Anaesthesia & Critical Care, Division of Clinical Neuroscience, School of Medicine, University of Nottingham, Nottingham, UK.
Department of Anaesthesia, Queen's Medical Centre, Nottingham University Hospitals NHS Trust, Nottingham, UK.
Br J Anaesth. 2018 Oct;121(4):867-875. doi: 10.1016/j.bja.2018.06.014. Epub 2018 Jul 31.
We examined the validity and reliability of the previously developed criterion-referenced assessment checklist (AC) and global rating scale (GRS) to assess performance in ultrasound-guided regional anaesthesia (UGRA).
Twenty-one anaesthetists' single, real-time UGRA procedures (total: 21 blocks) were assessed using a 22-item AC and a 9-item GRS scored on 3-point and 5-point Likert scales, respectively. We used one-way analysis of variance to compare the assessment scores between three groups (Group 1: ≤30 blocks in the preceding year; Group 2: 31-100; and Group 3: >100). The concurrent validity was evaluated using Pearson's correlation (r). We calculated Type A intra-class correlation coefficient using an absolute-agreement definition in two-way random effects model, and inter-rater reliability using an absolute agreement between raters. The inter-item consistency was assessed by Cronbach's α.
The greater UGRA experience in the preceding year was associated with better AC [F (2, 18) 12.01; P<0.001] and GRS [F (2, 18) 7.44; P=0.004] scores. There was a strong correlation between the mean AC and GRS scores [r=0.73 (P<0.001)], and a strong inter-item consistency for AC (α=0.94) and GRS (α=0.83). The intra-class correlation coefficient (95% confidence interval) and inter-rater reliability (95% confidence interval) for AC were 0.96 (0.95-0.96) and 0.91 (0.88-0.95), respectively, and 0.93 (0.90-0.94) and 0.80 (0.74-0.86) for GRS.
Both assessments differentiated between individuals who had performed fewer (≤30) and many (>100) blocks in the preceding year, supporting construct validity. It also established concurrent validity and overall reliability. We recommend that both tools can be used in UGRA assessment.
我们检验了先前开发的基于标准的评估检查表（AC）和总体评分量表（GRS）用于评估超声引导区域麻醉（UGRA）表现的有效性和可靠性。
21 位麻醉师的单次实时 UGRA 操作（总共 21 个阻滞）分别使用 22 项 AC 和 9 项 GRS 进行评估，评分分别采用 3 分和 5 分的 Likert 量表。我们采用单因素方差分析比较三组评估分数（第 1 组：前一年≤30 个阻滞；第 2 组：31-100 个阻滞；第 3 组：>100 个阻滞）。采用 Pearson 相关系数（r）评估同时效度。我们在双向随机效应模型中采用绝对一致定义计算 A 型组内相关系数，并采用评分者间绝对一致评估组内信度。采用 Cronbach's α 评估条目间一致性。
前一年 UGRA 经验较多与 AC [F（2, 18）=12.01；P<0.001]和 GRS [F（2, 18）=7.44；P=0.004]评分较好相关。AC 和 GRS 的平均评分之间存在强相关性[r=0.73（P<0.001)]，且 AC（α=0.94）和 GRS（α=0.83）的条目间一致性较强。AC 的组内相关系数（95%置信区间）和评分者间信度（95%置信区间）分别为 0.96（0.95-0.96）和 0.91（0.88-0.95），GRS 分别为 0.93（0.90-0.94）和 0.80（0.74-0.86）。
两种评估方法均能区分前一年完成阻滞操作较少（≤30 个）和较多（>100 个）的个体，支持结构有效性。同时还建立了同时效度和整体信度。我们建议这两种工具均可用于 UGRA 评估。
