Reply to Londhe et al.: CAPTURE: A Screening Tool for Chronic Obstructive Pulmonary Disease or Obstructive Airway Disease?
Reply to Londhe et al.: CAPTURE: A Screening Tool for Chronic Obstructive Pulmonary Disease or Obstructive Airway Disease?
复制标题
回复 Londhe 等人：CAPTURE：慢性阻塞性肺病或阻塞性气道疾病的筛查工具？
DOI：
10.1164/rccm.201707-1393le
复制
发表时间：
2018
影响因子：
24.7
通讯作者：
High-Risk-COPDScreeningStudyGroup
中科院分区：
文献类型：
--
作者：
Martinez,FernandoJ;Han,MeiLan;Leidy,Nancy;Make,Barry;Mannino,DavidM;Rennard,StephenI;Thomashow,ByronM;Yawn,BarbaraP;High-Risk-COPDScreeningStudyGroup
我们怀着极大的兴趣阅读了Martinez及其同事的文章（1），其中他们开发了一个五项问卷，除了呼气峰流速（PEFR）值，以筛选需要进一步评估的慢性阻塞性肺疾病（COPD）潜在病例。作者使用了一种创新的，多方法的方法来选择最好的，最小的一组问卷调查项目从一个全面的候选项目清单来自文献，现有数据的分析和定性研究。尽管这种方法的优点和独特性，我们发现一个主要的缺陷，在发展的捕获问卷。它没有一个呼吸道症状对照组（即另一个有呼吸道症状但没有COPD的对照组，例如，那些患有哮喘的人）。当我们对这里的一些受试者进行CAPTURE问卷调查时，许多哮喘患者的COPD筛查呈阳性。这并不奇怪，因为有些问题，如“由于季节或空气质量导致的呼吸变化"、“在香水、草或烟雾周围咳嗽”、“曾经呼吸急促”和“由于急性呼吸系统疾病而错过工作/活动”，哮喘受试者也会回答“是”。此外，患有哮喘的受试者也将具有低PEFR值。因此，CAPTURE可能会将大量哮喘受试者错误地标记为潜在COPD病例。我们认为这是CAPTURE的一个主要缺点，因此不值得将其标记为COPD筛查问卷。理想情况下，首先使用问卷和峰值流量计检测阻塞性气道疾病的存在似乎更合适，然后有一组额外的问题，这将有助于回答阻塞性气道疾病是否可能是哮喘或COPD。我们最近尝试使用这种方法取得了合理的成功（2），尽管我们的研究需要进一步验证。第二，CAPTURE对吸烟史的重视程度低于预期，作为帮助筛查COPD的重要问题，我们发现这很有趣。尽管在非吸烟者中筛查COPD确实很重要，但考虑到相当多的非吸烟者也会患上COPD（3），我们认为CAPTURE中遗漏了关于吸烟史的明确问题反映了研究设计的固有缺陷。作者将轻度COPD受试者纳入对照组，大概是为了平衡研究组和对照组的人数。但由于轻度COPD组中66%的人曾吸烟，这增加了对照组中吸烟者的比例，从而削弱了使用该问题发现病例和对照受试者之间差异的能力。我们认为这是研究设计中的第二个缺陷。
We read with great interest the article by Martinez and colleagues (1) wherein they developed a five-item questionnaire in addition to a peak expiratory flow rate (PEFR) value to screen potential cases of chronic obstructive pulmonary disease (COPD) that require further evaluation. The authors used an innovative, multimethod approach to select the best, smallest set of questionnaire items from a comprehensive list of candidate items derived from the literature, analysis of existing data, and qualitative research. Despite the merit and uniqueness of this approach, we find a major flaw in the development of the CAPTURE questionnaire. It did not have a respiratory symptoms control arm (ie, another control group of subjects who had respiratory symptoms, but no COPD—for example, those who had asthma). When we administered the CAPTURE questionnaire to some of our subjects here, many who had asthma screened positive for COPD. This is not surprising, because some of the questions, such as “change in breathing due to seasons or air quality,”“cough around perfume, grass, or smoke,”“ever short of breath,” and “missed work/activities due to an acute respiratory condition,” will also be answered “yes” by subjects with asthma. Moreover, subjects with asthma will also have low PEFR value. As a result, CAPTURE may falsely label a large number of subjects with asthma as potential cases of COPD. We believe this is a major drawback of CAPTURE, and therefore it does not merit being labeled as a COPD screening questionnaire. Ideally, it would seem more appropriate to first detect presence of obstructive airways disease using a questionnaire and a peak flow meter and then have an additional set of questions that will help answer if the obstructive airways disease is likely to be asthma or COPD. We made an attempt using this approach with a reasonable measure of success recently (2), although our study needs further validation.Second, CAPTURE gives lesser importance to history of tobacco smoking than might be expected as an important question to help screen COPD, which we found interesting. Although screening for COPD in nonsmokers is indeed important, given the fact that a significantly large number of nonsmokers also develop COPD (3), we believe that omission of an explicit question about smoking history in CAPTURE reflects an inherent flaw in the study design. The authors included subjects with mild COPD in the control arm, presumably to balance the numbers in the study and control arms. But because 66% of the people in the mild COPD group were former smokers, this increased the proportion of smokers in the control group, thereby weakening the ability to use this question to find differences between cases and control subjects. We believe this was a second flaw in the study design.■