AUTHOR=Otto-Yáñez Matías , Sarmento da Nóbrega Antônio José , Torres-Castro Rodrigo , Araújo Palomma Russelly Saldanha , Carvalho de Farias Catharinne Angélica , Dornelas De Andrade Armele de Fátima , Puppo Homero , Resqueti Vanessa Regiane , Fregonezi Guilherme Augusto de Freitas TITLE=Maximal Voluntary Ventilation Should Not Be Estimated From the Forced Expiratory Volume in the First Second in Healthy People and COPD Patients JOURNAL=Frontiers in Physiology VOLUME=11 YEAR=2020 URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2020.00537 DOI=10.3389/fphys.2020.00537 ISSN=1664-042X ABSTRACT=Purpose

To evaluate the concordance between the value of the actual maximum voluntary ventilation (MVV) and the estimated value by multiplying the forced expiratory volume in the first second (FEV1) and a different value established in the literature.

Methods

A retrospective study was conducted with healthy subjects and patients with stable chronic obstructive pulmonary disease (COPD). Five prediction formulas MVV were used for the comparison with the MVV values. Agreement between MVV measured and MVV obtained from five prediction equations were studied. FEV1 values were used to estimate MVV. Correlation and agreement analysis of the values was performed in two groups using the Pearson test and the Bland–Altman method; these groups were one group with 207 healthy subjects and the second group with 83 patients diagnosed with COPD, respectively.

Results

We recruited 207 healthy subjects (105 women, age 47 ± 17 years) and 83 COPD patients (age 66 ± 6 years; 29 GOLD II, 30 GOLD III, and 24 GOLD IV) for the study. All prediction equations presented a significant correlation with the MVV value (from 0.38 to 0.86, p < 0.05) except for the GOLD II subgroup, which had a poor agreement with measured MVV. In healthy subjects, the mean difference of the value of bias (and limits of agreement) varied between -3.9% (-32.8 to 24.9%), and 27% (-1.4 to 55.3%). In COPD patients, the mean difference of value of bias (and limits of agreement) varied between -4.4% (-49.4 to 40.6%), and 26.3% (-18.3 to 70.9%). The results were similar in the subgroup analysis.

Conclusion

The equations to estimate the value of MVV present a good degree of correlation with the real value of MVV, but they also show a poor concordance. For this reason, we should not use the estimated results as a replacement for the real value of MVV.