To the Editor:
I would like to thank Rutka, Palac, and Linel for their interest in making our research a consistent “source of reliable information” and highlighting some potential conflicting data related to our study results. As stated in the “Quick Look” section, our study1 supported 2 main findings: the first was the dramatic impairment of the diaphragmatic craniocaudal mobility correlated to lung function loss (in a cohort of 46 subjects with COPD and 16 healthy volunteers); the second was that, in the studied population, the 30 subjects with COPD who concluded the in-patient pulmonary rehabilitation presented improvements in the diaphragmatic motion and improvements in dynamic hyperinflation.
In their letter, Rutka and colleagues point out that the main problem with our paper is in the presentation of the associations between lung function and diaphragmatic mobility during rest breathing or deep inspiration, specifically that the text describes the opposite from what is presented in the respective graphics (Fig. 3 and 5), a “mistake” that is repeated in Figures 4 and 5. The Pearson correlation test is commonly used to demonstrate the monotonic association between 2 variables. Further, the graphic representation is highly recommended to avoid misunderstanding of the data, as we have done.2 Looking at the figures, 2 facts are clear: (1) The loss of lung function is associated with reductions of the diaphragmatic mobility during the deep inspiration maneuvers; (2) the loss of lung function is associated with the increases of the diaphragmatic mobility during rest breathing. Those associations were extensively discussed in paragraphs 1 and 5 of the discussion.
It is true that, at the fourth paragraph of the results, the following is written: “The correlation of FEV1% predicted and diaphragmatic mobility during rest breathing (r = –0.74, P < .001) and deep inspiration (r = 0.796, P < .001) (n = 45) are demonstrated in Figure 3.” There, it is possible to observe a minor typographical error of the minus sign, which is irrelevant considering some significant elements in our paper. First is the common understanding within our field that, in Pearson correlations, the x axis and the y axis must be crescent, so that, regarding the loss of lung function, this error of the minus sign would be confusing only for non-expert readers. Second, the figures demonstrated what they were designed for: the worse the lung function, the worse the diaphragmatic function. Further, it is clearly stated in the sixth paragraph of the methods section that the correlations were made with measurements from the healthy group (n = 16) and the patients with COPD who ended the rehabilitation (n = 30), for a total of 46 subjects, not 45 as written in the text.
Regarding the abstract, we agree and thank Rutka et al for highlighting that the text was confusing: the wording of “. . . 52 subjects with moderate to very severe COPD who underwent pulmonary rehabilitation and 16 healthy subjects” should be written as “. . . 52 subjects (ie, 46 subjects with moderate to very severe COPD who underwent pulmonary rehabilitation and 16 healthy subjects).” In any case, I strongly disagree that this information may confound the reader to understand the overall findings of this paper. To conclude, it seems evident that the percent of predicted FEV1 was used to determine the COPD severity. Still, in any case, if the authors Rutka and colleagues know another way to classify airway obstruction in patients with COPD, we will be happy to be enlightened. The author will update the information regarding the sample size registered at ClinicalTrials.gov. The assessments were concluded as we saw that the research goals were reached with 46 subjects.
As a physiotherapist, researcher, and lecturer, I have always encouraged my peers, colleagues, and students to discuss the meaning of the information related to medical science. It is undeniable that the information related to methods and results must be clear and precise, but, as I see it, to improve the debate’s quality we must avoid reducing the critical analyses to simply pointing out typographical errors.
Footnotes
- Correspondence: Camilo Corbellini PhD MSc PT. E-mail: camilo.corbellini{at}lunex-university.net
The authors have disclosed that they do not have any conflicts of interest.
- Copyright © 2021 by Daedalus Enterprises
Reference
- 1.
- 2.