Williams Marie, Cafarella Paul, Olds Timothy, Petkov John, Frith Peter
School of Health Sciences, University of South Australia, City East Campus, Adelaide.
Department of Respiratory Medicine, Repatriation General Hospital, Daw Park, Adelaide.
Chest. 2008 Sep;134(3):489-496. doi: 10.1378/chest.07-2916. Epub 2008 May 19.
If descriptors of the sensation of breathlessness are able to differentiate between medical conditions, the language of breathlessness could potentially have a role in differential diagnosis. This study investigated whether the language used to describe the sensation of breathlessness accurately categorized older individuals with and without a prior diagnosis of COPD.
Using a parallel-group design, participants with and without a prior diagnosis of COPD volunteered words and phrases and endorsed up to three statements to describe their sensation of breathlessness. Cluster analysis (v-fold cross-validation) was applied, and subjects were clustered by their choice of words. Cluster membership was then compared to original group membership (COPD vs non-COPD), and predictive power was assessed.
Groups were similar for age and gender (COPD, n = 94; 48 men; mean age, 70 +/- 10 years [+/- SD]; vs non-COPD, n = 55; 21 men; mean age, 69 +/- 13 years) but differed significantly in breathlessness-related impairment, intensity, and quality of life (p < 0.0001). Cluster membership corresponded accurately with original group classifications (volunteered, 85%; and up to three statements, 68% agreement). Classification based on a single best descriptor (volunteered [62%] or endorsed [55%]) was less accurate for group membership.
Language used to describe the sensation of breathlessness differentiated people with and without a prior diagnosis of COPD when descriptors were not limited to a single best word or statement.
如果呼吸困难感觉的描述词能够区分不同的医学状况,那么呼吸困难的语言表述在鉴别诊断中可能会发挥作用。本研究调查了用于描述呼吸困难感觉的语言是否能准确地将有和没有慢性阻塞性肺疾病(COPD)既往诊断的老年人进行分类。
采用平行组设计,有和没有COPD既往诊断的参与者自愿提供单词和短语,并认可至多三条陈述来描述他们的呼吸困难感觉。应用聚类分析(v折交叉验证),根据他们选择的单词对受试者进行聚类。然后将聚类成员与原始组成员(COPD组与非COPD组)进行比较,并评估预测能力。
两组在年龄和性别方面相似(COPD组,n = 94;男性48名;平均年龄,70±10岁[±标准差];非COPD组,n = 55;男性21名;平均年龄,69±13岁),但在与呼吸困难相关的损伤、强度和生活质量方面存在显著差异(p < 0.0001)。聚类成员与原始组分类准确对应(自愿提供的,85%;至多三条陈述的,68%一致)。基于单个最佳描述词(自愿提供的[62%]或认可的[55%])进行的分类对于组成员的准确性较低。
当描述词不限于单个最佳单词或陈述时,用于描述呼吸困难感觉的语言能够区分有和没有COPD既往诊断的人群。