Translation, adaptation, and validation of the Filipino version of the Caring Behaviors Inventory

Study Design

This study employed a descriptive, cross-sectional design guided by various published processes and approaches in the translation, adaptation, and validation of instruments (MAPI Research Institute, 2009; World Health Organization, 2016).

Translation, Adaptation, and Validation of the Instrument

Each of the CBI’s original subscales had a Cronbach’s alpha range of 0.81 to 0.92 (Wolf et al., 1998). The original CBI was revised through psychrometric processes, resulting in a 16-item instrument (CBI-16) (Wolf et al., 2017). The CBI-16 measures the perceptions of caring behaviors on a 6-point Likert scale. It has internal consistency reliability of 0.95, suggesting strong reliability. This version was then translated and adapted into Filipino by the investigators to become the new CBI-16-FIL. In this study, the original CBI-16 was administered to the nurses, while the CBI-16-FIL was administered to the patients. The primary author (ZW) gave written permission to use the original English 16-item Caring Behaviors Inventory (CBI-16).

Instrument Translation Process

The CBI-16 was forward translated into the Filipino language independently by a linguist of the Sentro ng Wikang Filipino (Filipino Language Center) in the University of the Philippines (UP) Manila and a bilingual nurse. To evaluate the content validity of the instrument and address discrepancies between the forward translation and the existing comparable previous versions, as well as inadequate expressions or concepts of the translation, a bilingual expert panel was assembled. It was composed of a nurse administrator, staff nurse involved in direct patient care, tool development expert, patient representative or advocate, and nursing instructor/lecturer in the academic setting. The expert panel was asked to evaluate (1) whether the wording of the items was clear, (2) the relevance of each item on the construct of nurse caring behavior on a 4-point Likert scale, and (3) whether the item will be retained, revised, or dropped. Any ambiguities and discrepancies were discussed and resolved by consensus.

The instrument was translated back into English by two independent translators using the same method as the forward translation. In order to assess how closely the two back-translated versions matched the original in terms of phrasing, sentence structure, meaning, and significance, they were combined and compared. In a pilot test, ten participants (five patients, five nurses) were given the CBI-16-FIL, the Filipino translation of the instrument.

The newly translated, adapted, and the cross-validated instrument was administered to a sample of the target population. In addition, the preliminary version was distributed to staff nurses and patients who are receiving nursing care. This research was carried out in a tertiary public hospital, one of the country’s largest, with a capacity of 1,500 beds. The hospital caters to about 600,000 patients every year with around 4,000 employees.

Content Validity

To determine if the items of the CBI were applicable to the Filipino context of caring, two rounds of content validation were done with two separate panels of five experts each. The experts were then asked to rate each item on clarity (yes or no), relevance to the concept (scale of 1 to 4, with 4 being the highest), and whether it should be retained, revised, or dropped.

Samples/Participants for Construct Validity

Participants in the study were chosen via purposive sampling. Nurses included in this study were at least 21 years old with at least three months of working experience in the hospital, while patients were at least 18 years old and admitted to the hospital for at least three days. Incapacitated patients and those unable to read and write were excluded from this study. In addition, nurses and patients in administrative offices, intensive care units, and other specialty areas were excluded. Due to the COVID-19 pandemic, only nurses and patients in non-COVID units were invited to participate in the psychometric validation of the instrument to limit exposure and reduce disease transmission. In addition, this study followed the general rule of thumb, stating that there should be at least 10 participants for each scale item to establish the first psychometric features of the newly translated instrument (Gunawan et al., 2021; Nunnally & Bernstein, 1978).

Data Collection and Analysis

Data for the study was gathered using a self-administered questionnaire. Data encoding accuracy was checked against the actual instrument. Duplicate entries and data with blank responses were removed as part of the data cleaning process.

STATA 14.1 (StataCorp, Texas) was used for data cleaning and analysis. Based on the consensus of the experts over whether each item will be retained, an item content validity index (I-CVI) was calculated. The average of the I-CVIs was computed as the scale CVI (S-CVI). A cut-off of 0.80 was set as an acceptable content validity index. Internal consistency was assessed with the use of Cronbach’s alpha.

The factor structure was identified using exploratory factor analysis with the principal factors method and a cut-off eigenvalue of 1.0. Using descriptive statistics, the characteristics of the respondents and their responses were examined. Inferential statistics such as the independent t-test, one-way ANOVA, and Pearson’s correlation coefficient were used to examine the differences in nurse caring behavior perspectives and the correlations between different research variables. Post-hoc pairwise comparisons with Bonferroni correction were done to determine pairwise differences in comparisons of three or more groups using One-way ANOVA.

The COSMIN Reporting Guideline for Studies on Measurement Properties of Patient-Reported Outcome Measures was followed in the reporting of this study (Gagnier et al., 2021).

Ethical Considerations

This study has been reviewed and approved by the UP Manila Research Ethics Board (UPMREB 2018-395-01) and the UP Philippine General Hospital (PGH) Expanded Hospital Research Office. Prior to taking part in this study, each participant was provided written informed consent and informed of their right to withdraw their participation at any time during or after the study.

Content Validity and Target Population Input on the Translated CBI-16-FIL

In the first iteration of the expert panel review, the S-CVI was 0.70, with eight items not reaching an item CVI (I-CVI) of 0.80. After revising the wording of these items following the panel comments, the second round of review resulted in an S-CVI of 0.825. Only four items had an I-CVI of 0.60 (items 6, 9, 13, and 14):

• CBI2: Giving instructions or teaching the patient

• CBI7: Being confident with the patient

• CBI11: Returning to the patient voluntarily

• CBI16: Relieving the patient’s symptoms

These items were reviewed again by the researchers, and a final wording was decided. All other items had I-CVIs greater than or equal to 0.80. Regarding the relevance rating, all items were rated with a score of 4.

Cognitive debriefing after the pilot testing with ten patients revealed that the CBI was easily understandable by the participants. There were no questions regarding the response options and the recall period in the instrument. Therefore, no further revisions were made to the translated instrument after this pilot. The final version of the translated CBI-16-FIL is available in a Supplementary Appendix.

Characteristics of the Participants

A total of 180 patients and 142 nurses participated in this study. There were 11 patients who had an incomplete response, resulting in 169 patients included in the foregoing analyses. The nurses (Table 1) had a mean age of 36.23 years old (SD = 8.79) and a mean work experience in the hospital of 10.77 years (SD = 8.36). They were mostly females (n = 112, 78.87%), single (n = 77, 54.23%), and had BSN as their highest educational attainment (n = 137, 96.48%). More than two-thirds of the nurses were from charity/service wards, where patient care is free for those indigents. Their areas of the assignment were also fairly distributed among medical, surgical, and mixed (medical and surgical) wards.

Patients (Table 2) had a mean age of 44.52 years old (SD = 19.19) and length of stay of 15.18 days (SD = 21.98). Majority were males (n = 92, 54.44%), single (n = 84, 49.70%) married (n = 77, 45.56%), high school graduates (n = 99, 58.58%), and unemployed (n = 95, 56.21%). Most patients came from the charity/ service wards (n = 156, 92.31%).

Psychometric Properties of the CBI-16-FIL

Table 3 shows the internal consistency and factor loadings of each CBI item from the nurses’ and patients’ responses. The original CBI-16 and CBI-16-FIL had excellent internal consistency, with overall Cronbach’s alphas of 0.93 and 0.95, respectively. None of the items would result in a considerable increase or decrease in Cronbach’s alpha if deleted.

The Kaiser-Meyer-Olkin Measures of Sampling Adequacy were 0.916 and 0.932 for the CBI-16 and CBI-16-FIL, respectively. Bartlett’s test of sphericity was not significant for both instruments (X² = 1189.3, p <0.001, and X² = 2059.30, p <0.001, respectively). These indicate the fit of the data for factor analysis. Exploratory factor analysis of the responses on the CBI-16 and CBI-16-FIL revealed a one-factor solution. This factor accounted for 85% of the eigenvalues for CBI-16 and CBI-16-FIL.

Perceptions of Nurse Caring Behaviors

While the original CBI-16 did not provide interpretations of the total instrument scores, the mean scores on the CBI-16 (M = 85.77, SD = 8.02) and CBI-16-FIL (M = 86.96, SD = 11.48) indicate a relatively high perception of nurse caring behaviors (89% of highest score for nurse responses and 91% for patient responses). The mean scores and comparisons are presented in Table 4. The difference between the overall perceptions of nurses and patients was not significant (p = 0.30). However, five individual items on the CBI were perceived differently by nurses and patients:

• CBI3: Treating the patient as an individual

• CBI4: Spending time with the patient

• CBI10: Treating patient information confidentially

• CBI13: Meeting the patient’s stated and unstated needs

• CBI15: Giving the patient’s treatments and medication on time

Relationships between Respondent Characteristics and Perceptions on Caring Behaviors

There were sex differences in terms of perceptions of nurse caring behaviors, with female nurses (p = 0.02) and patients (p = 0.05) reporting higher scores than males (Table 5). In addition, nurses from surgical wards reported higher views of nurse caring behaviors (p = 0.002), while patients who were separated had lower scores on the CBI-16-FIL (p = 0.01) (Table 6). However, other nurses’ and patients’ characteristics were not significantly related to perceptions of nurse caring behaviors.

Psychometric Properties of the CBI-16-FIL

In various countries, the psychometrics of the CBI with varied items and emergent factors have been evaluated. As shown in various studies, the Cronbach’s alpha of the tool ranged from 0.90 to 0.96 (Romero Martin et al., 2019). The CBI had strong validity with four components and good reliability with a Cronbach’s alpha of 0.93, according to a psychometric analysis of the 24-item assessment in Greece (Alikari et al., 2021). Wolf et al. (2017) reduced a 24-item tool to a 16-item tool. In Turkey, a psychometric investigation of the 42-item CBI found that a 30-item CBI with three variables had high reliability and validity (Gul & Dinc, 2020).

This Philippine study found a single-factor solution to the CBI-16 for patient and nurse samples. This is consistent with the first study that reported the CBI-16 in 2017, where items were removed from the CBI-24R following the frequency of missing responses and patient comments (Wolf et al., 2017). However, cross-cultural translation and validation of the CBI-16 among patients revealed contradicting solutions, with an Iranian version showing a two-factor structure while a Greek version was unidimensional (Alikari et al., 2021; Ghafouri et al., 2021). These differences in instrument form suggest that the idea of nurse caring behaviors as experienced by patients may be culturally variable. Leininger’s theory on transcultural nursing defines cultural diversity as the variations among and between groups resulting from differences in cultural aspects.

Given the wide range of factors documented in similar research conducted in many countries, it is reasonable to conclude that culture can influence patients’ views of nurses’ caring actions. The findings of our study also show that CBI-16-FIL has a high level of validity and reliability and that it can be used to assess nurses’ caring behaviors.

Perceptions of Nurse Caring Behaviors

The results of our study report significantly different scores between nurses and patients on the following CBI items: treating the patient as an individual, spending time with the patient, treating patient information confidentially, meeting the patient’s stated and unstated needs, and giving the patients treatments and medications on time.

According to data from a prior study, patients value technical nursing procedures more than other activities because they are acknowledged by patients more often than other activities, despite the fact that these tasks occupy the majority of nurses’ time at the bedside (Papastavrou et al., 2011). According to Watson (1988), giving the patient instrumental care is essential to developing a caring relationship, yet this does not entirely sum up caring. In fact, nurses tend to link caring more with expressive and relational activities than with instrumental care. It would be necessary to address these differences between how nurses and patients perceive acts of kindness.

The items of spending time with the patient and treating the patient as an individual yielded noticeably different results. It is posited that the difference in scores may be related to the nurses’ lack of time to provide nursing care. Nurses are often overloaded with work due to overcrowded departments and hospitals. In other words, how the healthcare system is set up results in impersonal nursing (Nelson & Watson, 2012). Another possible reason may be related to family-oriented culture, wherein it is expected that the family’s responsibility to spend time and connect with their family members. Therefore, if hospitalization occurs, family is considered an integral part of caregiving (Feliciano et al., 2022).

The CBI item on treating patient information confidentially also reported significantly different scores. This may be because, as health care professionals, nurses consider respect for patient confidentiality an absolute duty (Noroozi et al., 2018). Informed consent, consideration for a patient’s privacy, as well as the disclosure of critical information are all essential factors in improving the quality of patient care offered, according to several studies (Susilo et al., 2014). Future research should include variables related to other characteristics of nurses and patients and organizational aspects of health institutions.

As the findings show, there are disparities in how nurses and patients care for each other. This could be due to various variables, including cultural and socioeconomic background, religious beliefs, differing care standards, personal differences, and assumptions between nurses and patients. In these areas, more research may be required.

Relationships between Respondent Characteristics and Perceptions on Caring Behaviors

The results of this study show that gender is significantly related to nurse caring behaviors, with female nurses reporting higher CBI scores. This may be attributed to the fact that female nurses possess a maternal instinct crucially needed as a nurse. Moreover, this maternal instinct allows them to be more sensitive and receptive to patients’ needs compared to their male counterparts (Wang et al., 2012).

In this study, surgical nurses also had higher CBI ratings. This finding is consistent with the results of a Philippine study where it was found that nurses prioritize the medical, technical, and therapeutic aspects of care over other caring behaviors (Tamayo et al., 2022). These results could be explained by the fact that surgery patients require greater physical care throughout the recovery period. Additional investigation is required to identify other personal characteristics, such as knowledge, experience, and abilities, that can influence nurses’ tendency to care for others. According to earlier reports in the literature, caring behaviors can also be affected by other traits, including commitment, responsibility, religious convictions, and personal philosophy (Salimi & Azimpour, 2013). It is also important to remember that nursing practices are significantly impacted by health infrastructure and resources.

Few studies have examined the relationship between patient characteristics and their perceptions of caring behaviors. The most common patient characteristics associated with their perceptions of caring behaviors were found to be age, gender, education, and kind of admission (Jonsdottir, 2002; von Essen & Sjödén, 1991). The findings of this study showed that patients who were female reported higher CBI scores than patients who were male. This finding is consistent with a prior study that found that female patients felt better-taken care of and had a more positive hospital experience than male patients (Chan et al., 2015).

Our study found no significant correlation between patient age and nurse caring behaviors, in contrast to another study’s findings that older patients are more favorable in their assessments of those activities (Noroozi et al., 2018). Therefore, it is hypothesized that nurses prefer to pay more attention to senior patients than younger ones or that older patients have lesser expectations of their nursing care (Patiraki et al., 2014).

A related study conducted on nurses and patients in diverse clinical settings discovered a high perception of nurse caring behaviors (Patiraki et al., 2014). Additionally, this study found agreement between nurses’ and patients’ perceptions of caring actions. However, many investigations indicated the opposite (Modic et al., 2014; Muhammad Esmaiel et al., 2012). These results contribute to a better understanding of the local patient experience, notably the quality of patient care. Ordonez and Gandeza (2004) suggest that the Philippine culture continues to influence Filipino nurses’ health beliefs, behaviors, and practices. Filipinos’ flexibility, adaptability, openness, and understanding of other people lead to patient confidence and satisfaction (Ordonez & Gandeza, 2004).

Strengths and Limitations

This study followed the processes and approaches in the translation, adaptation, and validation of the CBI-16, a version of the Caring Behaviors Inventory with less respondent burden than the longer versions of the instrument. The use of a locally developed tool allowed for more internal validity in studying patients’ satisfaction with nursing care. While the sample size for patients was adequate based on the 1:10 rule of thumb, the smaller sample size for nurses was still found to satisfy the assumptions for factor analysis. It is also important to note that further hypothesis testing on differences in perceptions between nurses and patients and the relationships of various study variables were undertaken on the same respondents as with the validation sample.

Furthermore, results might only be generalizable to the present study setting and population and other government hospitals catering to patients with similar socioeconomic statuses. Therefore, more validation research must involve larger sample sizes and different study settings, such as private tertiary hospitals.

Implications for Nursing Practice

The CBI-16-FIL can be used to measure the patient perceptions of nurse caring behaviors, together with the original CBI-16 for nurses. Results of this study highlight the potential need for interventions focusing on male nurses and non-surgical clinical areas to improve their nurse caring behaviors and drive up patient satisfaction scores and quality of care.