Psychometric properties of the Brazilian version of the Sunnybrook Facial Grading System

Introduction: The Sunnybrook Facial Grading System (SFGS) is a scale to evaluate facial function in three domains, namely resting symmetry, voluntary movements, and synkinesis. It is commonly used in scientific research and clinical practice to assess and monitor people with facial paralysis. Objective: To translate and cross-culturally adapt the SFGS, develop a version for the Brazilian population (SFGS - Brazil) and analyze its psychometric properties, including validity, interrater reliability and responsiveness. Methods: A multidisciplinary panel translated and adapted the SFGS into Brazilian Portuguese, creating the SFGS-Brazil version. Next, content validation was carried out by a panel of four physical therapists with clinical experience in caring for people with facial paralysis, in addition to interrater reliability and scale responsiveness after physical therapy intervention. Results: For SFGD validation, committee agreement rate and the content validity index were greater than 90%. Agreement (interrater reliability) was excellent for most items and overall (intraclass correlation coefficient = 0.99; p < 0.000) and the scale proved to be responsive, indicating post-intervention improvement (t = 10.66; p = 0.000). Conclusion: The domains and items of the SFGS-Brazil are conceptually equivalent to those of the original version, and the instrument displays adequate psychometric properties, including validity, agreement and responsiveness. The SFGS-Brazil is suitable for the Brazilian population and can be used in scientific studies and clinical practice.


Introduction
Peripheral facial palsy (PFP) is partial or complete impairment of facial nerve function, the seventh cranial nerve. This nerve performs multiple functions, including innervation of the facial mimetic muscles, receiving taste sensations from the anterior two thirds of the tongue, and controlling the salivary and lacrimal glands. 1,2 AThe prevalence of facial palsy is approximately 15 to 40 cases per 100,000 people 3 and its main causes are traumatic, infectious, neoplastic, congenital, toxic and idiopathic, 4 with the last being responsible for more than 60% of cases, also known as Bell's palsy. 5 According to VanSwearingen, 6 neuromotor disorders resulting from PFP can be classified into four categories, two acute (initiation and facilitation) and two chronic phases (movement control and relaxation).
Patient recovery category depends on the type of nerve injury (neuropraxia, axonotmesis or neurotmesis), among other factors. It is known that 85% of individuals with PFP partially or completely recover facial movements within three weeks. 7 However, impaired facial expression and possible complications may remain, including severe functional and psychosocial issues. 7 Pinho 8 observed that physical therapists assess PFP based only on neurological signs and symptoms, do not use assessment scales translated and adapted for the Brazilian population, and are unaware of the possible complications of this condition. Furthermore, the proposed treatments are not based on the specific categories of PFP, as described by VanSwearingen. 6 The lack of structured assessment and specific instruments for PFP results in incomplete and ineffective evaluation in terms of tailoring the necessary treatment to each situation.
In addition, instruments should be objective in order to quantitatively assess the severity of facial dysfunction and, subsequently, the progression of the proposed therapy, allowing monitoring of the patient's evolution. 8 Other facial function classification systems have been proposed, such as the House e Brackmann, 9 Lacôte et al. 10

Translation and cross-cultural adaptation
The translation and cross-cultural adaptation of the SFGS for the Brazilian population was based on the guidelines described by Beaton et al. 16 and consisted of five stages, as described below. Prior authorization was obtained from the authors of the original scale to carry out cross-cultural adaptation.
Step 1: Translation into Portuguese The original scale was translated into two Brazilian Portuguese versions (SFGS-Brazil 1 and 2) by independent translators, with the aim of comparing the differences between them. To that end, two Brazilian translators fluent in English were chosen, one with and the other without experience in facial palsy.
Step 2: Synthesis of the translations In order to compile SFGS-Brazil version 3, the results of the translations were synthesized by the two translators and by an observer with a PhD in neuroscience and 15 years of experience.

Step 3: Back translation
Two translators blind to the original version of the questionnaire back translated SFGS-Brazil 3 to rule out possible conceptual errors in this version.
Step 4: Expert panel The  was set at α = 0.05. Figure 1 shows the study flowchart.

Translation and cross-cultural adaptation
Following translation, synthesis and back translation, the scale was submitted to a panel of four experts (step 4) who suggested the following modifications: change the word "excursion" to "muscle contraction", "whistle" to "puckered lips" and "eyelid surgery" to "droopy eyelid".
The word "excursion" was maintained because it refers to range of motion and not only a muscle contraction.
The expression "whistling" was also kept because the translated scale must match the original, and this activity is common among Brazilians. "Eyelid surgery" refers to suturing by a physician to prevent ophthalmologic complications in individuals unable to close their eyes.     6 (20.0%) partially understood at least one item, and 2 (6.6%) understood none. The main items that were partially or not understood were related to resting symmetry ("eyes" and "cheek") and the symmetry during movement and synkinesis ("raise the upper lip") domains.
Reliability Table 3 describes the sociodemographic and clinical characteristics of participants with PFP. Thirtyone individuals took part in this study, with an average age of 40.12 ± 15.53. Most were women (61.3%) whose right side was affected (71%), with 11 participants in the "initiation" category, seven "facilitation", eight "movement control", and five in the "relaxation" category.
Two trained examiners were selected to analyze agreement. Analysis of interrater reliability data ( According to Brazilian professionals, the items of the

Authors' contributions
TKMM and PFA performed sample selection and data collection. CCSCP was responsible for study conception and design, data analysis and interpretation.
All the authors contributed to writing the manuscript; and CCSCP and SRMM revised it.