Cross-Cultural Adaptation and Validation of the Malay Satisfaction Questionnaire for Osteoporosis Prevention in Malaysia

Introduction

Patient satisfaction is an important parameter to assess the quality of healthcare service delivery as it provides feedback and evaluation of services experienced by patients, ensures accountability among healthcare professionals and continued improvement in service delivery to meet patient needs.^1–3 Research has linked osteoporosis treatment effectiveness to patient satisfaction for postmenopausal women; as women showed better persistence to their osteoporosis treatment when satisfied.^4,5 Studies show that satisfied patients experience better patient-healthcare provider relationships, use health services more consistently, and register an improvement in clinical outcomes due to increased adherence to medical advice, treatments and medications.^1–3

Osteoporosis is underdiagnosed due to its asymptomatic nature and affects 1 in 4 women aged above 45 years in Malaysia.^6–8 Osteoporosis can have debilitating consequences when it causes an osteoporotic fracture. Patients suffer from pain, loss of mobility and independence, and increased mortality. In addition, health costs increase drastically due to hospitalization and rehabilitation.^6,8–10 Malaysia is projected to have the highest increase in hip fracture incidence when compared to other Asian countries due to its ageing population.^11–13 Hence, osteoporosis poses a serious medical concern for Malaysian healthcare.

Prevention is a key method in managing osteoporosis and fragility fractures.¹⁴ Prevention relies heavily on a patient’s motivation to frequent osteoporosis services provided, self-manage anti-osteoporotic medications and adhere to lifestyle advice on diet, supplementation and exercise.^7,15,16 It is therefore important to ensure that patients receive osteoporosis services that they are satisfied with so they continue to engage with their healthcare providers, attend clinic appointments and comply with treatment.

Currently, there are three tools that assess patient satisfaction in relation to osteoporosis including the Osteoporosis Patient Treatment Satisfaction Questionnaire (OPSAT-Q),¹⁷ and the Preference and Satisfaction Questionnaire (PSQ),¹⁸ which were developed in the United States, as well as the Osteoporosis Patient Satisfaction Questionnaire (OPSQ) which was developed in Malaysia.¹⁹ These osteoporosis satisfaction tools were developed in English and focus mainly on assessing patient satisfaction towards osteoporosis and its treatment. The SQOP however, focuses on evaluating patient satisfaction towards osteoporosis screening and prevention services in Malaysia.²

The SQOP was developed and validated in English to assess patient satisfaction towards osteoporosis screening and prevention services provided by pharmacists in Malaysia.² It was modified from the OPSQ and developed based on findings from a qualitative study assessing barriers and needs for an osteoporosis screening and prevention service in Malaysia.¹⁹ The SQOP can be used in any clinical setting.^2,19 However, in Malaysia, osteoporosis-related services are usually provided in hospitals with hospital pharmacists involved in providing osteoporosis-related patient education.^2,19

As Malay is the national language of Malaysia and widely spoken amongst Malaysians, a need was identified to cross-culturally adapt and validate the SQOP in Malay language. Additionally, this study shows global significance as Malay is an Austronesian language widely spoken by more than 200 million people in the South East Asian region.²⁰ Therefore, Malay-speaking patients will be able to use this instrument to evaluate their satisfaction regarding osteoporosis screening and prevention services received. This will allow healthcare professionals in this region to continuously improve the quality of osteoporosis screening and prevention services provided, tailoring it to patient needs and preferences. Overall, this may lead to better engagement from osteoporotic patients as they are more satisfied with the services provided. This study also outlines clear processes for conducting a validation study for the SQOP which can be replicated easily for researchers in other countries to validate the SQOP for use in their native regions and languages. This will allow for osteoporosis prevention and screening services to be assessed globally using the SQOP measure.

Patients and Methods

Aim of the Study

This study aimed to cross-culturally adapt and validate the SQOP in the Malay language (SQOP-M) among postmenopausal women in Malaysia. This study was divided into two phases: the cross-cultural adaptation of the Malay Satisfaction Questionnaire for Osteoporosis Prevention (SQOP-M) and its validation.

Phase 1: The Cross-Cultural Adaptation of the Satisfaction Questionnaire for Osteoporosis Prevention (SQOP) from English to Malay

During the content validity process, the SQOP was translated from English to Malay according to guidelines by The Professional Society for Health Economics and Outcomes Research (ISPOR).^21,22 This process is illustrated in Figure 1.

Figure 1 Translation process following the ISPOR guidelines.

The translated versions were reviewed and reconciled by an expert panel consisting of a doctor and three pharmacists with clinical experience who were fluent in English and Malay. The expert panel ensured that the meaning of the instrument was maintained during the forward (versions 1a and 1b) and backward (versions 3a and 3b) translation process. The final version (version 4) of the Malay SQOP (SQOP-M) was piloted amongst 10 postmenopausal women who spoke and understood Malay to assess for face and content validity.²² They were asked for their views regarding the phrasing, formatting and content of the instrument but did not encounter problems in answering the questionnaire.²² Hence, version 4 was finalized.

Phase 2: The Validation of the Malay Satisfaction Questionnaire for Osteoporosis Prevention (SQOP-M)

This prospective validation study was conducted from March 2018 to October 2018 at a primary care clinic located within a tertiary hospital in Kuala Lumpur, Malaysia.

Patients

Postmenopausal women aged ≥50 years who understood Malay were recruited. For the purpose of this study, postmenopausal was defined as having experienced a period of 12 consecutive months without menstruation after menopause. Patients were required to self-report and confirm if they were postmenopausal when screened by researchers. Any patient with a prior diagnosis of osteoporosis or who had a recent bone mineral density (DXA) scan (<2 years) were excluded.

Randomization

Patients were randomized into two groups (control and intervention) to assess the discriminative validity of the SQOP-M. They were asked to draw a piece of paper from an envelope containing pieces of blue and pink paper. If the patient selected a blue piece of paper, she was allocated to the control group, whilst those selecting the pink piece of paper were allocated to the intervention group.

Sample Size

Sample size was calculated based on a 10:1 patient ratio to calculate factor analysis.²³ The SQOP-M consisted of 23 items. Therefore, the total number of patients required was 230. This would include 115 patients in the control group and 115 patients in the intervention group.

Instruments Used

A baseline demographic questionnaire was used to assess baseline demographics such as age, weight, education and income level for both the control and intervention group.

The Malay Satisfaction Questionnaire for Osteoporosis Prevention (SQOP-M)

The SQOP-M consists of two domains: clinical services (18 items) and types of counselling (5 items). Each item consists of a statement that may be answered using a 5-point Likert scale. One indicated the lowest satisfaction with that item whereas five indicated the highest satisfaction with that item. The SQOP-M score was calculated by summing the scores of all items, and converting to a percentage from 0–100. The scores from each domain were also analyzed and summed up to provide the final domain score. A higher score indicated higher patient satisfaction with the osteoporosis services provided.

Intervention

The intervention consisted of patients receiving an osteoporosis prevention information booklet in Malay language as well as a 15-minute pharmacist-provided counselling session using the osteoporosis prevention information booklet which covered different lifestyle measures to prevent osteoporosis (dietary, supplementation, exercise and fall prevention strategies).²⁴ Patients were also able to clarify any further questions they had regarding osteoporosis and its prevention after the counselling session.

Control patients received standard care which involved a regular visit to the doctor. Any counselling provided by the doctor would be at the doctor’s discretion.

Procedure

Patients were approached individually by the researcher using the convenience sampling method whilst they were waiting to see their doctor. The convenience sampling method was used to accommodate for large volumes of patients in the primary care clinics and inconsistency of waiting times. The scattered seating layout of the primary care clinic with five different waiting areas across two floors also made it difficult to follow a systematic sampling process to approach patients at fixed periodic intervals due to difficulty in locating patients.

The purpose of the study was explained to patients who fulfilled eligibility criteria using the participant information sheet. For those who agreed to participate, written informed consent was obtained. Patients were informed that they could withdraw at any point if they felt uncomfortable or did not wish to continue. Baseline demographic information was obtained from the patients. Patients were then randomized to either the control or intervention group. The study procedure is further illustrated in Figure 2.

Figure 2 Randomized controlled trial process.

The SQOP-M questionnaire was administered to all patients at baseline. However, the intervention group patients received the intervention before the SQOP-M was administered at baseline. For patients who had difficulty reading the questions, the researcher assisted the patients in answering SQOP-M questions. The completed questionnaires were checked by the researcher to ensure that all questions had been answered.

The SQOP-M questionnaire was administered again two weeks later via phone for test-retest reliability assessment. The questionnaire was retested after two weeks as this time interval is accepted to be long enough so participants are unable to recall their original responses while still retaining knowledge of the subject they were tested on.^25,26 The control group received the intervention after the study had ended to ensure equity among both groups.

Ethics Approval

This study was approved by the Medical Research Ethics Committee of University Malaya Medical Centre (MREC ID NO: 2018115–5959) and the Medical Research Ethics Committee, Ministry of Health Malaysia (NMRR-18-1302-40112). This study was conducted in accordance with the Declaration of Helsinki.

Data Analyses

All data was entered into the IBM Statistical Package for the Social Sciences (SPSS) version 21 (IBM Corp, Armonk, NY). As normality could not be assumed, non-parametric tests were used. Continuous data was presented as median and interquartile range (IQR), whilst proportions were presented as number and percentage.

Factor Analysis

Confirmatory factor analysis (CFA) is used to confirm or reject hypothesized relationships based on theoretical constructs that are well understood for existing instruments, while exploratory factor analysis (EFA) is used to explore the patterns of a new tool to understand how different items and constructs relate to one another.²⁷ EFA was performed to explore the underlying relationships between the items as well as the appropriateness of the factor structure in the SQOP-M, as it is a new tool.^28,29 The factors were extracted using the principal component analysis method and rotated using the maximum likelihood and promax methods.^2,30 Factors with an eigenvalue of greater than 1 were retained.^2,30

Validity

Discriminative validity was assessed to determine if the SQOP-M could differentiate between responses from the control group and the intervention group. We hypothesized that the responses provided by both groups would be different, with intervention group scoring higher than the control group. To assess discriminative validity between both groups, the chi square test was used on categorical data of the individual items while the Mann Whitney U-test was used for continuous data of individual items and total domain scores.³¹

Reliability

The internal reliability of the SQOP-M was assessed using Cronbach alpha because it is used to measure internal consistency for measures with Likert scales.^32,33 Cronbach alpha values can be interpreted as: ≥0.80- high internal reliability, ≥0.51–0.79- adequate internal reliability, ≤0.50- low internal reliability.^34,35

Test-retest was assessed using the kappa measure of agreement (to define inter-rater reliability).³⁶ Kappa values were assessed as the following: 0.01–0.20- slight agreement, 0.21–0.40- fair agreement, 0.41–0.60- moderate agreement, 0.61–1.00- substantial to almost perfect agreement.^37,38 Wilcoxon signed rank test was used to assess continuous data of individual items and total domain scores for a difference between the patient group scores at test and retest.

Results

A total of 230/348 patients agreed to participate in this study (response rate=66.1%) (control group=115, intervention group=115). At retest, 82 (20.0%) patients could not be contacted. Hence, only 184 patients were included at retest (control group=98, intervention group=86, response rate=80.0%).

The demographic characteristics of the patients in this study are displayed in Table 1. Patients in both control and intervention groups shared the same median age of 64 years and similar characteristics for weight and income level. However, the level of education between both groups was significantly different with the intervention group having a far higher number of patients who were diploma holders or tertiary-level educated (50.5%) in comparison to control group (32.2%).

Table 1 Baseline Demographic Characteristics of Participants for the Malay Satisfaction Questionnaire for Osteoporosis Patients (SQOP-M)

Psychometric Properties of the Malay Satisfaction Questionnaire for Osteoporosis Prevention (SQOP-M)

Factor Analysis

The factor loading of items in the SQOP-M is shown in Table 2.

Table 2 Factor Loadings of Items in the SQOP-M

EFA extracted four domains with a criterion of eigenvalue > 1.0. These were similar to the domains extracted for the English SQOP.² The factor loading of the items ranged from 0.019–0.950. Three items had a factor loading of <0.3.

Item number 17 in the SQOP-M was excluded from factor loadings calculation as it was an optional question. Patients were asked in item number 16 if they would like to pay for a pharmacist service in osteoporosis prevention counselling. Item number 17 was then provided as an optional question to those that answered “Yes” for item number 16, where patients were allowed to select how much they would like to pay for a service like this. Many patients did not answer this question as they felt that this pharmacist service should be free.

Validity

The discriminative validity of the SQOP-M is shown in Table 3.

Table 3 Discriminative Validity of Satisfaction Scores of the Control and Intervention Groups at Test for SQOP-M

The overall total satisfaction score for patients in the intervention group was significantly higher than the control group [76.9 (47.6–53.9) vs 50.4 (47.6–53.9), p<0.001]. The intervention group scored significantly higher than the control group in all domains except the accessibility/convenience domain where the satisfaction scores were the same for both groups. The intervention group also scored significantly higher for 13/23 items of the SQOP-M.

Hence, similarly to the English SQOP, the SQOP-M was able to discriminate between a group with higher satisfaction and lower satisfaction. It also demonstrated that the intervention had an effect on the patient’s satisfaction levels.

Reliability

The reliability of the SQOP-M is shown in Table 4.

Table 4 Reliability of the SQOP-M

Cronbach’s alpha was calculated for each domain which ranged from 0.230 to 0.938. Five out of 30 items had corrected item-total correlations <0.2. The Kappa measurement of agreement values ranged from 0.124–0.627, where 2/23 (8.7%) items were in substantial agreement, 8/23 (34.8%) items were in moderate agreement, 7/23 (30.4%) items were in fair agreement and 5/23 (21.7%) items were in slight agreement. Four out of 23 items had Wilcoxon signed rank test values that were statistically significant, p<0.005. As item 17 was an optional question, it was excluded from the analysis.

Discussion

The SQOP-M had adequate psychometric properties and was able to significantly discriminate between the satisfaction scores of the patients in the control and intervention groups. The SQOP-M also performed similarly to the English SQOP.²

Demographically, the patients for both studies showed similar distribution in weight characteristics though SQOP-M participants were older with a median age of 64 years while patients for SQOP had a median age of 56 years.² Patients in the intervention group for SQOP-M also had significantly higher education levels than control groups. Previous research indicates that older patients tend to be more satisfied while highly educated patients tend to be less satisfied with the health-care services provided.^39–43 However, in this study regression analyses were not performed to show a link between variable and outcome as the aim of the study was to cross-culturally adapt and validate the SQOP-M.

EFA showed that there were four domains of satisfaction being measured within the SQOP-M to assess patients’ satisfaction towards osteoporosis prevention services in Malaysia. This differed to the SQOP which had been initially designed to extract seven domains.² However, the SQOP-M extracted similar domains to the SQOP which were deemed significant based on themes from a previous qualitative study that the SQOP was developed to address.⁴⁴ In comparison to domains measured by other satisfaction tools, there were similarities in the domains of convenience. However, it was difficult to compare all these tools as the premise of the SQOP-M was to measure satisfaction at osteoporosis prevention instead of treatment. The SQOP-M was also meant to mirror the SQOP, therefore no changes were made. Overall, The SQOP-M performed satisfactorily in the EFA indicating that it is suitable to assess patients’ satisfaction towards osteoporosis prevention services in Malaysia together with the SQOP.

The intervention group scored significantly higher than the control group for all items of the SQOP-M except in four items (item 2=During the session, what did you think about the time given to discuss your problems with the pharmacist?; item 4=How would you rate the comfort of the location?; item 5=If you have questions about osteoporosis, would you ask the pharmacist?; and item 16=Would you pay for a pharmacist counselling service?), which could be due to several reasons. Many patients expressed positively that they were satisfied at receiving counselling from the pharmacist but felt pharmacists should play a more prominent role in providing counselling sessions to improve their knowledge on health. However, patients still felt more comfortable discussing their health-related issues with their doctor as they were not used to having these types of conversations with a pharmacist whose core role was thought to be in supply of medications. Patients also expressed that having greater privacy and less background noise would have improved their satisfaction experience overall. A majority of patients also felt that counselling on health is an important service that should be provided freely to all. Overall, total satisfaction score for patients in the intervention group was significantly higher than the control group indicating that the SQOP-M was able to perform satisfactorily to discriminate between higher and lower satisfaction scores.

Cronbach’s alpha values were lowest for the “interpersonal relationship” and “accessibility/convenience” domains as these were the smallest domains containing less than four items. For the SQOP, the Cronbach alpha for the interpersonal relationship domain had not been calculated as it was considered too small.² Five items with corrected item-total correlations <0.2 were retained because excluding any of these items did not improve the overall Cronbach’s alpha value significantly. Overall, these domains and items were retained to maintain the similarity between the SQOP-M and the SQOP.

The Kappa measurement of agreement values were in slight agreement for 5/23 items. Three out of these items had Wilcoxon signed rank test values that were statistically significant, p<0.005 as patients were not “satisfied” that the SQOP-M was administered in a noisy clinic waiting room. Providing a private, consultation area or allocating a booth to the pharmacist for counselling purposes may result in improved patient satisfaction.^45–47 During administration of SQOP-M at both test and retest, patients expressed preference for discussing health issues with a doctor rather than a pharmacist which may indicate a need for pharmacists in Malaysia to embrace more patient-centered roles with a clinical focus at a primary care level in order to increase profile recognition amongst community-dwelling patients.^48,49 Patients may have been less satisfied with counselling on items 20 and 22, as the emphasis of the counselling session was on osteoporosis prevention. Hence, future counselling sessions should reflect this by educating patients on all aspects of osteoporosis equally. Overall, the SQOP-M performed satisfactorily at test–retest, indicating that the SQOP-M achieved stable reliability.

A limitation of this study as highlighted in the SQOP is that it was specifically designed to measure the satisfaction of patients towards an osteoporosis prevention service conducted by a pharmacist.² As a result, there are six items within the questionnaire worded specifically to assess the satisfaction of a pharmacist providing the service. No changes were made to the SQOP-M as it was meant to mirror the SQOP. However, minor modifications may need to be made to both the SQOP and the SQOP-M if it is used to assess osteoporosis prevention service satisfaction provided by other health care professionals.² This study was also administered using different modes where patients answered the SQOP-M themselves at test but it was later readministered over the telephone at retest.⁵⁰ This was carried out to minimize patient inconvenience in presenting physically to the clinic multiple times to answer the questionnaire as well as to maximize response rates.⁵⁰ The patients may have answered the items differently due to different modes of administration. However, as this would have been applicable to all participants in the study, its impact on the validation process would be negated.^26,50,51

Conclusion

The Malay Satisfaction Questionnaire for Osteoporosis Prevention (SQOP-M) was found to be a reliable and valid instrument to assess patient satisfaction on osteoporosis prevention services amongst Malay-speaking patients in Malaysia. It can be used together with the English version of the SQOP to assess the effectiveness of osteoporosis-related prevention services amongst the wider Malaysian population.