Table S3.
Procedure for handling missing data in the GLOW6 study
| Spirometry |
| • FEV1, FVC, and IC measurements within 6 hours of rescue-medication use or within 7 days of systemic corticosteroid use were set to missing. Trough FEV1 was defined as the average of the postdose 23 hour 15 minute and the 23 hour 45 minute FEV1 values. The allowable window for these scheduled measurements was 22 hours 45 minutes to 24 hours 15 minutes. Values measured outside of this window were set to missing. If one of the 23 hour 15 minute or 23 hour 45 minute values was missing, the remaining nonmissing value was taken as trough FEV1. If both values were missing, trough FEV1 was regarded as missing |
| • For the primary analysis, if trough FEV1 was missing at visit 8 (week 12) the latest nonmissing predose trough FEV1 (the mean of 45 and 15 minute predose measurements) from visits 5, 6, or 7 (days 29, 57, 84) was carried forward (LOCF). These measurements were required to have been taken before the next dose of study medication. |
| • Missing values for the analysis of secondary spirometry variables were not imputed. |
| Rescue-medication use |
| • The total number of puffs of rescue medication per day over the 12 weeks was calculated and divided by the total number of days with nonmissing rescue-medication data to derive the mean daily number of puffs of rescue medication taken for the patient. If the number of puffs was missing for part of the day (either morning or evening), then a half day was used in the denominator. |
Abbreviations: FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; IC, inspiratory capacity; LOCF, last observation carried forward.