区分孤立性药片吞咽困难与一般口咽吞咽困难：土耳其语版PILL-5的心理测量学特性

文献信息

中文摘要

背景： 尽管食物和液体吞咽功能保留，药片吞咽困难仍可能作为一种孤立表型出现，并且可能需要与一般口咽吞咽困难不同的临床路径。我们翻译、文化调适并验证了土耳其语版PILL-5（PILL-5(TR)），并检验其区分孤立性药片吞咽困难与健康成人及口咽吞咽困难的能力。方法：我们进行了正向-反向翻译、专家委员会评审和认知访谈（n=30）。成人（N=348）被分类为健康对照（n=175）、孤立性药片吞咽困难（功能性经口摄食量表，FOIS=7；n=107）或口咽吞咽困难（n=66）。我们评估了内部一致性、重测信度（n=48；7-14天）、使用探索性因子分析和基于共同因子的平行分析评估结构效度、与进食评估工具-10（EAT-10）、吞咽生活质量问卷（SWAL-QOL）和吞咽障碍 handicap 指数（DHI）的收敛效度，以及使用受试者工作特征（ROC）分析评估筛查区分能力。结果：PILL-5(TR)显示出很强的已知组间区分度（均值：0.37、5.81和10.47；p<0.001），并且联合因子分析结果支持以单维结构为主。信度良好（Cronbach's α=0.865；组内相关系数（ICC）=0.869）。在临床样本中，PILL-5(TR)与EAT-10强相关（Spearman's ρ=0.749），与DHI中度相关（ρ=0.387），与SWAL-QOL呈负相关（ρ=-0.482）（均p<0.001）。在研究样本中，区分孤立性病例与健康对照的表现极好（ROC曲线下面积[AUC]=0.992；初步阈值≥2；敏感性1.00；特异性0.87），而区分口咽吞咽困难与孤立性病例的表现可接受（AUC=0.793；初步阈值≥7；敏感性0.82；特异性0.71）。结论：PILL-5(TR)是一种有效且可靠的药片特异性患者报告测量工具。初步阈值可能有助于在土耳其语人群中开展以药片为重点的筛查、分诊和用药安全咨询，但需要在更具异质性的临床样本中进行外部验证。

英文摘要

Pill swallowing difficulty may occur as an isolated phenotype despite preserved food and liquid swallowing and may require different clinical pathways than general oropharyngeal dysphagia. We translated, culturally adapted, and validated the Turkish PILL-5 (PILL-5(TR)) and tested its ability to discriminate isolated pill swallowing difficulty from healthy adults and oropharyngeal dysphagia. We performed forward-back translation, expert committee review, and cognitive debriefing (n = 30). Adults (N = 348) were classified as healthy controls (n = 175), isolated pill swallowing difficulty (Functional Oral Intake Scale, FOIS = 7; n = 107), or oropharyngeal dysphagia (n = 66). We evaluated internal consistency, test-retest reliability (n = 48; 7-14 days), structural validity using exploratory factor analysis and common-factor-based parallel analysis, convergent validity with the Eating Assessment Tool-10 (EAT-10), the Swallowing Quality of Life questionnaire (SWAL-QOL), and the Dysphagia Handicap Index (DHI), and screening discrimination using receiver operating characteristic (ROC) analyses. PILL-5(TR) showed strong known-groups separation (means: 0.37, 5.81, and 10.47; p < 0.001), and the combined factor-analytic findings supported a predominantly unidimensional structure. Reliability was good (Cronbach's α = 0.865; intraclass correlation coefficient (ICC) = 0.869). In the clinical sample, PILL-5(TR) correlated strongly with EAT-10 (Spearman's ρ = 0.749), moderately with DHI (ρ = 0.387), and inversely with SWAL-QOL (ρ=-0.482) (all p < 0.001). In the study sample, discrimination between isolated cases and healthy controls was excellent (area under the ROC curve [AUC] = 0.992; preliminary threshold ≥ 2; sensitivity 1.00; specificity 0.87), while discrimination between oropharyngeal dysphagia and isolated cases was acceptable (AUC = 0.793; preliminary threshold ≥ 7; sensitivity 0.82; specificity 0.71). PILL-5(TR) is a valid and reliable pill-specific patient-reported measure. The preliminary thresholds may help inform pill-focused screening, triage, and medication-safety counselling in Turkish-speaking populations, but require external validation in more heterogeneous clinical samples.