Figures
Abstract
Background
The main behaviour change intervention available for coronary heart disease (CHD) patients is cardiac rehabilitation. There is little recognition of what the active ingredients of behavioural interventions for CHD might be. Using a behaviour change technique (BCT) framework to code existing interventions may help to identify this. The objectives of this systematic review are to determine the effectiveness of CHD behaviour change interventions and how this may be explained by BCT content and structure.
Methods and Findings
A systematic search of Medline, EMBASE and PsycInfo electronic databases was conducted over a twelve year period (2003–2015) to identify studies which reported on behaviour change interventions for CHD patients. The content of the behaviour change interventions was coded using the Coventry Aberdeen and London—Refined (CALO-RE) taxonomy. Meta-regression analyses examined the BCT content as a predictor of mortality. Twenty two papers met the criteria for this review, reporting data on 16,766 participants. The most commonly included BCTs were providing information, and goal setting. There was a small but significant effect of the interventions on smoking (risk ratio (RR) = 0.89, 95% CI 0.81–0.97). The interventions did not reduce the risk of CHD events (RR = 0.86, 95% CI 0.68, 1.09), but significantly reduced the risk of mortality (RR = 0.82, 95% CI 0.69, 0.97). Sensitivity analyses did not find that any of the BCT variables predicted mortality and the number of BCTs included in an intervention was not associated with mortality (β = -0.02, 95% CI -0.06–0.03).
Conclusions
Behaviour change interventions for CHD patients appear to have a positive impact on a number of outcomes. Using an existing BCT taxonomy to code the interventions helped us to understand which were the most commonly used techniques, providing information and goal setting, but not the active components of these complex interventions.
Citation: Goodwin L, Ostuzzi G, Khan N, Hotopf MH, Moss-Morris R (2016) Can We Identify the Active Ingredients of Behaviour Change Interventions for Coronary Heart Disease Patients? A Systematic Review and Meta-Analysis. PLoS ONE 11(4): e0153271. https://doi.org/10.1371/journal.pone.0153271
Editor: Yoshihiro Fukumoto, Kurume University School of Medicine, JAPAN
Received: December 22, 2015; Accepted: March 25, 2016; Published: April 22, 2016
Copyright: © 2016 Goodwin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The minimal data underlying the study findings are presented in Table 2 and additionally in Figs 3 to 5. The authors extracted additional data to this which is available on request by emailing Dr Laura Goodwin (email: laura.goodwin@kcl.ac.uk).
Funding: The author(s) received no specific funding for this work. LG and MHH are part funded by the National Institute for Health Research (NIHR) Mental Health Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London.
Competing interests: The authors have declared that no competing interests exist
Introduction
Coronary heart disease (CHD) is the leading cause of death worldwide, for both communicable and non-communicable diseases [1]. In the UK, CHD remains a stable killer with approximately 74,000 deaths per year [2]. Within high-income countries there are more CHD deaths in areas of greater deprivation [3] and in individuals in manual occupational groups [4].
There is a widespread public health campaign promoting healthy behaviour, such as physical activity, which is based upon primary prevention literature about risk factors for CHD [5]. However, much of the secondary prevention guidance, such as the Quality and Outcomes Framework [6], focuses on blood pressure and cholesterol control, which can also be achieved through pharmacological intervention. This may be due to a mixed evidence base regarding individual health behaviours in secondary prevention; for example a systematic review of interventions involving reduced fat diets did not find a significant reduction in the risk of future CHD events or mortality [7].
For those CHD patients who want to try and change behaviours, the main lifestyle intervention available in Europe is cardiac rehabilitation, which primarily comprises education on CHD, disease and stress management, and physical activity classes. Such interventions are not based on theories of behaviour change, and UK evidence suggests that the benefits may be limited [8]. European data from EUROASPIRE-IV also suggests that a large proportion of CHD patients are not meeting clinical targets, for example 60% reported little or no exercise [9]. Furthermore, only half of patients were referred to a cardiac rehabilitation programme and not all of these actually attended [9]. Issues of non-attendance and adherence may be more pronounced in those groups who are at greatest risk, for example, individuals with lower socio economic status [10].
Effective behaviour change requires more than simply providing information on what changes need be made. Michie and colleagues have stressed the importance of behaviour change interventions being placed in the context of psychological theory and developed a taxonomy to classify such interventions, called the CALO-RE taxonomy [11]. Behaviour change taxonomies can be helpful both for those describing an intervention they are administering, and for classification of existing interventions for review, using a comprehensive coding system.
In addition to the known issues regarding definition of intervention content, addressed through frameworks such as the TIDieR checklist [12], there are also problems with evaluation. This most commonly relates to issues with the primary outcomes selected, which is then reflected in the selection of outcomes for inclusion in previous meta-analyses. If interventions are aiming to target health behaviours, then the meta-analyses should reflect this. Existing meta-analyses of heterogeneous pools of RCTS suggest that psychosocial interventions for CHD may be effective(e.g. [13]), finding a weighted relative risk of 0.82 for mortality. However what these reviews cannot tell us is what components of these interventions may be effective and the effect of these interventions on other outcomes such as health behaviours.
The objectives of this systematic review are to identify psychosocial or lifestyle behaviour change RCTs for CHD patients and to: i) determine the effect on health behaviours, intermediate outcomes of blood pressure and BMI, and CHD events and mortality and ii) to code the content of these interventions using the CALO-RE behaviour change taxonomy [11] and examine how the content and structure (length, format, theoretical basis) can predict effectiveness.
Methods
This review was reported in accordance with PRISMA guidelines (see S1 Table).
Search Strategy
The search for appropriate literature was conducted using the OvidSP search engine in February 2016, searching Medline, EMBASE and PsycInfo electronic databases. The search aimed to identify studies which reported on lifestyle or psychosocial behaviour change intervention programmes for patients with CHD. The search included studies over a twelve year period from 2003–2015 to capture more recent developments in behaviour change theory and intervention [14] and was restricted to articles that were published in the English language.
A search of keywords, abstracts, and titles was conducted for the following search terms: (Intervention or rehabilitation or modification or prevention or promotion or management or programme or feedback) combined using the AND command with (behaviour or lifestyle), AND (myocardial infarction or coronary heart disease or coronary artery disease or heart attack or cardiac arrest or coronary infarction or cardiac infarction) AND (post or secondary).
After having performed the search, titles and abstracts were downloaded into an electronic database and duplicates were excluded. Two review authors (LG and NK) independently assessed the eligibility of each study. Disagreements were discussed with a third author (GO).
Inclusion criteria
The following inclusion criteria were applied to the articles:
- Randomised controlled trial design, with no restriction on the length of follow-up.
- Participants of 18 years of age or above, with a primary diagnosis of CHD, which included patients with unstable angina, patients who had undergone percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG), patients with acute coronary syndrome (ACS) or those who had suffered a myocardial infarction (MI).
- At least 100 participants in order to exclude very small randomised trials or pilots.
- Interventions should be psychosocial or lifestyle behaviour change interventions.
- Must report on quantitative analyses.
- Must report data on health behaviour outcomes, i.e. smoking, physical activity etc., although this did not have to be the primary outcome.
- Peer reviewed research study.
- Article written in the English language.
Data extraction and analysis
Data was independently extracted by two researchers (NK and GO) and data extraction was conducted by both researchers for all articles. Inter-rater agreement was assessed by a third researcher (LG) who checked through the full extraction, and any queries were discussed between all three researchers. The data extracted from the articles included: authors, date, country, population description, sample size and characteristics, type of randomisation, intervention overview and content, length, number of sessions, who administered by, timing and number of follow-ups, description of control arm, theoretical basis and a description of the primary outcome. The outcome data which was extracted included: health behaviours (smoking, physical activity, diet, medication adherence), intermediate outcomes (body mass index, blood cholesterol, blood pressure), cardiovascular events, mortality and total drop-outs. Further details on the results were extracted from the papers which is available from the authors on request.
Authors were contacted to request additional information for the meta-analysis when it was not reported in the paper. These requests were generally for continuous outcome data at follow-up for intermediate outcomes, when either change scores or the proportion of the sample meeting a specified clinical criterion was reported.
Content of the intervention
The content of the behaviour change interventions was coded using the Coventry Aberdeen and London—Refined (CALO-RE) taxonomy, which includes 40 behaviour change techniques [11]. This taxonomy includes a description of each of the techniques, with specific examples provided. Although there are items which appear similar, the taxonomy has been developed to avoid overlap; for example, technique 5 is ‘goal setting (behaviour)’ which is distinguished from technique 6 ‘goal setting (outcome)’, which relates to measurable outcomes such as blood pressure or weight loss. The BCT content of each intervention was rated by 3 researchers (NK, GO & LG) and any discrepancies were discussed and the final rating was agreed.
Assessment of study quality
The methodological quality of the included studies was assessed with the “Cochrane Risk of Bias Tool” [15], which aims to evaluate the risk for the most relevant biases, assigning a judgement of “low”, “unclear” or “high risk” for six different domains. These criteria included assessment of the internal validity of the trial and the quality of reporting Two authors (GO, LG) independently assessed the quality of each study. Disagreements were discussed and, if necessary, a third author (MHH) was consulted.
Statistical analysis
Stata v11.0 was used for all data analyses [16] and meta analyses were conducted to produce weighted estimates and to examine the between study heterogeneity. For continuous outcomes, the weighted mean difference estimates, indicating the difference in means between the intervention and control group, were computed using the sample size, mean and standard deviation (SD) at follow-up for the intervention and control groups. For categorical outcomes, the risk ratio was computed based upon the number of cases (i.e. events) and the number of non-cases (i.e. non-events) during the study period for both the intervention and control groups. For smoking, the outcome data was reported differently between studies. For the purposes of this review, the proportion of smokers and non-smokers at follow-up was calculated for all studies. The I2 statistic was used to assess between study heterogeneity [17]. Random effects models were conducted as it was predicted that there would be between study heterogeneity resulting from differences in the populations and the interventions. The meta-analyses for both mortality and smoking were also stratified by i) study length, ii) whether they were individual or group interventions, and iii) if the intervention had a theoretical basis for the outcomes. Sensitivity analyses were conducted examining the BCT content of the interventions as a predictor of mortality in meta-regressions. The most commonly applied BCTs were analysed in the meta-regression using the following categories: provide information—BCTs 1, 2, 20 & 21; goal setting/action planning—BCTs 5, 6 & 7; review of goals/self-monitoring—BCTs 10, 11, 16 & 17; stress management—BCT 36; social support—BCT 29, and provide feedback—BCT 19.
Results
Study selection
The original search revealed 1400 articles, 874 of which remained after duplicates had been removed and 157 remained after irrelevant articles were removed (see Fig 1). After full text review, twenty two papers met the inclusion criteria and were included in this review [8, 18–37] which in total reported data on 16,766 participants.
Overview of the interventions
See Table 1 for full details of the intervention studies. The intervention length differed between studies; five interventions lasted less than 3 months [8, 23, 25, 29, 38], ten were 3–6 months in length [18, 21, 22, 24, 30, 33–37], four lasted for 12 months [19, 27, 31, 32], one was 18 months [28] and two lasted as long as 3 years [20, 26]. The sample sizes ranged from 120 for the smallest study [35] to 3241 participants in the largest [20].
The interventions were delivered most commonly by nurses (six of the studies)[8, 21, 23, 24, 27, 38], followed by counsellors or coaches in 4 studies [18, 22, 32, 36], by doctors in 3 studies [19, 25, 26, 37], two were delivered by a cardiac rehabilitation team [20, 30], three through automated text messages [33–35] and the three remaining studies were delivered by pharmacists [31], by both GPs and nurses [28], and finally by both psychologists and nurses [29]. Seventeen of the interventions were delivered on an individual basis, either in person [23–28, 31, 38] or over the telephone [18, 19, 21, 22, 32–36], and five were group interventions [8, 20, 29, 30, 37].
The control condition differed between studies, but most commonly comprised “usual care”, which in seven studies involved a cardiac rehabilitation programme [20, 23, 24, 29, 33–35]. Only nine of the seventeen studies reported that they had a theoretical basis [18, 21–23, 28, 29, 33, 34, 36]. The primary outcome differed between studies. Four studies reported that the primary outcome was a derived CHD risk score or algorithm [24, 29, 30, 32], with a further five reporting multiple primary outcomes [22, 23, 27, 37, 38]. Four studies defined the primary as achieving a range of clinical targets [19, 25, 28, 34].
Inclusion of BCTs
Fig 2 displays the frequency of inclusion of the different BCTs. The most commonly included techniques were providing information on the consequences of behaviour (BCTs 1 &2), providing instruction on how to perform the behaviour (BCT 21), goal setting in relation to the outcome (BCT 6) and the behaviour (BCT 5), providing information on where and when to perform the behaviour (BCT 20) and prompt review of behavioural goals (BCT 10). Review of outcome goals (BCT 11) was less commonly reported. The mean number of BCTs included in the interventions was 8.3 (SD 3.1), ranging from 4 to 16 BCTs.
Effectiveness of the interventions (see Table 2)
Health behaviours.
Physical activity: Although physical activity was assessed by 20 studies as an outcome, it was measured in a number of different ways and the data could not be combined statistically. It was, for example, assessed using a physical activity questionnaire, or as the proportion of the sample who met a specified criteria, e.g. exercising 5 times per week. Twelve of the 20 studies reported a statistically significant finding, indicating a positive impact of the intervention on physical activity.
Diet: Fifteen of the studies measured diet as an outcome, but the definitions were very different; from a Mediterranean diet score, to servings of fruit and vegetables, to dietary fat intake. Nine of the 15 studies reported a statistically significant improvement in diet in the intervention compared to the control group.
Smoking: Fifteen studies reported on prevalence of smoking at follow-up and all of these were included in the meta-analysis (see Fig 3). The weighted risk ratio across these studies indicated a small but significant effect size (RR = 0.89, 95% CI 0.81–0.97), indicating that overall the interventions were more likely to result in smoking cessation compared to the control arms. Random effects models indicated that there was low heterogeneity between studies (I2 = 23%).
Medication adherence: Nine studies reported on medication adherence to cardioprotective medications [19, 23–25, 27, 31, 33, 34, 38] with four of these assessing self-reported medication adherence [19, 33, 34, 38]. Three of these studies reported that the intervention had a positive impact on adherence [27, 34, 38].
Intermediate outcomes.
BMI: Data was available for 8 studies, for both the intervention and control arms, on mean BMI at follow-up. The weighted mean difference between the intervention and control groups at follow-up was -0.39 kg/m2 (95% CI -1.03–0.25) suggesting that there was not a significant difference between the groups for BMI. There was high heterogeneity across studies (I2 = 89%).
Blood pressure: Data was available for 10 studies for means at follow-up for systolic blood pressure (SBP) and for 9 studies for diastolic blood pressure (DBP). The weighted mean difference for SBP suggested a significant difference between the intervention and control groups of -3.13 mmHg (95% CI -5.11 - -1.15) with moderate heterogeneity across studies (I2 = 69%). There was a small significant difference between the groups for DBP with a weighted mean difference of -1.12 mmHg (95% CI -2.10 –-0.13) and moderate heterogeneity across studies (I2 = 53%).
CHD events and mortality.
Five studies reported on the risk of CHD events between baseline and the final follow-up after the intervention. Four of these reported on the number of myocardial infarctions or total non-fatal cardiac events [8, 20, 23, 26], whereas one study reported only on the combined number of coronary deaths and non-fatal myocardial infarction events [27]. The weighted risk ratio indicated that the interventions did not have a significant effect in reducing the risk of CHD events (RR = 0.86, 95% CI 0.68, 1.09) and there was moderate heterogeneity across studies (I2 = 55%) (see Fig 4).
The total number of deaths was reported in 15 studies, even though for many it was not reported as an outcome and was only referred to in relation to the flow of participants. The weighted effect across all 15 studies indicated a small significant effect size of the interventions in reducing risk (RR = 0.82, 95% CI 0.69, 0.97), with low heterogeneity across studies (I2 = 6%) (see Fig 5).
Predictors of effects
Intervention length.
The 15 studies which reported on mortality were stratified by length, suggesting that the longer interventions may have better outcomes; less than 3 months (RR = 0.97, 95% CI 0.75–1.26), 3–6 months (RR = 0.79, 95% CI 0.37–1.66), and 12 months or longer (RR = 0.76, 95% CI 0.62–0.92) although the difference in estimates was not statistically significant. The same pattern was not shown for the length of the intervention on risk of smoking; less than 3 months (RR = 1.06, 95% CI 0.87–1.30), 3–6 months (RR = 0.81, 95% CI 0.70–0.95), and 12 months or longer (RR = 0.92, 95% CI 0.80–1.04).
Individual or group delivered.
The interventions were stratified by whether they were delivered on an individual basis or in a group setting, with the results for mortality suggesting that the individual interventions (RR = 0.76, 95% CI 0.61–0.95) may be more effective than the group interventions (RR = 0.94, 95% CI 0.73–1.22) although this difference was not statistically significant. The same analyses with smoking as the outcome showed no difference between these groupings; individual (RR = 0.89, 95% CI 0.80–1.00) and group interventions (RR = 0.89, 95% CI 0.76–1.05).
Theoretical basis.
Studies were stratified on the basis of whether the intervention had been developed based upon psychological theory. The results for mortality suggested that the interventions without a theoretical basis (RR = 0.79, 95% CI 0.67–0.94) were more effective than those with a theoretical basis reported (RR = 1.06, 95% CI 0.65–1.75), but the difference between these groups was not statistically significant. The results for smoking did not suggest that there was a difference between these groups: no theoretical basis (RR = 0.95, 95% CI 0.85–1.05) and theoretical basis (RR = 0.83, 95% CI 0.70–0.99).
Inclusion of BCTs.
Sensitivity analyses were conducted examining the different categories of BCTs as predictors of mortality in meta-regressions, in addition to a model examining the number of BCTs. The analyses did not show that any of the variables predicted mortality: goal setting/action planning (β = -0.06, 95% CI -0.28–0.15), review of goals/self-monitoring (β = -0.07, 95% CI -0.25–0.11), stress management (β = 0.08, 95% CI -0.13–0.30), social support (β = -0.06, 95% CI -0.25–0.14) and providing feedback (β = -0.03, 95% CI -0.20–0.15). The number of BCTs included in an intervention was also not associated with mortality (β = -0.02, 95% CI -0.06–0.03).
Quality assessment
The risk of bias for every domain for each study is reported in Fig 6. The majority of studies had a “low risk” for the selection bias domain, which includes random sequence generation and allocation concealment. Blinding was generally not possible for both patients and those who administered the intervention. For this reason most of the studies were considered to have an “unclear risk”. Attrition bias was one of the main issues in terms of quality. For almost half of the included studies the risk of attrition bias was considered to be “high”, as incomplete outcome data were not adequately addressed. The high prevalence of “unclear risk” judgments reflects the lack of clear reporting rather than a clear evidence of bias. This is in line with the finding of a general sub-optimal reporting of RCTs despite the large diffusion of instruments designed to help transparent reporting, such as the CONSORT Statement [39].
Discussion
This review identified twenty two behaviour change intervention studies, from a range of countries, evidencing an overall positive impact of these interventions on smoking, systolic and diastolic blood pressure and mortality, but no effect on BMI or CHD events. There was limited evidence that longer interventions may be more effective in reducing mortality risk than those lasting for three months or less, and that interventions delivered on an individual basis (rather than as a group) may also be more likely to reduce the risk or mortality. Although this review aimed to identify which BCTs may be most effective in CHD patients, there was no association between individual BCTs with mortality, nor with the number of techniques used in an intervention.
This review was unique in extracting data on a range of outcomes. However, there was such heterogeneity in the measurement of health behaviours, such as diet and physical activity, that it was not possible to combine the results across trials in meta-analyses. The trials did seem to be effective in reducing the risk of smoking, albeit only a small decrease in risk, even though all but one were not designed specifically as smoking cessation programmes. This is in agreement with existing findings that smoking cessation programmes are effective in CHD patients [40]. Although secondary prevention guidelines for CHD tend to focus on pharmacological management of risk factors (e.g. UK Quality and Outcomes Framework; [41]), only nine studies reported on medication adherence as an outcome, and only three of these reported a positive outcome. For many participants, adhering to medication may be a more tangible change than modifying his/her diet, or increasing frequency of exercise, so this may have been a missed opportunity in trials which did not target adherence. Even though there were only small changes in the intermediate risk factors of blood pressure and BMI, there was still a positive effect of the trials on the risk of mortality. The effect size for mortality was also in line with previous reviews of psychological interventions for CHD (e.g. [13]).
The different studies varied in how the primary outcome was defined, most likely as a result of the range of outcomes available to select in this type of trial. Although a number of studies tackled this choice by using a combined outcome, such as the Framingham algorithm for CHD risk, an equal number stated multiple primary outcomes which the CONSORT guidelines advise against [42]. Other studies reported the primary outcome as meeting clinical guidelines for a range of risk factors based upon a score or algorithm [24, 29–32], but this was also open to misinterpretation if the primary outcome was stated to be met if a single guideline was achieved [19]. One of the implications of this review is that complex interventions targeting multiple outcomes may still have to select a single outcome (such as medication adherence) as the primary outcome, as long as the other constituent outcomes are reported in detail as secondary outcomes.
One of the aims of this review was to use an existing taxonomy to code behaviour change interventions, and to identify which techniques may be most helpful to this patient group. Whilst we were able to code all of the interventions and examine which were the most commonly used techniques, such as providing information and goal setting, we did not find evidence that it was these or other techniques which contributed to the overall effectiveness of the interventions. This may be due to the fact that most interventions included a range of techniques and there may be a synergistic effect of combining different techniques, as opposed to single techniques working alone. One example of this was that goal setting was included more commonly than review of outcome goals, even though these strategies would be expected to be implemented together. There has also been debate as to whether including more or less BCTs may have the most positive outcomes; an intervention with fewer BCTs may be more coherent and therefore easier to ensure intervention fidelity [43]. Coding of inclusion of BCTs was also only based upon the written information provided in the papers, which may not have been complete descriptions.
A limitation of the current behaviour change taxonomies is that they do not provide detail on specifically how the BCTs were applied (i.e. the clinical competence attached to the technique) and additionally how long each technique was used for, which could be the crucial element in determining efficacy. For example, on paper, goal setting appears to be a simple technique which was used commonly across the interventions, and that is also evaluated favourably by clinicians (e.g. [44]). However, it may still be open to misuse and could have been implemented quite differently, e.g. if the goals were set by clinicians rather than patients.
The interventions in this review were most commonly delivered by nurses. In some studies this was specific cardiac nurses, but some studies did not seem to prioritise this illness specific experience. The recent competence framework, specific to psychological interventions for people with physical health conditions, highlights 7 domains including generic therapeutic competencies for psychological interventions and condition specific interventions [45]. This framework suggests that whilst there are psychological techniques that can be applied across any patient group, a level of expertise is required that is specific to a condition. Further issues such as disease specific training and supervision are also highlighted in this framework [45], which were not summarized in many of the studies included in this review.
Strengths and limitations
The strengths of this review are that it involved a comprehensive data extraction of a range of outcomes for CHD interventions, developing previous reviews both in the detail of the extraction and in undertaking BCT coding. One of the potential limitations is that there may not have been sufficient statistical power to examine the individual BCTs in meta-regression as predictors of outcome; however, the size of the effects found did not indicate that the null result was a consequence of lack of power. It was also not possible for us to assess publication bias, and methods such as assessing asymmetry in funnel plots are not recommended when there are a small number of studies [46]. Many of the studies included were assessed as having a high risk of attrition bias, which could have affected the preciseness of the estimates in the meta-analyses. The risk of bias assessment did not suggest that there was likely to be reporting bias due to selective reporting of outcomes.
There was considerable variation between studies in the measurement and definition of the health behaviour and biological outcomes. The majority of the health behaviour outcomes were also self-reported and there is potential that social desirability effects may have been greater if participants knew they were in the intervention group. This review applied broad inclusion criteria and patients who had experienced different treatment were pooled together. However, this was consistent with a pragmatic approach aimed at emphasising the generalisability of these results to healthcare settings more widely.
Clinical and research implications
One of the implications from this review for clinical practice relates to whether healthcare commissioners should be investing in new psychological behaviour change interventions, or if existing cardiac rehabilitation programmes could be further developed both in the content and through tackling issues which have led to poor uptake and adherence. Although the West study showed no positive effect of CR in a multi-centre RCT, other studies have argued that this may have been due to lack of power to detect change due to the trial terminating early [47] and a further Cochrane review of CR did find a positive effect on mortality [48]. Many of the studies in this review compared their intervention to usual care which included CR and showed positive outcomes, but in the main these effects were small. A number of the more recent interventions used automated text messaging for delivery [19, 33–35]. These methods seemed to be as effective as the face-to-face programmes, but are more likely to be cost effective and easily implemented on a wider scale.
We have identified a number of generic difficulties in reporting and synthesising intervention studies in this area, which should be addressed in future research. 1) The content of the intervention and the method of delivery were not clearly reported. The BCT taxonomies are helpful in defining the intervention techniques, but this will be most useful when this information is embedded within the content of the intervention (e.g. goals were set in relation to gradual increases in activity versus smoking cessation) and reports on how the techniques were delivered (e.g. using guided discovery and collaborative decision making versus a more didactic approach). Making treatment manuals publically available will enable better comparison of studies. 2) There was high heterogeneity and lack of specificity in the definition of ‘treatment as usual’. This issue may to some extent be unavoidable due to different practices in cardiac rehabilitation across regions and countries, but it can be improved through applying greater rigour and standardisation in the coding of treatment received by the control group. In some studies it was stated that control participants could be referred to cardiac rehabilitation, but whether referral/attendance actually occurred was not always taken into account in the evaluation. The use of BCT coding for the control treatment may additionally be helpful. 3) There were reporting issues for the primary outcome measure. In this area of research multiple outcomes are relevant and one health behaviour or physiological outcomes cannot necessarily be prioritised over another. Whilst CHD events or mortality may be an obvious primary outcome, many pilot or early stage intervention studies will not be powered to detect these outcomes. Reporting multiple endpoints may be the most appropriate option, even though this is problematic statistically, it would not fit with current reporting frameworks and there is scope for reporting bias. Furthermore there are some behavioural outcomes, for example smoking, which are not undertaken by all participants in a trial so additional reporting guidelines are required on the most appropriate evaluation approach.
Conclusions
Although this review found evidence for a positive effect of these secondary prevention interventions on smoking, systolic blood pressure and mortality, the major challenge is in defining the content of the interventions and identifying what might be the active components. Using a BCT taxonomy helped us to understand which were the most commonly used techniques, providing information and goal setting, but it was not possible to identify what were the essential ingredients of these behaviour change interventions.
Author Contributions
Conceived and designed the experiments: LG MHH RMM. Analyzed the data: LG GO NK. Wrote the paper: LG. Contributed to the data extraction for the review: LG GO NK.
References
- 1.
World Health Organisation. Global status report on noncommunicable diseases 2010: Description of the global burden of NCDs, their risk factors and determinants. Switzerland: WHO, 2011.
- 2.
Townsend N, Wickramasinghe K, Bhatnagar P, Smolina K, Nichols M, Leal J, et al. Coronary heart disease statistics 2012 edition. London: British Heart Foundation, 2012.
- 3. Bajekal M, Scholes S, Love H, Hawkins N, O'Flaherty M, Raine R, et al. Analysing Recent Socioeconomic Trends in Coronary Heart Disease Mortality in England, 2000–2007: A Population Modelling Study. PLoS Med. 2012;9(6):e1001237. pmid:22719232
- 4. Loucks EB, Lynch JW, Pilote L, Fuhrer R, Almeida ND, Richard H, et al. Life-Course Socioeconomic Position and Incidence of Coronary Heart Disease: The Framingham Offspring Study. American Journal of Epidemiology. 2009;169(7):829–36. pmid:19179358
- 5. Li J, Siegrist J. Physical Activity and Risk of Cardiovascular Disease—A Meta-Analysis of Prospective Cohort Studies. International Journal of Environmental Research and Public Health. 2012;9(2):391–407. pmid:22470299
- 6.
Prescribing and Primary Care Services. Quality and Outcomes Framework—Prevalence, Achievements and Exceptions Report, England 2013–14. Leeds, England: Health and Social Care Information Centre, 2014.
- 7. Schwingshackl L, Hoffmann G. Dietary fatty acids in the secondary prevention of coronary heart disease: a systematic review, meta-analysis and meta-regression. BMJ Open. 2014;4(4).
- 8. West RR, Jones DA, Henderson AH. Rehabilitation after myocardial infarction trial (RAMIT): multi-centre randomised controlled trial of comprehensive cardiac rehabilitation in patients following acute myocardial infarction. Heart. 2012;98(8):637–44. pmid:22194152
- 9. Kotseva K, Wood D, De Bacquer D, De Backer G, Rydén L, Jennings C, et al. EUROASPIRE IV: A European Society of Cardiology survey on the lifestyle, risk factor and therapeutic management of coronary patients from 24 European countries. European journal of preventive cardiology. 2015.
- 10. Lane D, Carroll D, Ring C, Beevers DG, Lip GYH. Predictors of attendance at cardiac rehabilitation after myocardial infarction. Journal of Psychosomatic Research. 2001;51(3):497–501.
- 11. Michie S, Ashford S, Sniehotta FF, Dombrowski SU, Bishop A, French DP. A refined taxonomy of behaviour change techniques to help people change their physical activity and healthy eating behaviours: The CALO-RE taxonomy. Psychology & Health. 2011;26(11):1479–98.
- 12. Hoffmann TC, Glasziou PP, Boutron I, Milne R, Perera R, Moher D, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ (Clinical research ed). 2014;348:g1687. Epub 2014/03/13. pmid:24609605.
- 13. de Waure C, Lauret GJ, Ricciardi W, Ferket B, Teijink J, Spronk S, et al. Lifestyle interventions in patients with coronary heart disease: a systematic review. American journal of preventive medicine. 2013;45(2):207–16. Epub 2013/07/23.
- 14. Michie S, Fixsen D, Grimshaw J, Eccles M. Specifying and reporting complex behaviour change interventions: the need for a scientific method. Implementation Science. 2009;4(1):40.
- 15.
Higgins J, Altman D, Sterne J. In: Higgins J, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions Version 510 (updated March 2011): The Cochrane Collaboration; 2011.
- 16.
StataCorp. Stata Statistical Software: Release 11. College Station, TX: Stata Corporation; 2009.
- 17. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ (Clinical research ed). 2003;327(7414):557–60.
- 18. Berndt N, Bolman C, Froelicher E, Mudde A, Candel M, de Vries H, et al. Effectiveness of a telephone delivered and a face-to-face delivered counseling intervention for smoking cessation in patients with coronary heart disease: a 6-month follow-up. J Behav Med. 2014;37(4):709–24. pmid:23760610
- 19. Blasco A, Carmona M, Fernandez-Lozano I, Salvador CH, Pascual M, Sagredo PG, et al. Evaluation of a telemedicine service for the secondary prevention of coronary artery disease. Journal of cardiopulmonary rehabilitation and prevention. 2012;32(1):25–31. Epub 2011/11/25. pmid:22113368.
- 20. Giannuzzi P, Temporelli P, Marchioli R, Maggioni AP, Balestroni G, Ceci V, et al. Global secondary prevention strategies to limit event recurrence after myocardial infarction: Results of the gospel study, a multicenter, randomized controlled trial from the italian cardiac rehabilitation network. Archives of Internal Medicine. 2008;168(20):2194–204. pmid:19001195
- 21. Hanssen TA, Nordrehaug JE, Eide GE, Hanestad BR. Improving outcomes after myocardial infarction: a randomized controlled trial evaluating effects of a telephone follow-up intervention. European Journal of Cardiovascular Prevention & Rehabilitation. 2007;14(3):429–37.
- 22. Hawkes A, Patrao T, Atherton J, Ware R, Taylor C, O’Neil A, et al. Effect of a Telephone-Delivered Coronary Heart Disease Secondary Prevention Program (ProActive Heart) on Quality of Life and Health Behaviours: Primary Outcomes of a Randomised Controlled Trial. IntJ Behav Med. 2013;20(3):413–24.
- 23. Jolly K, Taylor R, Lip G, Greenfield S, Raftery J. The Birmingham Rehabilitation Uptake Maximisation Study (BRUM). Home-based compared with hospital-based cardiac rehabilitation in a multi-ethnic population: cost-effectiveness and patient adherence. Health Technology Assessment. 2007;11(35):118.
- 24. Jorstad HT, von Birgelen C, Alings AMW, Liem A, van Dantzig JM, Jaarsma W, et al. Effect of a nurse-coordinated prevention programme on cardiovascular risk after an acute coronary syndrome: main results of the RESPONSE randomised trial. Heart. 2013;99(19):1421–30. pmid:23813851
- 25. Muniz J, Gomez-Doblas J, Santiago-Perez M, Lekuona-Goya I, Murga-Eizagaetxebarria N, de Teresa-Galvan SZ E, et al. The effect of post-discharge educational intervention on patients in achieving objectives in modifiable risk factors six months after discharge following an episode of acute coronary syndrome, (CAM-2 Project): a randomized controlled trial. Health and Quality of Life Outcomes. 2010;8(1):137.
- 26. Munoz M-A, Vila J, Cabañero M, Rebato C, Subirana I, Sala J, et al. Efficacy of an intensive prevention program in coronary patients in primary care, a randomised clinical trial. International Journal of Cardiology. 2007;118(3):312–20. pmid:17261336
- 27. Murchie P, Campbell NC, Ritchie LD, Simpson JA, Thain J. Secondary prevention clinics for coronary heart disease: four year follow up of a randomised controlled trial in primary care. BMJ: British Medical Journal. 2003;326(7380):84-. PMC139939. pmid:12521974
- 28. Murphy AW, Cupples ME, Smith SM, Byrne M, Byrne MC, Newell J, et al. Effect of tailored practice and patient care plans on secondary prevention of heart disease in general practice: cluster randomised controlled trial. BMJ: British Medical Journal. 2009;339:b4220. PMC2770592. pmid:19875426
- 29. Murphy BM, Worcester MUC, Higgins RO, Elliott PC, Le Grande MR, Mitchell F, et al. Reduction in 2-Year Recurrent Risk Score and Improved Behavioral Outcomes After Participation in the “Beating Heart Problems” Self-management Program: RESULTS OF A RANDOMIZED CONTROLLED TRIAL. Journal of cardiopulmonary rehabilitation and prevention. 2013;33(4):220–8. 01273116-201307000-00004. pmid:23595004
- 30. Otterstad JE. Influence on lifestyle measures and five-year coronary risk by a comprehensive lifestyle intervention programme in patients with coronary heart disease. European Journal of Cardiovascular Prevention & Rehabilitation. 2003;10(6):429–37.
- 31. Bond C, Community Pharmacy Medicines Management Project Evaluation Team. The MEDMAN study: a randomized controlled trial of community pharmacy-led medicines management for patients with coronary heart disease. Family Practice. 2007;24(2):189–200. pmid:17272285
- 32. Wister A, Loewen N, Kennedy-Symonds H, McGowan B, McCoy B, Singer J. One-year follow-up of a therapeutic lifestyle intervention targeting cardiovascular disease risk. CMAJ: Canadian Medical Association Journal. 2007;177(8):859–65. PMC1995136. pmid:17923653
- 33. Chow CK, Redfern J, Hillis GS, Thakkar J, Santo K, Hackett M, et al. Effect of lifestyle-focused text messaging on risk factor modification in patients with coronary heart disease: A randomized clinical trial. JAMA. 2015;314(12):1255–63. pmid:26393848
- 34. Dale LP, Whittaker R, Jiang Y, Stewart R, Rolleston A, Maddison R. Text Message and Internet Support for Coronary Heart Disease Self-Management: Results From the Text4Heart Randomized Controlled Trial. Journal of medical Internet research. 2015;17(10):e237. pmid:26490012
- 35. Varnfield M, Karunanithi M, Lee C-K, Honeyman E, Arnold D, Ding H, et al. Smartphone-based home care model improved use of cardiac rehabilitation in postmyocardial infarction patients: results from a randomised controlled trial. Heart. 2014;100(22):1770–9. pmid:24973083
- 36. Yan J, You L-m, Liu B-l, Jin S-y, Zhou J-j, Lin C-x, et al. The effect of a telephone follow-up intervention on illness perception and lifestyle after myocardial infarction in China: A randomized controlled trial. International Journal of Nursing Studies. 2014;51(6):844–55. pmid:24211192
- 37. Melamed RJ, Tillmann A, Kufleitner H-E, Thürmer U, Dürsch M. Evaluating the Efficacy of an Education and Treatment Program for Patients with Coronary Heart Disease: A Randomized Controlled Trial. Deutsches Ärzteblatt International. 2014;111(47):802–8.
- 38. Zhao Y, Wong FKY. Effects of a postdischarge transitional care programme for patients with coronary heart disease in China: a randomised controlled trial. Journal of Clinical Nursing. 2009;18(17):2444–55. pmid:19619203
- 39. Turner L, Shamseer L, Altman DG, Weeks L, Peters J, Kober T, et al. Consolidated standards of reporting trials (CONSORT) and the completeness of reporting of randomised controlled trials (RCTs) published in medical journals. The Cochrane database of systematic reviews. 2012;11:Mr000030. Epub 2012/11/16. pmid:23152285.
- 40. Barth J, Jacob T, Daha I, Critchley JA. Psychosocial interventions for smoking cessation in patients with coronary heart disease. The Cochrane database of systematic reviews. 2015;7:Cd006886. Epub 2015/07/07. pmid:26148115.
- 41. Doran T, Kontopantelis E, Valderas JM, Campbell S, Roland M, Salisbury C, et al. Effect of financial incentives on incentivised and non-incentivised clinical activities: longitudinal analysis of data from the UK Quality and Outcomes Framework. British Medical Journal. 2011;342.
- 42.
Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ, et al. CONSORT 2010 Explanation and Elaboration: updated guidelines for reporting parallel group randomised trials [Journal Article]. 2010 [updated 2010-03-24 00:06:15; cited 340]. Available: http://www.bmj.com/bmj/340/bmj.c869.full.pdf.
- 43. Michie S, Jochelson K, Markham WA, Bridle C. Low-income groups and behaviour change interventions: a review of intervention content, effectiveness and theoretical frameworks. Journal of Epidemiology and Community Health. 2009;63(8):610–22. pmid:19386612
- 44. MacGregor K, Handley M, Wong S, Sharifi C, Gjeltema K, Schillinger D, et al. Behavior-Change Action Plans in Primary Care: A Feasibility Study of Clinicians. The Journal of the American Board of Family Medicine. 2006;19(3):215–23. pmid:16672674
- 45.
Roth AD, Pilling S. A competence framework for psychological interventions with people with persistent physical health conditions. Available: http://www.ucl.ac.uk/CORE/2015.
- 46. Sterne JAC, Sutton AJ, Ioannidis JPA, Terrin N, Jones DR, Lau J, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ (Clinical research ed). 2011;343.
- 47. Doherty P, Lewin R. The RAMIT trial, a pragmatic RCT of cardiac rehabilitation versus usual care: what does it tell us? Heart. 2012;98(8):605–6. pmid:22505460
- 48. Heran BS, Chen JM, Ebrahim S, Moxham T, Oldridge N, Rees K, et al. Exercise-based cardiac rehabilitation for coronary heart disease. The Cochrane database of systematic reviews. 2011;(7):Cd001800. Epub 2011/07/08. pmid:21735386; PubMed Central PMCID: PMCPmc42299