https://orcid.org/0000-0002-7950-8979
Figures
Abstract
Objective
This study aimed to adapt and validate an Evidence-Based Practice (EBP) tool for healthcare providers in Hong Kong, addressing the need for reliable and culturally relevant tools.
Methods
The original EBP measure, developed for recent Canadian physiotherapy and occupational therapy graduates, was systematically adapted for use by healthcare providers in Hong Kong following established cross-cultural adaptation guidelines. The adaptation process consisted of two phases: Phase I involved forward and backward translation, expert panel review, and cognitive debriefing interviews with a purposive sample of 36 healthcare providers to ensure linguistic and cultural relevance. In Phase II, the finalized instrument was administered to a convenience sample of 248 registered healthcare providers in Hong Kong. The structure of the adapted instrument remained consistent with the original, comprising two models encompassing six constructs: (1) a formative model (use of EBP and EBP activities) and (2) a reflective model (knowledge, self-efficacy, attitudes, and resources). Construct validity was assessed using the Rasch model, internal consistency reliability was evaluated with the Person Separation Index (PSI), and differential item functioning (DIF) was examined.
Results
The formative model only required linguistic modifications. The Rasch model was applied to the reflective component. For knowledge, 45% (5/11) items fit the Rasch model with a chi-square fit statistic (χ² = 17.90, p = 0.268; PSI = 0.84). For self-efficacy, 89% (8/9) items fit the Rasch model with χ² = 26.48, p = 0.33; PSI = 0.93. The attitudes construct was divided into positively and negatively worded subscales due to multidimensionality. The resources construct showed a good fit with χ² = 26.60, p = 0.49; PSI = 0.84. DIF was not observed in the final measures.
Conclusions
The adapted EBP measure demonstrated evidence of construct validity and internal consistency reliability among healthcare providers in Hong Kong. However, further research is needed to assess additional aspects of validity, such as test–retest reliability and responsiveness, as well as to evaluate its applicability in broader healthcare settings.
Citation: Al Zoubi FM, Bussières A, Cheung JPY, Chu EC-P, Abu-Odah H, Wong AYL, et al. (2026) Cross-cultural adaptation of evidence-based practice measure among Hong Kong healthcare providers. PLoS One 21(6): e0351754. https://doi.org/10.1371/journal.pone.0351754
Editor: Majed Sulaiman Alamri, University of Hafr Al-Batin, SAUDI ARABIA
Received: May 18, 2025; Accepted: June 1, 2026; Published: June 26, 2026
Copyright: © 2026 Al Zoubi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data underlying the results presented in the study are available in the paper, tables, figures, and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Evidence-Based Practice (EBP) is a systematic approach to healthcare that integrates the best available research evidence with clinical expertise and patient preferences to inform care [1]. It involves formulating precise questions, searching for relevant evidence, critically appraising that evidence, applying it in the clinical context, and evaluating outcomes to improve future practice [2]. In this study, the term “measure” refers to a validated questionnaire or psychometric tool designed to assess various EBP-related competencies and perceptions among healthcare providers, rather than an audit or compliance indicator.
EBP is a key competency for healthcare professions [3,4] such as nursing, physical therapy, and occupational therapy and is integrated into entry-level curricula. While the integration of EBP is widely recognized as a core competency in healthcare professions internationally [4–8], with explicit expectations articulated in the professional codes of conduct and practice of international federations, confederations, and regulatory councils, the situation in Hong Kong is less clear. To our knowledge, current professional codes of conduct for healthcare providers in Hong Kong do not explicitly require the implementation of EBP. This absence of a formal mandate may influence healthcare providers’ engagement with EBP and their responses to EBP-related measures. It also underscores the importance of culturally and contextually relevant assessment tools, as local professional norms and regulatory frameworks can significantly shape both attitudes toward and the adoption of EBP.
Research on how best to teach and evaluate EBP remains evolving [9–12]. Reviews of EBP educational interventions have shown that critical appraisal and the application of evidence in practice are often underemphasized [9–12]. In parallel, an umbrella review identified 204 distinct EBP measures, but many lacked thorough psychometric validation and failed to encompass the entire EBP process [13]. Existing tools tend to focus on selected domains rather than providing a comprehensive assessment of EBP [14], including both individual competencies and organizational supports [9,10,12]. These limitations highlight the need for a more robust and contextually relevant EBP measure.
Study context
To address these measurement gaps, we previously developed a validated 68-item EBP measure in English and French for recent Canadian physiotherapy and occupational therapy graduates [15]. The measure includes both formative and reflective model constructs and captures individual as well as organizational dimensions of EBP. Although it has proven reliable and valid in its original context, cross-cultural adaptation is necessary before use in other cultural and professional settings [16].
We recently conducted a cross-cultural adaptation of the measure for use with Hong Kong physiotherapy students, demonstrating its reliability and validity in capturing both individual and organizational dimensions of EBP [17]. However, healthcare providers often differ significantly from students or recent graduates in terms of clinical exposure and professional responsibilities. Therefore, there is a need to expand the validation of the EBP measure to a broader group of healthcare providers in Hong Kong, including physiotherapists, occupational therapists, medical doctors, chiropractors, and nurses. This broader approach enables future comparisons of EBP competencies across different healthcare professions and better reflects the multidisciplinary nature of clinical practice in Hong Kong. Accordingly, the current study aimed to cross-culturally adapt the EBP measure for Hong Kong healthcare providers and rigorously evaluate its psychometric properties, ensuring its applicability across a broader spectrum of professional experience and practice contexts.
Materials and methods
This study adhered to the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) checklist to ensure methodological rigor and transparency in reporting [18]. Ethical approval was obtained from the Institutional Review Board of The Hong Kong Polytechnic University (reference number: HSEARS20201208001). All participants gave electronic informed consent.
Design
This study used a cross-sectional validation design to adapt and evaluate the psychometric properties of the Canadian EBP measure for use among Hong Kong healthcare providers. The research followed a two-step procedure. The first step focused on the cross-cultural validation of the EBP measure. This involved first translating the measure into Traditional Chinese, followed by conducting focus groups and pilot testing to assess and refine its linguistic clarity and cultural appropriateness for healthcare providers in Hong Kong. In the pilot-testing and focus meetings, we recruited medical doctors, registered nurses, physiotherapists, and chiropractors. The second step entailed the cross-sectional administration of an e-survey to a sample of practicing Hong Kong providers to assess the psychometric properties of the translated EBP measure, with a particular emphasis on evaluating structural validity and internal consistency reliability. Data collection was conducted in two phases. Phase I (linguistic and cultural validation) took place from March 29, 2021, to June 28, 2021, while Phase II (psychometric validation) was conducted from June 29, 2021, to March 28, 2022.
EBP measure
We adapted the EBP measure originally developed in English and validated among Canadian occupational therapy and physiotherapy graduates [15]. Since Hong Kong healthcare providers are trained and practice in English, the Canadian version was considered suitable for adaptation. However, it is important to note that while English is widely used in Hong Kong, there are differences in vocabulary, usage, and cultural context compared to Canadian English. Therefore, the measure was further reviewed and adapted to ensure appropriateness for Hong Kong English and local clinical practice. The measure consists of six constructs—use of EBP, EBP activities, knowledge, self-efficacy, attitudes, and resources—encompassing a total of 68 items. Each construct is described in Table 1.
Phase I: Linguistic and cultural validation
To ensure cultural relevance and linguistic clarity, the measure was adapted from North American English to Hong Kong English, following guidelines for cross-cultural adaptation [16]. We utilized a six-step adaptation process:
- i. Forward translation: Three local translators, fluent in Hong Kong English, independently reviewed the original North American English EBP measure. Their task was to identify and adjust terms or phrases to enhance comprehension among local healthcare providers. One translator had a background in physiotherapy with 15 years of experience and familiarity with the research topic and EBP questionnaires, which provided subject-matter insight into the conceptual meaning of the items. The other two translators did not have medical backgrounds, and they have degrees in English translation with 2 and 10 years of experience and were chosen to provide a lay perspective and to suggest alternative phrasing to improve clarity and understanding. All translators were bilingual in English and traditional Chinese and were selected based on language proficiency and relevant experience. Each translator first reviewed the measure independently, after which they discussed their suggestions as a group.
- ii. Translation synthesis: After the independent review, the three translators convened for a synthesis meeting. During this meeting, each translator presented their suggested translations and the rationale behind their choices. The group systematically compared each item, discussed any differences in wording or interpretation, and collaboratively resolved discrepancies through discussion and consensus. If disagreements arose, the group considered the clarity, cultural appropriateness, and intended meaning of each item in order to reach a decision. A physiotherapy professor (AW) acted as the synthesis recorder and facilitator, guiding the discussion, ensuring that all viewpoints were considered, and mediating any unresolved disagreements. The outcome of this process was a single, harmonized draft version of the EBP measure that reflected both professional and lay perspectives.
- iii. Back translation: This step was deemed unnecessary, as the adaptation involved transitioning between two forms of English rather than distinct languages.
- iv. Expert committee review: An expert committee, comprising the forward translators, the synthesis recorder (AW), two North American English-speaking EBP experts (AT, AB), and an experienced researcher in outcome measures and EBP (FAZ), reviewed the draft. During two separate meetings, the committee evaluated the measure for semantic and idiomatic accuracy, ensuring conceptual and experiential equivalence. The revised preliminary version of the measure was approved by all members of the expert committee.
- v. Test of the pre-final version: This pre-final English Hong Kong version was tested with a purposive sample of healthcare providers, including physiotherapists, occupational therapists, medical doctors, chiropractors, and nurses. In line with established recommendations for cross-cultural validation of measurement instruments [16], 30–40 participants were invited to review the measure for clarity, relevance, and cultural appropriateness. The sample was selected to ensure representation across a range of healthcare professions that would use the EBP measure in practice. Selection criteria included being an adult (18 years or older), having graduated from a healthcare program, and being registered with the local licensing board. Participants were contacted via email or professional networks. Email contact details were obtained through the researchers’ existing professional networks and contacts within the local healthcare community. Upon providing informed consent, participants were first asked to independently complete the pre-final version of the EBP measure. Following completion, a trained research assistant with 7 years of experience in quantitative and qualitative research as a psychologist conducted a semi-structured cognitive debriefing interview with each participant, either in person or via videoconference. Before data collection, the research assistant received study-specific training from the research team on informed consent procedures, administration of the semi-structured interview guide, and techniques for conducting cognitive debriefing interviews consistently and neutrally. During the interview, participants were asked to rate each item’s relevance to EBP on a 5-point Likert scale (ranging from ‘not relevant at all’ to ‘extremely relevant’) and to assess the comprehensibility of each item on a 5-point scale (from ‘very difficult to understand’ to ‘very easy to understand’). Participants were also encouraged to discuss any items they found unclear, confusing, or culturally inappropriate and to suggest alternative wording where necessary. This process followed established recommendations for cross-cultural validation [16], ensuring both quantitative and qualitative feedback was systematically collected to inform further refinement of the measure. We recognize that participants’ judgments of item relevance during cognitive debriefing may have been influenced by their own level of EBP knowledge and confidence. To mitigate this, we recruited a diverse sample and instructed participants to consider the broader relevance of each item to the corresponding EBP construct in their profession. The expert panel also reviewed all feedback to ensure that item retention was guided by both participant input and the conceptual framework of EBP. Interviews were audio-recorded, transcribed verbatim, and systematically analyzed using deductive thematic analysis [19] to identify issues related to the measure’s relevance, comprehensiveness, and cultural appropriateness. Key themes and participant suggestions were extracted and summarized.
- vi. Final version and appraisal of the adaptation: The summarized findings from the cognitive debriefing interviews —including both quantitative ratings and qualitative themes— were presented to the expert panel for review. The panel discussed the results and made recommendations for final adjustments to the measure. Based on these recommendations, necessary revisions were incorporated. The finalized Hong Kong English version was then sent to the original North American developers for review and approval, ensuring its alignment with the intended purpose and conceptual framework of the EBP measure.
Phase II: Psychometric validation
The psychometric properties of the EBP measure were rigorously evaluated in accordance with established guidelines [20–22]. Structural validity was specifically assessed using Rasch analysis, a robust statistical method for analyzing ordinal data and ensuring the measure’s unidimensionality and reliability [15].
Participants.
A convenience sampling approach was used to recruit a diverse group of healthcare providers in Hong Kong. Eligible participants were registered healthcare professionals—including physiotherapists, occupational therapists, medical doctors, chiropractors, nurses, and other allied health providers—who had attended institutions where English was the primary language of instruction. Recruitment was conducted through professional bodies and associations, such as the Hong Kong Physiotherapy Association and Chiropractic Doctors’ Association of Hong Kong, along with postings in academic departments (e.g., Department of Rehabilitation Sciences at The Hong Kong Polytechnic University and Department of Orthopaedics and Traumatology, The University of Hong Kong) and clinics. Invitations were distributed via email lists, newsletters, and online platforms managed by these organizations. Inclusion criteria were (1) currently registered and practicing in Hong Kong during the study period (2021–2022) and (2) able to read and understand English. There were no restrictions regarding years of experience, clinical specialty, or workplace setting. Following methodological guidelines, a minimum sample size of 200 participants was targeted to ensure sufficient statistical power for Rasch analysis to yield stable and reliable estimates of the measure’s psychometric properties [23–25].
Recruitment and data collection.
Participants were recruited through three main methods: the graduate member lists of healthcare programs; bulk email invitations from the Hong Kong Physiotherapy Association and the Chiropractic Doctors’ Association of Hong Kong; along with postings in academic departments (e.g., Department of Rehabilitation Sciences at The Hong Kong Polytechnic University and Department of Orthopaedics and Traumatology, The University of Hong Kong) and clinics. All three methods outlined the study’s objectives and the estimated 10-minute survey completion time. The survey was created using Qualtrics software (Qualtrics Survey2020, Utah, USA; https://www.qualtrics.com/). It began with two pages for the consent form and demographic data, followed by six pages focused on evaluating each construct within the EBP measure. Participation in both Phase I and Phase II was voluntary, and no financial or non-financial compensation was provided.
Data analysis.
Descriptive statistics were used to summarize the data. Continuous variables were reported as means and standard deviations (SD), while categorical variables were presented as frequencies and percentages. All statistical analyses were conducted using the Statistical Package for the Social Sciences (SPSS v.26) [26]. Additionally, Rasch analyses were performed using the Rasch Unidimensional Measurement Model (RUMM) software (version 2030) [27] to evaluate the structural validity and reliability of the EBP measure.
Rasch analysis.
The reliability and validity of the EBP measure were evaluated using Rasch analysis [15]. Rasch analysis is a robust psychometric technique within the framework of item response theory, designed to transform ordinal data into interval-level data, enabling more precise measurement [28]. This model organizes items hierarchically based on their difficulty, aligning them with the ability range of respondents. Specifically, individuals with higher abilities (e.g., those who practice EBP more frequently) are more likely to select advanced response options, while those with lower abilities tend to choose less advanced options. This hierarchy is visually represented through an item map, which illustrates the distribution of items along a continuum from least to most difficult [28]. Rasch analysis assumes unidimensionality; therefore, differential item functioning (DIF) evaluations are paramount to ensure that items operate consistently across different subgroups. Applying these psychometric properties is essential to verify the reliability, validity, and scoring methods of any instrument [29].
The Masters’ partial credit Rasch polytomous model was employed, as it is well-suited for analyzing ordinal response options [30]. As outlined in our initial study [15], two constructs—EBP usage and EBP activities—are formative in nature and thus do not require Rasch analysis. Conversely, the four reflective constructs—attitudes, self-efficacy, knowledge, and resources—were subjected to Rasch analysis to assess their psychometric properties. Table 2 details the Rasch analysis steps we applied to its assumptions.
Results
Phase I: Linguistic and cultural validation
The translation and cultural adaptation of the EBP measure were systematically carried out through steps I to IV of phase I. During this process, forward translators proposed several modifications to simplify the language, which were subsequently reviewed and approved by the expert committee. In phase V, cognitive debriefing was conducted with 36 participants to assess the relevance and comprehensibility of the adapted items. The demographic characteristics of the cognitive debriefing sample are presented in Table 3. Feedback from the interviews indicated that the EBP items were relevant to their respective constructs, easily understandable, and required an average completion time of 15 minutes. A detailed summary of the changes made to the EBP constructs’ items, along with additional insights from the cognitive debriefing process, is further detailed in S1 File.
Phase II: Psychometric validation
Of the 640 users who accessed the study link, 248 (119 PTs, 45 OTs, 36 DCs, 32 medical doctors, and 16 registered nurses) successfully completed the e-survey. Detailed demographic and professional characteristics of the participants are presented in Table 3.
Formative model (use of EBP and EBP activities).
For the Use of EBP construct, items remained unchanged during the linguistic and cultural adaptation phase, ensuring consistency with the original North American instrument. The scoring approach for this construct was retained from the original Canadian EBP measure [15], in which Use of EBP was conceptualized as a formative construct. Each of the 9 items includes 5 frequency-based response options referring to the past 6 months; however, for scoring purposes, responses are dichotomized and assigned a score of either 0 or 1. Specifically, selecting “Never” (indicating no use) is scored as 0, while selecting any of the other options (indicating one or more instances of use) is scored as 1. This total score reflects the number of EBP behaviors performed rather than the intensity or exact frequency of performance. This cutoff was adopted from the original instrument to indicate whether each core EBP behavior had been undertaken at least once during the reference period. Accordingly, the total score reflects the number of EBP behaviors performed, rather than their exact frequency or intensity. The total score is obtained by summing the scores for all 9 items, yielding a possible range from 0 to 9. A detailed breakdown of the scoring methodology is provided in Table 4.
For the EBP Activities construct, the original scoring system was preserved, including the method for calculating the total score to maintain consistency with the source instrument [15]. Unlike the Use of EBP construct, these items were not dichotomized. Instead, response options were weighted to approximate the frequency with which each activity was performed over the previous month. Specifically, the categories were scored as follows: “Never” = 0, “Monthly or less” = 1, “Bi-weekly” = 2, “Weekly” = 4, and “Daily” = 20. These weights were designed to reflect the approximate number of occasions per month on which the activity occurred, with “Daily” corresponding to an estimated 20 working days in a 4-week month. As this construct is formative, the total score should be interpreted as a practical index of engagement in EBP-related activities rather than as a reflective scale with equally spaced psychometric intervals. During the pilot study (Phase I), participants agreed that the response options accurately reflected their frequency of engagement in the listed activities. The expert committee reviewed this feedback and endorsed it without any further adjustments. A summary of the results for the EBP Activities measure is provided in Table 5.
Reflective constructs
Knowledge about EBP.
Table 6 presents the Rasch analysis results for the original 11 items assessing EBP knowledge. Initially, seven items exhibited disordered thresholds, prompting the rescoring of these items by merging the response categories “Never heard the term” and “Have heard it but don’t understand” (items 34, 35, 36, 39, 41, 42, and 44). Following this adjustment, six items were removed due to a misfit with the Rasch model. Three items (39, 40, and 42) were excluded based on fit residuals, while the remaining three (34, 41, and 44) were removed due to significant χ² and F-statistics.
Local dependency was initially observed between items 34 and 35, as well as between items 41 and 44. However, these dependencies were resolved in the final analysis after the removal of misfitting items. DIF was identified for item 37 (Meta-analysis) by profession, but the item was retained due to its relevance and contribution to the construct.
The final 5-item measure demonstrated excellent fit to the Rasch model (χ² = 17.90, df = 15, p = 0.268), with no significant concerns regarding unidimensionality (% of significant t-tests = 4.99, below the 5% threshold). The threshold map, illustrated in Fig 1a, revealed a clear hierarchy in the difficulty of the knowledge items. Understanding confidence intervals was the least challenging, while grasping treatment effect size was the most difficult. Fig 1b shows that participants were reasonably well-targeted by the items, with a mean person location of 0.90 (expected 0) and an SD of 2.02 (expected 1). The measure also exhibited strong reliability, with a Person Separation Index (PSI) of 0.84 and a Cronbach’s α of 0.88, indicating high internal consistency.
Self-efficacy towards EBP.
The self-efficacy items related to EBP are outlined in Table 7. Initially, all items displayed disordered thresholds, requiring rescoring to ensure proper alignment with the Rasch model. For seven items (46, 47, 48, 50, 51, 52, and 53), the “90%” and “100%” response categories were merged. Item 45 was adjusted to two binary categories after rescoring. Item 49 had the “80%,” “90%,” and “100%” categories consolidated. Six items (46, 48, 50, 51, 52, and 53) involved combining the “0%,” “10%,” “20%,” and “30%” categories. Two items (47 and 49) merged the “0%,” “10%,” “20%,” “30%,” and “40%” categories. Following these modifications, item 53 was excluded due to its misfit with the Rasch model. The final 8-item measure showed no evidence of local item dependency or DIF, confirming its robustness.
The model demonstrated excellent fit (χ² = 26.48, df = 24, p = 0.33), with no significant concerns regarding unidimensionality (% of significant t-tests = 4.87, below the 5% threshold). Fig 2a illustrates the progression of item thresholds, reflecting the hierarchical structure of the self-efficacy items. Fig 2b indicates that participants were well-targeted by the items, with a mean person location of 0.49 (expected 0) and an SD of 2.39 (expected 1). The final measure exhibited strong reliability, with a PSI of 0.93 and a Cronbach’s α of 0.95, underscoring its high internal consistency.
Attitudes towards EBP
The attitudes construct was significantly multidimensional (% of significant t-tests = 25.85), comprising two distinct sets of items: positively worded (items 17–25) and negatively worded (items 26–33). Despite efforts to rescore and eliminate misfitting items, maintaining both sets within a single construct proved impractical. A bi-factor Rasch analysis confirmed their lack of unidimensionality, necessitating separate analyses for each set.
Positively worded attitudes towards EBP.
Among the initial nine positively worded items, three exhibited disordered thresholds and were rescored by merging the ‘Strongly disagree’ and ‘Disagree’ categories (items 21, 23, and 25). These same items were subsequently removed due to a misfit with the Rasch model. The final 6-item measure showed no evidence of local item dependency or DIF. It demonstrated good fit to the Rasch model (χ² = 15.08, df = 18, p = 0.66), with no significant dimensionality concerns (% of significant t-tests = 4.98). Fig 3a illustrates the item-threshold gradient, reflecting the hierarchical structure of the items. However, Fig 3b indicates that participants were not well-targeted by the items, with a mean person location of 2.65 (expected 0) and an SD of 1.98 (expected 1). Despite this, the measure exhibited good reliability, with a PSI of 0.84 and a Cronbach’s α of 0.86. Table 8 presents the Rasch analysis results for the positively-worded attitudes towards EBP.
Negatively-worded attitudes towards EBP.
All eight negatively worded items were reverse-scored due to their phrasing. None exhibited disordered thresholds, but item 31 was removed due to misfit. The final 7-item measure showed no local item dependency or DIF and demonstrated a good fit to the Rasch model (χ² = 15.23, df = 21, p = 0.81), with no significant dimensionality issues (% of significant t-tests = 4.96). Fig 4a displays the item-threshold gradient, while Fig 4b indicates that participants were well-targeted by the items, with a mean person location of 0.01 (expected 0) and an SD of 1.40 (expected 1). The measure also exhibited good reliability, with a PSI of 0.84 and a Cronbach’s α of 0.86. Table 9 presents the Rasch analysis results for the negatively worded attitudes towards EBP.
Resources
Item 65 was reverse-scored due to its negative phrasing (Table 10). Initially, all items in the resources construct exhibited disordered thresholds, which were addressed by merging the ‘Strongly disagree’ and ‘Disagree’ response categories. Seven items (55, 57, 59, 64, 65, 67, and 68) were subsequently removed due to a misfit with the Rasch model. The final 8-item measure showed no evidence of local item dependency or DIF. The final model demonstrated a good fit to the Rasch model (χ² = 26.60, df = 27, p = 0.49), with no significant dimensionality concerns (% of significant t-tests = 4.89). Fig 5a illustrates the gradient across item thresholds, reflecting the hierarchical structure of the items. Fig 5b indicates that participants were reasonably well-targeted by the items, with a mean person location of 0.42 (expected 0) and an SD of 1.37 (expected 1).
Discussion
This study aimed to cross-culturally adapt and validate an EBP measure for healthcare providers in Hong Kong, ensuring its linguistic, cultural, and psychometric appropriateness. The findings demonstrate that the adapted EBP measure is a robust tool for evaluating EBP-related constructs among these providers, with strong reliability and validity across its six constructs.
Minimal modifications were required to simplify language and align with local terminology, reflecting the shared professional language of English in Hong Kong’s healthcare system. High relevance and comprehensibility ratings from cognitive debriefing further support the measure’s cultural appropriateness.
The use of EBP and EBP activities constructs, which follow a formative model, requires no substantive changes beyond linguistic adjustments. These constructs provide valuable insights into the frequency and types of EBP activities undertaken by healthcare providers, offering a practical framework for evaluating EBP integration into clinical practice. Psychometric validation through Rasch analysis revealed strong construct validity and reliability for most constructs of the reflexive model.
The knowledge construct underwent significant refinement, with six misfitting items removed. The final 5-item measure showed excellent reliability and unidimensionality, with a clear hierarchy of item difficulty. This suggests that while participants were generally familiar with basic EBP concepts, more advanced topics, such as treatment effect size, posed greater challenges. However, it is important to consider that healthcare providers may not need to master these intricate details to engage in EBP effectively [36]. Instead, focusing on fundamental competencies that support the practical application of EBP in daily clinical settings may be more beneficial. These findings highlight potential gaps in EBP knowledge among Hong Kong healthcare providers, which could inform future educational interventions by emphasizing essential skills that enhance clinical practice.
Self-efficacy and resources also demonstrated strong psychometric properties. The rescoring of disordered thresholds and removal of misfitting items improved model fit, ensuring that the measures accurately reflect participants’ confidence in applying EBP and their access to EBP-related resources. The reliability scores for these constructs further validate their utility in assessing EBP implementation.
However, the attitudes construct was found to be multidimensional, requiring its division into positively and negatively worded subscales. This was not necessary when validating the measure among physiotherapy and occupational therapy graduates [15] or physiotherapy students [17]. Despite these findings, both subscales demonstrated good fit and reliability, supporting their use as distinct measures.
This finding suggests that attitudes toward EBP may not be adequately captured as a single unidimensional construct in this population. Rather, the positively and negatively worded items appear to reflect two related but distinct dimensions of attitudes. In practical terms, this supports interpreting the attitudes domain as two separate subscales rather than a single overall score when used among Hong Kong healthcare providers. At the same time, this result also indicates that further refinement of the attitudes construct may be warranted in future research to determine whether a more coherent unidimensional scale can be developed for this population.
This division likely reflects differences in professional experience between practicing clinicians and students or recent graduates. As Thomas et al. [37] discuss, attitudes toward EBP are complex and shaped by factors such as social desirability and exposure to real-world constraints. While students’ perspectives are often shaped by academic training and emphasize the idealized benefits of EBP [38,39], practicing clinicians navigate barriers like time constraints, resource limitations, and organizational challenges [40]. These challenges may lead to more nuanced or critical attitudes, particularly in response to negatively worded items [37]. Cognitive load differences may also impact how clinicians interpret these items, potentially introducing measurement error [41,42]. Negatively worded items were intentionally retained from the original measure in part to reduce acquiescence bias; however, such items may also introduce wording-related method effects. Because items were presented in construct-based sections rather than randomized order, some response-style bias may still have been present. Given these factors, treating positively- and negatively-worded attitudes as distinct constructs improves measurement accuracy in heterogeneous populations [43]. This approach provides a more detailed understanding of how attitudes influence EBP implementation, allowing for targeted educational and support strategies.
The multidimensionality of the attitudes construct is consistent with previous research that used different attitude scales, demonstrating that combining positively and negatively worded items can introduce artificial factors and undermine unidimensionality [44–47]. Weijters and Baumgartner (2012) [48] found that combining positively and negatively worded items in the same scale can result in multidimensionality. To improve validity and interpretability, it is recommended that such items be separated [44]. The literature supports our decision to divide the attitudes construct into two subscales in order to accurately represent a single underlying dimension.
These findings have practical implications for future use of the measure. The adapted instrument can be used to assess key EBP-related constructs among healthcare providers in Hong Kong, but the attitudes domain may be more appropriately interpreted as two separate subscales rather than as a single overall score. This distinction is important for both researchers and educators, as it allows a more nuanced understanding of clinicians’ endorsement of EBP and their perceived reservations or barriers. Future research should examine whether refinement of item wording or construct structure can improve coherence while preserving conceptual breadth.
The Rasch analysis revealed a notable lack of middle-range items across most EBP constructs, except for negatively worded attitudes toward EBP. This gap suggests that the measure may not fully capture the spectrum of EBP competencies, particularly for individuals with moderate engagement or understanding. The absence of middle-range items could result from a skewed item pool or contextual differences in Hong Kong’s healthcare system that shape how EBP is practiced and perceived [49]. While this issue was not observed in the original [15] or student [17] versions of the measure, it highlights the need for further item development. Future research should focus on expanding the item pool, potentially through qualitative methods such as interviews or focus groups, to better reflect the diverse experiences of healthcare providers.
Recent studies emphasize dynamic frameworks that account for evolving clinical environments, shifting patient demographics, and the increasing complexity of healthcare systems in assessing attitudes toward EBP [50,51]. While quantitative measures provide valuable standardized assessments, there is growing recognition that they may not fully capture the contextual and experiential factors influencing EBP adoption [52]. Some scholars advocate for mixed-method approaches, arguing that qualitative data—such as interviews and focus groups—can reveal deeper insights into barriers, facilitators, and real-world applications of EBP that quantitative scales alone might overlook [37,53]. The field may need to move toward a more integrated approach, combining multiple tools and data sources to gain a fuller and more nuanced understanding of EBP [54]. By leveraging both qualitative and quantitative methods, we can develop more comprehensive measurement strategies that better reflect the complexities of clinical practice and support meaningful advancements in EBP implementation.
Our study employed self-reported data to evaluate participants’ EBP knowledge, skills, and activities. While self-reported measures are practical and offer valuable insights, they inherently depend on participants’ perceptions, which may not always perfectly align with their actual abilities. This is a common consideration in survey-based research. In addition, self-reported Rasch ordinal data may be affected by potential biases, such as social desirability bias due to overreporting positive behaviors [55–57], recall bias [58], or certain response styles, such as participants of certain cultures consistently choosing extreme options [59]. Such biases can influence the normal distribution of responses and may affect the accuracy of Rasch model estimates, leading to measurement error [60].
Future studies could improve the evaluation of EBP education by integrating self-reported data with performance-based assessments, particularly for tasks that require specific skills, such as developing research keywords and conducting literature searches. However, within the scope and constraints of our study, self-reported data remains a valuable tool for capturing participants’ self-assessed competencies and experiences.
Strengths and limitations
This study had several strengths. First, its rigorous cross-cultural adaptation and comprehensive psychometric validation ensure the measure’s cultural appropriateness and reliability for Hong Kong healthcare providers. The use of Rasch analysis confirmed the measure’s unidimensionality and construct validity while addressing misfitting items and local dependencies. Second, its methodological rigor, transparency, and adherence to the STROBE checklist enhance the credibility and reproducibility of the findings [16,18]. Third, strong internal consistency reliability, with Cronbach’s α and PSI values exceeding 0.7 for all constructs, underscores the measure’s robustness [33,34]. Fourth, by capturing both individual and organizational factors, the measure provides a holistic assessment of EBP competencies, supporting continuous improvement in healthcare delivery and ultimately enhancing patient outcomes.
However, several limitations that warrant consideration. First, the sample size, while sufficient for Rasch analysis, may not fully represent the diversity of healthcare professions in Hong Kong, potentially limiting the generalizability of the findings. Second, the study’s cross-sectional design, focused on cross-cultural validation and psychometric assessment, precludes the evaluation of test-retest reliability, which is essential for assessing the stability of the measure over time. Third, the exclusion of certain healthcare professions or those with limited English proficiency may further restrict the measure’s applicability across all provider groups such as traditional Chinese medicine practitioners. Fourth, while the measure was adapted for Hong Kong, its use in other cultural or linguistic contexts would require additional validation to ensure its relevance and accuracy. Fifth, some items removed due to misfit in the Rasch analysis might be important to EBP, although similar concepts may be covered by other items in the measure. While their deletion was necessary to maintain the psychometric integrity of the scale, we acknowledge that these changes may have resulted in a narrower representation of certain EBP dimensions. Sixth, the scoring of the formative constructs. For Use of EBP, the dichotomous cutoff of “at least once” was retained from the original instrument, while for EBP Activities, weighted response options were used to approximate the number of activity occasions per month. Although these approaches are conceptually consistent with the original formative model, they may still reduce granularity or affect score distributions. We did not conduct sensitivity analyses using alternative thresholds or scoring structures in the present sample, and this should be explored in future studies. Future research could explore alternative item formulations or supplementary approaches to ensure comprehensive coverage of all relevant aspects of EBP.
Conclusions
This study successfully adapted and validated an EBP measure for use among Hong Kong healthcare providers. The measure’s strong psychometric properties and cultural relevance make it a valuable tool for assessing and enhancing EBP competencies in this population. By identifying gaps in knowledge, self-efficacy, and resource availability, this measure can inform targeted interventions to promote evidence-based care, ultimately improving patient outcomes and healthcare quality in Hong Kong.
Supporting information
S1 File. Changes to the original questionnaires and the pilot study results.
https://doi.org/10.1371/journal.pone.0351754.s001
(DOCX)
S1 Appendix. List of item hierarchy of each construct.
https://doi.org/10.1371/journal.pone.0351754.s002
(PDF)
References
- 1. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn’t. BMJ. 1996;312(7023):71–2. pmid:8555924
- 2. Straus SE, Sackett DL. Using research findings in clinical practice. BMJ. 1998;317(7154):339–42. pmid:9685286
- 3.
Institute of Medicine. Evidence-based medicine and the changing nature of healthcare: 2007 IOM annual meeting summary. Washington, DC: Institute of Medicine. 2008.
- 4.
World Health Professions Alliance. EB 144 Agenda Item 5.5 on Universal Health Coverage – Community Health Workers Delivering Primary Health Care: Opportunities and Challenges. https://www.whpa.org/eb-144-agenda-item-55-universal-health-coverage-community-health-workers-delivering-primary-health
- 5.
World Confederation for Physical Therapy. Policy statement: Evidence-based practice. 2019. https://world.physio/policy/ps-ebp
- 6.
International Council of Nurses. About International Classification for Nursing Practice (ICNP). 2022. https://www.icn.ch/what-we-do/projects/international-classification-for-nursing-practice
- 7.
World Federation of Occupational Therapists. Guiding Principles for the Use of Evidence in Occupational Therapy. 2021. https://wfot.org/resources/guiding-principles-for-the-use-of-evidence-in-occupational-therapy
- 8.
World Medical Association. Policy tag: Professional practice. 2022. https://www.wma.net/policy-tags/professional-practice/
- 9. Albarqouni L, Hoffmann T, Glasziou P. Evidence-based practice educational intervention studies: a systematic review of what is taught and how it is measured. BMC Med Educ. 2018;18(1):177. pmid:30068343
- 10. Sapri ND, Ng YT, Wu VX, Klainin-Yobas P. Effectiveness of educational interventions on evidence-based practice for nurses in clinical settings: A systematic review and meta-analysis. Nurse Educ Today. 2022;111:105295. pmid:35144204
- 11. Saunders H, Gallagher-Ford L, Kvist T, Vehviläinen-Julkunen K. Practicing Healthcare Professionals’ Evidence-Based Practice Competencies: An Overview of Systematic Reviews. Worldviews Evid Based Nurs. 2019;16(3):176–85. pmid:31074582
- 12. Portela Dos Santos O, Melly P, Hilfiker R, Giacomino K, Perruchoud E, Verloo H, et al. Effectiveness of Educational Interventions to Increase Skills in Evidence-Based Practice among Nurses: The EDITcare Systematic Review. Healthcare (Basel). 2022;10(11):2204. pmid:36360544
- 13. Roberge-Dao J, Maggio LA, Zaccagnini M, Rochette A, Shikako K, Boruff J, et al. Challenges and future directions in the measurement of evidence-based practice: Qualitative analysis of umbrella review findings. J Eval Clin Pract. 2023;29(1):218–27. pmid:36440876
- 14. Alavi SM, Karami H, Khodi A. Examination of factorial structure of Iranian English language proficiency test: An IRT analysis of Konkur examination. Curr Psychol. 2021;42(10):8097–111.
- 15. Al Zoubi F, Mayo N, Rochette A, Thomas A. Applying modern measurement approaches to constructs relevant to evidence-based practice among Canadian physical and occupational therapists. Implement Sci. 2018;13(1):152. pmid:30563550
- 16. Beaton DE, Bombardier C, Guillemin F, Ferraz MB. Guidelines for the process of cross-cultural adaptation of self-report measures. Spine (Phila Pa 1976). 2000;25(24):3186–91. pmid:11124735
- 17. Al Zoubi FM, Bussières A, Chan HW, Leung KY, Ng YY, Lau KC, et al. Refining and adapting the measurement properties of evidence-based practice measures for physiotherapy students. PLoS One. 2024;19(3):e0298611. pmid:38451977
- 18. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453–7. pmid:18064739
- 19. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006;3(2):77–101.
- 20. Terwee CB, Bot SDM, de Boer MR, van der Windt DAWM, Knol DL, Dekker J, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007;60(1):34–42. pmid:17161752
- 21. Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, et al. The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. J Clin Epidemiol. 2010;63(7):737–45. pmid:20494804
- 22. Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, et al. The COSMIN checklist for assessing the methodological quality of studies on measurement properties of health status measurement instruments: an international Delphi study. Qual Life Res. 2010;19(4):539–49. pmid:20169472
- 23. Smith AB, Rush R, Fallowfield LJ, Velikova G, Sharpe M. Rasch fit statistics and sample size considerations for polytomous data. BMC Med Res Methodol. 2008;8:33. pmid:18510722
- 24. Hagell P, Westergren A. Sample Size and Statistical Conclusions from Tests of Fit to the Rasch Model According to the Rasch Unidimensional Measurement Model (Rumm) Program in Health Outcome Measurement. J Appl Meas. 2016;17(4):416–31. pmid:28009589
- 25. Linacre J. Sample size and item calibration stability. Rasch Meas Trans. 1994;7:328.
- 26.
IBM Corp. IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY: IBM Corp. 2021.
- 27.
Andrich D, Sheridan B, Luo G. Interpreting RUMM2030. Perth: RUMM Laboratory Pty Ltd. 2009.
- 28. Boone WJ, Staver JR, Yale MS. Wright maps: First steps. Rasch analysis in the human sciences. 2014. 111–36.
- 29. Khodi A. The affectability of writing assessment scores: a G-theory analysis of rater, task, and scoring method contribution. Lang Test Asia. 2021;11(1).
- 30. Masters GN. A Rasch Model for Partial Credit Scoring. Psychometrika. 1982;47(2):149–74.
- 31.
Andrich D, Sheridan B. RUMM2030 manual. Perth, Australia: RUMM Laboratory. 2009.
- 32.
Christensen KB, Kreiner S, Mesbar M. Rasch models in health. Hoboken: Wiley. 2013.
- 33. Gibbons CJ, Mills RJ, Thornton EW, Ealing J, Mitchell JD, Shaw PJ, et al. Rasch analysis of the hospital anxiety and depression scale (HADS) for use in motor neurone disease. Health Qual Life Outcomes. 2011;9:82. pmid:21955749
- 34. Pallant JF, Tennant A. An introduction to the Rasch measurement model: an example using the Hospital Anxiety and Depression Scale (HADS). Br J Clin Psychol. 2007;46(Pt 1):1–18. pmid:17472198
- 35. Smith EV Jr. Detecting and evaluating the impact of multidimensionality using item fit statistics and principal component analysis of residuals. J Appl Meas. 2002;3(2):205–31. pmid:12011501
- 36. Thomas A, Iqbal MZ, Roberge-Dao J, Ahmed S, Bussières A, Debigaré R, et al. Perspectives on how evidence-based practice changes over time: A qualitative exploration of occupational therapy and physical therapy graduates. J Eval Clin Pract. 2025;31(3):e14200. pmid:39415504
- 37. Thomas A, Chin-Yee B, Mercuri M. Thirty years of teaching evidence-based medicine: have we been getting it all wrong?. Adv Health Sci Educ Theory Pract. 2022;27(1):263–76. pmid:34559327
- 38. Iqbal MZ, Rochette A, Mayo NE, Valois M-F, Bussières AE, Ahmed S, et al. Exploring if and how evidence-based practice of occupational and physical therapists evolves over time: A longitudinal mixed methods national study. PLoS One. 2023;18(3):e0283860. pmid:37000834
- 39. Thomas A, Al Zoubi F, Mayo NE, Ahmed S, Amari F, Bussières A, et al. Individual and organizational factors associated with evidence-based practice among physical and occupational therapy recent graduates: A cross-sectional national study. J Eval Clin Pract. 2021;27(5):1044–55. pmid:33314562
- 40. Ayoubian A, Nasiripour AA, Tabibi SJ, Bahadori M. Evaluation of Facilitators and Barriers to Implementing Evidence-Based Practice in the Health Services: A Systematic Review. Galen Med J. 2020;9:e1645. pmid:34466560
- 41. Zeng B, Wen H, Zhang J. How Does the Valence of Wording Affect Features of a Scale? The Method Effects in the Undergraduate Learning Burnout Scale. Front Psychol. 2020;11:585179. pmid:33117248
- 42. Coleman C. Effects of negative keying and wording in attitude measures: A mixed-methods study. 2013.
- 43. Zeng B, Jeon M, Wen H. How does item wording affect participants’ responses in Likert scale? Evidence from IRT analysis. Front Psychol. 2024;15:1304870. pmid:39430906
- 44. Mahat M. The development of a psychometrically-sound instrument to measure teachers’ multidimensional attitudes toward inclusive education. International Journal of Special Education. 2008;23:82–92.
- 45. Garcia SF, Hahn EA, Magasi S, Lai J-S, Semik P, Hammel J, et al. Development of self-report measures of social attitudes that act as environmental barriers and facilitators for people with disabilities. Arch Phys Med Rehabil. 2015;96(4):596–603. pmid:25045803
- 46. Upegui-Arango LD, Mainz V, Gecht J, Mueller C-A, Quack V, Heinemann AW, et al. Development of the German social attitude barriers and facilitators to participation-scales: an analysis according to the Rasch model. BMC Musculoskelet Disord. 2022;23(1):423. pmid:35524254
- 47. Jiang W, Fenwick EK, Lamoureux EL, Zhang Z, Feng Y, Wang Y, et al. Linguistic and Cultural Validation of the Diabetic Retinopathy Knowledge and Attitudes Scale (DRKA) in a Chinese Population. Transl Vis Sci Technol. 2023;12(6):17. pmid:37358494
- 48. Weijters B, Baumgartner H. Misresponse to Reversed and Negated Items in Surveys: A Review. Journal of Marketing Research. 2012;49(5):737–47.
- 49. Hagquist C, Bruce M, Gustavsson JP. Using the Rasch model in nursing research: an introduction and illustrative example. Int J Nurs Stud. 2009;46(3):380–93. pmid:19059593
- 50. Landsverk NG, Olsen NR, Brovold T. Instruments measuring evidence-based practice behavior, attitudes, and self-efficacy among healthcare professionals: a systematic review of measurement properties. Implement Sci. 2023;18(1):42. pmid:37705031
- 51. Santesson AHE, Holmberg R, Bäckström M, Gustafsson P, Jarbin H, Perrin S. Clinician attitudes towards adoption of evidence-based practice: a nationwide multiprofessional cross-sectional study of child and adolescent mental health services in Sweden. BMC Health Serv Res. 2024;24(1):1432. pmid:39558358
- 52.
Patton MQ. Qualitative research & evaluation methods: integrating theory and practice. 4th ed. Thousand Oaks, CA: SAGE Publications, Inc. 2015.
- 53. Thomas A, Roberge-Dao J, Iqbal MZ, Salbach NM, Letts LJ, Polatajko HJ, et al. Developing multisectoral strategies to promote evidence-based practice in rehabilitation: findings from an end-of-grant knowledge translation symposium. Disabil Rehabil. 2024;46(11):2449–63. pmid:37399539
- 54. Engle RL, Mohr DC, Holmes SK, Seibert MN, Afable M, Leyson J, et al. Evidence-based practice and patient-centered care: Doing both well. Health Care Manage Rev. 2021;46(3):174–84. pmid:31233424
- 55.
Graeff TR. Response Bias. In: Kempf-Leonard K. Encyclopedia of Social Measurement. New York: Elsevier. 2005. 411–8.
- 56. Van de Mortel T. Faking it: Social desirability response bias in self-report research. Thea van de Mortel. 2008;25.
- 57. Teh WL, Abdin E, P V A, Siva Kumar FD, Roystonn K, Wang P, et al. Measuring social desirability bias in a multi-ethnic cohort sample: its relationship with self-reported physical activity, dietary habits, and factor structure. BMC Public Health. 2023;23(1):415. pmid:36859251
- 58. Sedgwick P. What is recall bias?. BMJ. 2012;344:e3519.
- 59. Kreitchmann RS, Abad FJ, Ponsoda V, Nieto MD, Morillo D. Controlling for response biases in self-report scales: forced-choice vs. psychometric modeling of Likert items. Frontiers in Psychology. 2019;10:2019.
- 60.
Eid M, Zickar MJ. Detecting response styles and faking in personality and organizational assessments by mixed Rasch models. Multivariate and mixture distribution Rasch models: Extensions and applications. New York, NY, US: Springer Science Business Media. 2007. 255–70.