Introduction
Touch comfort refers to the level of comfort nurses feel internally when touching patients [1]. During hospitalization, nurses frequently need to touch patients [2]. As a part of caring activities, touch can not only improve patient outcomes but also alleviate their physical and mental discomfort [3‐5]. Although the effectiveness of touch has been widely recognized in international studies, there is still limited research on nurses’ perceptions of touch. A report indicates that 34.17% of clinical nurses experience low comfort when touching patients, leading to severe occupational burnout [6]. This underscores the importance of understanding nurses’ own touch comfort.
Studies have found that touch comfort in nurses is closely related to the frequency of touch behaviors, quality of nursing practice, patient experiences, and nurses’ physical and mental health [7‐10]. Good touch comfort is crucial for nurses to effectively implement humanistic care measures [11]. Higher touch comfort in nurses correlates with greater job satisfaction, professional recognition, and sense of professional benefit [12], which facilitates building harmonious nurse-patient relationships and improves the effectiveness of touch interventions [13]. However, research shows that nurses have many concerns and discomforts when touching patients, with overall touch comfort levels being moderate to low, significantly lower than in other professions [14]. Low touch comfort can become a source of stress, leading to occupational burnout, compassion fatigue, depression, depersonalization, and even psychophysical imbalance in nurses [15‐17]. This not only increases sick leave and turnover rates in nursing teams but also has long-term negative impacts on patient safety and holistic nursing quality [18]. The World Health Organization (WHO) calls for ensuring that nurses, as the backbone of healthcare systems and key to achieving universal health coverage, can contribute optimally in service environments [19]. Therefore, accurately measuring nurses’ touch comfort levels is essential to caring for both patients and nurses’ well-being, and to creating a good humanistic touch environment.
Anzeige
However, there are currently few relevant assessment tools. The nurses’ comfort with touch scale developed by Pedrazza et al. is the only tool available for nurses [1, 20]. This scale was developed in a western cultural context and includes five dimensions: physical comfort, emotional containment, task-oriented contact, reassurance, and personal care, with a total of 23 items [1]. Since touch is a form of proxemic behavior, researchers believe that it has cultural roots [21]. Different cultural backgrounds lead to differences in healthcare systems, patient needs, and medical behaviors, and therefore, nurses’ understanding and expression of touch also differ. Thus, applying a scale developed in a western context to Chinese nurses may prove unsuitable. For instance, when Chinese scholars attempted to revise the Chinese version of the scale, they found that the emotional containment dimension did not exist in the Chinese context. In a preliminary study, our research team used the Chinese version of this scale [7] to survey 766 nurses, who reported that items such as “massaging the patient’s feet, face, and back” were never part of their nursing duties, and they had to imagine responses, indicating that foreign scales cannot accurately and comprehensively reflect the touch comfort of Chinese nurses.
Accordingly, there is a pressing need to develop a nurses’ touch comfort evaluation scale tailored to the Chinese nursing cultural context. Such a tool is essential to assess the current state of nurses’ touch comfort, identify areas for improvement, and provide a foundation for implementing targeted measures to enhance touch comfort.
The aims of the research were twofold: (1) to develop a nurses’ touch comfort evaluation scale specifically designed for use in China and (2) to evaluate the scale’s performance, including its reliability and validity, to provide a high-quality tool for nursing managers to assess the touch comfort levels of Chinese nurses during caregiving.
Methods
Study design and methods
Following the Cosmin guidelines [22] and the scale development procedures recommended by DeVellis [23] and Wang Yuanyuan [24], this study was conducted in three stages: initial scale construction, item optimization, and scale performance evaluation. In stage 1, an initial item pool was developed based on theoretical analysis, literature review, and team discussion. The initial scale was formed by refining the contents of the items through the Delphi method and a pre-survey. In Stage 2, a questionnaire survey was conducted with 231 nurses. Item analysis was employed to optimize the scale items, while exploratory factor analysis (EFA) was used to identify the scale’s structure, resulting in the creation of a predictive scale. In Stage 3, data were collected from 355 nurses to validate the scale’s performance. Confirmatory factor analysis (CFA) and content validity were used to assess the scale’s validity, while Cronbach’s alpha coefficient, split-half reliability, and test-retest reliability were applied to evaluate its reliability. The flowchart of the study process is presented in Fig. 1.
Anzeige
Fig. 1
The development procedure of the nurses’ touch comfort evaluation scale
×
![]()
Stage 1 initial scale construction
Formation of the research team
A research team was formed, consisting of a head nurse from the infection department, a surgical nurse, a psychiatric nurse, two internal medicine nurses, a statistics teacher, and three nursing graduate students. The team members are experts, graduate students, and clinical nurses (subjects) in relevant fields of this study, providing professional advice from different perspectives at each research stage.
Construction of the item pool
Rationale At different moments of touch, nurses may experience different levels of touch comfort. To better understand nurses’ touch comfort throughout the entire touch process, this study selected Watson’s human caring theory as the foundational framework. Watson’s human caring theory, based on humanism, is a significant nursing theory that includes transpersonal caring relationship, caritas processes, and caring moments [25]. This study established a nurse touch comfort assessment model based on the three aspects of Watson’s theory. Watson categorizes nursing caring behaviors into instrumental activities and expressive activities [26]. Following this classification method, this study divided the moments when nurses touch patients into instrumental activities (task-oriented treatments, operations providing physiological comfort, and operations offering individualized assistance) and expressive activities (activities providing emotional support). The initial item pool was preliminarily divided into four dimensions: treatment task, physiological comfort, individualized assistance, and emotional support.
Literature review Using search terms such as “nurse,” “touch,” “comfortable,” “experience,” “care,” and “scale,” this study retrieved data from official government websites, Chinese nursing professional books, Chinese electronic databases(such as Wanfang Data, CNKI, and SinoMed) and English electronic databases (such as Web of Science, PubMed, and MEDLINE). The research team thoroughly reviewed theories, professional knowledge, assessment tools, policy documents, and scale research methods related to the study topic. Combining literature analysis [27] and clinical practice experience, this study found significant differences between pediatric and adult nursing tasks, with notable differences in certain touch aspects. The research team decided the scale’s measurement target as clinical nurses caring for adult (≥ 15 years old [28]) inpatients after discussion.
Team discussions and item pool content determination To ensure comprehensiveness, the team referred to various electronic literature, standards related to nursing humanistic care practices [29, 30], all nursing procedures in basic nursing [31], and items from the Chinese version of the nurses’ comfort with touch scale [7]. Based on Watson’s “caring moments” and integrating a holistic nursing view centered on humanism, the goal was to cover all touch moments in adult patient care. Under brainstorming method, the team initially drafted 57 items suitable for the scale dimensions. After four team discussions, representative items were selected, and unclear, highly correlated, and similar items were removed, resulting in an initial item pool of 40 items.
Determining item response format The Likert scale is commonly used to assess attitudes and behaviors on a given topic, and the 5-point rating scale is most suitable for health measurements [32]. Therefore, this study used a Likert 5-point scale, where 1 = very uncomfortable, 2 = uncomfortable, 3 = neutral, 4 = comfortable, and 5 = very comfortable.
Delphi method
From October to December 2022, two rounds of expert consultations were conducted via email and WeChat. The ideal number of experts is between 8 and 23 [33], and this study selected 20 experts. The selection criteria included: at least 10 years of experience in clinical nursing, nursing management, psychological nursing, humanistic nursing, or scale research; a title of intermediate or higher; a bachelor’s degree or higher; and voluntary participation in the study. Exclusion criteria included the inability to complete the expert consultation form on time. After each round of consultation, data were organized and expert opinions summarized by two individuals. The response rate of the expert consultation forms, expert authority coefficient (Cr), coefficient of variation (CV), and Kendall’s W coefficient were calculated. A recovery rate of ≥ 70% indicates high expert enthusiasm; Cr ≥ 0.7 suggests high expert authority; CV > 0.25 indicates low expert coordination; the higher the W value, and if W test significance level P < 0.05, the better the expert opinion coordination [24]. Indicator screening method: items with a mean ≤ 3.50 and a full score ratio ≤ 20% were deleted; items with CV > 0.25 and dimensions and items suggested by experts were screened and revised after discussion by the research team [24].
Pre-survey
In December 2022, we conducted a pre-survey using a convenience sampling method with 15 clinical nurses working in a tertiary hospital in Changsha, Hunan Province, China, to test the readability of the items. The inclusion criteria for the survey subjects were: voluntarily participating in the survey; clinical nurses working in hospitals in Changsha. The exclusion criteria were: pediatric nurses (clinical nurses in departments caring for hospitalized children under 15 years old); Nurses caring for non-hospitalized patients. Based on the feedback from the participants, we adjusted the wording and phrasing of some items to form the initial draft of the scale.
Stage 2 item optimization and formation of the predictive version of the scale
In February 2023, the research team conducted the first round of questionnaire surveys using a convenience sampling method with clinical nurses working in 17 hospitals in Changsha. The inclusion and exclusion criteria were the same as in the pre-survey. Electronic questionnaires were distributed and collected via the platform named the “Wen juan xing”. The survey tools included a general information form for nurses and the initial draft of the scale.
Item analysis
We used classical test theory methods, including item discrimination analysis, Cronbach’s alpha coefficient method, communality and factor loading, and correlation coefficient method, to quantitatively evaluate and screen the scale items [24]. The screening criteria were as follows: (1) item discrimination analysis: the subjects were divided into high-score groups (top 27%) and low-score groups (bottom 27%) based on the total scores of the scale items. Two-sample t-test was used to compare the significance of differences between the two groups. Items with critical ratio < 3 and P ≥ 0.05 were deleted [34]. (2) Cronbach’s alpha coefficient method: if the overall Cronbach’s alpha coefficient of the scale significantly increased after deleting a particular item, that item was deleted [34]. (3) commonality and factor loading: items with commonality < 0.30 and factor loading < 0.4 were deleted [35]. (4) correlation coefficient method: items with a correlation coefficient < 0.3 with the total score of the scale and the dimension it belongs to, and less than its correlation with other dimensions, were deleted [24, 36].
Exploratory factor analysis
We used principal component analysis and orthogonal rotation with varimax to conduct exploratory factor analysis to evaluate the structure of the scale. Generally, if the Kaiser-Meyer-Olkin (KMO) value is > 0.6 and Bartlett’s sphericity test is statistically significant (P < 0.05), factor analysis is considered feasible [24]. Items with factor loadings < 0.40 on their associated factors and those belonging to common factors containing only one item were deleted [24]. Through item analysis and exploratory factor analysis, the scale items were optimized to form a predictive version of the scale.
Stage 3 Performance evaluation and formation of the final version of the scale
Study subjects
In February 2023, the research team conducted a second round of online questionnaire surveys using a convenience sampling method in 16 hospitals in Changsha. The subjects were clinical nurses working in hospitals in Changsha, with the same inclusion and exclusion criteria as the pre-survey. The survey tools included a general information form for nurses and the predictive version of the scale.
Reliability testing
Internal consistency reliability was assessed using Cronbach’s alpha coefficient and split-half reliability for the total scale and each dimension. External stability reliability was determined by having 30 nurses who participated in the second questionnaire survey complete the survey again two weeks later, and test-retest reliability was calculated [37]. Generally, Cronbach’s alpha coefficient, split-half reliability, and test-retest reliability values of ≥ 0.7 are considered acceptable [24].
Validity testing
Content validity Based on the authority coefficients from the first round of expert consultations, seven experts with high authority coefficients were invited to evaluate the content validity. Item-level content validity (I-CVI) and scale-level content validity (S-CVI) were calculated [38, 39]. S-CVI includes universal agreement S-CVI (S-CVI/UA) and average S-CVI (S-CVI/Ave). I-CVI ≥ 0.78 [39] and S-CVI/UA and S-CVI/Ave ≥ 0.9 [38] indicate good content validity.
Construct validity Confirmatory factor analysis was used to verify the fit between the actual data and the model obtained from exploratory factor analysis, and to calculate convergent validity and discriminant validity. Acceptable standards for scale model fit are standardized chi-square statistics (χ²/df) < 5, root mean square error of approximation (RMSEA) < 0.08, goodness-of-fit index (GFI) > 0.9, adjusted goodness-of-fit index (AGFI) > 0.9, comparative fit index (CFI) > 0.9, incremental fit index (IFI) > 0.9, Tacker-Lewis index (TLI) > 0.9, parsimonious comparative fit index (PCFI) > 0.5, and parsimony normed fit index (PNFI) > 0.5 [40]. Convergent validity was assessed using average variance extracted (AVE) and combined reliability (CR), and discriminant validity was evaluated using the heterotrait-monotrait ratio (HTMT). Generally, AVE values > 0.5 and CR values > 0.7 indicate high convergent validity [41], and HTMT values < 0.9 indicate good discriminant validity between factors [42].
Data collection and quality control
Both rounds of the questionnaire survey were distributed and collected using the “Wen juan xing” platform. The questionnaire design included detailed instructions at the beginning, at sections, and on separate pages to guide respondents in filling out the questionnaire. The distribution process involved first sending a uniform instruction message followed by the questionnaire link to ensure that respondents fully understood the content. To ensure the quality of the questionnaires, the questions were set as mandatory, and each device could only submit the questionnaire once. Trained researchers distributed the questionnaires and were available to answer any questions from the respondents. Data were double-checked and organized, and questionnaires were deleted if the response time was less than 90 s, if the hospital name was outside Changsha, if age, work experience, and professional title were clearly mismatched, or if answers followed a repetitive pattern.
Statistical analysis
Data were organized and analyzed using Microsoft Excel, IBM SPSS 26.0, and Amos 22.0 software. Measurement data following a normal distribution were described using mean and standard deviation, while non-normal distribution data were described using median and quartile range. Categorical data were described using frequency and composition ratio. The reliability of the expert consultation was evaluated using the response rate, expert authority coefficient, coefficient of variation, and Kendall’s W coefficient. Item analysis was used to screen items, exploratory factor analysis was used to explore the structure of the scale, and confirmatory factor analysis was used to test the validity of the scale structure. The reliability of the scale was assessed using Cronbach’s alpha coefficient, split-half reliability, and test-retest reliability. A P-value of less than 0.05 was considered statistically significant.
Ethical considerations
This study was approved by the ethics committee of Xiangya Hospital of Central South University (202212303). Before the survey began, the purpose, methods, process, risks, and benefits of the study were explained to the participants. Participation was completely voluntary, and respondents could choose to withdraw at any time without any impact on their nursing work. All data and personal information collected in this study will be kept confidential and anonymous within the legal limits. The identification of study subjects was only for sample coding, and they will not be identifiable in the presentation of the research results.
Results
Stage 1 initial scale construction
Delphi method results
A total of 20 experts were consulted, representing 12 tertiary hospitals and medical schools across 7 provinces in China, ensuring geographic representativeness. Among them, 19 experts (95%) were over 40 years old, 14 (70%) had a master’s degree or higher, 16 (80%) had associate senior titles or above, and 16 (80%) had over 20 years of work experience. Their research fields included clinical nursing (7 experts, 35%), humanistic nursing (4 experts, 20%), psychological nursing (4 experts, 20%), nursing management (4 experts, 20%), and scale research (1 expert, 5%), ensuring professional representativeness. The effective response rates for the two rounds of expert consultation questionnaires were 100% and 95%, respectively. In the first round, 16 experts provided 131 modification suggestions, and in the second round, 11 experts provided 30 suggestions, indicating high enthusiasm. The Cr for the two rounds were 0.837 and 0.861, and Kendall’s W were 0.107 and 0.203 (P < 0.05), indicating high authority and good coordination among experts’ opinions. All dimensions and items met the criteria of an average score > 3.5 and a full score ratio > 30%, with 33 items having CV > 0.25, indicating concentrated expert opinions.
Based on comprehensive screening criteria and expert suggestions, 20 items were modified, 9 were merged, 2 were transferred, 8 were deleted, and 10 new items were added. The initial scale draft included 36 items across 4 dimensions: treatment task, physiological comfort, individualized assistance, and emotional support.
Pre-survey results
During the pre-survey, respondents found the items clear and easy to understand, with no modifications or additional content needed.
Stage 2 item optimization and formation of the predictive version of the scale
General information of study subjects
For factor analysis, the item-to-sample ratio is typically 1:5 to 1:10 [41]. With 36 items in the initial scale draft, at least 180 samples were needed for the first-round survey. A total of 245 questionnaires were distributed, with 14 invalid and 231 valid responses. The general information of the respondents is shown in Table 1.
Table 1
General demographic data
Characteristics | Categories | First questionnaire (n = 231) | Second questionnaire(n = 355) | ||
|---|---|---|---|---|---|
Number(n) | Percentage(%) | Number(n) | Percentage(%) | ||
Department | Obstetrics and gynecology | 14 | 6.1 | 117 | 33.0 |
Department of infectious diseases | 29 | 12.5 | 118 | 33.2 | |
Emergency department | 15 | 6.5 | 23 | 6.5 | |
Internal medicine | 78 | 33.8 | 39 | 11.0 | |
Surgery department | 67 | 29.0 | 44 | 12.4 | |
Intensive care unit | 28 | 12.1 | 14 | 3.9 | |
Sex | Male | 20 | 8.7 | 19 | 5.4 |
Female | 211 | 91.3 | 336 | 94.6 | |
Age (year) | < 25 | 35 | 15.1 | 62 | 17.5 |
25 ∼ 29 | 38 | 16.4 | 78 | 22.0 | |
30 ∼ 34 | 90 | 39.0 | 116 | 32.7 | |
35 ∼ 39 | 48 | 20.8 | 52 | 14.6 | |
≥ 40 | 20 | 8.7 | 47 | 13.2 | |
Career(year) | < 5 | 46 | 19.9 | 91 | 25.6 |
5 ∼ 9 | 56 | 24.2 | 95 | 26.8 | |
10 ∼ 14 | 90 | 39.0 | 100 | 28.2 | |
15 ∼ 19 | 23 | 10.0 | 33 | 9.3 | |
≥ 20 | 16 | 6.9 | 36 | 10.1 | |
Marital status | Separated/Divorced/Widowed | 4 | 1.7 | 3 | 0.8 |
Unmarried | 72 | 31.2 | 117 | 33.0 | |
Married | 155 | 67.1 | 235 | 66.2 | |
Education level | Below bachelor degree | 29 | 12.6 | 38 | 10.7 |
Bachelor degree | 186 | 80.5 | 282 | 79.4 | |
Master’s degree or above | 16 | 6.9 | 35 | 9.9 | |
Professional title | Nurse | 39 | 16.9 | 75 | 21.1 |
Nurse practitioner | 53 | 22.9 | 76 | 21.4 | |
Nurse-in-charge or higher | 139 | 60.2 | 204 | 57.5 | |
Position | General Nurse | 226 | 97.8 | 347 | 97.7 |
Nursing supervisor | 5 | 2.2 | 8 | 2.3 | |
Item analysis results
(1) Item discrimination analysis: The critical ratio for each item ranged from 9.947 to 17.062, with statistically significant differences (P < 0.05). (2) Cronbach’s alpha coefficient method: The overall Cronbach’s alpha coefficient for the initial scale draft was 0.980, with no items showing a significant increase in Cronbach’s alpha coefficient if deleted. (3) Commonality and factor loading: Item A5 (touching the patient while handling a corpse) had a commonality value of 0.296 (< 0.3) and was deleted. (4) Correlation coefficient method: The correlation coefficients between each item and the total scale ranged from 0.565 to 0.848, and the correlation coefficients with their respective dimensions ranged from 0.641 to 0.948, all higher than those with other dimensions. Therefore, item A5 was deleted, and 35 items were retained.
Exploratory factor analysis results
The KMO value for the scale was 0.960, and Bartlett’s sphericity test yielded χ² = 9919.847 (P = 0.000 < 0.001), indicating suitability for factor analysis. Using principal component analysis and orthogonal rotation with varimax, 4 common factors with eigenvalues ≥ 1 were extracted, with a cumulative variance contribution rate of 76.661%. All item factor loadings after rotation were > 0.4 (range 0.531 to 0.859) and aligned with the theoretical model dimensions, with no items deleted. The final predictive version of the scale consisted of 4 dimensions and 35 items. Factor loadings for each item are shown in Table 2, and the scree plot is shown in Fig. 2.
Table 2
Results of exploratory factor analysis rotated component matrix (n = 231)
Item | D | B | C | A |
|---|---|---|---|---|
A1 Touching the patient to take vital signs (temperature, pulse, respiration, blood pressure). | 0.310 | 0.247 | 0.171 | 0.759 |
A2 Touching the patient for ECG monitoring. | 0.287 | 0.297 | 0.232 | 0.786 |
A3 Touching the patient during treatment-oriented (e.g., injection, blood sampling, blood sugar testing, nebulization, oxygen inhalation). | 0.312 | 0.242 | 0.247 | 0.788 |
A4 Touching the patient to assess condition (e.g., touching chest, abdomen, arteries, lymph nodes). | 0.256 | 0.312 | 0.203 | 0.755 |
A5 Touching the patient for hot/cold compresses (including cooling sponge baths). | 0.282 | 0.492 | 0.220 | 0.608 |
A6 Touching the patient during CPR. | 0.232 | 0.190 | 0.354 | 0.611 |
B1 Washing and drying the patient’s face. | 0.344 | 0.736 | 0.131 | 0.346 |
B2 Washing and drying the patient’s feet. | 0.107 | 0.819 | 0.196 | 0.193 |
B3 Washing and drying the patient’s hands. | 0.374 | 0.773 | 0.190 | 0.186 |
B4 Touching the patient during oral care. | 0.228 | 0.663 | 0.364 | 0.354 |
B5 Touching the patient during perineal care. | 0.141 | 0.641 | 0.364 | 0.332 |
B6 Touching the patient during hair cleaning care. | 0.202 | 0.799 | 0.270 | 0.241 |
B7 Touching the patient to assist with bowel movements (including stoma care, enemas). | 0.033 | 0.675 | 0.438 | 0.176 |
B8 Touching the patient during skin care (including trimming fingernails/toenails). | 0.198 | 0.697 | 0.374 | 0.256 |
B9 Touching the patient to relax tense muscles, such as massaging the patient. | 0.503 | 0.563 | 0.340 | 0.217 |
B10 Touching the patient to relieve tension and promote sleep, such as massaging the head of an insomniac patient. | 0.471 | 0.681 | 0.212 | 0.146 |
B11 Massaging bedridden patients to make them comfortable, such as massaging the head, arms, or legs. | 0.499 | 0.670 | 0.228 | 0.216 |
B12 Performing aromatherapy for terminal patients to relieve pain (aromatherapy is a natural therapy that uses plant essential oils through massage, incense, bathing, etc., to help patients restore health). | 0.433 | 0.623 | 0.217 | 0.231 |
C1 Touching the patient to help them eat or tube feed. | 0.361 | 0.354 | 0.680 | 0.214 |
C2 Touching the patient to assist bedridden patients in changing positions (e.g., turning, walking, sitting in a wheelchair). | 0.433 | 0.263 | 0.675 | 0.201 |
C3 Touching the patient during drainage tube care (e.g., changing drainage bags, squeezing drainage tubes). | 0.346 | 0.366 | 0.696 | 0.288 |
C4 Touching the patient during beating expectoration or sputum suctioning. | 0.344 | 0.301 | 0.668 | 0.267 |
C5 Touching the patient to help change bed linens or clothes. | 0.400 | 0.432 | 0.541 | 0.299 |
C6 Touching the patient to apply protective restraints to restless patients. | 0.295 | 0.357 | 0.669 | 0.229 |
C7 Touching the patient during wound care. | 0.214 | 0.364 | 0.676 | 0.325 |
D1 Touching the patient to relieve pain, such as holding the patient’s hand. | 0.628 | 0.569 | 0.208 | 0.169 |
D2 Providing calming touch to agitated patients. | 0.531 | 0.513 | 0.328 | 0.212 |
D3 Listening and accompanying lonely and helpless patients, such as placing a hand on the patient’s shoulder or arm. | 0.726 | 0.236 | 0.342 | 0.231 |
D4 Comforting sad patients, such as wiping tears and patting the shoulder. | 0.747 | 0.345 | 0.228 | 0.248 |
D5 Encouraging depressed patients, such as holding the patient’s hand to show understanding. | 0.823 | 0.236 | 0.262 | 0.277 |
D6 Soothing anxious patients, such as holding the hand of a preoperative patient. | 0.859 | 0.198 | 0.211 | 0.280 |
D7 Calming worried and fearful patients, such as holding the patient’s hands or gently patting the shoulder. | 0.855 | 0.211 | 0.253 | 0.245 |
D8 Helping pessimistic and hopeless patients build faith and hope, such as holding the patient’s hands. | 0.852 | 0.204 | 0.269 | 0.213 |
D9 Providing humanistic care to terminal patients. | 0.726 | 0.263 | 0.254 | 0.175 |
D10 Touching the patient to share joy when the patient’s condition improves. | 0.838 | 0.138 | 0.107 | 0.244 |
Eigenvalue | 8.518 | 8.335 | 5.073 | 4.905 |
Variance contribution rate(%) | 24.337 | 23.814 | 14.494 | 14.016 |
Cumulative variance contribution rate(%) | 24.337 | 48.151 | 62.645 | 76.661 |
Factor naming | Emotional support | Physiological comfort | Individualized assistance | Treatment task |
Fig. 2
Scree plot of the nurses’ touch comfort evaluation scale
×
![]()
Stage 3 Performance evaluation and formation of the final version of the scale
General information of study subjects for scale performance testing
The predictive scale had 35 items. For the second-round survey, at least 175 samples were required. We collected 383 questionnaires, eliminating 28 invalid ones, resulting in 355 valid responses, meeting the sample size requirements for factor analysis. The general information of the respondents is shown in Table 1.
Reliability testing
The overall Cronbach’ s alpha coefficient for the scale was 0.981; The Cronbach’ s alpha coefficients for each dimension ranged from 0.928 to 0.972; The split-half reliability coefficient for the overall scale was 0.927; The split-half reliability coefficients for each dimension ranged from 0.903 to 0.956; The test-retest reliability for the overall scale was 0.886, with each dimension ranging from 0.767 to 0.875. Detailed results are shown in Table 3.
Table 3
Aggregate reliability and validity of the scale
Factor | Item | Standardized factor loading | S.E. | P | AVE | CR | Cronbach’s alpha | Half reliability | Test-retest reliability |
|---|---|---|---|---|---|---|---|---|---|
A.Treatment task | A1 | 0.878 | - | - | 0.700 | 0.933 | 0.928 | 0.903 | 0.767 |
A2 | 0.910 | 0.041 | *** | ||||||
A3 | 0.893 | 0.044 | *** | ||||||
A4 | 0.863 | 0.046 | *** | ||||||
A5 | 0.770 | 0.050 | *** | ||||||
A6 | 0.682 | 0.061 | *** | ||||||
B.Physiological comfort | B1 | 0.824 | - | - | 0.671 | 0.961 | 0.960 | 0.920 | 0.802 |
B2 | 0.845 | 0.051 | *** | ||||||
B3 | 0.831 | 0.051 | *** | ||||||
B4 | 0.754 | 0.055 | *** | ||||||
B5 | 0.796 | 0.065 | *** | ||||||
B6 | 0.779 | 0.054 | *** | ||||||
B7 | 0.872 | 0.055 | *** | ||||||
B8 | 0.821 | 0.060 | *** | ||||||
B9 | 0.847 | 0.055 | *** | ||||||
B10 | 0.777 | 0.063 | *** | ||||||
B11 | 0.833 | 0.061 | *** | ||||||
B12 | 0.842 | 0.047 | *** | ||||||
C.Individualized assistance | C1 | 0.824 | - | - | 0.703 | 0.943 | 0.943 | 0.942 | 0.830 |
C2 | 0.827 | 0.048 | *** | ||||||
C3 | 0.862 | 0.050 | *** | ||||||
C4 | 0.825 | 0.054 | *** | ||||||
C5 | 0.857 | 0.053 | *** | ||||||
C6 | 0.811 | 0.054 | *** | ||||||
C7 | 0.863 | 0.051 | *** | ||||||
D.Emotional support | D1 | 0.831 | - | - | 0.780 | 0.972 | 0.972 | 0.956 | 0.875 |
D2 | 0.867 | 0.050 | *** | ||||||
D3 | 0.740 | 0.056 | *** | ||||||
D4 | 0.873 | 0.049 | *** | ||||||
D5 | 0.913 | 0.048 | *** | ||||||
D6 | 0.908 | 0.049 | *** | ||||||
D7 | 0.940 | 0.046 | *** | ||||||
D8 | 0.953 | 0.045 | *** | ||||||
D9 | 0.934 | 0.047 | *** | ||||||
D10 | 0.850 | 0.052 | *** |
Validity testing
Content validity The I-CVI, S-CVI/UA, and S-CVI/Ave were both 1, indicating strong content validity.
Construct validity Using AMOS 22.0 software for confirmatory factor analysis, the model fit indices deviated somewhat from ideal values due to the model’s complexity and high modification indices (MI). After adjusting the residuals with high MIs, the model fit indices were as follows: χ²/df = 3.432, RMSEA = 0.083, GFI = 0.767, CFI = 0.911, IFI = 0.911, TLI = 0.902, PCFI = 0.830, PNFI = 0.801. These indices indicate a good model fit. The path diagram for the scale model is shown in Fig. 3. The CR values ranged from 0.933 to 0.972, the AVE values ranged from 0.671 to 0.780, and the HTMT values ranged from 0.780 to 0.886 for each dimension. Detailed values are shown in Tables 3 and 4. The final version of the scale is shown in Supplementary Material 1.
Fig. 3
Confirmatory factor analysis model diagram of the nurses’ touch comfort evaluation scale
×
![]()
Table 4
HeterotraitMonotrait ratio of discriminant validity evidence
Heterotrait-monotrait ratio (HTMT) | 1 | 2 | 3 | 4 |
|---|---|---|---|---|
1.Treatment task | - | |||
2.Physiological comfort | 0.791 | - | ||
3.Individualized assistance | 0.826 | 0.886 | - | |
4. Emotional support | 0.780 | 0.785 | 0.835 | - |
Discussion
Innovative and significant scale development
With the advancement of the “Healthy China” initiative, humanized nursing has become a focus for building a high-quality and efficient nursing service system. In China, nurses are the primary group in contact with patients. Focusing on nurses’ touch comfort can significantly improve the level of humanized nursing services and greatly impact patient outcomes [11]. Existing assessment tools [1, 7] were developed based on Western cultural orientations of openness and freedom in interpersonal communication. While these tools address various moments of touch within nursing procedures, they do not encompass most touch-related scenarios in the Chinese clinical nursing context. Additionally, the content of the items differs significantly from the realities of Chinese clinical practice and the reserved, subtle interpersonal communication habits of Chinese people, which can lead to biased results when applied to assess the touch comfort of Chinese nurses. To address cultural differences and make the assessment more comprehensive, this study, guided by Watson’s human caring theory and drawing from the clinical nursing experiences of Chinese nurses, deeply explored the various moments when nurses touch patients. It extracted culturally specific measurement items for touch comfort and developed the nurses’ touch comfort evaluation scale in China. The dimensions of the scale developed in this study align with the framework of caring behaviors in Watson’s caring theory. It not only includes all the elements measured by existing tools [1, 7] for nurses’ touch comfort (treatment task, physiological comfort, and emotional support) but also adds the dimension of individualized assistance. This addition integrates the essence of Chinese caring culture, which emphasizes respecting patients’ individualized needs, and provides a more comprehensive framework for research in this field. When designing the items, particular attention was paid to cultural and regional differences. For example, in the treatment task dimension, some items from previous tools [1, 7], such as “Touching the patient’s arm to measure blood pressure.” and “Touching the patient to take his or her pulse.”, are similar to items in this scale. However, we recognized that during electrocardiogram (ECG) monitoring, exposure of the patient’s chest, potentially touching private areas such as the breasts, could cause embarrassment for both Chinese nurses and patients. This differs significantly from the evaluation of routine vital signs. Therefore, we specifically added the item “Touching the patient for ECG monitoring.”. In the physiological comfort dimension, Pedrazza’s scale [1] focuses on the physical relaxation brought about by massage, while this scale emphasizes a combined state of cleanliness comfort and physical relaxation, highlighting the Chinese holistic nursing perspective centered on humanity. Additionally, we incorporated touch concepts that are less easily understood by Chinese nurses in existing tools, such as “Massaging”, into specific clinical nursing contexts in China. For instance, we added the item “Touching the patient to relieve tension and promote sleep, such as massaging the head of an insomniac patient.” These revisions improve the practicality of the scale during assessment and enhance the accuracy of evaluation results. In the emotional support dimension, the impact of positive emotions on touch comfort was well-considered in our study. For example, we included the item “Touching the patient to share joy when the patient’s condition improves.” This important aspect appears to have been overlooked in previous tools [1]. Ultimately, the scale developed in this study provides a precise and culturally appropriate description of nurses’ touch comfort in the context of Chinese nursing culture. The content addresses the provision of care that meets diverse physiological and psychological needs, encompassing care for individuals across different life stages, from illness to health and even end-of-life care. It represents a touch-centered care framework that spans the entire human life cycle. This scale not only enriches the content of China’s health management system, which covers the full life cycle [43] but also offers a new perspective for applying caring theory to guide humanistic nursing practices.
Moreover, this study has significant practical implications. By self-assessing touch comfort, clinical nurses can promptly identify and improve their uncomfortable touch experiences, maintain a positive mindset for nursing work, and actively safeguard their mental and physical health. It also allows them to recognize deficiencies in their humanistic care, enhance their humanistic qualities, improve nursing practice quality, build harmonious nurse-patient relationships, and increase patient satisfaction. On the other hand, evaluating the touch comfort of clinical nurses helps nursing managers understand the psychological state of nurses, implement effective psychological interventions for those with negative emotions, stabilize the nursing team, identify deficiencies in care from different perspectives, and develop scientific care management plans. This approach fosters nurses’ care abilities, motivates proactive care behavior, and ensures high-quality nursing services, ultimately safeguarding patient safety.
Reliability evaluation of the scale
The overall Cronbach’s alpha coefficient for the scale was 0.981, and the split-half reliability was 0.927. For each dimension, Cronbach’s alpha coefficients and split-half reliability coefficients were both greater than 0.900, indicating excellent internal consistency and high precision [24]. The test-retest reliability for the overall scale was 0.886, with each dimension’s test-retest reliability exceeding 0.700, suggesting good stability [24]. In summary, the scale demonstrates excellent reliability.
Validity evaluation of the scale
In the method of exploratory factor analysis, four factors with eigenvalues greater than 1 were extracted, with a cumulative variance contribution rate of 76.661%. All item factor loadings after rotation were greater than 0.5. The four determined dimensions were treatment task, physiological comfort, individualized assistance, and emotional support, which align well with the theory of care and are similar to the dimensions identified in Pedrazza et al.’ s [1] touch comfort scale and common touch classification methods [44‐46]. This indicates a reasonable structure of the scale [34].
Anzeige
In the method of confirmatory factor analysis, the model fit indices such as χ²/df, CFI, IFI, TLI, PCFI, and PNFI met acceptable standards, and RMSEA was close to the acceptable threshold, indicating a good fit between the empirical data and the model [40]. Each dimension’s CR value was greater than 0.7 [41], the AVE value was greater than 0.5 [41], and the HTMT value was less than 0.90 [42]. This suggests good convergent and discriminant validity of the scale.
At the item level, the I-CVI was 1, and at the scale level, both the S-CVI/UA and S-CVI/Ave were 1, indicating excellent content validity.
Limitations of the study
The scale developed in this study provides a tool to gather raw data on the current state of touch comfort among Chinese nurses. Based on the research findings, the limitations and recommendations for future research are as follows. The first is sampling bias. This study utilized convenience sampling, with participants limited to nurses from hospitals in Changsha, which may introduce bias to the results. Future studies should consider conducting multi-center investigations across hospitals in different provinces and cities to enhance the generalizability of the findings. The second is longitudinal validation. Future research could implement longitudinal studies to assess the reliability of the scale across different time points, providing a deeper understanding of its stability over time. The third is expanded validity assessment. Future studies should evaluate other validity scores, such as concordant validity or discriminant validity, to further strengthen the psychometric assessment of the scale. These improvements will contribute to further validating and enhancing the utility of the scale for broader applications.
Conclusion
The nurses’ touch comfort evaluation scale developed in this study comprises 35 items across four dimensions: treatment task, physiological comfort, individualized assistance, and emotional support. This scale demonstrates acceptable reliability and validity, making it a reliable and effective tool for measuring the touch comfort of clinical nurses in China. Future studies can employ this scale for multicenter surveys across different provinces and hospitals to comprehensively understand the current state of nurse touch comfort, providing a reference for formulating touch management intervention strategies.
Anzeige
Acknowledgements
We would like to thank all the experts and clinical nurses who participated in this study.
Declarations
Ethics approval and consent to participate
This study was approved by the Clinical Medical Ethics Committee of Xiangya Hospital, Central South University (202212303). All participants provided informed consent. All methods were carried out in accordance with relevant guidelines and regulations.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Anzeige
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
