Introduction
Health-related quality of life (HRQoL) is an important and relevant concept in patients with cardiac disease, allowing for a more comprehensive assessment of health status as perceived by the patient. HRQoL presents the patient's individual perspective of the burden and trajectory of his or her illness as well as of his or her overall health (Rumsfeld et al., 2013). Thus, careful assessment of this concept may provide valuable information about the patient and help guide clinical decisions and treatment. Several generic and disease-specific tools have been used to assess HRQoL in cardiac patients. Disease-specific tools tend to better reflect the impact of a disease on various aspects of a patient's life (Pavy et al., 2015) and are more sensitive to changes during recovery from cardiac events such as myocardial infarction (MI) and related treatments (Nakajima, Rodrigues, Gallani, Alexandre, & Oldridge, 2009).
Despite significant improvements in intervention and treatment, MI remains one of the leading causes of mortality and morbidity in South Korea (Kook et al., 2014). Increasing attention is being paid to improving the experience of patients with MI and the impact that such a life-threatening event has on patient well-being and quality of life (Rumsfeld et al., 2013). In Korea, the 36-item Short Form Health Survey is the most commonly used generic tool for assessing HRQoL in cardiac patients, followed by the Seattle Angina Questionnaire, the Quality of Life Index-Cardiac, and the Padilla and Grant's Quality of Life Index (Lee, Tak, & Song, 2005). However, the MacNew Heart Disease Health-related Quality of Life Questionnaire (MacNew) has been shown to reflect the experiences of patients with cardiac diseases more comprehensively than any of the abovementioned measures. The MacNew requires that patients answer the items about their "heart problem" and the impact of these experiences on different aspects of their life, whereas the Seattle Angina Questionnaire and the Minnesota Living with Heart Failure Questionnaire refer to "chest pain, chest tightness or angina" and "your heart failure," respectively, in the questions, which may not sufficiently describe types of cardiac events such as heart attacks (Hofer et al., 2012; Pavy et al., 2015).
The MacNew is the modified version of the Quality of Life after Myocardial Infarction Questionnaire that was developed by Hillers et al. (1994) and the Quality of Life after Myocardial Infarction-2 Questionnaire that was developed by Valenti, Lim, Heller, and Knapp (1996). It is a valid and reliable questionnaire for assessing HRQoL in patients with a broad range of cardiac diseases, including angina, heart failure, and MI (Hofer, Lim, Guyatt, & Oldridge, 2004). The MacNew has been translated into a wide range of languages and is currently used in more than 50 countries. The psychometric properties of the tool have been validated in patients with MI and other cardiac conditions in 20 languages (MacNew.org, 2016). However, the Korean MacNew had not yet been validated. The aim of this study was to translate the MacNew into Korean and then assess the reliability, validity, and factor structure of the tool in terms of measuring HRQoL in Korean patients with MI.
Methods
Settings and Subjects
This study was implemented as part of a larger observational longitudinal study that aimed to examine changes in HRQoL after MI and to identify factors affecting HRQoL in the patient population with MI. The study was conducted at the cardiovascular centers of two major tertiary referral hospitals in the southern part of South Korea. Patients admitted to these centers were consecutively recruited from August 2015 to February 2016 and were followed up for 3 months after discharge. The inclusion criteria were patients who were (a) admitted to a cardiac department with a diagnosis of MI (either STEMI[ST-Elevation MI] or non-STEMI), (b) able to understand and speak Korean, (c) living in Korea, and (d) able to understand the study and provide informed consent.
Two hundred fifteen patients were screened in accordance with the inclusion criteria. One hundred fifty patients (69.8%) gave informed consent to participate and completed the study questionnaires at baseline and after 3 months. Of the screened patients, 65 were not recruited because of poor health conditions (n = 23), having declined without providing a reason (n = 19), inadequate hearing (n = 17), being discharged before enrollment (n = 5), or being unconscious (n = 1). The categories of poor health condition included experiencing dyspnea, pain on the site of intervention, severe tremors, and tiredness. By the time of the follow-up, four of the 150 participants had died and 10 were lost to follow-up. As a result, 136 participants completed the study questionnaires, including the Korean MacNew, at the 3-month follow-up.
Study Procedure
The ethics approval of each institutional review board was obtained before recruiting participants. After participants had signed the consent form, they were asked to complete the self-report questionnaires in the Korean language. Clinical data on the participants were retrieved from medical records. Nursing staff and cardiologists collaborated and provided counseling on recruiting study participants and collecting medical records. At 3 months after hospital discharge, either a follow-up telephone interview was carried out with each participant or participants were asked to complete the follow-up questionnaires in a face-to-face session when they attended the outpatient department as part of their routine care.
Measurements
The study questionnaires included key sociodemographic questions such as age, gender, marital status, and self-evaluated income (excellent, good, only fair, and poor) and questions about clinical characteristics, including recurrent MI (yes/ no) and physical activity (active/relatively active/not active).
This study used the Korean MacNew. The original MacNew consists of 27 items, which assess the perceived emotional (14 items), physical (13 items), and social functioning (13 items) status of cardiac patients over the previous 2-week period, with some items included in multiple subscales. Each item is scored on a 7-point Likert scale ranging from 1 to 7, with higher scores indicating better HRQoL. The total score of the MacNew is calculated as the average of the 27 items, and domain scores are the average of the items in each subscale, with a possible score range between 1 and 7. Missing items are excluded from scoring, and the 27th item may be excluded in the physical domain (Dixon, Lim, & Oldridge, 2002).
For the purpose of this study, the tool was translated into Korean following the guidelines suggested by Guillemin, Bombardier, and Beaton (1993). To obtain a quality translation, the principal researcher, who has significant experience translating English texts into Korean, first translated the instrument into Korean. Back-translation was then carried out by two bilingual experts who had not seen the questionnaire previously. Differences in the translations were discussed, and agreement was reached on the final version. The Korean MacNew was next reviewed for face validity by three Korean health professionals in the field of cardiovascular disease and five laypersons.
The Depression Anxiety and Stress Scale (DASS 21), the single-item fatigue scale, and the single-item global quality of life scale were used to assess the concurrent construct validity of the Korean MacNew. It was hypothesized that the MacNew total score would correlate significantly and closely with the DASS 21 (McDonnell, Mackintosh, Hillier, & Bryan, 2014). The DASS 21, the short form of the DASS 42, is designed to measure the severity of the core symptoms of depression, anxiety, and stress. This scale consists of 21 items, with each item scored from 0 = does not apply to me at all to 3 = applies to me very much or most of the time. Higher total scores on the DASS 21 represent greater emotional distress. Past research shows a strong positive relationship between the experience of depression and poorer HRQoL in patients with MI (McDonnell et al., 2014; Morys, Bellwon, Hofer, Rynkiewicz, & Gruchala, 2016). The DASS 21 has been widely used in Asian countries, and a Korean version is available (Cha, 2014). This study further hypothesized that the MacNew total score would correlate significantly and closely with the single-item fatigue scale (Alsen & Brink, 2013; Casillas, Damak, Chauvet-Gelinier, Deley, & Ornetti, 2006; Hwang, Liao, & Huang, 2014). The single-item fatigue scale is a valid tool with response options ranging from 0 = no fatigue to 10 = greatest possible fatigue (H. J. Kim & Abraham, 2017).
In addition, this study expected to find a significant correlation between the MacNew global scale with the single-item quality of life scale (de Boer et al., 2004), as both measure the same construct. The single-item quality of life scale has proved to be a valid tool with response options ranging from 0 = the worst it has ever been to 10 = the best it has ever been (de Boer et al., 2004). On the basis of past literature, this study hypothesized that the MacNew scores would correlate negatively with the DASS (McDonnell et al., 2014; Morys et al., 2016) and the fatigue scores (Alsen & Brink, 2013; Casillas et al., 2006; Hwang et al., 2014), but positively with the single-item quality of life score (Alsen & Brink, 2013; de Boer et al., 2004; Wang, Thompson, Ski, & Liu, 2014).
Ethical Considerations
Ethical approvals were obtained from the relevant research ethics committees before commencement of the study (PNUH-IRB no. H-1505-008-029, PNUYH IRB no. 05-2015-072, and UTS HREC approval no. 2015000254). The researcher provided verbal and written information about the study and its objective to the participants and assured them of their voluntary participation, confidentiality, and privacy. Informed consent was obtained from all participants.
Statistical Analyses
The follow-up data of 136 participants were analyzed for the present validation study using IBM SPSS Statistics 24. Characteristics of the participants were described in terms of frequencies, means, and standard deviations (Table 1). The psychometric properties of the Korean MacNew were assessed in accordance with the recommendations of the Scientific Advisory Committee of the Medical Outcomes Trust (2002). Specifically, two steps of factor analysis were conducted. First, partial confirmatory factor analysis (PCFA) with direct oblimin rotation (maximum likelihood) was performed to determine if the items loaded similarly to the theoretically clear structure in three factors of the original version (Valenti et al., 1996). Second, exploratory principal component factor analysis (EFA) with varimax rotation was carried out to identify a better structure for the Korean MacNew. Twenty-six items were included, and Item 27 was excluded in the factor analysis based on the original report (Valenti et al., 1996). The suitability of data for factor analysis was assessed by .30 and higher in the correlation matrix, the recommended value of .60 in the Kaiser-Meyer-Olkin value, and the significance (p < .001) in the Bartlett's test of sphericity. The values of evaluating the CFA model fit required higher than .950 of the normed fit index, comparative fit index, and Tucker-Lewis index as well as < .06 or < .08 of root mean square error of approximation and standardized root mean square residual.
The Cronbach's [alpha] coefficient was used to examine the internal consistency of the overall tool and of each dimension. The emotional subscale included Items 1-8, 10, 12, 13, 15, 18, and 23; the physical subscale included Items 1, 6, 9, 12, 14, 16, 17, 19, 20, 21, and 24-26; and the social subscale included Items 2, 11-13, 15, 17, and 20-26. The concurrent construct validity of the Korean MacNew was assessed by calculating the Pearson correlation coefficients of the total score of the MacNew with the DASS, with the single-item fatigue scale, and with the single-item global quality of life scale, respectively. The strength of the correlations was considered weak when r = .10-.29, medium when r = .30-.49, and strong when r = .50-1.0 (Pallant, 2016). Discriminant validity was determined via assessment of the ability of the Korean MacNew to discriminate between men and women, different age groups, and levels of physical activity. We hypothesized that the HRQoL of the participants would be poorer if they were female, older in age, and less active. Gender (Lim et al., 1993; Valenti et al., 1996) and age (H. M. Kim, Kim, & Hwang, 2015; Oginska-Bulik, 2014; Valenti et al., 1996) have been used in previous studies to assess the discriminant construct validity of the MacNew. In addition, evidence has shown that patients who are less physically active tend to have a poorer quality of life than their more active peers (Hawkes et al., 2013).
Results
The sociodemographic and clinical characteristics of the study participants are presented in Table 1. The sample had a mean (SD) age of 64.35 (11.61) years. Most of the sample were male (73.5%), were married (87.5%), and earned a "fair" level of income (60.3%). Moreover, most were experiencing MI for the first time (77.9%) and were physically not active or only relatively active (71.3%).
The PCFA with direct oblimin rotation was applied to help interpret the factor loading of each of 26 items of the MacNew on the three factors in the original version (Valenti et al., 1996), explaining 50.2% of the total variance (see Table 2). Most of the items loaded on the similar factor of the original study at more than .4, although six items loaded at between .3 and .4. Specifically, four of the 26 items in the Korean MacNew loaded on unexpected factors. Item 16 (aching legs) loaded on the social factor instead of the physical factor, whereas Item 21 (unsure about exercise), Item 22 (overprotective family), and Item 24 (excluded) loaded on the emotional factor rather than the physical or social factor. In addition, except for standardized root mean square residual (.055), other values of normed fit index (.784), comparative fit index (.879), Tucker-Lewis index (.843), and root mean square error of approximation (.082) did not meet the requirements in the three-factor structure.
To examine the factor structure of the Korean MacNew, the data on the 26 items were subjected to EFA. The inspection of the correlation matrix revealed the presence of many coefficients at higher than .30 and .89 of the Kaiser-Meyer-Olkin value and that Bartlett's test of sphericity reached significance (p < .001), supporting the factorability of the data. The results of rotation sums of squared loadings in the EFA presented the five-factor structure, explaining 64.9% of the total variance (Table 3). Half of the 26 items loaded on more than one factor, and the first three factors explained about 50% of the variance.
The internal consistency of the global scale was high, with a Cronbach's alpha coefficient of .93. In addition, coefficients of emotional, physical, and social subscales, allocated as the original study (Valenti et al., 1996), were examined as indicated by .92, .88, and .91, respectively.
The concurrent construct validity of the Korean MacNew was supported by showing strong negative correlations between the global MacNew and DASS 21 (r = -.81, p < .001) and between the global score and the single-item fatigue scale (r = -.51, p < .001). Furthermore, there were significant positive correlations between the total MacNew score and the single-item global quality of life scale (r = .73, p < .001).
The discriminant validity of the Korean MacNew was further supported by examining the discriminant function of the tool across different age groups, gender groups, and physical activity groups. Patients >= 65 years old showed lower HRQoL than those < 65 years old (5.34 vs. 5.74, respectively; p = .002). Moreover, the Korean MacNew discriminated well between female and male, with female patients showing poorer HRQoL than male patients (5.20 vs. 5.66, respectively; p = .002). The differences in the MacNew scores among patients who were more active, relatively active, and not active (5.82, 5.81, and 5.19, respectively; p < .001) were statistically significant, indicating that the discriminant concurrent validity for the Korean MacNew was well confirmed.
Discussion
The self-evaluation of patients regarding the impact of disease on their daily functionality and quality of life is important to facilitate patient-centered care and to improve disease and patient outcomes. HRQoL has been an important patient-reported health outcome in consideration of its prediction of mortality, recurrence of cardiovascular events, and rehospitalization among patients with cardiovascular diseases, particularly MI (Anker et al., 2014). The MacNew is one of the most popular disease-specific questionnaires for assessing HRQoL in cardiac patients (MacNew.org, 2016). The current study showed that the Korean MacNew is also reliable and valid for measuring HRQoL in patients with MI.
The Cronbach's alpha coefficients of the Korean MacNew in the current study were high, with .93 for the overall score and .92, .88, and .91 for the emotional, physical, and social subscales, respectively. These results are consistent with prior internal consistency studies, with the average internal consistency reliability coefficients for the total, emotional, physical, and social domains reported as .93, .92, .86, and .88, respectively, across 23 validation studies conducted on different language versions of the MacNew (Hofer et al., 2004, 2012; Pavy et al., 2015; Wang, Lau, Palham, Chow, & He, 2015). The concurrent construct validity and discriminate construct validity of the Korean MacNew were also supported. Therefore, the Korean MacNew showed high reliability and validity for assessing HRQoL in patients with MI.
When performing PCFA with the direct oblimin rotation solution to determine if items loaded similarly to the theoretically clear structure on the three factors in the original version (Valenti et al., 1996), all of the 26 items in the `Korean MacNew met the threshold standard for item retention and most loaded on the same factors as in the original study. The physical and social factors explained only 6.7% and 4.4% of the variance, respectively, whereas the emotional factor explained 39.1% of the variance. A few items loaded on a factor that differed from the original validation study. Thus, EFA was implemented to identify a better structure for the Korean MacNew, with the results revealing that the 26 items of the Korean MacNew were likely grouped into five factors.
In general, the results of previous validation studies on both English-language (Dempster, Donnelly, & O'Loughlin, 2004) and non-English-language patients with cardiac disease failed to support the item loading pattern (Gramm, Farin, & Jaeckel, 2012) that was reported in the original study. For example, Dempster et al. (2004) established a five-factor solution that included the factors of emotion, restriction, symptoms, perception of others, and social with a population of cardiac patients in Ireland (Dempster et al., 2004). These findings imply that the factor structure of the MacNew may need to be reviewed further.
Overall, the results of the current validation study suggest that the Korean MacNew is a valid and reliable tool for assessing HRQoL in patients with MI. However, we recommend that only the total score for the Korean MacNew be used at this stage, unless future studies with bigger sample sizes provide more consistent results on the pattern of item loadings on the individual subscales. Our study sample size of 136 may not be large enough to produce reliable results. Although some authors suggest that five cases for each item are adequate for a factor analysis in most cases, the typical recommendation has been that the larger sample size, the better the reliability and validity of analysis results (Tabachnick & Fidell, 2013).
Conclusions
The Korean MacNew showed consistently acceptable psychometric properties of reliability and validity in patients with MI. Therefore, this instrument may be used to assess HRQoL in Korean patients with MI to develop a better understanding of the health conditions of patients after MI and to evaluate the effectiveness of interventions and related treatments based on actual patient experiences. However, caution should be taken in using the subscale scores.
References