The European Heart Failure Self-care Behaviour scale revised into a nine-item scale (EHFScB-9): a reliable and valid international instrument
1 Department of Cardiology, Thoraxcenter, University Medical Centre, University of Groningen, PO Box 30.001 9700 RB, Groningen, the Netherlands
2 School of Human Sciences, University of Kalmar, Kalmar, Sweden
3 Department of Nursing Science, School of Nursing Sciences, Jönköping University and Unit of Research and Development in Primary Care, Jönköping, Sweden
4 School of Nursing, University of California, San Francisco, CA, USA
5 Department of Cardiology, Linköping University Hospital and Department of Medicine and Care, Faculty of Health Sciences, Linköping University, Sweden
* Corresponding author. Email: t.jaarsma{at}thorax.umcg.nl
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingReferences |
|---|
Aims: Improved self-care is the goal of many heart failure (HF) management programmes. The 12-item European Heart Failure Self-Care Behaviour Scale (EHFScB scale) was developed and tested to measure patient self-care behaviours. It is now available in 14 languages. The aim of this study was to further determine reliability and validity of the EHFScB scale.
Methods and results: Data from 2592 HF patients (mean age 73 years, 63% male) from six countries were analysed. Internal consistency was determined by Cronbach’s alpha. Validity was established by (1) interviews with HF experts and with HF patients; (2) item analysis; (3) confirmatory factor analysis; and (4) analysing the relationship between the EHFScB scale and scales measuring quality of life and adherence. Internal consistency of the 12-item scale was 0.77 (0.71–0.85). After factor analyses and critical evaluation of both psychometric properties and content of separate items, a nine-item version was further evaluated. The reliability estimates for the total nine-item scale (EHFScB-9) was satisfactory (0.80) and Cronbach’s alpha varied between 0.68 and 0.87 in the different countries. One reliable subscale was defined (consulting behaviour) with a Cronbach’s alpha of 0.85. The EHFScB-9 measures a different construct than quality of life (r = 0.18) and adherence (r = 0.37).
Conclusion: The 12-item EHFScB scale was revised into the nine-item EHFScB-9, which can be used as an internally consistent and valid instrument to measure HF-related self-care behaviour.
Key Words: Heart failure • Self-care • Instrument development
Received March 4, 2008; Revised September 1, 2008; Accepted September 1, 2008
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Patient education and support for patients with heart failure (HF) to improve self-care behaviour is the goal of many HF management programmes. The effectiveness of these programmes is often evaluated based on their effects on readmission, costs, and quality of life.1,2 Since there is an increasing call for a tailored approach in HF management, it is important to evaluate programmes on outcomes such as symptom relief and self-care behaviour.
HF-related self-care behaviour reflects the actions that an HF patient undertakes to maintain life, healthy functioning, and well being. This definition includes behaviours like adherence to medication, diet and exercise, as well as self-management of symptoms, but it also refers to behaviours such as daily weighing to assess fluid retention and seeking assistance when symptoms occur.3,4 To evaluate the effectiveness of interventions aimed at improving self-care, it is important to know if and how patients changed their self-care behaviour as a result of such interventions. Identification of deficits in HF-specific behaviours can help health-care professionals improve patient education or support behavioural change.5,6
To evaluate the effectiveness of interventions aimed at improving self-care behaviours of HF patients, a valid, reliable, and user-friendly scale is needed. For reasons of comparison, it is important that such a scale can be used in different countries.
Two HF-specific scales are currently available to measure self-care behaviours specially related to HF. The Self-Care of Heart Failure Index (SCHFI)4 is a self-report scale comprising 15 items rated on a four-point response scale and divided into three subscales measuring self-care maintenance, self-care management, and self-care confidence.4
The European Heart Failure Self-Care Behaviour Scale (EHFScB scale), comprising 12 items rated on a 5-point scale between 1 (I completely agree) and 5 (I completely disagree), was published in 2003.3 The EHFScB scale is available in 14 languages (British and American English, Dutch, Swedish, Italian, Spanish, Catalan, German, Finnish, Danish, Hebrew, Lithuanian, Chinese, and Japanese). The scale is considered easy to administer and practical to use. The scale has been found to measure change in behaviour over time. Gonzalez et al.7 used the Spanish version to measure the effects of an HF clinic and described a significant correlation between EHFScB score and the time of follow-up in the unit, where patients were followed every 3 months by an HF nurse who provided specific information about HF. The highest scores, indicating worse self-care behaviours, were obtained in patients evaluated at the first visit and the lowest scores were obtained in patients evaluated at 15 months after their first visit. The investigators also showed on an item-level that most items changed significantly over time.
Pulignano et al.8 also found, using the Italian version, that when patients with and without HF education were compared, HFScB was significantly better in those that had received education.
The British English version of the scale was recently evaluated and demonstrated a good test–retest reliability. A somewhat lower internal consistency than in other European populations was found and further exploration was advised.9 Although the scale is used in different HF clinics and research studies evaluating effectiveness of HF interventions, little data are published on the reliability and validity of the EHFScB scale in different European countries.
The aim of this study was therefore to further determine reliability and validity of the EHFScB scale in eight different patient samples from six countries.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Sample and setting
The EHFScB was tested in a pooled convenience sample of patients with HF from two sites in Sweden (n = 69 and n = 92), two sites in the Netherlands (n = 249 and n = 994), and from single sites in the United Kingdom (n = 177), Italy (n = 173), Germany (n = 285), and Spain (n = 553). Data from the other translated versions were not available to the researchers. Data were collected for research purposes in six sites and clinical purpose in two sites. For this analysis no repeated measurement data were used. In four sites the scales were completed during patients’ hospitalization and in the other sites in outpatient clinics.
Procedure
In this study the relevant translated versions of the European Heart Failure Self-care Behaviour Scale were used. The full version and data on the first testing has been previously published.3 All 14 official versions of the scale have been translated and back translated by bilingual persons knowledgeable about self care in HF and the intentions and basis of the scale. Some versions have been previously tested.7–9 The scale was administered by a nurse and either completed by the patient using the paper and pencil form or by interview. All patients gave informed consent to participate in the study. The study conforms to the principles outlined in the declaration of Helsinki. Ethics approval was obtained at each site according to the local regulations.
Data on the scale were submitted to the first author as excel or SPSS files. Thereafter all data were merged into a new SPSS file and translated to SPSS version 15.0.1. Out of 2592 individuals, there were missing data from 94, with no large differences between the different data sets. In 57 of these scales one item was missing; in 11 scales there were two items missing; and in three scales three items were missing. In these scales the missing items were replaced by the number 3 as decided by the constructors and described in a previous publication.3 The other 23 scales with more than three items missing were omitted from the analysis.
Content validity
Content validity can be obtained from the literature, representatives of the relevant population, and content experts. In the initial development of the scale, the literature was analysed and experts consulted.3 In the current study, additional information on content validity was collected from additional HF experts who used the scale, by a short five-item survey, by review of guidelines on HF treatment10,11 and from HF patients. Ten HF experts were consulted and open-ended questions were asked concerning their opinion on the completeness of the scale, the user friendliness, and on the items they experienced problems with. A subgroup of 16 patients selected from the Swedish HF patients was interviewed about their self-care behaviours. The sampling of these patients was strategic and the individuals varied with regard to age, gender, educational level, HF severity (NYHA class), and aetiology. The patients were asked to describe what they defined as self-care behaviours when suffering from HF and what self-care behaviours they conducted. The interviews were transcribed verbatim and analysed using content analysis.12 Initially the whole context was reread several times in order to find HF-related self-care behaviours. Meaningful units were taken from the transcripts, condensed, and labelled with a code. The codes were sorted into areas and compared with the items of the EHFScB scale.
Construct validity
A previous exploratory factor analysis of the EHFScB scale only partly reproduced the three assumed theoretical dimensions (asking for help, adapting activities, and complying with the regimen). However, this factor analysis was made on an earlier version of the EHFScB and major revisions of the scale have been made since then, e.g. item reduction and a revised scoring format.3 To establish construct validity of the scale, there was reason to evaluate the dimensionality further.
Construct validity of the EHFScB scale was determined using item analysis,13 confirmatory factor analysis (CFA)14 and convergent and discriminant validity.15 The item analysis was performed by calculating the distribution of the responses. The item total correlation of the items and Cronbach’s alpha if the item was deleted. A series of CFAs were performed on the pooled sample to evaluate the assumed theoretical dimensions behind the EHFScB scale and to explore improvement of the instrument. In the first step, the original 12-item version of the EHFScB scale was evaluated with a one factor model for the total scale and a three factor model for the assumed subscales. Results from these analyses together with the initial item analysis were used to improve the scale by item reduction. This revised version of the scale was evaluated in a second step, with a one factor model for the total scale and a two factor model for the revised subscales. An asymptotically distribution-free (ADF) estimation method was used for all CFAs as the assumption of multivariate normality was not fulfilled. The goodness-of-fit of the CFA models was evaluated with 
2 goodness-of-fit, root mean square error of approximation (RMSEA),16 comparative fit index (CFI),17 normed fit index (NFI),18 goodness-of-fit index (GFI), and adjusted goodness-of-fit index (AGFI).19 An acceptable model fit in the present study was defined as 
0.08 for RMSEA and 
0.90 for GFI, AGFI, NFI, and CFI. The differences between the rival models were also analysed with the 2 difference test.20
Convergent and discriminant validity was evaluated by analysing the relationship (Pearson’s correlation coefficient) between the EHFScB total scale with the revised HF Compliance Questionnaire21 and the Minnesota Living with Heart Failure Questionnaire (MLwHFQ).22 A moderate to strong correlation to the revised HF Compliance Questionnaire was expected to support convergent validity as they capture similar concepts while a weak correlation to the MLwHFQ was expected to support discriminant validity as they measure different concepts.
Both parametric and non-parametric statistics were used.23 Cronbach’s alpha was used to estimate internal consistency reliability. The level for statistical significance was set at P < 0.05. All statistical analyses were conducted using the SPSS 15.0.1 and Amos 7.0 software.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Participants
Pooled sample
The participants consisted of 2592 patients with HF from eight sites in six countries. Mean age of patients was 71 (SD 12) years ranging in the eight data sets from 64 years (German data set) to 79 years (Swedish data sets). Both male and female patients were included in all data sets, ranging from a Swedish data set with 51% males to the German data set with 75% males, with 64% males in the pooled sample of 2592 patients. Patients from all NYHA classifications were included in the total pooled data set with a mean NYHA classification of 2.6 ranging from a mean of 2.3–3.3 in the separate data sets. The mean LVEF was 34% (SD 14). This was comparable in the six of the eight data sets, with the two Swedish data sets not having data on LVEF available.
Swedish subgroup
The subgroup of 16 Swedish HF patients interviewed on their self-care behaviours were equally distributed according to gender and had a mean age of 75 ± 9 (range 66–95) years. Time since the diagnosis of HF varied between 6 months and 10 years (mean 29 months). The majority of the individuals (n = 12) had an elementary school education; four had college or university education. In 12 individuals, HF was caused by ischaemic heart disease; two had dilated cardiomyopathy and two had heart-valve disease as the aetiology of HF.
Content validity
Patients
Eleven of the 12 items on the EHFScB scale were spontaneously mentioned by HF patients in the interviews as relevant self-care behaviours. The only exception was ‘getting a flu shot every year’, which was not mentioned by any patient as an HF-related self-care behaviour. Some new self-care behaviours that were not on the scale were described as relevant for HF by the patients. These behaviours were all related to general and not HF-specific self care such as maintaining a healthy diet, losing weight, having a moderate intake of alcohol, smoking cessation, avoiding emotional distress by trying to stay positive or talking to someone when worried or depressed and being able to sleep well.
Researchers
Researchers who used the scale commented on the content of the scale by completion of a short survey. Items they identified as important but missing related to smoking and alcohol use. Others commented on the possibility that national or local guidelines might not be in line with the items on fluid restriction and diet.
Guidelines
In reviewing international guidelines,10,11 all recommendations were reflected in the items except for one. The scale does not include an item on ‘drugs to avoid’ (e.g NSAIDs).
Construct validity
Item analysis
The initial item analysis of the 12-item version of the EHFScB total scale on the pooled sample demonstrated that most items had a positively skewed distribution. The exception was items 1 and 12 which were uniformly distributed. The item-total correlation varied between 0.14 and 0.65. Five items demonstrated satisfactory item-total correlation 0.4 (items 3, 4, 5, 6, 8). Three scored between 0.3 and 0.4. The rest of the items demonstrated correlations <0.3 with the item about flu shots having the lowest correlation (Table 1). The frequency of missing data varied between 0.2 and 1.2% across items. The items in the different samples followed a similar response pattern with some exceptions. Patients in Germany and Spain scored overall lower median values on the items compared with the other countries.
|
Confirmatory Factor Analysis
The CFA on the original 12-item version demonstrated ambiguous results concerning the fit between the model and the data for both the one-factor model and the three-factor model (Table 2). While acceptable model fit was demonstrated with RMSEA (0.06), GFI (0.94), and AGFI (0.91), poor fit was demonstrated with NFI (0.62) and CFI (0.64). Another concern was that the factor concerning ‘adapting activities’ correlated weakly with the two other factors, ‘asking for help’ (0.37) and ‘complying with the regimen’ (0.20) in the three-factor model. As the ‘adapting activities’ factor contained two items (items 2 and 7) which previously demonstrated weak item-total correlations, this factor and its items were removed for further analyses. Item 11 was also removed as it demonstrated weak factor loadings to the total scale (0.05) and to the ‘complying with the regimen’ factor (0.09), as well as a low item-total correlation (0.14) in the item analysis. Although, item 12 demonstrated a weak item-total correlation (0.27), it was not removed as it demonstrated factor loadings >0.3 across the factor analysis.
|
The CFA on the revised nine-item version of the EHFScB scale with one factor and three items removed (items 2, 7, and 11), demonstrated a significant improvement over the 12-item version, although the ambiguous results concerning model fit still remained (Table 2). The one-factor model based on the nine-item version demonstrated a significant improvement compared with the one-factor model based on the 12-item version [
2difference(27)=287.7, P < 0.001]. Also the revised two-factor model demonstrated a significant improvement over the original three factor model [2difference(25)=278.6, P < 0.001]. The goodness-of-fit statistics demonstrated an improvement as well. Although RMSEA did not demonstrate any differences in model fit across the CFA, the revised nine-item version demonstrated an improvement over the 12-item version based on GFI, AGFI, NFI, and CFI. Even if NFI and CFI increased, the scale could probably be further improved as both of these goodness-of-fit indices did not exceed the level for an acceptable model fit (>0.9).
Convergent and discriminant validity
Convergent and discriminant validity were established for the EHFScB scale. Both the 12-item version and the revised nine-item version of the EHFScB scale correlated moderately to the revised HF Compliance Questionnaire (r = 0.32 and 0.37, respectively, P < 0.001). As expected, poor correlation was demonstrated for the 12-item version or the revised nine-item version of the EHFScB scale and the MLwHFQ (r = 0.01 and 0.18, respectively, ns).
Reliability of the EHFScB-9
The reliability estimate for the nine-item total scale was satisfactory (0.80) and higher than the 12-item version (0.77) (Table 1). Reliability estimates for the nine-item scale for the separate data sets were: Sweden 0.78 and 0.77, the Netherlands 0.87 and 0.73, United Kingdom 0.68, Italy 0.78, Germany 0.77, and Spain 0.85. For the EHFScB-9, the reliability estimates for the subscale ‘consulting behaviour’ (0.85) in the pooled sample was satisfactory and was 0.80 in all data sets except for the UK data set, with a Cronbach’s alpha of 0.72. However, the subscale ‘adherence to regimen’ was less internally consistent in the pooled sample (0.56) as well as in the separate samples (0.17–0.78).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
It is increasingly important to focus on the effects of self-care management of HF and therefore it can be useful to include outcomes other then morbidity, mortality, and quality of life in clinical trials. The EHFScB scale can be used to measure the effectiveness of education and support. After factor analyses and critical evaluation of both psychometric properties and content of separate items, a nine-item version was further evaluated. The revised scale (EHFScB-9) in several of its translations was shown to be an internally consistent and valid scale to measure self-care behaviour related to HF.
Although the EHFScB scale was related to patient adherence, the correlation was only moderate and we conclude that a different construct is measured. The concept HF self-care behaviour is broader than adherence. It not only includes adherence to life style changes but also poses questions about the patient's consultation with a health-care professional for changes in symptoms.
The scoring of the scale overall was consistent in the different translations of the scale and no large cultural differences seemed to play a role in this European population. Only the reliability in the UK sample was somewhat lower. This was also reported by Shuldham et al.9 who concluded that the use of the English language seems to be appropriate and underlines the importance of continuous assessment of the applicability of this scale in a range of cultures and settings. Data from non-European samples need to be collected to validate the scale in other populations.
Revision of the scale
The EHFScB scale is based on international guidelines for HF management10,11 as well as on the opinions of representatives of the relevant population and content experts. As a result of the analysis conducted for the current study, three items could be deleted from the scale. Although the deleted items are recognized as important issues in patients’ lives (taking rest if dyspnoea occurs and resting during the day), these activities are often not deliberate actions patients take as part of their self care but are probably forced upon them by their deteriorating condition or worsening symptoms. Thus the items were deleted from the scale. The topic of activity and rest is important to discuss in daily patient care. Likewise, the item addressing the flu shot is important for daily patient education; however, due to psychometric properties the item was deleted from the scale. First, because ‘a flu shot’ is not recognized as HF-specific self-care behaviour by patients, secondly, because the item had a low factor loading and low item total correlations; and, thirdly because the scoring system with five categories is not suitable for an action patients undertake once a year. Although the item analyses as well as the CFAs indicated that the EHFScB-9 has a better construct validity and reliability compared with the EHFScB-12, problems concerning the homogeneity of some items in the scale still remained. This problem is above all related to the items in the ‘adherence to the regimen subscale’ reflected by low item-total correlations. This fact can probably explain the low NFI and CFI indices. Even if all fit indices indicated an improvement between the data and the model using the EHFScB-9 before the EHFScB-12, NFI and CFI still indicate poor fit between the data and the model. However, different goodness-of-fit indices have their own strength and weaknesses, e.g. GFI, AGFI, and NFI can be affected by large sample size while CFI favours more complex models.26 Since neither NFI nor CFI indicated good fit between the data and the model and as ‘adherence to the regimen’ demonstrated low internal consistency, this subscale is not recommended to be used separately.
Based on the factor structure of the scale, the total score of the EHFScB-9 can be used as well as the subscale ‘consulting behaviour’. Denollet and coworkers25 found comparable results in which they described that scores on this subscale (alpha 0.86) were related to Type D personality.
Limitations of the study and weaknesses of the scale
Although the pooled sample consisted of patients from six countries, the samples from the different countries were quite different in size. The samples from the Netherlands comprised 48% of the total sample and therefore had a strong impact on the results. It is therefore important to confirm the validity of the new scale and the two-factor model in other larger international samples.
The factor correlations for item 10 and item 12 are still weak in all factor models and have low item-total correlations. However, both item 10 and 12 are important for describing self-care behaviour, and although this is not supported by the psychometric analysis, based on content validity, these two items were retained. One explanation might be the response pattern for item 10 (taking medication as prescribed) is, as expected, heavily skewed in the positive direction. Responses to item 12 (exercise regularly) are uniformly distributed across the five response categories. Exercise behaviour is a broad and complex concept which may comprise quite different activities, especially in a population of elderly persons with chronic conditions. It is possible that the exercise question is too generally formulated and that more specific items on exercise behaviour are required. Including exercise as an HF-specific self-care behaviour is important since systematic reviews and meta-analyses have shown that physical conditioning by exercise training reduces mortality and hospitalization.24,27
Future steps to improve the scale might be to improve the construct validity of the ‘adherence to regimen’ subscale by possibly adding new HF-specific items relating to this domain.
Implications
The EHFScB-9 is a user-friendly, reliable, and valid instrument to use in research. In the future, the applicability of the nine-item scale in different languages needs to be further studied.
Additionally, the value of a cut-off score needs to be explored. Such a score might be used to target certain patients for additional tailored self-care management interventions, such as increased monitoring or a more intensive educational approach or follow-up.28,29 Further, a clinically significant change should be defined from prospective data in order to have a target effect size when calculating sample size and power and when interpreting results.
In clinical practice, the scale can be used to assess self-care behaviours and as a baseline for the mutual decision between the patient and nurse or physician regarding self-care. It can also be used in following patients to evaluate changes in their self-care behaviour. There are ongoing discussions about how to define when self-care behaviour reaches a satisfactory level and whether the scale could be used in clinical practice as a profile to evaluate each separate item instead of the total score of the nine items. There may be a need for a more extended version to be used in clinical practice including smoking, avoiding use of NSAIDs and healthy diet.
In conclusion, the revised nine-item EHFScB scale was shown to be an internally consistent and valid scale to measure self-care behaviour related to HF. It can be used in research studies to assess changes in self-care behaviours, as well as in clinical settings.
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
This work has been supported by Netherlands Heart Foundation (TJ) Östergötland County Council (TJ, AS), Swedish Research Council (AS) and Vardal Foundation (AS).
|
Acknowledgement |
|---|
The authors would kindly acknowledge the following researchers who worked with the scale and provided data: J. Lupón, B. González, G. Cleuren, M. Holst, G. Pulignano, R. Holland, N. Holzapfel.
Conflict of interest: none declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
- McAlister F, Stewart S, Ferrua S, McMurray J. Multidisciplinary strategies for the management of heart failure patients at high risk for admission. A systematic review of randomised trials. J Am Coll Cardiol (2004) 44:810–819.
[Abstract/Free Full Text] - Phillips CO, Singa RM, Rubin HR, Jaarsma T. Complexity of program and clinical outcomes of heart failure disease management incorporating specialist nurse-led HF clinics. A meta-regression analysis. Eur J Heart Fail (2005) 7:333–341.
[Abstract/Free Full Text] - Jaarsma T, Strömberg A, Mårtensson J, Dracup K. Development and testing of the European Heart Failure Self-Care Behaviour Scale. Eur J Heart Fail (2003) 5:363–370.
[Abstract/Free Full Text] - Riegel B, Carlson B, Moser D, Sebern M, Hicks F, Roland V. Psychometric testing of the self-care of heart failure index. J Card Fail (2004) 10:350–360.[Medline]
- van der Wal MH, Jaarsma T, Moser DK, Veeger NJ, van Gilst WH, van Veldhuisen DJ. Compliance in heart failure patients: the importance of knowledge and beliefs. Eur Heart J (2006) 27:434–440.
[Abstract/Free Full Text] - van der Wal MH, Jaarsma T, van Veldhuisen DJ. Non-compliance in patients with heart failure; how can we manage it? Eur J Heart Fail (2005) 7:5–17.
[Abstract/Free Full Text] - Gonzalez B, Lupon L, Parajon T, Urrutia A, Herreros J, Valle V. Use of the European Heart Failure Self-care Behaviour Scale (EHFScBS) in a heart failure unit in Spain. Rev Esp Cardiol (2006) 59:166–170.[Medline]
- Pulignano G, Del Sindaco D, Jaarsma T. Translation and validation of the Italian version of the European self-care behaviour scale. Eur J Heart Fail (2004) 6:264.
- Shuldham C, Theaker C, Jaarsma T, Cowie M. Evaluation of the European heart failure self-care behaviour scale in a United Kingdom population. J Adv Nurs (2007) 60:87–95.[Medline]
- Swedberg K, Cleland J, Dargie H, Follath F, Komajda M, Tavazzi L, Smiseth OA, Gavazzi A, Haverich A, Hoes A, Jaarsma T, Korewicki J, Lévy S, Linde C, Lopez-Sendon JL, Nieminen MS, Piérard L, Remme WJ, Task Force for the Diagnosis and Treatment of Chronic Heart Failure of the European Society of Cardiology. Guidelines for the diagnosis and treatment of chronic heart failure: executive summary (update 2005): The Task Force for the Diagnosis and Treatment of Chronic Heart Failure of the European Society of Cardiology. Eur Heart J (2005) 26:1115–1140.
[Free Full Text] - Heart Failure Society of America. Executive summary: HFSA 2006 Comprehensive Heart Failure Practice Guideline. J Cardiac Fail (2006) 12:10–38.[CrossRef][Web of Science][Medline]
- Graneheim U, Lundman B. Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness. Nurse Educ Today (2004) 24:105–112.[CrossRef][Web of Science][Medline]
- Nunally JC, Bernstein IH. Psychometric Theory (1994) 2nd ed. New York: McGraw-Hill.
- Thompson B. Exploratory and Confirmatory Factor Analysis: Understanding Concepts and Applications (2004) Washington: American Psychological Association.
- Fayers PM, Machin D. Quality of Life: Assessment, Analysis and Interpretation (2000) London: John Wiley & Sons.
- Steiger J. Structural model evaluation and modification: An interval estimation approach. Multivariate Behav Res (1990) 25:173–180.[CrossRef][Web of Science]
- Bentler P. Comparative fit indices in structural models. Psychol Bull (1990) 107:238–246.[CrossRef][Web of Science][Medline]
- Bentler P, Bonett D. Significance tests and goodness of fit in the analysis of covariance structures. Psychol Bull (1980) 88:588–606.[CrossRef][Web of Science]
- Jöreskog K. Testing Structural Equation Models (1993) Newbury Park, CA: Sage.
- Kelloway EK. Using LISREL for Structural Equation Modeling: A Researcher’s Guide (1998) Thousand Oaks: SAGE.
- Evangelista LS, Berg J, Dracup K. Relationship between psychosocial variables and compliance in patients with heart failure. Heart Lung (2001) 30:294–301.[CrossRef][Web of Science][Medline]
- Rector TS, Kubo SH, Cohn JN. Validity of the Minnesota Living with Heart Failure questionnaire as a measure of therapeutic response to enalapril or placebo. Am J Cardiol (1993) 71:1106–1107.[CrossRef][Web of Science][Medline]
- Altman DG. Practical Statistics for Medical Research (1991) London: Chapman & Hall.
- Piepoli M, Davos C, Francis D, Coats A, ExTraMATCH Collborative. Exercise training meta-analysis of trials in patients with chronic heart failure (ExTraMATCH). BMJ (2004) 328:189.
[Abstract/Free Full Text] - Schiffer AA, Denollet J, Widdershoven JW, Hendriks EH, Smith OR. Failure to consult for symptoms of heart failure in patients with a type-D personality. Heart (2007) 93:814–818.
[Abstract/Free Full Text] - Sun J. Assessing goodness of fit in confirmatory factor analysis. Measurement and Evaluation in Counseling and Development (2005) 37:240–256.[Web of Science]
- Rees K, Taylor R, Singh S, Coats A, Ebrahim S. Exercise based rehabilitation for heart failure. Cochrane Database Syst Rev (2004) 3:CD003331.[Medline]
- Jaarsma T, van der Wal M, Lesman-Leegte I, Luttik MLA, Hogenhuis J, Veeger NJGM, Sanderman R, Hoes AW, Gilst WH van, Lok DJA, Dunselman PHJM, Tijssen JGP, Hillege HL, Veldhuisen DJ van, for the COACH study group. Effects of moderate or intensive disease management program on outcome in patients with heart failure. The Coordinating study evaluating Outcomes of Advising and Counseling in Heart Failure (COACH). Arch Intern Med (2008) 168:316–324.
[Abstract/Free Full Text] - Jaarsma T, van Veldhuisen DJ. When, how and where should we ‘coach’ patients with heart failure: the COACH results in perspective. Eur J Heart Fail (2008) 10:331–333.
[Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. M. Clark, C. N. Freydberg, F. A. McAlister, R. T. Tsuyuki, P. W. Armstrong, and L. A. Strain Patient and informal caregivers' knowledge of heart failure: necessary but insufficient for effective self-care Eur J Heart Fail, June 1, 2009; 11(6): 617 - 621. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||