Cross-Cultural Adaptation of the Instrument “Nurse–Physician Relationship Survey: Impact of Disruptive Behavior in Patient Care” to the Spanish Context

Disruptive behavior in the healthcare context has an impact on patient care, healthcare personnel, and the health organization, and it also influences the therapeutic relationship, communication process, and adverse events. However, there is a lack of instruments that could be used for its analysis in the hospital care environment in the Spanish context. The objective of the study was to culturally adapt and perform a content validation of the tool “Nurse–Physician Relationship Survey: Impact of Disruptive Behavior on Patient Care”, to the Spanish content (Spain). An instrumental study was conducted, which included an analysis of conceptual and semantic equivalence. A panel of experts analyzed the translations, by analyzing the Content Validity Index (CVI) of the group of items in the scale through the Relevance Index (RI) and the Pertinence Index (PI). Only a single item obtained an RI value of 0.72, although with PI value of 0.81, with consensus reached for not deleting this item. The CVI of all the items was >0.80 for the mean value of the RI, as well as the PI. The instrument was adapted to the Spanish context and is adequate for evaluating the disruptive behaviors on nurse–physician relationships and its impact on patient care. However, the importance of continuing the analysis of the rest of the psychometric properties in future studies is underlined.


Introduction
Disruptive behavior is defined as any inadequate behavior, confrontation, or conflictranging from verbal abuse (abusive, intimidating, disrespectful, or threatening behavior), to physical or sexual abuse-that could negatively affect work relations, communication efficiency, and the caregiving process and its results [1][2][3].
In the healthcare context, the relationship between disruptive behavior and the negative consequences for the patient, for the healthcare personnel, and for the organization have been demonstrated, having an influence on the therapeutic relationship, communication process, and adverse events and dissatisfaction and compromised patient safety [4][5][6][7][8][9][10]. Moreover, Keller et al. [5] stressed that at the organizational level, the physical and emotional workload in services with a high variety of processes, can be a predictor of disruptive behaviors. For example, if communication is incorrect, or if there is some type of disruptive behavior, there is a greater probability that a mistake in care can be produced [11]. In 2006, the Joint Commission International (JCAHO) named disruptive behavior as a Sentinel Event Alert 40 due to its deconstructive effect on the culture of safety [6].
The study conducted by Rosenstein et al. [3] to analyze the disruptive behavior with respect to the nurse-physician relationship and its impact on patient care, is considered an important reference for addressing disruptive behaviors in area of healthcare. In this specific case, through the application of the original instrument used in our transcultural

Materials and Methods
Instrumental study of transcultural adaptation and validation of content of an instrument to the Spanish version (Spain).
For the methodological process of the study, 3 phases were implemented, following international guidelines [21]: First Phase: Transcultural adaptation (initial translation, synthesis of the translations, back-translation) [22]. Second Phase: Analysis of content validity through a panel of experts and pilot study. Third Phase: Test-retest reliability ( Figure 1).

Phase 1: Transcultural Adaptation
For the initial translation, a translation was made from English to Spanish (Spain) by three bilingual translators (mother tongue Spanish/Spain, with English language certification), independently. One of them knew the objective of the study, and a request was made that the translation had a semantic perspective (semantic equivalence to the Spanish context-without adding any new concept) [22,23].
In this phase, the three translators were asked to indicate dubious words and their corresponding suggested changes, and if the question was well formulated.
The stage of back-translation from Spanish (Spain) to English was performed by three bilingual translators (mother tongue English with a certification for the Spanish (Spain) language), independently, and began after the synthesis of the translations from the first stage.

Phase 1: Transcultural Adaptation
For the initial translation, a translation was made from English to Spanish (Spain) by three bilingual translators (mother tongue Spanish/Spain, with English language certification), independently. One of them knew the objective of the study, and a request was made that the translation had a semantic perspective (semantic equivalence to the Spanish context-without adding any new concept) [22,23].
In this phase, the three translators were asked to indicate dubious words and their corresponding suggested changes, and if the question was well formulated.
The stage of back-translation from Spanish (Spain) to English was performed by three bilingual translators (mother tongue English with a certification for the Spanish (Spain) language), independently, and began after the synthesis of the translations from the first stage.

Phase 2: Analysis of the Content Validity through a Panel of Experts and Pilot Study
For this phase, an adaptation of the conventional Delphi technique was utilized [24,25]. To provide a response to this analysis, a panel of experts was composed by utilizing two strategies. On the one hand, to recruit the experts, a search for information was conducted with managers in healthcare (titles, functions, years of service, participation in panels in related areas), who were identified with the help of databases of scientific evidence published.
After these steps, 28 health professionals were invited (nurses and/or physicians) by phone. An explanation was given as to the object of study, and they were told that participation was on an individual basis, without information provided on the composition of the panel [25]. Moreover, they were informed that that the confidentiality and protection of data were guaranteed, according to the Organic Law 3/2018, from 5 December, of Protection of Personal Data and guarantee of digital rights [26].

Phase 2: Analysis of the Content Validity through a Panel of Experts and Pilot Study
For this phase, an adaptation of the conventional Delphi technique was utilized [24,25]. To provide a response to this analysis, a panel of experts was composed by utilizing two strategies. On the one hand, to recruit the experts, a search for information was conducted with managers in healthcare (titles, functions, years of service, participation in panels in related areas), who were identified with the help of databases of scientific evidence published.
After these steps, 28 health professionals were invited (nurses and/or physicians) by phone. An explanation was given as to the object of study, and they were told that participation was on an individual basis, without information provided on the composition of the panel [25]. Moreover, they were informed that that the confidentiality and protection of data were guaranteed, according to the Organic Law 3/2018, from 5 December, of Protection of Personal Data and guarantee of digital rights [26].
The inclusion criteria were (a) ≥5 years of experience in hospitals (professionals from different units and different hospitals), (b) having previously participated in other research studies or publication in patient safety (associated in some manner to the subject under study: patient safety, disruptive behavior, and/or communication between health professionals), (d) capacity to provide opinions and suggestions, and (e) motivation and availability to participate in the study.
To evaluate the suitability of the panel of experts, the Competence Coefficient (K), represented by (K) = 0.5 (Kc + Ka), was calculated, with (Kc) being the Knowledge Coefficient, defined as the self-evaluation of the experts about their knowledge with respect to the subject studied, and (Ka) the Argumentation Coefficient, defined the arguments (sources) used to prove their knowledge [27]. Table 1 shows the score applied for Ka. After evaluating the value of the Coefficient of Competence (K), 11 experts were selected [28], who were provided with the translated versions of the scale for them to judge each of the items and the scale in general. This number of experts was considered sufficient for evaluating the content validity [27,29].
This evaluation was performed in two rounds. In the first round, the experts evaluated if the item was clear and comprehensible (consensus from at least 70% of the judges) [30]. In the second round, they evaluated the Relevance Index (RI), and the Pertinence Index (PI) of the items of the scale, with 0 = not pertinent/not relevant, and 1 = pertinent/relevant. The formula for its calculation was Number of experts who scored with a 1 divided by the total number of experts who evaluated the item. Lastly, the Content Validity Index (CVI) of the entire set of items of the scale was calculated (mean of the RI/PI of the items as a set). A value of ≥8 was considered as an accepted validity for each of the items, as well as for the Scale as a whole [28,30]. In this phase, the qualitative observations of the experts were collected, for each of the items that shaped the original instrument.
A pilot study was conducted with 10 nurses. The objective was to evaluate is understandability, acceptability, and time for completion of the scale for the evaluators.

Phase 3: Test-Retest Reliability
Finally, the test-retest reliability was analyzed using the Kappa index [31], the Intraclass Correlation Coefficient (ICC) [32] and the general concordance index according to the nature of the variable in a sample of 50 participants (nurses and physicians) from different hospital units/services, carrying out clinical or management activities. It has been applied in an interval of 15 days between test and retest.
As for the ethical aspects, the ethics and copyright principles were followed, with authorization solicited from Alan H. Rosenstein (the main author of the original instrument), and a commitment was made to provide information on the phases of the study and the evaluation of the final adapted version (before its publication).

Results
After the process of translation, the three translators created an online version where they indicated the dubious words. So that the best transcultural version was selected, the panel of experts, after analyzing the items in which these words were found, created a matrix of composition of the items, according to the analysis of clarity and comprehension of these items, and their cultural and idiomatic semantics ( Table 2). The panel of experts stressed the importance of adding the definition "disruptive behavior" at the start of the Scale (before the first question). Table 3 shows that for the most part, the experts obtained a high Coefficient of Competences (K), with values 0.8 < K < 0.9. However, one of them (expert 4) was excluded from the panel for not answering the item Argumentation coefficient (Ka), so that the composition was finally comprised by 11 experts. Table 4 shows the composition of the items after the content validity analysis. For the RI, the only item that obtained an RI value < 0.8 was item 5, "Is there a specialization in which disruptive behaviors are produced frequently?" being susceptible for elimination. However, this item obtained a PI of 0.81, and a consensus was made to not delete the item, according to the evaluation of the suggestions and opinions of the experts and study researchers. No items were eliminated or added with respect to the original instrument. Likewise, no changes were made to the response scales of the items. The CVI calculated for the complete scale was RI = 0.89, and PI = 0.94. "Disruptive behavior" was defined as any inappropriate behavior, altercation or conflict ranging from verbal abuse to physical or sexual abuse. One of the potential consequences of disruptive behavior is its effect on the collaboration and communication between physicians and nurses, which can lead to negative results in patient care. The present questionnaire was designed to identify the potential impact of disruptive behavior on adverse events, medical mistakes, patient safety, quality, and other aspects associated to care. Justification: There is a 100% consensus between the experts, who selected the term "disruptive behavior" considering the reference studies, including the original study on the instrument and the analysis of its concept. Moreover, an indication was made to add all the information before the first question (concept of disruptive behavior, objectives, and indications).

Médicos y enfermeros
Médicos y enfermeros/as Justification: Although a consensus of 70% was not found between the translators, a 95% consensus was found in the panel of experts, who decided to identify the nursing category with the translation "enfermeros/as", as in Spanish, the ending "os" and "as" refer to male or female nurses, respectively. Moreover, this is the idiomatic expression used in healthcare. However, in the physicians category, the word "medicos" is maintained, independently of the sex of the healthcare professional. Unidad en concreto ¿Hay alguna unidad en concreto en la cual el comportamiento disruptivo sea más frecuente? (Are there any particular settings where disruptive behavior is most prevalent?) Justification: there was a 95% consensus in the panel of experts leading to the decision of using the term "unidad" (unit), because the term "zona" (area) is a limited space that is most frequently used in geography or public administrations. On the other hand, the term "lugar" (place) can create confusion in its interpretation, for example lugar-puesto (place-position), lugar-cargo (place-load). In the Spanish context, it is more frequent to use "unidad" to identify a physical space, where at the level of hospitals, we find equipment and personnel for the monitoring and treatment of patients. indicate that healthcare institutions/organizations must implement a system of reporting that is inherent to a non-punitive culture and which adheres to the confidentiality of the personal data of those who report. nc * = self-evaluation of the experts about their level of knowledge, Scale from 1 to 10; fa * = sources of argumentation for items 1 to 6; NA * = No answered; K high * = 0.8 < K < 1.0; K medium * = 0.5 < K < 0.8.  After the cognitive pilot study, it was verified that all the nurses understood the writing and sense of the items. The questionnaire completion time was 10 min.
The sociodemographic and professional characteristics of the 50 participants used to measure temporal stability (test-retest reliability) were women (60%) with a mean age of 42 (SD = 11.6) years, 52% being physicians, and 74% of the participants performing their care work in the clinical setting.
As we can see in Table 4, most of the items presented a Kappa Index with values that indicated a concordance of the responses between moderate and almost perfect (16 items), ICC with values equal to or greater than 0.75 (indicating a reproducibility excellent) and general agreement indices of 100% in all items ( Table 4).
As the result of the study, a version that was transculturally adapted to Spanish (Spain) was created, given the consensus by the panel composed of 11 experts (Appendix A, Table A1).

Discussion
Like in other studies investigating the disruptive behavior [4][5][6][7][8][9][10], our results indicate that, after translating and adapting the "Nurse-Physician Relationship Survey: Impact of disruptive behavior on patient care" [1] to Spanish, the tool is a valid, ready to be used in Spain, and opening possibilities to study disruptive behaviors in Spain and compare the results internationally.
After a systematic review [12] on disruptive behavior, it was discovered that no studies existed in Spain on the construction or transcultural adaptation of an instrument on disruptive behavior in healthcare, and its impact on safe care. However, when the instrument "Nurse-Physician Relationship Survey: Impact of disruptive behavior on patient care" [1] was identified, it was observed that its approach could provide an answer to our questions, to identify aspects related with disruptive behaviors in healthcare, and its association to adverse events, satisfaction, communication, and improvement strategies. Moreover, we considered the international standing and prestige of the authors of the original version with respect to the subject studied, with this questionnaire used in various studies and in different specializations [2][3][4][5][6][7][8][9][10]20]. None of the studies found carried out a cross-cultural adaptation with the methodological rigor used in our study, and none of them carried out a reliability study. This makes it difficult to compare the results of our study with other similar studies. This justified our study, as it was agreed that a process of adaptation would be more efficient instead of the creation of a new instrument, which would be more time-consuming and costly.
The process of transcultural adaptation in the healthcare environment dates to the 1960s in mental health. However, perhaps with a different process of adaptation with "less rigor" and specificity for each instrument, given the scientific and methodological advancements in that period [33]. Presently, the process of cultural adaptation consists of a rigorous methodological process that is not solely based on linguistic translation.
Steps are taken to achieve a high quality linguistic and transcultural version. The method by Beaton et al. [22] utilized in the present study, is considered one of the most utilized for transcultural adaptation, as it recommends a rigorous follow-up and monitoring of systematized stages and the participation of the author of the original questionnaire. In our case, the author (original version), aside from authorizing the use of the instrument, also monitored the stages of the study and approved the final version created by the panel of experts.
It was not easy to define the methodological process of adaptation and validation, given the variability of the response scales. However, the high Competence coefficient (K) obtained by the panel of experts ensured an adequate level of competence when evaluating the instrument [27]. It is important for an expert to be familiar with the area studied, and to possess a good level of knowledge to evaluate a guide or protocol [20].
With respect to the composition of the items, all the items, or dubious words found in the translation process obtained a 95% consensus from the panel of Experts, justifying every decision. An aspect that must be underlined was the "resounding" decision of the panel of experts with respect to the reporting of adverse events, indicating the exclusion of the term "non-punitive", as it is inherent and intrinsic to a reporting system. Effectively, the evidence shows that confidentiality, non-punitive actions, and anonymity, are basic principles of a reporting system, which at the same time, improves the degree of safety of the patient and the quality of care, creating a positive and collaborative environment, one of non-rejection and under reporting [34,35].
The only item that obtained an RI < 0.80 was item 5 "Are there any particular specialties where disruptive events occur most?" which was not eliminated after the evaluation by the panel of experts and researchers. The justification was that this item could not be substituted by or be part of item 4, as it was understood that it is not only important to identify the setting (unit), but also to discover if there is a more prevalent specialties with respect to disruptive behavior. Moreover, this item obtained a PI of 0.81, whose value indicated that the item was appropriate, opportune, and convenient for the subject proposed. The PI of the items is their degree of pertinence with respect to the object of study and the research objectives as well [36]. The results have shown that the items have an adequate content validity, which ensures that the items from the instrument are adequate and a representative sample of the content to be evaluated [29].
The Delphi method used in the study allowed the instrument adaptation proposal to be subjected to the judgement of experts, who assessed and evaluated the quality criteria necessary for the instrument to be a consistent and coherent scale, and that the concepts explored from another culture could be significantly reproduced to the culture found in the healthcare system in the Spanish context.
According to Kyle et al. [37], organizations do not understand how disruptive behaviors appear in the healthcare context, or how they can be measured, as there is little research on the subject. For this reason, stress is placed on the importance of conducting studies that identify disruptive behaviors and that analyze improvement strategies, to promote a culture that is proactive in the prevention and monitoring of these behaviors.
This study is not without limitations. The use of a sample of 11 experts, 10 nurses and 50 physicians/nurses could create possible bias with respect to the subjectivity and generalization of the results. However, carrying out the three phases of the study, with the different samples, gave the study methodological rigor. Another limitation could be that the pilot was only carried out with nurses. Lastly, one more limitation could be that in the third phase of the study some relevant background of the respondents (public, rural, urban hospitals, etc.) were not collected.

Conclusions
The instrument Nurse-Physician Relationship Survey: Impact of disruptive behavior on patient care was adapted to the Spanish context and is adequate for evaluating the disruptive behaviors on nurse-physician relationships and its impact on patient care. The use of a systematic and rigorous methodology allowed us to obtain a conceptual version that is linguistically equivalent to the original.
The clinical applicability of the instrument to measure disruptive behavior in the relationships of health professionals would be to identify the factors associated with said behaviors and their impact on health care. The evaluation of disruptive behaviors can be used to design prevention strategies and improve the quality of care and patient safety. Nevertheless, the importance of continuing with the analysis of the rest of the psychometric properties in future studies is underlined. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.

Versión Original Versión Adaptada y Validada
Title: Nurse-Physician Relationship Survey: Impact of Disruptive Behavior on Patient Care (versión original) Título: Escala sobre la relación Enfermero/a-médico: impacto del comportamiento disruptivo en la atención al paciente "Disruptive behavior" is defined as any inappropriate behavior, confrontation or conflict ranging from verbal abuse to physical or sexual harassment. One potential consequence of disruptive behavior is its effect on collaboration and communication between physicians and nurses that may result in an adverse effect on patient care. The current survey is designed to assess the potential impact of disruptive behavior on adverse events, medical errors, patient safety, quality and other outcomes of care.
Please choose only one answer for each question unless otherwise stated. If you are completing the survey electronically, you can double click on any box and a prompt will appear. Under default value, click on "checked" and an X will appear in the box.
Por favor, escoja solo una de las respuestas por cada pregunta, salvo que se indique lo contrario. En caso de contestar el cuestionario de forma electrónica, puede hacer doble clic en cualquier casilla y aparecerá un aviso de confirmación. Debajo de cada valor, haga clic sobre la respuesta y una X aparecerá en la casilla.