1. Harvey B-G, Strulovici-Barel V, Kaner RJ, Sanders A, Vincent TL, Mezey JG, Crystal RG. Risk of COPD with obstruction in active smokers with normal spirometry and reduced diffusion capacity. Eur Respir J 2015; 46: 1589-97.[PubMed]

This paper misleads the reader about the relationship between gas transfer (TLco) and risk of developing COPD in smokers.
The authors hid within an on-line supplement the way they chose the two groups of subjects for their study. One group had TLco above 80% predicted (see my section under 'Issues' on 'What is Abnormal?' as to why this is a misleading threshold). Their other group had TLco below a lower 95% confidence limit. They showed that this second group tended to develop COPD over time.

The misleading error was then to state that a TLco below 80% predicted was related to developing COPD. This ignores the fact that a very large number of subjects will have a TLco below 80% predicted but above their lower 95% confidence limit. The outcome of this large number of people was not investigated. This was a fatal flaw in their analysis and completely invalidates their conclusions.

Sadly this error was compounded by an Editorial about this paper from David Mannino and Barry Make [PubMed] who did not spot this serious flaw in the paper. In my opinion this paper should be withdrawn.

2. Bhatt SP, Balte PP, Schwartz JE, Cassano PA, Couper D, Jacobs DR, et al. Discriminative Accuracy of FEV1 :FVC Thresholds for COPD-Related Hospitalization and Mortality. JAMA. 2019; 321: 2438–47.[PubMed]

Bhatt and colleagues published a paper looking at participants in large population studies in middle aged people dating back to the late 1980s and 1990s. They used ICD-10 codes for COPD in relation to hospital admissions and death and related these to spirometry results. It was not clear when the authors obtained their spirometry values. They conclude that because GOLD criteria on their spirometry best correlated with the ICD-10 COPD diagnosis this proves the criterion is correct. However, the ICD-10 codes were derived by clinicians who were not part of the study and had access to their own spirometry. It was disingenuous of the authors to use the ICD codes for COPD as if they were an independent verification of COPD diagnosis, that is, independent of any spirometric threshold applied by the clinicians looking after the patients.

All that can be deduced from this study is that the majority of clinicians in the USA use GOLD criteria to diagnose COPD.

In their reply to this challenge [PubMed] the authors state [PubMed] that the clinicians were blinded to the authors' spirometry and that clinicians underutilize spirometry for diagnosing COPD. Since it is unsafe and incorrect to diagnose COPD without spirometry either the authors think their clinicians are incompetent and so the ICD-10 codes they used are not secure for diagnosing COPD or it is true that the majority of clinicians use the GOLD criterion for defining airways obstruction. Either way the conclusion of the paper misleads readers and in my opinion this paper should be withdrawn.

Again the accompanying Editorial from Jorgen Vestbo and Peter Lange [PubMed] did not comment on this flaw in the paper.