The MacNew is used to assess HRQOL in patients with CAD and comprises three dimensions: physical, emotional, and social [7]. Since being validated at the time of development for use in patients with heart disease, it has been translated from English into several other languages (macnew.org). In this study, we tested the construct validity, criterion validity, convergent validity, and reliability of the K-MacNew to determine whether the instrument can be used among patients with CAD in Korea.
As the MacNew comprises three dimensions, we also performed an EFA to verify the construct validity of the K-MacNew. Generally, an eigenvalue greater than 1 has been used as the criterion to determine the number of factors in factor analysis in studies testing the construct validity of multidimensional instruments. However, owing to concerns that this may overestimate the number of factors, parallel tests have been performed in recent studies to determine the appropriate number of factors [28]. Hence, we determined that the K-MacNew has three factors based on theoretical evidence, PCA with varimax rotation, and a parallel test. The EFA yielded an explanatory power of 57.92% of the total variance
The MacNew was used with a factor loading threshold of > 0.40 in the factor structure analysis [13, 14]. Therefore, as the MacNew allows cross-loading of items, one item is often cross-loaded onto different factors during factor analysis to test construct validity. The original tool has 12 items that are cross-loaded onto two or more factors [13]. Furthermore, although it varies across studies, previous studies reported that at least three items are cross-loaded onto two or more factors [6, 7, 12, 26]. The K-MacNew also features seven items that are cross-loaded onto two factors, and one item is cross-loaded onto three factors. However, some researchers have questioned the validity of including all items with a factor weight > 0.40 [12, 29]. This is because the results vary across studies, in which 27 items are loaded onto different factors depending on the language, and some items are loaded onto dimensions not explained in the original instrument.
Of the 27 items included in this study, 15 were loaded onto the emotional dimension. All 14 items from the emotional dimension of the original MacNew were included in the emotional dimension of the Korean version, along with one additional item (Item 11, more dependent). Item 11 was loaded onto the social dimension in the original version and the physical dimension in the Sri Lankan version [12], and was not loaded onto any of the three dimensions in the Chinese [6] and Turkish versions [27]. The factor loadings of Item 11 across studies may be attributable to the cultural differences that occur when interpreting the statements of the instrument. In other words, similar to Sri Lanka, in Korean patients, becoming more physically dependent owing to a heart problem may have been perceived as becoming an emotional burden for significant others such as one’s family [12]. Of the 15 items, the loading value was the highest for Item 4 (down in the dumps), followed by Items 1 (frustrated), 2 (worthless), and 10 (tearful). This is similar to the results of a previous study in which the emotional dimension addresses with emotional responses generally experienced by patients with CAD, such as depression [26].
Nine out of the 27 items were loaded on the physical dimension. Of these items, eight items were consistent with the original MacNew, and Item 15 (lack self-confidence) was added. Item 15 was cross-loaded onto the emotional and social dimensions in the original version, but cross-loaded onto the emotional and physical dimensions in the Korean version. This may be attributable to the ambiguity of the original tool. For instance, Item 3—loaded onto the emotional dimension—specifically asks about one’s confidence and decision-making regarding the management of heart problems. However, Item 15 asks about overall lack of confidence, without specifying to what that confidence is related. Therefore, it is possible that participants understood Item 15 to be asking about self-confidence of symptom management; however, considering that Item 15 was cross-loaded onto the emotional and social dimensions in previous studies [6, 7, 12], it is necessary to conduct replication studies with a larger sample to further confirm the construct validity of the K-MacNew.
The physical dimension assesses the level of physical symptoms of heart disease and resulting physical limitations. The items with the highest loading values for the physical dimension in our study were items pertinent to physical symptoms (Item 9: shortness of breath, Item 14: chest pain, Item 16: aching legs, and Item 19, dizzy or lightheaded). However, five items that had been loaded onto the physical dimension in the original instrument (Items 12, 24, 25, 26, and 27) were instead included in social dimension in our study. This results was similar to the versions from other Asian countries, including the Sri Lankan [12] and Chinese versions [6]. This suggests that patients perceive items related to physical limitations to be pertinent to engaging in less of social activity as a result of such limitations. Thus, subsequent studies should make multinational comparisons of how items related to physical symptoms are perceived among patients with CAD in various Asian countries.
Finally, 11 of the 27 items of the K-MacNew were loaded onto the social dimension. Ten of these items were consistent with the original, and Item 27 (sexual activity) was included. A number of previous studies excluded Item 27 when testing construct validity owing to cultural reasons [15]. However, sexual activity is important to consider when examining the HRQOL of patients with heart disease. This is because it can affect one’s physiological needs as well as the intimate relationship with spouse. Item 27, which was classified into the physical originally, loaded into the social dimension in this study. This is similar to the results of a previous study that tested the construct validity of the English version of the MacNew in patients with angina and ischemic heart failure [15]. The loading of Item 27 onto the social dimension is related to the fact that this dimension of the MacNew assesses social situations experienced in a physical or emotional context [15]. However, it is necessary to confirm the appropriateness of the K-MacNew factor loadings via a confirmatory factor analysis. In the present study, Items 24 (excluded), 26 (physical activity), and 25 (unable to socialize) had the highest factor loadings in the social dimension. These items also had high factor loadings and were cross-loaded in the physical and social dimensions in the original version. However, these items were only loaded onto the social dimension in this study, presumably owing to cultural differences in how disease impact was perceived. Therefore, participants seemed to perceive their physical symptoms as physical health problems and disease-related limitations to social and other activities as social health problems.
In the analysis of known-group validity, the low KASI class, which have little limitation physical activities, scored significantly higher for the K-MacNew compared to the high KASI class, and the MID of the K-MacNew between the two groups exceeded 0.5 [17]. This is similar to the finding of a previous study in which the severe symptoms group reported significantly higher MacNew scores compared to the non-severe symptoms group among patients with ischemic heart disease, and that the difference was beyond the MID [8]. Consequently, the K-MacNew would be a useful instrument in clinical settings to examine HRQOL according to the level of physical activity and symptom severity in patients with CAD, as well as to assess cardiac rehabilitation therapies.
To test the concurrent validity of the K-MacNew, we analyzed the physical and emotional dimensions of the K-MacNew and the physical health and mental health domains of the SF-36, respectively. The correlation coefficients ranged from 0.53 to 0.75, showing a moderate correlation. This finding is in line with prior studies [7, 15]. Additionally, the K-MacNew had a moderate or higher negative correlation with depression and anxiety, thus establishing convergent validity. This is consistent with previous reports that HRQOL decreases as depression and anxiety increase [6, 8, 27].
Regarding reliability, Cronbach’s alphas ranged from 0.88–0.93, and the Cronbach’s alphas for the global and dimensional scores were similarly high. This is similar to the level of internal consistency reported by previous studies that applied the MacNew in Asia [6, 12]. The K-MacNew is a homogeneous tool for measuring HRQOL in patients with CAD in Korea. However, we could not verify its test-retest reliability; therefore, subsequent studies to ensure scoring stability in addition to item homogeneity are needed.
This study had a few limitations. As participants were selected using convenience sampling, the sample is not representative of all patients with CAD. Second, most participants were women and those with unstable angina; thus, caution should be taken when interpreting the validity of the K-MacNew with men and those with myocardial infarction. However, despite these limitations, this study is significant in that it helps to cross-culturally validate the MacNew. Therefore, accumulating further evidence on the validity and reliability of the K-MacNew through replication studies would enhance the sensitivity of the K-MacNew to assess HRQOL in patients with CAD and promote its wide use in clinical settings.