RESUMEN
BACKGROUND: Pain assessment of patients with traumatic brain injury is a challenge because they are unable to self-report their pain experience. AIMS: To investigate the psychometric properties of validity, reliability, and responsiveness of the Brazilian version of the Behavioral Pain Scale (BPS-Br) in patients with traumatic brain injury. METHODS: This was an observational, cross-sectional, repeated-measure and analytical study. This study was developed at the medical and surgical ICUs in a high-complexity public hospital at Aracaju, Sergipe, Brazil. Thirty-seven adult patients with moderate or severe TBI were included. This study was completed with 444 independent observations, a pairwise comparison, and was performed simultaneously before, during, and after eye cleaning and endotracheal suctioning of 37 adult patients with moderate to severe traumatic brain injury. RESULTS: The BPS-Br had good internal consistency (.7 ≤ α ≤ .9), good discriminant validity (p < .001), moderate to excellent reliability based on inter-rater agreement (intraclass correlation coefficient = 0.66-1.00; κ = 0.5-1.0), and high responsiveness (0.7-1.7). The upper limbs subscale had the highest score during the nociceptive procedure (1.8 ± 0.9). Deep sedation affected the increase of grading during painful procedures (p < .001). CONCLUSIONS: Our results suggest the BPS-Br is a useful tool for clinical practice to evaluate the pain experienced by patients with traumatic brain injury. Further studies of different samples are needed to evaluate the benefits of systematic pain assessment of critically ill patients.