Cross-Cultural Validation of the Global Interprofessional Therapeutic Communication Scale© in the Korean context

Background: Person-centered nursing is pivotal to the high-quality and safe practice of nursing, and therapeutic communication plays an essential role in this process. Therapeutic communication by healthcare professionals is vital in developing and maintaining constructive relationships with patients. The Global Interprofessional Therapeutic Communication Scale© (GITCS©) measures the therapeutic communication skills of healthcare providers. This scale is useful for assessing the verbal, non-verbal, and culturally sensitive therapeutic communication abilities of healthcare providers in various clinical situations. The purpose of this study was to evaluate the cross-cultural validity and reliability of the Korean version of the Global Interprofessional Therapeutic Communication Scale © (K-GITCS). Methods: A cross-cultural validation of a psychometric evaluation is conducted in the present study. The instrument was translated into Korean using the original developers’ translation process. A convenience sample of 300 registered nurses with more than one year of experience was recruited from a tertiary hospital. Validity was evaluated through a conrmatory factor analysis, and the instrument was tested for psychometric reliability. Results: The three-factor structure of the K-GITCS was validated. A conrmatory factor analysis of the K-GITCS was conducted, and the results satised the statistical criteria with a standardized root mean square residual of 0.06, a non-normed t index of 0.88, a root mean square error of approximation of 0.07, and a comparative t index of 0.82. The following factors were correlated: trust and rapport building, power-sharing, and empathy. Cronbach’s alpha was 0.94, indicating good internal consistency. Initial testing with experienced registered nurses one institution indicated that the K-GITCS © instrument identify communication skills this The K-GITCS can used assess healthcare interprofessional specic

global nature allows it to be translated into other languages for use. Few studies have con rmed the validity and reliability of GITCS© in other countries and cultures. However, it is important to note that communication depends on the cultural background. For example, in East Asia, people say "no thanks," it sometimes means "yes." People from western countries often require much time to understand this culture [18]. Hence, developing different language versions and psychometric evaluations of the GITCS© would be important initial steps promoting its wider use. This study can be used to identify the phenomenon of TC, and develop the suitable program to improve TC skills for RNs.

Aims
The purpose of the study was to translate the GITCS© into Korean, elucidate its validity and reliability in a cross-cultural context, and clarify its tness for use with RNs to assess their self-reported TC skills. This study was carried out international regulations based on the STROBE guidelines [19].

Design
This study uses a methodological research design to con rm the validity and reliability of the Korean version of the GITCS© (K-GITCS).

Participants
The participants in this study consisted of a convenience sample recruited from a tertiary hospital in Seoul, South Korea. Survey questionnaires were distributed to 300 RNs who were working at that time, and 271 (90.33%) were returned. Finally, 249 participants (83.00%) were included in the analyses after excluding 22 incomplete questionnaires. The appropriate number of participants was considered to be more than ve times the number of questions [20]. In this respect, the sample size of 249 was acceptable for a psychometric assessment (35 items x 6 = 210).

Data collection
For nal con rmation validity and reliability of the K-GITCS, the researchers administered the survey to participants during their shift breaks from April to May of 2019. The RNs were asked to ll out the questionnaires in a private space, place the completed surveys in envelopes provided by the research team, and seal them to ensure con dentiality. The K-GITCS took 10 to 15 minutes to complete.
The participants' demographic traits, i.e., age, sex, educational status, clinical experience, were used to check their introductory information. Then, a rubric of the K-GITCS© was used to evaluate the TC skills of the RNs. The GITCS© was developed by Campbell and Aredes [17] in Canada. The original scale was developed as a checklist to be used during simulation education. The researchers used the GITCS© as a self-reported psychometric questionnaire after con rmation by the original authors. A ve-point Likerttype scale ranging from 1 ("strongly disagree") to 5 ("strongly agree") was used for the 35 items. The questionnaire has three domains: trust and rapport building, empowering, and empathy. A higher score indicates a higher level of TC. The negatively worded items (9, 12, and 16) were reversely scored. The internal consistency reliability coe cient was 0.95 for the original scale [17].

Ethical considerations
This study was approved by the Institutional Review Board of Severance hospital (Y-2019-0001). The purpose of the study was explained to the participants by the researchers, and they were assured of con dentiality before they provided informed consent. In order to ensure that the surrounding environments did not affect their voluntary participation, unit managers were not involved in any decisions with regard to the participation of the RNs. Researchers visited wards at the end of the corresponding working hours and held a brie ng session with RNs who wished to participate in this study.

Data analysis
The data in this study were analyzed using the SPSS 25.0 (IBM, Armonk, NY, USA) and STATA 13.0 (StataCorp, College Station, TX, USA) software packages. The participants' demographic factors were determined by means of descriptive statistical analyses. To con rm the content validity, the items were analyzed using corrected item-total correlations. The Cronbach's alpha coe cients were determined to test the internal consistency of the K-GITCS, and a con rmatory factor analysis (CFA) was used to test the validity through structural equation modeling. Like the original GITCS©, the K-GITCS was considered to have three factors.

Validity, reliability, and rigor
The validation procedure involved two steps: translation and back-translation, and psychometric testing. Translation was conducted according to the GITCS© translation process, adapted by Campbell and Aredes [17]. Figure 1 presents the multiple phases of the translation process.
First, the scale was independently translated from English into Korean by two translators. These two translators and another local researcher, who is responsible for the K-GITCS, synthesized the two translators' results. Second, the GITCS translated into Korean was back-translated into English by a bilingual researcher of nursing. Third, the scale's translations were evaluated semantically and then adjusted by an expert panel, who veri ed the nal version of the K-GITCS.
For psychometric testing, the content validity of the translated GITCS© was examined by experts. The rst version of the K-GITCS was checked by an expert panel, i.e., one nursing professor and two RNs with more than two years of clinical experience. The linguistic validity of the scale was then reviewed independently by a researcher of Korean literature. After the rst version of the K-GITCS was completed, our research team discussed a semantic adjustment with Campbell and Aredes, the original authors of GITCS©. This discussion was conducted based on the back-translation of the K-GITCS. The mutual agreement with the original authors was reached without revision.
The construct validity and internal consistency reliability of the K-GITCS were also evaluated. First, a CFA was conducted to identify its structure. Second, Cronbach's alpha was used to examine whether the K-GITCS was internally consistent, elucidating how appropriately the items test the same construct [21].

Participants' demographics and characteristics of the K-GITCS
The average age of the participants was 31.92 ± 7.10 years. Most of the participants were women (n = 232, 93.17%), and the majority had a bachelor's degree in nursing (n = 199, 79.91%). Their workplaces included the medical-surgical ward (n = 159, 63.85%) and the intensive care unit (n = 90, 36.14%). Table 1 shows the number of responses to each item by the participants. Most participants responded with scores of either 4 or 5 on the scale. The highest score was for item number 15, "Describes what they are going to do before doing it." The lowest score was for item number 31, "Encourages feedback and input from the patient."

CFA
To measure the elements' t to the data, a CFA was conducted on 35 items of the K-GITCS (Table 2 and Figure 2). The result from a chi-square test was statistically signi cant (χ 2 = 11298.07, df = 557, p < .001).
Hooper, Coughlan, and Mullen [22] suggested that when reporting t indices, the standardized root mean square residual (SRMR) should be included along with the non-normed t index (NNFI), root mean square error of approximation (RMSEA), or the comparative t index (CFI). The CFA's results were as follows: SRMR = 0.06, NNFI = 0.81, RMSEA = 0.07 and CFI = 0.82. An SRMR value lower than 0.08 is deemed acceptable for well-tted models [23]. Further, it is recommended that the value of the NNFI, which is also known as the Tucker-Lewis index, should exceed 0.80 as the cutoff for a good t [22]. The RMSEA is sensitive when choosing a model with fewer parameters, and the upper limit for a well-tting model is 0.08 [22]. Meanwhile, it is argued that the CFI should exceed 0.70 [24], while another source proposes that a value above 0.95 is proper to ensure a good t [23].
In addition to a model evaluation, the RMSEA and CFI should consider the number of estimates for the variables and the multifactor structure [25]. As a result, the CFA showed a good t for the three-factor structure, which was acceptable and statistically signi cant at p < .001.

Criterion validation
The three subscales were correlated; however, the Pearson's correlation coe cients differed (0.26-0.71). This variation indicates that the three subscales evaluate different domains [26]. Thus, these ndings suggest that the K-GITCS has good criterion validity (Table 3).

Descriptive statistics for the scores of the three domains
The three factors of the K-GITCS' scores were measured to assess ceiling and oor effects. There were no large clusters of cases at the end of the statistical distribution for any factor according to tests of additional histograms. In this respect, there was no evidence of skewness or kurtosis (Table 4).

Reliability analysis
The Cronbach's alpha was applied to inspect the internal consistency of the K-GITCS. In this study, the Cronbach's alpha for the entire scale was 0.94, indicating good internal consistency ( Table 5). The elimination of some items resulted in reduced alphas, except for three items: 9, 12, and 16. For example, the removal of item number 12 in the K-GITCS reduced the coe cient alpha to 0.94.
The reliability coe cient for each of the three domains ranged from 0.49 to 0.94. Two of the K-GITCS subscales-trust and rapport building (23 items, Cronbach's alpha = 0.94) and power-sharing (seven items, Cronbach's alpha = 0.83)-showed high reliability. In comparison, empathy ( ve items, Cronbach's alpha = 0.49) had low reliability. The K-GITCS empathy subscale did not show good internal consistency in the self-reported questionnaire.
The corrected item-total correlation, i.e., the correlation between each item and the total score of the scale, was > 0.40 after excluding three items (9, 12, and 16). An item-total correlation coe cient exceeding 0.40 suggests a suitable correlation between the item and the total score [27].

Discussion
Mutual and goal-directed communication among healthcare professionals and patients is TC, which promotes the quality of care [1]. TC can encourage person-centered and integrated care because RNs can contribute to good health outcomes through timely verbal and nonverbal communication [1]. Hence, appropriate instruments to measure TC and suitable programs that can be used to identify and enhance the TC skills of RNs are required. This cross-cultural validation study presents a psychometric evaluation of the K-GITCS and certi es its tness for RNs. Through this study, the hypothesis that the K-GITCS is a valid and reliable instrument for measuring the TC of RNs was con rmed. We investigated the construct validity, criterion validity, and reliability of the K-GITCS through a sample of RNs working at a tertiary hospital.
The goodness of t of the K-GITCS was demonstrated by means of a CFA. The current study adopted the three-factor structure of the original instrument [17]. The CFA revealed that it is a good t for the data and that it has a three-factor structure. However, no previous studies investigated the exploratory factor analysis (EFA) of the GITCS©. Further discussion about the instrument's factor structure and its use in actual clinical settings is necessary, as is additional research. More speci cally, an EFA can con rm the latent variables and identify groups of items [21].
The reliability analysis evaluated the translated instrument's internal consistency. The suggested Cronbach's alpha for an evaluation tool is above 0.70 [21]. The K-GITCS© had a Cronbach's alpha value of 0.94 for the overall questionnaire, demonstrating good internal consistency. The values for trust and rapport building and for power sharing all exceeded 0.80, which are considered acceptable [21]. However, empathy, which consists of ve items, had a Cronbach's alpha value of 0.49. When translating an instrument into other languages, cultural adaptation should be taken into account. Further, historical, social, and present situations can affect people's thinking and meaning interpretations [28]. Empathy is a key concept of genuine TC, and it ultimately depends on the relationships between practitioners, patients, and family members [29,30]. Thus, empathy is a culturally and linguistically distinct concept. Depending on the culture or language, people can understand the same sentence in different ways. For example, item numbers 5 and 30 of the K-GITCS could be perceived as contradictory items. Item 5 states "maintain an appropriate distance" whereas item 30 is "stay beside the patient." Subjects may have thought that maintaining an appropriate distance is the opposite of staying beside the patient, although these two items are both TC competencies.
Cultural differences are imperative factors for communication instruments. Existing references demonstrated that the cultural background is considerable for validity and reliability as a measurement invariance [31][32][33]. Nursing researchers have to substantiate the measurement invariance after identifying the validity and reliability of a cross-cultural developmental instrument [34]. Further studies need to evaluate the measurement invariance between groups or longitudinal designs. Additionally, differences in clinical environments, such as those caused by the different patient-RN relationships between South Korea and Canada, the atmospheres of different hospitals, and the scopes of nurses' authority, can also affect the validity and reliability of the GITCS©. In this context, North American and East Asian clients have unique views of healthcare providers of each other. Asian cultures emphasize politeness or respect; communication is more earnest and weighed than in other cultures [1]. Depending on this cultural background, interpersonal relationships appear different, and RNs are required to consider the context to communicate with patients effectively. As empathy can occur in therapeutic relationships between RNs and their clients, we should accept the differences in atmospheres between these cultures. For example, Asian and Hispanic cultures attach signi cance to the family. It is therefore crucial to understand the dynamics of family interactions with empathy for patients and accept familycentered decision-making for Asians and Hispanics [1]. The core of person-centered care is to respect the values and preferences of patients or caregivers [35]. When nurses listen carefully to patients or caregivers in challenging clinical situations, it will help in the decision making regarding care plans with healthcare providers. It is clear that advanced communication skills, such as empathy, are pivotal for a person-centered care process [36]. Therefore, more studies are needed to learn about empathy in various cultures and their effects on TC between patients and healthcare providers.
Patients can feel left alone in a hospital, especially when they need unfamiliar healthcare services in a strange place [37]. The curing of diseases can be evidence-based without the individual, but the caring should consider individualization [38]. Person-centeredness has become a central concept in nursing and healthcare services, re ecting the shift in focus from healthcare providers to patients. The key to nursing care is the caring practice itself considering the interests of the patients and their characteristics [38].
Kawandi [39] noted that TC was the most important aspect provided by healthcare professionals in hightech medical environments. Among healthcare professionals, RNs who remain at the patient's side for 24 hours are given much weight in patients' therapeutic relationships. In terms of the importance of TC, the abilities of RNs to acquire and interpret patients' psychosocial status or experiences are essential. Furthermore, advanced communication is paramount in person-centered nursing for optimal clinical nursing practice. Nurses must involve patients in their care process through positive, attentive interaction, and information sharing. Person-centered communication can improve patient satisfaction with the quality of nursing and reduce issues related to patient safety.

Limitations
This instrument-development study has several limitations. First, this study was surveyed in only one tertiary hospital with a group of experienced RNs. Consequently, careful should be exercised when generalizing the scope of the results. Second, the GITCS© was initially developed to evaluate the effects of simulation education. Although Campbell and Aredes, the developers of the GITCS©, explained that the GITCS© can be used as a self-reported scale, differences in results cannot be ruled out. Finally, additional research is needed to con rm the validity and reliability of the K-GITCS in different settings, including learning environments that use simulations.

Conclusion
Overall, this study found that the K-GITCS has suitable validity and reliability as an instrument to measure the TC skills of RNs working in a tertiary hospital. This study has the signi cance of being the rst effort to exhibit the psychometric properties of the Korean version of the GITCS. The original instrument was developed to enable instructors to assess their students' TC skills, but this study con rmed that the instrument can be used for self-reported assessments. The K-GITCS can accurately evaluate the TC skills of RNs and provide useful insights and can serve as the basis for appropriate and competent interventions. Future research can verify the K-GITCS's utility by testing its psychometric properties, such as patients' perspectives toward the care they receive, with various groups, including nursing students and other healthcare providers. Additionally, the next step for future studies can be to assess and explore the factors associated with TC by using the K-GITCS to provide a well-structured and tailored educational intervention program for healthcare professionals. Written consent was procured from all registered nurses prior to the start of the survey. All methods were carried out in accordance with relevant guidelines and regulations.  Factor 1=trust and rapport building, Factor 2=Power-sharing, Factor 3=Empathy ** p<.001 Table 4 Descriptive statistics for the scores on the three factors (N=249) Factor 1=trust and rapport building, Factor 2=Power-sharing, Factor 3=Empathy  Items and factors' Cronbach's alpha coe cients.