RESUMO
Violence is increasing on medical-surgical units as a "silent epidemic." This quality improvement project employs a small non-experimental, single-group, pre- and post-test design (N = 11) to determine the effectiveness of de-escalation training on medical-surgical nurses' confidence levels when dealing with agitated patients. Regardless of age, education, or years of experience, scores improved for each question on Thackrey's (1987) Confidence in Coping with Patient Aggression Instrument after implementing Ten Domains of De-escalation by Richmond et al. (2012). A paired-sample two-tailed t-test significantly increased from Time 1 pre-test (M = 49.82, SD = 10.11) to Time 2 post-test (M = 72.82, SD = 14.41), t(10) = 4.46, p <.001. The mean increase was 23.00 [95% CI, 11.51-34.49]; d = 1.84 indicating a large effect size (Pilot, 2010). A sensitivity analysis (Wilcoxon Signed Rank Test) showed a median difference among the matched pairs with a significant increase in confidence levels post-training, z = -2.847, p <.004. The median score increased from the pre-test (Md = 51) to the post-test scores (Md = 71) (Pallant, 2013).