Plain English summary

Specific patient-rated outcome measurement instruments are important for both clinicians and researchers in order to evaluate wrist/hand function and severity in patients with wrist/hand injuries or disorders. In Thailand, there is still lack of the specific wrist/hand outcome measurement which is valid, has ability to repeat and detect clinically significant changes over time. Patient-Rated Wrist/Hand Evaluation is one of the widely used patient-rated outcome measurement instruments but cannot apply to use in Thai patients without the translation and cross-cultural adaptation processes. The study was conducted to create a Thai version of the Patient Rated Wrist/Hand Evaluation (Thai PRWHE), including standard translation and cross-cultural adaptation. Furthermore, this Thai PRWHE was proved to be valid, had ability to repeat and detect clinically significant changes over time in high numbers of patients who had wrist/hand injuries or disorders.

Introduction

Patient-rated outcome measurements are widely used in the evaluation of healthcare and research systems for patients with musculoskeletal injuries or disorders [1,2,3,4,5,6]. These represent different aspects of outcomes as assessed by patients which have been determined to be valid, reliable, and responsive to minimal clinically important changes. They provide great benefits that differ from traditional evaluations, e.g., range of motion, muscle strength, and radiographic reports, which are subjectively evaluated by a clinical examiner [7,8,9,10].

The Patient-Rated Wrist Evaluation (PRWE) questionnaire was originally a region-specific questionnaire reflecting wrist pain and function that provided reliability, validity, and responsiveness [5, 11]. This patient-based self-report instrument has now been translated into many different languages and has been cross-culturally adapted [10, 12,13,14,15,16,17,18,19,20]. It has been slightly modified for the wrist and hand as “The Patient-Rated Wrist/Hand Evaluation PRWHE” and also includes an esthetic subscale [21].

Some studies comparing the psychometric properties of the PRWE and other self-reported instruments have found that the PRWE is more responsive than the Disability of the Arm, Shoulder, and Hand questionnaire (DASH) and the Short Form-36 (SF36) [21, 22]. Since the PRWHE has fewer items than the DASH, it is considerably more user-friendly [21].

Currently, the cross-cultural adaptation of PRWHE in Thai version has never been reported. Consequently, we aim to develop the valid, reliable, and responsive patient-rated questionnaire to evaluate Thai patients who had wrist/hand injuries or disorders. The purposes of this study are to translate the PRWHE into Thai version and to examine its validity, reliability, and responsiveness.

Methods

This observational study was approved by the local institutional research ethics committee. The study was composed of two stages. The first stage was to translate into Thai and cross-culturally adapt the original PRWHE. The second stage was to investigate the psychometric properties of the Thai version, i.e., the validity, reliability, and responsiveness of the PRWHE questionnaire.

The original PRWE was developed in 1996 by MacDermid et al. as a scoring tool to rate the scale of wrist pain and disability [11]. The validity, reliability, and responsiveness of that questionnaire were assessed in patients who had wrist injuries [5, 22]. The PRWE was modified in 2004 as the Patient-Rated Wrist/Hand Evaluation (PRWHE) for use with both wrist and hand problems [21]. The PRWHE questionnaire consists of 3 subscales: Pain (5 items), Function (6 items for Specific Activities, 4 items for Usual Activities), and Appearance (optional questions) (2 items). Each Pain item is rated from 0 (none) to 10 (worst). Each function item is rated from 0 (no difficulty) to 10 (unable to do). The Total Pain score is the sum of the individual Pain subscale scores (maximum = 50). The Total Function score is the sum of the Function subscale scores divided by 2 (maximum 50). Hence, the total possible score is 100. Higher scores are correlated with more severe pain and disability [5].

Translation and cross-cultural adaptation

After receiving the permission from MacDermid et al. to translate the original PRWHE into a Thai version, the authors adhered to the process of translation and cross-cultural adaptation proposed by the Institute for Work & Health (IWH) which also provided standard recommendations for the cross-cultural adaptation of the DASH and the QuickDASH outcome measures in other languages. The translation and adaptation process consisted of 5 stages as follows [3, 23, 24].

  • Stage 1: Forward translation. Two independent bilingual translators (a Thai Orthopaedic resident and an academic English language lecturer) individualistically prepared forward translations into Thai (T1 and T2 versions) from the original PRWHE [21].

  • Stage 2: Synthesis. The translators discussed translation differences in the two versions and created one common translation (T12). Disagreements regarding translation were discussed to with an Orthopaedic staff member (PA) to arrive at a consensus.

  • Stage 3: Back translation. The T12 version was independently translated back into the original language (BT1 and BT2 versions) by bilingual native English speakers (missionaries) in order to warrant the concepts of original version had been preserved.

  • Stage 4: Review by a committee of experts on hand surgery and measurement development. The pre-final version of the questionnaire was consolidated from all the translated versions (T1, T2, T12, BT1, BT2) by an expert committee consisting of 3 hand surgeons, 1 rehabilitation medicine specialist, 1 psychiatrist, and 2 linguistic experts. They will evaluate the agreement of each item using item-objective congruence (IOC) rating. The scoring system for each item was described as + 1 = clearly measuring, 0 = degree to which it measures the content area is unclear, − 1 = clearly not measuring. The IOC value in each item was assessed by the summation of scores from each expert divided by the number of experts. All items that received an IOC value of below 0.5 were modified and retested until the IOC value was as a minimum of 0.5.

  • Stage 5: Test of the Pre-Final version. 15 healthy volunteers and 15 patients who had wrist or hand injuries were recruited for the field test. All were native Thai speakers who were able to read, write, and understand the Thai language. They were requested to read the instructions and all rating items and to provide the IOC rating for each and were then probed for suggestions. The IOC rating was accepted if the value was at least 0.5. Finally, the pre-final version was translated back into English and referred to MacDermid et al. for their endorsement.

Psychometric testing of the Thai PWWHE

The Thai PRWHE was evaluated for internal consistency, construct validity, test–retest reliability, and responsiveness in patients recruited prospectively at the Orthopaedic Hand Clinic, Academic University Hospital from May 2018 to May 2019. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist [25]. Inclusion criteria were patients who had wrist or hand musculoskeletal disorders, whose first language was Thai and were able read and understand Thai, were between 15 to 70 years old, and could finish the questionnaire without major assistance. Patients who had a musculoskeletal disorder above wrist level, an active cerebral disorder, or communication problems were excluded. After receiving informed consent, patients were registered in study. The rights of the patients were also protected. Demographic data including age, sex, dominant hand, injured hand, and diagnosis were recorded.

Internal consistency

By definition, internal consistency is the degree of interrelationship between the items of the measurement. Cronbach’s alpha, a parameter of internal consistency, was used to evaluate each subscale as well as the total score of the Thai PRWHE. The Cronbach alpha value range is from 0 to 1, with greater values representing higher interrelatedness between items. The values of at least 0.70 were considered adequate [26].

Construct validity

Construct validity is an inter-instrument measure of the correlation between theoretical hypotheses concerning the concepts being evaluated [27, 28]. The Thai version of the Disabilities of the Arm, Shoulder, and Hand (Thai DASH) questionnaire [3, 29] and the Thai EQ-5D-5L [30,31,32] were the alternative tools for evaluating the concepts being measured in this study. Spearman’s rank correlation coefficient (r) was used to evaluate the subscales among the Thai PRWHE, the Thai EQ-5D-5L and the Thai DASH. We hypothesized that the similar or related subscales (Pain and Total Function subscales, Usual and Specific Function subscales in Thai PRWHE, and Pain subscale between the Thai PRWHE, Thai DASH, and the Thai EQ-5D-5L) should have a high correlation, while different or unrelated subscales (Total Function subscale in the Thai PRWHE and the Mobility subscale in the Thai EQ-5D-5L) should present a weak correlation.

Thai DASH

The Thai DASH questionnaire contains 30 self-reported questions to evaluate symptoms and disability status. It also includes two optional modules, a work and sports module and a performing arts module [4]. The Thai version of DASH was translated by Tongprasert et al. [3]. The DASH questionnaire can be categorized into 5 subscales as follows: common activities (20 items, items 1–12 and 16–23), self-care activities (3 items, items 13–15), pain symptoms (2 items, items 24–25), other symptoms including numbness, joint stiffness, weakness, and sleep problems (4 items, items 26–29), and psychological effects (1 item, item 30). The range of scores is 0 to 100, with higher scores indicating greater disability. A resent publication reported that the Thai DASH questionnaire has good psychometric properties for evaluating patients with carpal tunnel syndrome [29].

Thai EQ‐5D‐5L

The EQ-5D-5L is a commonly used instrument for evaluating health status. The original version was translated into Thai and the psychometric properties were tested [30,31,32]. This questionnaire has two parts. The first part comprises 5 subscales including Mobility, Self-care, Usual Activities, Pain/Discomfort, and Anxiety/Depression. Each subscale has five levels of severity ranging from no problems to extreme problems. We used the first part to assess the construct validity of the Thai PRWHE by measuring convergent and discriminant validity. The Thai PRWHE should show good correlation among similar subscales, but weak or zero correlation among the unrelated subscales of the EQ-5D. Correlations were rated as weak (r < 0.25), moderate (0.25 < r < 0.50), good (0.50 < r < 0.75), and excellent (r > 0.75). The second part evaluates general health condition on a scale ranging from 0 to 100. Higher scores indicate a better health condition. However, we considered the second part to be an optional scale.

Construct validity was additionally assessed using Confirmatory Factor Analysis (CFA) to assess the nature of and relationship between latent constructs. Maximum likelihood was used to identify suitable estimators. A hypothesized three-factor model of the Thai PRWHE was analyzed. Indices used for evaluating goodness of fit included the Comparative Fit Index (CFI) ≥ 0.95, the Non -Normed Fit Index (NFI) or Tucker–Lewis Index (TLI) ≥ 0.9, the root-mean-square error of approximation (RMSEA) ≤ 0.6 (0.08 was considered acceptable fit) [33, 34], and c2/df < 3 [35].

Analysis was conducted using Mplus 8 Software (1998–2017, Muthén & Muthén). Modification indices were used after initial analysis and error term correlation was used as indicated [36].

Test–retest reliability

Test–retest reliability measures the reproducibility of a measurement, e.g., the ability to obtain similar results when repeating the same test with a group after a period of time. The recommended time period between the initial and repeat administrations is one week which is long enough to prevent recall and short enough to certify that clinical change has not occurred. In this study, the Thai PRWHE was assessed twice with a 7-day interval during which clinical symptoms remained stable. The value of intraclass correlation coefficients (ICC) ranges from 0 to 1. We accepted coefficients > 0.7 as representing good reliability [28].

Responsiveness

Responsiveness is the ability of an instrument to identify clinically important changes over time [28]. Responsiveness of the Thai PRWHE was assessed by comparing the scores at baseline and at follow-up periods using the standardized response mean (SRM) which is calculated as the observed mean change divided by the standard deviation of the observed change [21, 37]. SRM is the preferred value to use in comparing paired data measurements made at different time points. Values of 0.8, 0.5, and 0.2 were demonstrated as large, moderate, and small, respectively [2].

Statistical analysis

Categorical demographic data are described as frequencies and percentages. Continuous variables are presented as means and standard deviations. Statistical significance was agreed at P < 0.05. For multiple comparisons, the P-value was adjusted using the Bonferroni method. For factor analysis, the minimum sample size should be at least 10 times of number of the items in the questionnaire [38], thus at least 160 patients needed to be included in this study. For test–retest sample size, it was calculated using confidence interval for mean difference between paired samples [39]. Confidence level was set at 95%, the expected standard deviation of the differences (20.3) and the total width of the confidence interval (10.4) were obtained from the related study [17]. The minimum sample size required for the pairs was 61.

Results

Translation and cross-cultural adaptation

Some items were translated with minor cross-cultural adaptation, e.g., the item “Cut meat using a knife in my affected hand” was translated in Thai directly at the first time. However, the IOC rating of this item by the experts was 0.29. Some experts realized that there are various kinds of food in Thailand consist of rice, meat, vegetables and fruit. Therefore, this item was changed to as “Cutting meat, vegetables or fruit with a knife” and the final IOC value was 1. The item “carry a 10 lb. object in my affected hand” was modified to “carry a 5 kg object in my affected hand” as the metric system is the standard in Thailand. A major adaptation occurred in the last optional item “Rate how dissatisfied you were with the appearance of your wrist/hand during the past week” with the answers ranging from “No Dissatisfaction” to “Complete Dissatisfaction.” The IOC evaluation of this item by the experts was 0 because most experts considered that the Thai translation of the double negative phrase “No Dissatisfaction” might result in the respondent misunderstanding the question. The committee of experts decided to change the answers to “Complete Satisfaction” and “Complete Dissatisfaction.” The final IOC evaluation of this item was 1. In the test of the pre-final version process, 30 volunteers (15 healthy volunteers and 15 patients) completed the questionnaire. All items had an IOC value ˃ 0.5, an acceptable level (Supplement 1). The final Thai version was approved by Professor Joy C. MacDermid.

Demographic data of the participants are shown in Table 1. Two hundred and ninety-two patients who had an injury or disorder of the wrist/hand were enrolled in our study. The average age was 47 years. Most were female (66%) and right-handed (89%). The most frequently affected side was the left (47%). The most common causes of injury or disorder were tendon entrapment (30%), wrist fracture (17%), and nerve entrapment (16%).

Table 1 Demographic data of patients (n = 292)
Full size table

Internal consistency and test–retest reliability

The internal consistency of Thai PRWHE in each subscale was evaluated for 292 patients. All subscales [Pain, Specific Function, Usual Function, Appearance (optional)] had high Cronbach’s alpha values ranging from 0.91 to 0.96 (Table 2). After excluding patients who could not follow up after 1 week and patients who could not finish all items, the test–retest reliability was conducted with 61 patients. Each subscale (Pain, Specific Function, Usual Function, Total Function) was found to have excellent reliability with ICC values between 0.94 and 0.96. The ICC of the complete Thai PRWHE was 0.96 (Table 3).

Table 2 Cronbach’s Alpha of subscales in Thai PRWHE (n = 292)
Full size table
Table 3 Test–retest reliability of subscales in Thai PRWHE (n = 61)
Full size table

Construct validity

Construct validity of the Thai PRWHE was evaluated using responses of 292 patients. For each subscale, the correlation between the Thai PRWHE (Table 4) and the Thai PRWE, and Thai DASH and Thai EQ-5D-5L was compared (Table 5). Related dimensions of the Thai PRWHE demonstrated good correlation between Pain and Total Function (r = 0.67, P < 0.0001) and excellent correlation between Usual and Specific Function (r = 0.76, P < 0.0001). Evaluation of the Pain subscale between the Thai PRWHE, Thai DASH, and the Thai EQ-5D-5L showed good correlation between the Thai PRWHE and the Thai DASH (r = 0.55, P < 0.0001) as well as between the Thai PRWHE and the Thai EQ-5D-5L (r = 0.57, P < 0.0001). Excellent correlation (r = 0.77, P < 0.0001) was found between the Total Function subscale of the Thai PRWHE and Common Activities in the Thai DASH. Comparison of the Total Function subscale in the Thai PRWHE and Self-care Activities in the Thai EQ-5D-5L showed good correlation (r = 0.56, P < 0.0001). The only weak correlation (r = 0.09, P < 0.12) was between unrelated dimensions, the Total Function subscale in the Thai PRWHE, and the Mobility subscale in the Thai EQ-5D-5L. All measurements indicated convergent and discriminant validity.

Table 4 Spearman's correlation between Subscales in Thai PRWHE (n = 292)
Full size table
Table 5 Spearman's Correlation between subscales of the Thai PRWHE and (A) the Thai DASH and (B) Thai EQ-5D-5L (n = 292)
Full size table

A three-factor solution fitted the data best when some error terms in each subscale, especially the Pain subscale, were correlated. The goodness of fit statistics were as follows: χ2 = 188.43, df = 78, χ2/df = 2.42, RMSEA = 0.070 (90% CI 0.057–0.082), CFI = 0.963, TLI = 0.950. These results confirm that the Thai PRWHE has three-factor structure as designated.

Responsiveness

Fifty-four patients completed all items in the Thai PRWHE questionnaire both before treatment and at follow-up. The mean follow-up time was 56 ± 27 days. The three most common treatments were steroid injection (34%), surgical fixation (26%), and surgical release (22%). The SRM of Thai PRWHE was large (0.94) indicating high responsiveness (Supplement 2).

Discussion

The aim of this study was to construct a Thai version of the PRWHE through a process of translation and cross-cultural adaptation and to test the psychometric properties of the translation including validity, reliability, and responsiveness.

During the translation and cross-cultural adaptation process, minor modifications to the questionnaire were made, e.g., “Cut meat using a knife in my affected hand” became “Cutting meat, vegetables or fruit with a knife” as numerous kinds of foods in Thailand are cut with a knife. Also, this item was modified as “cutting with a knife with my affected hand” with the same reason in Danish version [18]. The Hindi version modified this item to be “cutting vegetables using a knife in my affected hand” because the majority of patients are vegetarian [13]. Similarly, the Korean version modified this item to “cut food using a knife” since a knife is rarely used to cut meat in Korean culture [40]. The item “carry a 10 lb. object in my affected hand’’ was changed to the metric system as “5 kg.,” the same as in the Hindi, Korean, Brazilian, Turkish, and Spanish versions [12, 13, 20, 40]. A major modification in this version was changing the range of the answer scale from “No Dissatisfaction” to “Complete Dissatisfaction” to be “Complete Satisfaction” to “Complete Dissatisfaction” to avoid potential misunderstanding. This change has not been reported elsewhere.

The Thai PRWHE had Cronbach alpha values ranging from 0.91 to 0.96 for all modules. This result is comparable to that of the Brazilian version which was tested in 61 patients with orthopedic wrist injuries (range 0.85 to 0.92), the Japanese version which was assessed with 117 patients who had a wrist or hand disorder (range 0.86 to 0.94), the Swedish version which analyzed 124 patients who had a wrist or hand injury (range 0.93 to 0.97), the Arabic version which included 48 patients who had hand disabilities (range 0.93 to 0.96) [41], the Turkish version with 166 patients who had a wrist or hand disorder (range 0.79 to 0.92), and the Hong Kong version which included in 47 patients who had a wrist or hand injury (range 0.78 to 0.92) [10, 12, 14, 17, 42]. Internal consistency was also examined in patients who had had a distal radius fracture in the Spanish version (40 patients), the Hindi version (50 patients), the Korean version (63 patients) and the German version (44 patients), all of which demonstrated high values (Cronbach alpha between 0.76 and 0.98 [13, 15, 20, 40]. A Cronbach alpha of 0.8 indicates good internal consistency and 0.9 indicates excellent internal consistency [20, 43]. The Thai PRWHE had excellent internal consistency in all modules. However, this high internal consistency raises some concerns with regard to the duplication of items [44].

In the analysis of test–retest reliability, Thai PRWHE presented excellent correlations in all modules (ICC in the PRWE Pain and Function subscales were 0.96, 0.95, and 0.95, respectively). These results are comparable to the original and other versions (ICC range 0.71 to 0.99). However, the Hindi version showed lower ICC than other versions (PRWE, Pain, and Function subscales were 0.81, 0.76, and 0.85, respectively) [10, 13, 15, 17, 18, 20, 40,41,42, 45,46,47].

Two methods were employed to evaluate the construct validity of the Thai PRWHE. The first method was the correlation of each of the subscales among the common instruments (Thai DASH, Thai EQ‐5D‐5L) which had been previously used to evaluate construct validity in patient-rated questionnaires [10, 12, 20, 29, 40, 42, 46]. The second method was Exploratory Factor Analysis (EFA). The Thai PRWHE showed good correlation in the Pain subscale with the Thai DASH and the Thai EQ‐5D‐5L. These results are in concordance with the Arabic version (PRWHE vs DASH, r = 0.44) [41], the Korean version (PRWE vs DASH, r = 0.63) [40], and the German version (PRWE vs DASH, r = 0.55) [15] for the Pain subscale. For the function subscale, the Thai PRWHE had excellent correlation between Total Function and Common Activities (r = 0.77) as well as between Total Function and Self-care Activities (r = 0.7261) which relates with other versions, e.g., the Korean version (r = 0.76) [40]. The German version and the Arabic version demonstrated good correlation between the Function subscales of the PRWHE and the DASH (r = 0.55 and r = 0.66, respectively) [15, 41]. We expected to find a weak correlation between the different aspects of the module. In fact, a weak correlation was found not only between the Total Function subscale of the Thai PRWHE and the Mobility subscale in the Thai EQ-5D-5L but also between the Pain subscale of the Thai PRWHE and the Mobility subscale of the Thai EQ-5D-5L. This concept was applied as described in a previous study [29]. CFA results demonstrated that the items included in the designated subscales of the PRWHE clearly indicated a three-factor model, which is consistent with other related studies [14, 17, 48]. However, these results raise concern regarding using the PRWHE as a total score because of its dimensionality and the assumption that it represents “pain and disability.” A similar concern was mentioned in a previous study [49]. A subscale analysis is generally favored over a total score, especially when the changes of the scores are targeted.

Analysis of the responsiveness of the Thai PRWHE showed a large standardized response mean (SRM = 0.94) which was comparable to the SRM of the Thai DASH (0.91) (Supplement 2). The level of responsiveness of the PRWE was dependent on the type of injury, type of treatment and the period of time before follow-up. As a result, the responsiveness of the different versions of the PRWE were varied. The original PRWE reported an SRM of 2.27 at 3 months and SRM of 0.74 at between 3 and 6 months following various kinds of treatment in 59 patients with a distal radius fracture [22]. The Swedish version had an SRM of the PRWE of 1.4 after 6 months of follow-up with 50 patients who had a distal radius fracture, the majority of whom were treated with casting [16]. The reported SRM of the PRWE in 50 patients who had a wrist or hand disorder in the Japanese version was 1.55 at 3 months after open surgery [17]. An SRM of 0.89 at 3 and 6 months postoperatively in 63 patients with a distal radius fracture who underwent plate fixation was reported for the Korean version [40].

The present study has multiple strengths. First, the number of patients in this study is large (n = 292), adequate for psychometric properties testing. Second, many psychometric properties, including validity, reliability, and responsiveness, were examined in the same cohort in the Thai PRWHE while other publications reported only some properties [10, 12, 14, 18,19,20]. Third, we have excluded questionnaires which had missing values. That suggests the results are more reliable than if missing values had been dealt with by using multiple imputation or if missing values had been replaced with mean values. Fourth, this cohort includes many types of wrist/hand injuries or disorders, so the results of the psychometric testing can generally be applied with patients having various different types of wrist/hand injuries or disorders.

Limitations of the study include the following. First, in the test–retest reliability and responsiveness evaluations, patients who were not followed up or who did not finish the entire questionnaire were excluded; nevertheless, the number of patients remaining was still comparable to prior publications [5, 10, 13, 15, 18, 20, 40, 41, 47]. Second, the responsiveness to treatment for each diagnosis was not reported for the cohort since our objective focused on generalized wrist/hand injuries or disorders. To evaluate the responsiveness with specific diagnoses, treatments and follow-up periods would be preferred in further study. Finally, the Thai PRWHE should be further examined using item response theory, e.g., the Rasch measurement model, to demonstrate further information on both persons and item calibration.

The Thai PRWHE created in this study exhibits excellent internal consistency in all modules as well as good construct validity and reliability for patient-rated outcome measurements in Thai patients with wrist/hand injuries or disorders and provides a large standardized response mean after treatment. In clinical practice, this Thai PRWHE would be one of the valuable outcome measurements for evaluation of the functioning of the wrist/hand.