Psychometric properties of the Chinese version of the recovery locus of control (RLOC) Scale among patients with myocardial infarction

Abstract Aim To translate the RLOC Scale into Chinese and test its psychometric properties in the Chinese patients with myocardial infarction (MI). Design A cross‐sectional design was used. Methods A convenience sample of 233 patients with MI who have undergone percutaneous coronary intervention and were ready for discharge were recruited in a level A tertiary hospital of Shandong Province from January 2019 to April 2019. Data were analysed using item analysis, internal consistency reliability and exploratory factor analysis. Results Two factors—external RLOC and internal RLOC—were extracted, accounting for 70.5% of the variance. Cronbach's alpha for the Chinese version of RLOC Scale was 0.80 and for the two factors was 0.92 and 0.76, respectively. The Chinese version of RLOC Scale showed satisfactory reliability and validity, which can be used to measure the ability of recovery locus of control in Chinese patients with MI.


| INTRODUC TI ON
Perceived control of recovery is an individual belief that patients have about their recovery process, which affects their involvement in health-related behaviours considered to enhance the recovery process (Johnston, Morrison, Macwalter, & Partridge, 1999).
Studies have shown that greater perceived personal control, that is an internal locus of control, is associated with more beneficial outcomes. For example, in the health filed, this result has been found for Parkinson's disease (Rizza et al., 2017), for medication adherence (Nafradi, Nakamoto, & Schulz, 2017) and for health-promoting behaviours in Chinese patients with coronary heart disease (Zou, Tian, Chen, Cheng, & Fan, 2017).
Myocardial infarction (MI) is now a major cause of morbidity and mortality in China, with a rapid increase in the number of affected patients between 1999-2015 (Zhao et al., 2017) Cardiac remodelling and subsequent heart failure remain critical issues after MI despite improved treatment and reperfusion strategies (Haubner et al., 2016). The burden of MI remains high, currently causing one million deaths annually (Li et al., 2016). Therefore, a comprehensive strategy for secondary prevention is warranted among patients, after an initial MI, with the need for cardiac rehabilitation beginning soon after discharge from the hospital (Reed, Rossi, & Cannon, 2017). In the literature on coping and health, patients' self-management, particularly their perception of the extent of control over processes and outcomes (in other words, perceived personal control), is shown to play an important role in enhancing recovery.
Recovery locus of control (RLOC) is a personality trait, which affects patients' involvement in health-related behaviours, particularly self-management (Partridge & Johnston, 1989;Thakral, Bhatia, Gettig, Nimgaonkar, & Deshpande, 2014). A previous study found that stroke patients who were able to positively perceive their level of self-control and overcome negative thoughts towards stroke had faster recovery (Thompson, 1991). Stroke patients with a higher RLOC also had increased physical functioning (Mohd Zulkifly, Ghazali, Che Din, Desa, & Raymond, 2015). However, to our knowledge, there are no published papers about RLOC among patients with MI.
Various measures of health-related locus of control exist. For example, Wallston, Wallston, Kaplan, and Maides, (1976) and Wallston, Wallston, and Devellis, (1978) Health Locus of Control (HLOC) Scale, which was later developed into the Multidimensional Health Locus of Control (MHLC) Scale, is widely used. Yet, they focus on preventive health behaviours make them unsuitable for measuring perceived control over existing severe physical disease. Partridge and Johnston (1989) developed a situation-specific measure, the RLOC Scale, to predict the behaviour cognitions of individuals in the context of a physical disability (e.g. patients with stroke). Patients' perceived control over the recovery process is presented as the RLOC, characterized by internal recovery locus of control (IRLOC) and external recovery locus of control (ERLOC); IRLOC is the patient's belief that their health condition depends on themselves, while ERLOC is the belief that it is determined by external environmental factors (Partridge & Johnston, 1989). Studies have shown that patients with higher IRLOC have faster recovery (Hanusch, O'Connor, Ions, Scott, & Gregg, 2014;Shaw, McColl, & Bond, 2003). The post-MI recovery period can be a confusing, emotional time. However, MI patients' behaviour cognitions of self-care over the recovery process play an important role in improving outcomes, preventing hospital readmission and another MI. Thereby, it is important to choose a situation-specific evaluation instrument to estimate patients' recovery beliefs before health caregivers effectively empower patients and their families, and engage patients in self-management and health behaviour change. Hence, the RLOC Scale is an ideal assessment tool to evaluate MI patients' recovery locus of control.
The original version of the RLOC Scale, developed by Partridge and Johnston (1989), has been tested in European countries with good psychometric properties, in patients with physical disabilities or stroke. However, little is known about how the RLOC Scale performs in Chinese patients with MI. In fact, to date, no study examining the RLOC Scale in Chinese samples has been published in an Englishlanguage journal. It is important for nurses to know the self-efficacy regarding personal care in patients with myocardial infarction (MI), to identify individuals at risk and to make care plans. Assessment using the RLOC Scale and consequent treatment is expected to reduce psychological effects such as depression and loss of personal control due to MI, as it is believed to encourage patients to have positive perceptions towards their recovery. In addition to treatment, establishing an understanding of patients' RLOC can not only reduce their psychological burden but also promote recovery from disease.
Thus, patients can improve their physical functioning and avoid critical issues in recovery. The use of a reliable and valid instrument that measures the RLOC may stimulate further research related to health promotion and chronic disease self-management. Therefore, the aim of this study was to test the validity and reliability of the Chinese version of the RLOC Scale among patients with MI.

| Study design and sample
This study used a cross-sectional survey design. The participants were 285 patients with MI who had undergone percutaneous coronary intervention (PCI) and were ready for discharge. They were recruited from a level A tertiary hospital of Shandong Province between January and April 2019. Of them, 52 were excluded from the study for the following reasons: 28 patients were diagnosed by their doctor to have cognitive impairment; nine patients had auditory dysfunction and/or dyslexia; one patient had a malignant tumour; three patients had severe liver and kidney disease; and 11 patients refused to take part in this survey or were eliminated for other reasons. Finally, 233 patients were included in the study.
The sample size was estimated based on the criterion for psychometric assessment of an instrument requiring 5-10 subjects per item (Nunnally & Bernstein, 1994). For the nine-item Chinese version of the RLOC Scale, a sample of at least 90 subjects would be required; therefore, the sample size used in this study was reasonable.
Inclusion criteria were as follows: (a) patients with MI who underwent PCI and were ready for discharge; (b) aged between 18-75 years; and (c) able to read and speak Chinese. Exclusion criteria were as follows: (a) patients with malignant tumours, such as colorectal cancer, oesophageal cancer, gastric cancer or liver cancer; (b) severely impaired renal function (with estimated glomerular filtration rate < 30 ml/min/1.73 m 2 or on dialysis); (c) female patients in the pregnancy or suckling period; (d) patients with auditory dysfunction and/or dyslexia; and (e) patients with cognitive impairment and/or mental disorders.

| Instruments of data collection
The nine-item RLOC Scale is a self-report tool with five internal and four external items (Partridge & Johnston, 1989). IRLOC measures the belief that patients' health condition depends on themselves, whereas ERLOC (4 items) measures the belief that patients' health condition is determined by external environmental factors. Each item is rated on a five-point Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree), providing a total score in the range of 9-45, with a higher score indicating a better RLOC. The construct validity for the two-factor model has been demonstrated, and the original scale has been shown to have internal consistency of, respectively, 0.50-0.77 (Johnston et al., 1999). The sociodemographic variables and clinical data obtained from the participants included sex, age, marital status, residence, educational level, occupation and the primary caregiver.

| Translation procedures
The RLOC Scale was translated into Chinese using Brislin's (1986) forward and backward translation method. The translation of the RLOC Scale from the original English to Chinese was first performed by two independent and professional translators each (A and B) in the research team. The two translated versions (RLOC Scale-A and RLOC Scale-B) were merged into a single forward translation version (RLOC Scale-C) by a third professional and native Chinese speaker (C). This version was then translated back into English by a fourth, bilingual researcher (D) who was not exposed to the scale previously.
Discrepancies between the original and the back-translated versions were reviewed for equivalence of meaning. Finally, the Chinese version of the RLOC Scale was modified and refined.

TA B L E 1
The items in the Chinese version of the RLOC Scale clarity, comprehension and interpretability were discussed. For example, the item "It is often best to just wait and see what happens" was difficult for patients to understand and we discussed how to express it more clearly and comprehensibly. The Chinese version of the RLOC Scale was finalized when no substantial disagreements remained, as shown in Table 1.

| Data analysis
Statistical analyses were conducted using SPSS version 22.0.
Significance levels were set at p-value < .05. Descriptive statistics, including frequencies, percentages, means and standard deviations, were used to summarize the sociodemographic characteristics of the patients. The Shapiro-Wilk method was carried out to test the normality of the RLOC Scale data. The upper and lower 27% rule is commonly used in item analysis based on Kelley's (1939) derivation, and any item with ≥70% of the patients choosing the same extreme response option was considered non-discriminative (Juniper, Guyatt, & Jaeschke, 1996).
Cronbach's alpha and correlated item-to-total correlation coefficients were calculated to determine the internal consistency of the Chinese version of RLOC Scale. Values >0.7 and 0.3 of the respective parameters indicate adequate internal consistency (Nunnally, 1978).
Validity for the Chinese version of the RLOC Scale was evaluated by exploratory factor analysis with Promax. We first conducted the Kaiser-Meyer-Olkin (KMO) test and Bartlett's test to determine whether there were statistically significant correlations among items to perform this analysis.
Additionally, Pearson's correlation coefficients were used to examine the correlations between total RLOC and its subscales. A series of t tests and one-way ANOVA tests were conducted to examine the relationships between sociodemographic variables and the Chinese version of the RLOC Scale.

| Sample characteristics
A total of 233 hospitalized patients with MI were recruited in this study. The mean age of the sample was 61.5 (SD = 12.0) years. Fiftyseven per cent of them were male and 45.9% came from the city.
The primary caregivers were mainly the spouses (48.1%) and children (45.9%). More characteristics of the subjects are presented in Table 2. There were no statistically significant relationships among the sample characteristics and the total score of the Chinese version of RLOC Scale (p > .05).

| The item analysis
Results showed that there was a significant difference in RLOC total score between the upper 27% group and the lower 27% group (t = 26.72, p < .001). Table 3 lists the distribution of the responses for each item in the Chinese version of the RLOC Scale.
None of the items had more than 70% of the patients choosing the same extreme response option. All items were therefore considered discriminative.

| Reliability
The internal consistency coefficient (Cronbach's alpha) for the had satisfactory internal consistency reliability (DeVellis, 2017). All items were above the level of 0.40, which indicated the item's homogeneity in measuring the concept of recovery locus of control (Table 3). The skewness value for all items ranged from −0.08 to −0.57. The Shapiro-Wilk normality statistic for the total score of the Chinese version of RLOC Scale was 0.99 (p = .25), and the total score of the Chinese version of the RLOC Scale and the subscale scores were significantly correlated (p < .01).

| Validity
The exploratory factor analysis showed that the KMO measure was 0.82 and the approximate chi-square value for Bartlett's test was 1,190.61 (df = 36, p < .001). Then, the unweighted least squares (ULS) was used to extract factors and the rotation method Promax was used, as the two factors were significantly correlated. The two factors explained a total of 70.50% of the total variance, and the per cent variances for the two factors were 47.82% and 22.68%, respectively, and were extracted with eigenvalues > 1. The factor loadings of all nine items ranged from 0.58 (item 8) to 0.94 (item 7), as shown in Table 4.  (Nunnally, 1978). These results were better than those found in a previous study with stroke patients (Johnston et al., 1999). By using exploratory factor analysis, all items were found to load on two distinctive factors: IRLOC and ERLOC, as proposed for the original version (Partridge & Johnston, 1989). These results demonstrated that the RLOC was validated in post-MI populations other than stroke patients. The Chinese version of RLOC Scale will be useful instruments to estimate post-MI patients' recovery beliefs, which, in turn, could help health caregivers provide strategies to build self-confidence in the recovery process and engage patients in self-management.

| D ISCUSS I ON
Additionally, the Chinese version of RLOC will help researchers to conduct quantitative study to explore factors related to post-MI patients' recovery locus and provide more specific interventions. Though the Chinese version of the RLOC Scale proved to be a reliable and valid instrument in assessing the RLOC in Chinese patients with MI, there were limitations in the current study. First, recruiting a convenience sample of patients with MI from a level A tertiary hospital may limit the generalizability of the findings to patients in the county hospitals. Moreover, there were no statistically significant relationships among the sample characteristics and the total score of the Chinese version of the RLOC Scale, which might be explained by sampling error. Second, the cross-sectional data collection design and the recruitment of patients soon to be discharged did not allow for the evaluation of the test-retest reliability. Finally, the face validity and the convergent validity of the Chinese version of RLOC Scale were not tested in the current study, which may limit the estimates of validity. In the future, the convergent validity of the Chinese version of the RLOC Scale could be tested by comparing the scale to other self-efficacy or RLOC scales.

| CON CLUS IONS
The Chinese version of RLOC Scale had satisfactory reliability and validity and can be used to measure the RLOC in Chinese patients with MI. This opens up opportunities for further research to develop interventions aiming to improve the RLOC or self-efficacy regarding personal care in patients with MI, with the aim of increasing their involvement in health-related behaviours designed to enhance recovery process, especially in the personal care context.

ACK N OWLED G EM ENTS
We would like to thank the participants of the study and the au-