the approval of the Local Committee on Medical Research Ethics to allow participants a minimum of 4 hours to consider their participation in the study, which improved the number of patients recruited and reduced the number of patients being missed. However capturing more patients earlier in their hospital stay also increased the number of patients who refused to participate in the study.

The groups were similar at baseline (Table 1). Mean change scores were calculated for mood, knowledge and misconceptions, cardiovascular symptoms, QoL and cardiovascular risk. The results of the ITT analysis are presented in Table 2 (parametric testing) and Table 3 (nonparametric testing). The sequence x group interaction for SF-36 energy and vitality (F (2,217) = 3.24, P= 0.04) suggested that any significant results seen with this measure were influenced by the sequence of the questionnaire administration and should be interpreted with caution.

Table 1.  Characteristics of the groups at baseline
 SC n = 109AP n = 109
Mean age (SD)65.94 (9.96)64.8 (10.04)
n (%)n (%)
Number of males71 (65%)78 (72%)
 Females38 (35%)31 (28%)
Lives alone31 (28%)34 (31%)
Lives with partner78 (72%)75 (69%)
Deprivation category 1–2 (most affluent)30 (28%)26 (24%)
Deprivation category 3–560 (55%)64 (59%)
Deprivation category 6–7 (least affluent)19 (17%)19 (17%)
Presence of CHD and/or angina87 (80%)94 (86%)
Previous diagnosis of unstable angina51 (47%)57 (52%)
Previous myocardial infarction48 (44%)50 (46%)
Procedure performed in the past
 •PTCA21 (19%)25 (23%)
 •CABG20 (18%)22 (20%)
Presence of peripheral vascular disease9 (8%)6 (6%)
Previous cerebrovascular event11 (10%)12 (11%)
Previous attendance of cardiac rehabilitation33 (30%)29 (27%)
Diabetes24 (22%)18 (17%)
 Current25 (23%)21 (19%)
 Ex-smoker51 (47%)53 (49%)
 Nonsmoker33 (30%)35 (32%)
Family history of CHD78 (72%)67 (62%)
Hypertension diagnosis70 (64%)59 (54%)
Systolic blood pressure – mean (SD)127.17 (19.97)125.61 (18.14)
Diastolic blood pressure – mean (SD)68.98 (11.15)69.10 (10.41)
Total cholesterol
 Up to 4.8 mmol/L6359
 4.81 mmol/L and above4340
Body mass index mean (SD)28.52 (4.66)28.48 (4.80)
Units of alcohol/week mean (SD)8.12 (11.37)7.52 (9.35)
Table 2.  Main ITT analysis with mean baseline, follow-up and difference between T1 and T2 scores for the two groups of patients
 Time 1-baselineTime 2-follow-upt2–t1 mean (SD)Group test
SC Mean (SD)AP Mean (SD)SC Mean (SD)AP Mean (SD)SCAP
  1. * Transformed by square root.

  2. ** Transformed by log+1.

  3. *** SAQ exertional capacity scale measures physical limitations.

Misconceptions/knowledge28.45 (7.23)29.69 (8.02)26.43 (6.81)22.15 (7.38)–2.01 (6.39)–7.51 (7.76)F(1,216) = 33.88, P≤ 0.000
HADS anxiety*2.65 (0.76)2.51 (0.85)2.41 (0.95)2.16 (1.08)–0.24 (0.84)–0.35 (0.92)t(216) = 0.32, P= 0.32
HADS depression*2.07 (0.79)2.07 (0.84)2.15 (0.86)2.00 (0.93)0.79 (0.77)–0.07 (0.87)t(216) = 1.34, P= 0.18
BMI28.52 (4.66)28.48 (4.80)28.89 (4.77)28.35 (4.84)0.37 (1.24)–0.13 (1.39)t(216) = 2.83, P= 0.005
BP systolic127.17 (19.97)125.61 (18.14)129.81 (18.62)126.69 (18.95)2.65 (17.17)1.13 (16.06)t(214) = 0.67, P= 0.50
BP diastolic68.98 (11.15)69.10 (10.41)72.57 (11.21)71.16 (10.49)3.59 (10.88)2.06 (9.65)t(215) = 1.10, P= 0.27
CLASP angina9.44 (2.82)9.95 (2.81)8.22 (2.56)8.77 (2.85)–2.44 (3.23)–1.64 (2.87)F(1,217) = 1.24, P= 0.27
CLASP SOB8.70 (2.89)9.06 (2.77)8.51 (3.15)8.33 (2.90)–.85 (3.69)–1.23 (2.87)F(1,188) = 0.24, P= 0.62
CLASP ankle swelling5.61 (2.41)5.85 (2.34)5.17 (2.34)6.26 (2.22).07 (2.23).00 (2.44)F(1,38) = 0.08, P= 0.78
CLASP tiredness6.91 (1.59)7.06 (1.73)6.76 (1.68)6.57 (1.88)–.25 (1.61)–.59 (1.60)F(1,118) = 1.10, P= 0.29
CLASP mobility9.81 (3.29)10.10 (3.48)9.25 (3.35)9.07 (3.30)–.55 (2.79)–1.02 (3.04)F(1,217) = 1.59, P= 0.29
CLASP social/leisure4.72 (1.52)4.88 (1.55)4.67 (1.53)4.39 (1.52)–.05 (1.47)–.48 (1.62)F(1,217) = 4.21, P= 0.04
CLASP concerns**2.57 (0.54)2.61 (0.49)2.29 (0.44)2.24 (0.44)–0.08 (0.16)–0.11 (0.17)F(1,217) = 1.12, P= 0.29
CLASP sex6.32 (2.99)7.15 (3.53)6.81 (3.32)6.00 (3.15).24 (3.10)–.79 (2.60)F(1,52) = 0.97, P= 0.34
SAQ exertional capacity***56.75 (26.24)53.67 (26.92)59.39 (23.77)63.51 (26.16)2.35 (24.11)10.01 (23.03)F(1,212) = 5.15, P= 0.02
SAQ anginal frequency52.48 (25.21)45.78 (26.64)70.81 (28.35)70.32 (27.92)18.33 (29.11)24.54 (31.29)F(1,217) = 2.46, P= 0.12
SAQ disease perception46.79 (23.97)50.61 (24.79)66.21 (23.73)71.77 (23.93)19.43 (22.51)21.16 (28.20)F(1,217) = 0.33, P= 0.56
SF-36 physical function55.64 (29.08)54.27 (28.98)55.66 (28.13)57.96 (28.33).02 (23.22)3.69 (21.77)F(1,217) = 1.67, P= 0.19
SF-36 energy and vitality47.20 (23.24)44.63 (24.51)48.51 (22.93)50.45 (23.59)1.30 (21.34)5.82 (20.35)F(1,217) = 3.09, P= 0.08
SF-36 pain55.45 (28.35)49.54 (26.65)59.47 (28.58)61.43 (28.27).02 (31.15)11.89 (27.75)F(1,217) = 3.59, P= 0.06
SF-36 GH perception46.19 (24.02)46.64 (23.34)50.53 (23.45)53.01 (24.79)1.34 (20.10)6.37 (16.74)F(1,217) = 4.75, P= 0.03
SF-36 change in health39.22 (21.62)34.17 (19.74)49.21 (26.01)49.41 (24.76)9.99 (31.20)15.24 (27.19)F(1,217) = 2.17, P= 0.14
SEIQoL-DW QoL score68.24 (15.30)68.37 (16.80)73.36 (16.76)73.55 (15.19)4.83 (16.57)6.53 (15.02)t(88) =–0.51, P= 0.61
Table 3.  Main ITT for the two groups of participants – nonparametric tests
 T1-baseline medianT2-follow-up mediantest
Standard careAngina PlanStandard careAngina Plan
Risk factors
 Total cholesterol4.664.594.394.18U = 2704.5, z =–1.79, P= 0.073
 Alcohol intake per week2. = 4060.5, z =–1.12, P= 0.27
Quality of life
 SF-36 role physical25.0025.0062.5075.00U = 5430.0, z =–1.11, P= 0.27
 SF-36 role emotional100.00100.00100.00100.00U = 5620.5, z =–0.73, P= 0.47
 SF-36 soc. functioning66.6766.6777.78100.00U = 5806.0, z =–0.29, P= 0.77
 SF-36 mental health76.0072.0080.0084.00U = 5825.5, z =–0.25, P= 0.80
 CLASP home score6.006.005.845.00U = 5319.0, z =–1.36, P= 0.17
 SAQ treatment satisfaction93.7593.7587.5093.75U = 5624.0, z =–0.69, P= 0.49
 SAQ anginal stability25.0025.0050.0050.00U = 5703.0, z =–0.55, P= 0.60

There was no significant difference in between-group change scores for anxiety or depression (Table 2). There was a highly significant difference in change scores between groups in knowledge and misconceptions with the AP group demonstrating improved knowledge and fewer misconceptions (Table 2) than the control group.

Analysis of cardiovascular symptoms demonstrated that there was a significant difference in change scores between-groups in the social and leisure component of the CLASP [24]. Physical limitation due to angina measured by the Seattle Angina Exertional Capacity Scale [23] showed a significant between-group change score. A single significant difference was shown in change scores between groups in general health perception in the SF-36. No reliable differences were found in any other domains of the SF-36 nor the total QoL scores in the SEIQoL-DW.

In relation to cardiovascular risk factors, a significant difference in change scores between-groups was found for BMI. Categorical assessment of BMI showed more AP participants within the normal BMI category at follow-up in comparison with fewer SC participants (Table 4). Differences in patient movement between categories were significant between-groups with more participants in the AP group reducing their BMI and more in standard care increasing their BMI (U = 5327.0, z =–2.27, P= 0.023) (see Table 4). Participants’ motivation to make behavioral changes was compared examining shifts between active (maintenance and action) and nonactive (preparation, contemplation and precontemplation) stages. A significant difference in change scores between-groups was found for exercise, with more AP participants moving from the nonactive to active stage and more SC participants moving from active to nonactive stage over time (U = 5100.5, z =–2.28, P= 0.02). Dietary data although collected using an adaptation from the Healthy Eating Quiz (HEQ) [32] have not been presented due to the low reliability of the scale totals.

There were no significant between-group differences in hospital admissions as assessed by either screening of records (t(188) = 0.26, P= 0.79) or self-report (t(182) =–0.73, P= 0.47), contacts with primary care services using records (t(186) = 0.87, P= 0.39) or patient self-report (t(184) = 1.75, P= 0.0.82), nor in participants satisfaction with the care they received (U = 3897.0, z =–0.05, P= 0.96).

Sensitivity Analyses

During this trial, 65 participants underwent further intervention through CABG or PCI. A sensitivity analysis was conducted to control for the potential effects of these interventions on outcomes. Once scores from these participants had been removed, analysis on 77 SC participants and 76 AP participants confirmed the significant between-group differences on knowledge and misconceptions and BMI. In addition between-group difference in depression approached significance (MAP=–0.16, SD = 0.78; MSC= 0.08, SD = 0.69; t(151) = 1.97, P= 0.051). Although a significant result emerged in energy and vitality (SF-36) for the AP participants (MAP= 8.87, SD 19.82; MSC= 1.14 SD = 20.55; F(1,152) = 6.36, P= 0.013), this result is likely to be an artifact of the sequence by group effect reported previously. This sensitivity analysis also showed that the significant between-group differences seen in SAQ physical limitation, CLASP social and leisure activities, SF-36 general health perception disappeared. Thus suggests that some improvements reported in the ITT analysis may have been influenced by the positive effects of CABG/PCI on symptoms and recovery.

When reviewing participants with probable presence of clinical depression, that is, those with scores above 11, 13 participants (6%) were identified 6 in the AP group and 7 in the SC group. When reviewed individually, these participants had reported ongoing major life events, including bereavement, caring for a terminally ill relative and considering euthanasia, clinical depression on treatment and factors unrelated to cardiac disease and that may have affected their mood. After removing those with elevated depression at baseline, a significant between-group difference was seen in change scores in depression. Depression increased in the SC group (MSC= 0.77, SD = 3.02) and decreased in the AP group (MAP=–0.20, SD = 3.18), and the between-group changes scores were reliably different t(203) = 2.23, P= 0.027). These results suggest that depression levels in AP participants improved marginally whilst depression levels in SC participants deteriorated at 6 months although this effect was initially obscured by participants reporting high depression scores at baseline.


This pragmatic trial evaluated the effectiveness of the AP in an unselected population of patients admitted to secondary care with an acute exacerbation of angina. The results of this trial demonstrated that participants receiving the AP reported improved knowledge and fewer misconceptions, improvements in some cardiac risk factors (BMI, self-reported exercise) and improvements in physical limitation, general health perceptions and social and leisure activities. There were no reliable effects in the current trial on anxiety and depression in participants in the ITT analysis.

The lack of significant difference in between-group change scores for anxiety or depression in the current trial was disappointing (Table 2). The current trial showed no significant differences in between -group changes for either anxiety, the primary outcome measure, or depression despite the trial having sufficient power to detect these effects which was based on data in a similar population of CHD patients [33]. Previous research has suggested that depression is a strong and independent risk factor for mortality, recurrent cardiac events and lower functional status in patients with unstable angina [6] and myocardial infarction [5]. The prevalence of anxiety and depression in this trial was similar to those reported by other researchers [34]. One possible explanation for this may have been the reduction in anxiety reported in the standard care groups, a finding not seen in the Lewin et al. (2002) [9] trial. This improvement in mood in the standard care arm may be explained by improved follow-up and management of angina patients in primary care, which was not seen in the past.

No reliable differences were found in depression levels in the ITT analysis, yet Lewin et al. [9] previously reported that the most distressed participants showed a significant improvement in mood. In contrast those with elevated depression at baseline in the current study did not show improvement at follow-up. One possible explanation was that their depression may not have been related to their cardiac condition therefore this intervention had a limited effect in this group. After removing participants with elevated depression at baseline a significant between-group difference in change scores was seen in depression, supporting this argument. Further research may illustrate whether the additional identification and targeting of the underlying depression in those with elevated depression at baseline would allow such people to benefit from the AP intervention.

The impact of knowledge and misconceptions in cardiac patients on functional and psychological status and as a potential barrier to lifestyle changes have been increasingly investigated by researchers [35–38]. Adaptive change in angina beliefs over 1 year was the most significant predictive factor for physical functioning at 12-month follow-up [16]. Such belief based cognitive interventions can enable patients to participate actively in recovery and lifestyle change thus improving self-management of CHD [38]. National guidelines now recommend targeting misconceptions as part of CR and recovery programs for cardiac patients as this has the potential to further enhance recovery [7]. The current trial showed a highly significant improvement in knowledge and a reduction in misconceptions for AP participants in the ITT and sensitivity analyses, supporting previous studies showing improvements in knowledge following the delivery of a targeted intervention [35,36]. This suggests that the AP may be useful in clinical practice to allow the rapid identification of misconceptions and guide intervention to correct these beliefs, facilitating patient engagement in recovery and lifestyle modification.

Differences in cardiovascular symptoms were restricted to the social and leisure dimension of the CLASP and physical limitation as measured in the SAQ. A mean difference of 10.01 in the SAQ physical limitation between baseline and 6 months for the intervention group demonstrated clinically significant improvement with reduction in physical limitation in addition to the statistical difference seen in the between group change scores [23]. However these effects were lost after the sensitivity analysis of interventions was completed. This suggests the effect on symptoms and physical limitations may have strongly influenced by the effects of PCI or surgery which themselves produce improvement in symptoms. Alternatively the smaller sample used for this analysis may then have been underpowered to detect the ITT effects with CABG/PCI patients removed.

The importance of QoL assessment in CR and recovery programs is well recognized [39]. Significant improvements in QoL as measured by the SF-36 have been demonstrated in other studies of angina patients awaiting cardiac surgery, however the intervention [40] was delivered over a longer time period and involved face-to face contact with the patient which may have affected results. Both groups in the current trial showed improved scores on some domains of the SF36 as had been reported in other groups with Acute Coronary Syndrome [41]. General Health Perception was the only significant between group difference shown in the current trial with the participants in the intervention group reporting better general health perception.

Cardiovascular risk profiles of the participants in this trial were similar to those reported internationally, with similar baseline levels of obesity [42]. Previous work suggested that a 6-month period may be too short to detect modification in such risk factors [9]. This trial, however, demonstrated a comparative reduction over 6 months in BMI for the AP group, both as a continuous and categorical variable. This supports findings of a previous trial which demonstrated improvement in BMI following an intervention delivered on average for a 9-month period for angina patients on the waiting list for by pass surgery [40]. The reduction in BMI in the current trial may have been enhanced by the increase in self-reported activity in the AP participants. In addition the AP participants were more likely to be in the active stages of behavior change at follow-up suggesting that the use of the AP may be a useful strategy to support client engagement in positive behavioral change. Since obesity is becoming a major national public health challenge [42] this type of intervention may be important in helping patients with CHD to modify cardiovascular risk factors and prevent secondary events [43].

Although the previous trial on the AP did not report on health service use [9], this outcome was measured in the current trial by screening medical records and by patient self-report and revealed no difference in health service use between-groups. The current trial results differ from previous research evaluating a self-help, home-based program, based on the principles of CBT, on post-MI patients which demonstrated significantly less use of hospital services by the intervention groups [33]. One possible explanation for the lack of difference seen in the current trial may have been because the AP facilitator identified ongoing problems resulting in increased but appropriate contact with health services, a finding reported in previous studies.

During the completion of this trial, 65 participants underwent further intervention through CABG or PCI. The results of the sensitivity analysis showed even after removing these participants the significant between-group differences on knowledge and misconceptions, and BMI were sustained, suggesting that the effects on these factors are a result of the AP intervention. Interestingly a near significant reduction of depression was shown. The disappearance of the significant between-group differences seen in physical limitation, social and leisure activities, general health perception suggests that these improvements reported in the ITT analysis may have been influenced by the positive effects of CABG/PCI on symptoms and recovery and not resulting from the AP intervention alone.

There is a number of limitations to this trial. This trial was conducted in one UK, NHS Health Board Area and therefore may not be typical of other settings. However, as these sites included a large teaching hospital and a smaller district general setting, this may strengthen the applicability of these findings to a range of healthcare settings. The evaluation of this trial was confined to 6-month follow-up therefore we are unable to determine if any positive effects were sustained beyond this time. Longer-term evaluation would be useful. The SEIQoL-DW completion rate was lower than in other studies [44], which may have resulted from the number of questionnaires administered. Self-reported physical activity data within the trial may also have been strengthened by additional objective measurement. A total of 56% of eligible patients did not participate in the study as they were either missed or refused to participate (Figure 2). Approaching patients earlier for inclusion in the study resulted in less patients being missed due to early discharge. However, including more patients who were staying 24 hours or less increased the number of patients who refused to participate. From anecdotal information, a common reason for refusal to participate appeared to be the patients’ imminent discharge and desire to leave hospital rather than participate in the trial.

One limitation of the trial potentially arises from the use of multiple testing. The Bonferroni adjustment was not applied generally across this study. This conservative method reduces the risk of falsely rejecting a null hypothesis by dividing the significance level of 0.05 by N (the number of comparisons being undertaken). Antony [45] contends that if multiple tests are addressing different research questions, in this instance intervention effects on mood, BMI, and Health-Related Quality of Life, this adjustment is not required. However, within our analysis of between-group differences in Health Related Quality of Life scores using the CLASP, SAQ, and SF 36, multiple testing may remain an issue. It is therefore possible, but unlikely that some of the four significant results in this element may have resulted by chance.

In conclusion, although the main results of this trial did not detect any reliable effects on anxiety and depression in participants at 6-month follow-up, the results have shown that angina patients, with new and established CHD, admitted to an acute hospital setting who received the AP reported improved knowledge, fewer misconceptions, improvements in some cardiac risk factors (BMI, self-reported exercise) improvements in physical limitation, general health perceptions and social and leisure activities. Past research has indicated that such people often do not engage in CR programs, even when they are available and may not wish to attend a hospital-based program [46]. One benefit of the AP is that unlike other models of rehabilitation and support, the follow-up interventions may be delivered to the patients where it suits them either at home by telephone support, as in the current trial, or at their GP practice [9] and thus is not resource intensive.

These trial findings should be considered by clinicians involved in the redesign and delivery of CR services locally and nationally. The AP offers a short intervention (2.5 hours per patient) for the management of a currently undertreated population of cardiac patients and has shown benefits, which are sustained up to 6-month follow-up. Further research is needed to explore whether these benefits are sustained over a longer time period and whether such effects are comparable to those that might be obtained through other CR programs offered to angina patients. Further qualitative research is needed to explore participants’ experiences of the AP in more detail. This process evaluation would enhance the current findings and add to the understanding of patient experiences after admission to hospital with angina.


The authors would like to thank Connie Dunbar and Catherine Ryan, the Angina Plan facilitators for delivering the intervention and Dr. Simon Ogston for his statistical advice. The study was supported by a grant provided by the Chief Scientist Office [CZH/4/99]. Stella Zetta was in receipt of doctoral studentship provided by the School of Nursing and Midwifery, University of Dundee.

Conflict of Interest

The authors have declared no conflict of interests.