RESUMO
OBJECTIVE: To evaluate an observational-behavioral pain tool among individuals with acute poststroke aphasia. METHODS: We performed a randomized, double-blind, controlled study of experimental pain assessment among 36 adult patients with acute poststroke aphasia. Patients were administered 3 levels of mechanical pain, including placebo. The behavioral responses were video recorded and then evaluated by 3 neurology nurses using the Pain Assessment Checklist for Seniors With Limited Ability to Communicate (PACSLAC-II). Pain-specific facial action units were quantified with FaceReader version 6.1. RESULTS: Median PACSLAC-II ratings for 0-, 2-, and 4.5-lb weight stimuli were 2 (0, 3), 1 (0, 3), and 2 (1, 5), respectively. Overall, differences were not detected (p = 0.06). Pairwise comparisons with the Wilcoxon method demonstrated significance in differentiating PACSLAC-II ratings of patients experiencing the 4.5-lb stimulus vs either the 2-lb weight (p = 0.03) or placebo (p = 0.05). Overall interrater reliability by the Cronbach α was strong at 0.87, 0.94, and 0.96 for weights of 0, 2, and 4.5 lb, respectively. Pain-specific facial activation and negative valence were observed similarly in placebo and experimental pain groups. CONCLUSIONS: Among our cohort with acute poststroke aphasia, the PACSLAC-II was not able to overall differentiate patients experiencing experimental mechanical pain, although differences in those experiencing the strongest pain stimulus were significant. The detection of pain-specific facial activation and negative valence in the placebo group indicates that pain and distress are unmet needs among stroke patients who are unable to verbally communicate.