Response was defined as change from baseline. Baseline FEV1 was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FEV1 AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Saint George's Respiratory Questionnaire (SGRQ) measures the impact of COPD on overall health, daily life, and perceived well-being ranging from 0 (no limitations) to 100 (most limitations).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Saint George's Respiratory Questionnaire (SGRQ) measures the impact of COPD on overall health, daily life, and perceived well-being ranging from 0 (no limitations) to 100 (most limitations).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Saint George's Respiratory Questionnaire (SGRQ) measures the impact of COPD on overall health, daily life, and perceived well-being ranging from 0 (no limitations) to 100 (most limitations).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Saint George's Respiratory Questionnaire (SGRQ) measures the impact of COPD on overall health, daily life, and perceived well-being ranging from 0 (no limitations) to 100 (most limitations). This is a combined analysis of the data from NCT00793624 and NCT00796653 showing adjusted values using a MMRM model.

Full analysis sets (FAS) of the trials NCT00793624 and NCT00796653. FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FEV1 was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FEV1 AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FEV1 was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FEV1 AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FEV1 was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FEV1 AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FEV1 was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FEV1 AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FEV1 was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FEV1 is defined as the FEV1 performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FEV1s if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline peak FEV1 was defined as the mean of the available pre-dose peak FEV1 values prior to first dose of randomized treatment. Peak FEV1 (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FEV1 was defined as the mean of the available pre-dose peak FEV1 values prior to first dose of randomized treatment. Peak FEV1 (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FEV1 was defined as the mean of the available pre-dose peak FEV1 values prior to first dose of randomized treatment. Peak FEV1 (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FEV1 was defined as the mean of the available pre-dose peak FEV1 values prior to first dose of randomized treatment. Peak FEV1 (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FEV1 was defined as the mean of the available pre-dose peak FEV1 values prior to first dose of randomized treatment. Peak FEV1 (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FVC was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FVC AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FVC was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FVC AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FVC was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FVC AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FVC was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FVC AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline FVC was defined as the mean of the -1 h and -10 min measurements performed just prior to administration of the first am dose of randomized treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit interaction as fixed categorical effects, baseline and baseline-by-visit interaction as fixed continuous covariates and patient as random effect. FVC AUC 0-3h was calculated from 0-3 hours post-dose using the trapezoidal rule, divided by the observation time (3h) to report in litres.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline trough FVC was defined as the mean of the -1 hour and -10 minute measurements performed just prior to first dose of randomized treatment. Trough FVC is defined as the FVC performed at -10 mins prior to study drug inhalation as the end of the dosing interval or the mean of -1h and -10 min FVCs if both available. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

FAS

Response was defined as change from baseline. Baseline peak FVC was defined as the mean of the available pre-dose peak FVC values prior to first dose of randomized treatment. Peak FVC (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FVC was defined as the mean of the available pre-dose peak FVC values prior to first dose of randomized treatment. Peak FVC (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FVC was defined as the mean of the available pre-dose peak FVC values prior to first dose of randomized treatment. Peak FVC (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FVC was defined as the mean of the available pre-dose peak FVC values prior to first dose of randomized treatment. Peak FVC (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Response was defined as change from baseline. Baseline peak FVC was defined as the mean of the available pre-dose peak FVC values prior to first dose of randomized treatment. Peak FVC (0-3h) values were obtained within 0 - 3 hours after treatment. Means are adjusted using a mixed effects model with treatment (trt), tio stratum, visit, trt-by-visit as fixed categorical effects, baseline and baseline-by-visit as fixed continuous covariates and patient as random effect.

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Weekly mean pre-dose morning and evening PEFR. Results are from non-MMRM ANCOVA models by week, with Last observation carried forward (LOCF) up to each week. Fixed effects include treatment, tiotropium, strata and baseline.

FAS

Mean number of puffs of rescue medication used per day (daytime/nighttime/total)

FAS

Patient's Global Rating (PGR) was a patient assessment of their health (respiratory condition) at each visit (compared to the day before they started study drug) and ranged from 1 (very much better) to 7 (very much worse).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Patient's Global Rating (PGR) was a patient assessment of their health (respiratory condition) at each visit (compared to the day before they started study drug) and ranged from 1 (very much better) to 7 (very much worse).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Patient's Global Rating (PGR) was a patient assessment of their health (respiratory condition) at each visit (compared to the day before they started study drug) and ranged from 1 (very much better) to 7 (very much worse).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Patient's Global Rating (PGR) was a patient assessment of their health (respiratory condition) at each visit (compared to the day before they started study drug) and ranged from 1 (very much better) to 7 (very much worse).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis set (FAS). FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.

Qualifying events of COPD were specifically pre-defined in the protocol. Other respiratory related events were evaluated by the investigator to see if they met the pre-defined criteria. Time to event was measured from the beginning of treatment. Cox regression analysis of treatment effect using tiotropium stratum as a stratification factor.

Qualifying events of COPD were specifically pre-defined in the protocol. Other respiratory related events were evaluated by the investigator to see if they met the pre-defined criteria. These exacerbations required hospitalization. Time to event was measured from the beginning of treatment. Cox regression analysis of treatment effect using tiotropium stratum as a stratification factor.

Qualifying events of COPD were specifically pre-defined in the protocol. Other respiratory related events were evaluated by the investigator to see if they met the pre-defined criteria. These exacerbations did not lead to hospitalization but included treatment with antibiotics and/or systemic steroids. Time to event was measured from the beginning of treatment. Cox regression analysis of treatment effect using tiotropium stratum as a stratification factor.

Qualifying events of COPD were specifically pre-defined in the protocol. Other respiratory related events were evaluated by the investigator to see if they met the pre-defined criteria.

Treated set- all patients who received at least one dose of study medication

Qualifying events of COPD were specifically pre-defined in the protocol. Other respiratory related events were evaluated by the investigator to see if they met the pre-defined criteria. These exacerbations required hospitalization.

Treated set- all patients who received at least one dose of study medication

Qualifying events of COPD were specifically pre-defined in the protocol. Other respiratory related events were evaluated by the investigator to see if they met the pre-defined criteria. These exacerbations did not lead to hospitalization but included treatment with antibiotics and/or systemic steroids.

Treated set- all patients who received at least one dose of study medication

Occurence of cardiac disorders and investigations related to treatment.

Treated set.

Absolute plasma concentrations of Olodaterol. Values presented are across visits and summarised into geometric means.

Pharmacokinetic set includes all patients in the treated set who had at least one valid olodaterol plasma concentration measurement after initial administration of study drug. This set is restricted to patients in either the Olodaterol 5 μg or 10 μg dose group.

This outcome measure describes the combined analysis of the trials NCT00793624 and NCT00796653. Mahler Transitional Dyspnea Index (TDI) focal score measures 3 components of dyspnea that evoke dyspnea in daily living: Functional Impairment, Magnitude of Task, and Magnitude of Effort. The TDI measures the change from the baseline assessment ranging from -9 (most deterioration) to +9 (most improvement).

Full analysis sets (FAS) of the trials NCT00793624 and NCT00796653. FAS is defined as all randomized patients who received at least one dose of treatment and had both baseline data and at least one post-baseline measurement at or before 24 weeks for any of the co-primary efficacy variables.