Whaley Megan, Awad Nouf Bin, Warholak Terri, Axon David Rhys
Department of Pharmacy Practice and Science, College of Pharmacy, University of Arizona, Tucson, AZ 85721, USA.
Pharmacy (Basel). 2021 Oct 15;9(4):170. doi: 10.3390/pharmacy9040170.
Validation studies of pain interference instruments used among student pharmacists are rare yet essential for understanding their appropriate use and interpretation in pharmacy education and practice. This study conducted validation and reliability assessments of a five-item Pain Interference Scale previously administered to student pharmacists. Construct validity was assessed using Rasch analysis. Unidimensionality was measured using: point-biserial measure correlations; percent of raw variance explained by items; difference between expected; variance modeled by items; and Rasch model fit. To assess scale functioning, response frequency distribution, observed average and sample expected logit distribution, Andrich logit distribution, item separation, and item reliability were assessed. Visual examination of the Item-Person Map determined content validity. Items explained 64.2% of data raw variance. The difference between raw variance modeled and observed was 0.6. Point-biserial measure correlations were >0.77. Item mean-square infits were 0.7-1.3 while outfit measures were 0.72-1.16. There were >10 responses per response category, response frequency and Andrich thresholds progressively advanced, and observed average and sample expected logits advanced monotonically, Andrich logits = -2.33-1.69, item separation = 2.61, and item reliability = 0.87. Item probability curves indicated response categories were minimally yet adequately distinct. Cronbach's alpha = 0.93. The Item-Person Map had a ceiling effect indicating content gaps. In conclusion, the pain interference instrument has acceptable construct validity yet contains content gaps. Additional difficult items should be added to the instrument to better capture pain interference among student pharmacists.
针对学生药剂师使用的疼痛干扰工具的验证研究很少见，但对于理解其在药学教育和实践中的恰当使用及解读至关重要。本研究对先前施用于学生药剂师的一个五项疼痛干扰量表进行了验证和可靠性评估。使用拉施分析评估结构效度。使用点二列相关系数、项目解释的原始方差百分比、预期差异、项目建模方差以及拉施模型拟合度来测量单维性。为评估量表功能，对响应频率分布、观察到的平均和样本预期对数分布、安德里奇对数分布、项目区分度和项目可靠性进行了评估。通过对项目 - 人员图的目视检查确定内容效度。项目解释了64.2%的数据原始方差。建模方差与观察到的方差之差为0.6。点二列相关系数>0.77。项目均方内拟合值为0.7 - 1.3，而外拟合值为0.72 - 1.16。每个响应类别有超过10个响应，响应频率和安德里奇阈值逐渐提高，观察到的平均和样本预期对数单调递增，安德里奇对数=-2.33 - 1.69，项目区分度=2.61，项目可靠性=0.87。项目概率曲线表明响应类别有最小但足够的区分度。克朗巴哈α系数=0.93。项目 - 人员图有天花板效应，表明存在内容差距。总之，疼痛干扰工具具有可接受的结构效度，但存在内容差距。应在该工具中添加更多难度较大的项目，以更好地捕捉学生药剂师中的疼痛干扰情况。
