Getting the Most out of Spirometry: A Tool to Guide Dry Powder Inhaler Use

Background: Dry powder inhaler (DPI) use requires sufficient peak inspiratory flow over the DPI internal resistance (PIFR). Objectives: We examined whether spirometric peak inspiratory flow (PIFspiro) could serve to predict PIFR in patients with obstructive lung disease. Method: Thirty healthy nonsmokers and 140 stable outpatients (70 COPD, 70 asthma) performed spirometry according to the 2019 ERS/ATS spirometry update, yielding PIFspiro. Using a PIFR measurement device with varying orifices, all subjects' PIFR values were recorded for 5 predefined resistance levels, characterized by 5 orifice cross sections (S-R). A test group including all healthy subjects, 30 of the asthma, and 30 of the COPD patients was used to establish the relationship between PIFR and both PIFspiro and S-R by multiple regression. A validation group including the remaining 40 asthma and 40 COPD patients, served to verify whether their predicted PIFR value corresponded to the measured PIFR for each resistance level. Results: The asthma (FEV1 = 78 +/- 17 [SD] %pred) and COPD (FEV1 = 46 +/- 17 [SD] %pred) patients under study had varying airway obstruction. In the test group, PIFR could be predicted by ln[PIFspiro] (p < 0.0001), S-R (p < 0.0001), and S-R(2) (p = 0.006), with an adjusted R-2 = 0.71. In the validation group, estimated PIFR did not significantly differ from measured PIFR (p > 0.05 for the 5 resistance levels). Conclusions: We propose a simple method to predict PIFR for a range of common DPI resistances, based on the device characteristics and on the patient's characteristics reflected in PIFspiro. As such, routine spirometry can serve to estimate a patient's specific PIFR without the need for additional testing.

Automated summary

This study aimed to investigate whether the spirometric peak inspiratory flow (PIFspiro) could predict peak inspiratory flow rate (PIFR) in patients with obstructive lung diseases. The results showed that PIFR could be predicted by PIFspiro and resistance values, and the prediction results were consistent with the measured values.