Translation, cross-cultural adaptation, and validation of the Behavioral Activity Rating Scale (BARS) for the Brazilian population

Abstract Introduction Few instruments are available in Brazil to evaluate psychomotor activity in psychiatric emergency, clinical, and research settings. This study aimed to perform a cross-cultural adaptation of the Behavioral Activity Rating Scale (BARS) into Brazilian Portuguese and assess the adapted scale’s psychometric properties. Method An expert consensus committee conducted a translation and back-translation of the original scale, resulting in the BARS-BR. Four pairs of physicians administered the BARS-BR and the Sedation-Agitation Scale (SAS) to patients in a hospital psychiatry emergency room and patients in the hospital’s psychiatric wards. The BARS-BR was compared to the SAS to assess concurrent validity and internal consistency was evaluated with the Bland-Altman technique. Results In the emergency room, the correlation coefficients between the first and second assessments were rho = 0.997 and rho = 1.0, respectively. In the hospital wards, the correlation coefficient between the pair of evaluators was rho = 0.951. There were strong correlations between the BARS-BR score of the first examiner and the SAS score of the second examiner (rho = 0.903) and between the SAS score of the first examiner and the BARS-BR score of the second examiner (rho = 0.893). Conclusion The BARS-BR showed good psychometric properties, and we recommend its use because it constitutes an easy method for assessment of changes in psychomotor activity. Further studies are suggested to evaluate adoption and comprehension of the BARS-BR scale by all classes of healthcare professionals.


Introduction
A medical emergency is defined as a situation involving imminent risk to life and requiring immediate and unavoidable intervention. 1,2 Psychiatric emergencies entail behavioral changes that result in risk to the patient or to others and require immediate therapeutic intervention (in minutes or a few hours) to prevent harmful effects. The most common psychiatric emergencies include suicidal behavior, depressive episodes, manic episodes, self-mutilation, markedly compromised critical judgment, severe self-neglect, intoxication or withdrawal states, psychomotor agitation, and aggressiveness. [2][3][4][5][6] Psychomotor agitation or abnormal behavioral activity in patients with psychiatric disorders is a frequent phenomenon and constitutes a clinically relevant condition, not only in emergency situations, but also during hospitalization or in outpatient settings. 6,7 There is scant data on the current prevalence of agitation in medical settings in the world and in Brazil. In the United States, more than 1.7 million patients with agitation are seen at medical emergency services 8,9 in urban centers and the prevalence of agitation in emergency departments is 2.6%. 10 In Europe, the prevalence of agitation in psychiatric emergencies is 4.6%. 11 Worldwide, agitation accompanied by aggressiveness are present in 2.6 to 52% of the patients seen in psychiatric emergencies. 8 In Brazil, the estimated prevalence of agitation is 24% of psychiatric emergencies, 12 although up-to-date data are needed in this field.
Approximately 10% of patients seen in psychiatric emergencies can become agitated and/or violent during the assessment process, 2,6 which shows that psychomotor agitation is a dynamic situation. 6 Although the literature implies that in most cases violent behavior occurs without warning signs, 13,14 some authors have indicated that episodes of aggressiveness may be associated with specific risk factors and preceded by behavioral warning signs. 15 The objective of care for patients with psychomotor agitation, with or without aggression, is to protect them and the people around them by adopting attitudes that aim to reassure them. 2,6 The first step is evaluation, during which the physician should perform an initial mental status examination as soon as possible to determine the most likely cause of the condition and to guide preliminary interventions to calm the patient.
Verbal intervention or voluntary medication (medication administered with the patient's consent) is recommended before moving on to other strategies. 16 When neither option is possible, administration of medication on an involuntarily basis may become necessary. To this end, the concept of rapid tranquilization should be used: calming the patient without excessive sedation, or even without sedation, with fast-acting medications, with the fewest possible side effects. [17][18][19] Once the patient is calm, a broader psychiatric evaluation can be completed. In this context, caring for the patient in such a situation requires quick decisions and therefore demands training, professional experience, and technical and scientifically based decisions. It is possible to use scales that allow objective and equal assessments across the team that can also be used to follow-up the approach's effectiveness. 2 The American Association of Psychiatric Emergencies (AAEP) proposes use of the Behavioral Activity Rating Scale (BARS) for triage and management outside the emergency room 20 because it is based on clinical observation, measuring the severity of agitated behavior using a single item that describes seven levels of severity (from a state of sedation to a state of agitation), and is easy to apply when assessing the motor activity of individuals with mental disorders. 21 Thus, we chose the BARS for translation, cross-cultural adaptation, and validation in Brazilian Portuguese. It will be the first scale designed to assess motor activity in patients with primary mental disorders to be validated for Brazilian Portuguese. The importance of this study is that it will result in availability of this instrument adapted for use in clinical practice by psychiatrists in Brazil.

Study design
This study was carried out in two phases: 1) translation and cross-cultural adaptation of the BARS into Brazilian Portuguese (resulting in the BARS-BR) and 2) assessment of the psychometric properties of the BARS-BR, Phase 1 was based on recommendations in the literature on how to translate and validate a scale, 22 carried out in five steps: translation, synthesis, back-translation, review by a committee of experts, and pre-test.
Phase 2 consisted of a cross-sectional study. Figure 1 shows a flowchart illustrating this study. Two hundred patients in a psychiatric hospital were divided into two groups and their psychomotor activity was evaluated.
The first group comprised psychiatric patients seen in the emergency room (n = 100). Each patient had their psychomotor activity level evaluated at initial presentation and again after 60 minutes. These two measures were conducted to evaluate administration of the BARS-BR during two phases of psychomotor activity, from agitation at admission to sedation after To estimate sample size, it is recommended that ten participants should be analyzed per item on a scale. 23 Since the BARS has one item scored from 1 to 7, we decided to recruit at least 70 participants. We had access to a large sample in a psychiatric hospital and so we were able to assess 200 patients.

Eligibility criteria
The patients included in the study were at least 18 years old and agreed to participate in the study.
Consent forms were provided after evaluations with the patients or their legal guardians. Individuals in the emergency room with signs of psychomotor agitation were included. In the psychiatric wards, the severity of psychomotor activity was not an inclusion criterion, enabling assessment of patients with several levels of psychomotor activity (from sedation to agitation).
Based on the DSM-5 24 criteria, patients with unstable clinical diseases, diagnoses of organic disorders, anxiety disorders, or personality disorders were excluded.

Assessments and Instruments
Two instruments were used in this study: the translated version of the BARS (BARS-BR), to assess the level of behavioral agitation, and the Riker Agitation-Sedation Scale (SAS), to assess patients' behavioral activity and test convergent validity of BARS-BR. Patients in the emergency room were evaluated at two points in time: when first seen at presentation and after 60 minutes.
Patients in the psychiatric wards were evaluated once.

Behavioral Activity Rating Scale (BARS)
The BARS 21 comprises a single item that assesses psychomotor activity across seven intensity categories ranging from score 1 (difficult or unable to rouse) to score 7 (violent, requires restraint). The scale allows evaluation of states ranging from profound sedation to severe agitation. The scale is simple and intuitive and the physician can assess behavioral activity in seconds.
The BARS-BR was used to evaluate patients in the emergency room and in the wards.

The Riker Sedation-Agitation Scale (SAS)
The SAS 25 is a single item scale with seven progressive points of severity ranging from agitation (7 = dangerous agitation) to sedation (1 = unarousable). 25,26 The scale is the most compatible scale for evaluation of psychomotor activity available in Brazilian Portuguese 27 and it was therefore the best choice to compare with the BARS-BR to assess convergent validity.

Procedures
Step 1: Translation and cross-cultural adaptation of the

BARS into Brazilian Portuguese
Based on the work of Beaton et al., 22  trained these eight physicians.
The physicians were allocated to pairs, which simultaneously evaluated ten patients in the emergency room and then ten patients in the wards. After the evaluations, each evaluating physician gave a rating between 1 and 10, indicating how easy they found it to understand the scale. The physicians also discussed the application and the scale with the principal investigator; agreeing that no changes were required. Finally, translation distortions were discussed, resulting in a final version of the scale in Brazilian Portuguese called BARS-BR (shown in Table 1), ready for assessment of its psychometric properties.
Step In both of the groups, we assessed the internal consistency of the BARS-BR and tested its concurrent validity by evaluating its correlation with the SAS.
We used the results of simultaneous administrations of the BARS-BR by different examiners to test interrater agreement (between the results of BARS-BR administered by Evaluator 1 and SAS administered by Evaluator 2 and between BARS-BR administered by Evaluator 2 and SAS administered by Evaluator 1).

Statistical analysis
For validation studies, it is recommended that at least 10 participants are recruited per item on the scale being validated. 28 A sample size of at least 70 subjects   Table 2 shows this process.

CVI of the cross-cultural adaptation
The CVI score for the seven items of the final version

Pre-test phase
In the pre-test phase, inter-rater correlations were

Psychometric properties of the BARS-BR
We assessed a total of 200 patients in the hospital's emergency room and psychiatric wards. Tables 3 and 4 show the sociodemographic and clinical characteristics of the sample.

Convergent validity
In the initial emergency room evaluation (n = 100), In the initial assessment, the BARS-BR scores were as follows: item 5 = 68%, item 6 = 15%, and item 7 = 17%. Hence, it was possible to assess the items corresponding to agitated behavior. In the second assessment, item 4 dominated, with 67%, which signals unchanged motor activity; however, sedation categories were also endorsed.
In the assessments of the psychiatric wards sample (n = 100), the pairs of raters selected the same BARS-BR score in 90% of cases. There was a strong interrater correlation between the BARS-BR scores recorded by each member of the pairs (Spearman's rho = 0.951; p < 0.0005). A 90% match was observed between the BARS-BR option selected by rater 1 and the corresponding SAS item selected by rater 2, and there was an 89% match between the SAS option selected by rater 1 and the BARS-BR option selected by rater 2.
In the wards, we analyzed the inter-rater correlations  Table 5 shows Spearman's coefficients in the main correlations.

Differences between the BARS-BR and SAS
We used the BA technique to determine the difference between the scores generated in the evaluations Table 5 -Intra and inter-rater Spearman's correlation coefficients (rho) for the assessments in the wards where both scales were employed; we placed them on a regression graph as being dependent on the mean obtained from the two scales. Agreement between the scales was analyzed for 300 assessments. In the initial emergency room assessment (n = 100) and the final assessment (n = 100), the same rater's scores were utilized. In the wards, the BARS-BR scores from the first rater were compared to the SAS scores from the second rater (n = 100), evaluating the 300 scenarios where the scales were applied simultaneously.

Intra-rater Spearman's correlation
The line of difference between the scales was close to 0, as shown in Figure 2, demonstrating no deviation trend. Thus, we can conclude that correlations were strong for the concurrent validity and reliability of the BARS-BR.

Discussion
As part of the cross-cultural adaptation, 22 Inter-rater reliability was observed across a broad spectrum of patients, demonstrated by the excellent degree of agreement between evaluations. 36 The BARS-BR is a simple-to-use instrument for assessment of motor activity in psychiatric patients. 21 It has well-defined criteria and enough levels to assess initial motor activity, as well as to monitor response to pharmacological therapies used in the management of psychomotor agitation. Although two different conditions (agitation and sedation) are assessed on a single scale, the sequential approach establishes a single score, appraising agitation first and then sedation.
The BARS-BR is an excellent option for a wide range of uses in different clinical settings, because it is a single-item scale with seven levels of ascending severity ranging from sedation to agitation. 25 The scale can also be administered

Limitations and perspectives
The validation of any scale is an evolving process.  In this study, the number of patients who had sedation scores was considerably lower than the number of agitated patients. There was, however, an important agreement between BARS-BR and SAS in evaluation of subjects who had sedation scores. Considering that the SAS was originally configured for use in the intensive care setting, where the prevalence of sedated patients is possibly higher than in psychiatric units, we can deduce that the BARS-BR is useful for assessing individuals with slowed motor activity.

Conclusion
The present study presents evidence that the translation and cross-cultural adaptation of the BARS were satisfactory and successful, considering that content validity and cross-cultural validity were demonstrated for the Brazilian Portuguese version. We followed steps suggested in the literature 22 to make the BARS-BR possible.
We found significant correlations between the scores derived from this scale and the SAS scale, supporting these theoretical positions. Hence, this psychometric instrument is capable of measuring motor activity in acute psychiatric patients consistently across time, raters, and items. Additional studies are needed to fill certain methodological gaps, however, in order to make a categorical statement.
This work provides the Brazilian population with an easily applicable scale, capable of assessing initial changes in psychomotor behavior and sensitive to changes related to pharmacological treatment, as observed in validation studies of the original scale.
Based on the present study, the BARS-BR may be used in formulation of protocols for the care of patients with altered motor activity secondary to mental illness, thus providing individualized and effective management in this context.

Disclosure
No conflicts of interest declared concerning the publication of this article.