Comparative Study of Error Rates Between ELLIPTA® Dry Powder Inhaler (DPI) and Other DPIs

• Sponsor: GlaxoSmithKline (Industry)
• Overall Status: Completed
• CT.gov ID: NCT03114969
• Collaborator: (none)
• Enrollment: 450
• Locations: 13
• Actual Duration (Months): 9
• Patients Per Site: 34.6
• Patients Per Site Per Month: 3.8

Study Details

Study Description

Brief Summary

For effective drug delivery using inhalation route, it is important to use the inhalers correctly. This open-label study will evaluate the error rates during the use of ELLIPTA DPI, alone or in combination, in comparison with other DPIs. The study aims to provide clinical evidence in subjects with COPD that the reduced number of steps required to use the ELLIPTA DPI could result in fewer errors made by subjects, and therefore a more consistent treatment. Approximately 450 subjects prescribed with either of RELVAR® ELLIPTA, ANORO® ELLIPTA, INCRUSE® ELLIPTA, SYMBICORT® TURBUHALER®, SERETIDE® DISKUS®, SPIRIVA® HANDIHALER®, ULTIBRO® BREEZHALER® or SEEBRI® BREEZHALER will be included in the study and will have 2 clinical visits. At Visit 1, subjects will take their maintenance DPIs and the critical and overall errors made by subjects will be assessed. After the assessment, subjects will be instructed on correct use or informed of their correct use of their DPIs. The total duration of the study is approximately 6 weeks. ELLIPTA, SERETIDE and DISKUS are registered trademarks of the GSK group of companies. SYMBICORT and TURBUHALER are registered trademarks of the AstraZeneca group of companies. SPIRIVA and HANDIHALER are registered trademarks of Boehringer Ingelheim Pharmaceuticals. ULTIBRO, BREEZHALER and SEEBRI are registered trademarks of the Novartis group of companies.

Condition or Disease	Intervention/Treatment	Phase
Pulmonary Disease, Chronic Obstructive	Device: Relvar ELLIPTA; Device: Symbicort TURBUHALER; Device: Seretide DISKUS; Device: Spiriva HANDIHALER; Device: BREEZHALER; Device: Incruse ELLIPTA; Device: Anoro ELLIPTA

Study Design

• Study Type: Observational
• Actual Enrollment: 450 participants
• Observational Model: Cohort
• Time Perspective: Prospective
• Official Title: An Open-label, Low Interventional Clinical Study Investigating Error Rates (Critical and Overall) Prior to Any Retraining in Correct Use of the ELLIPTA Dry Powder Inhaler (DPI) Compared to Other DPIs Including; DISKUS, Turbuhaler, HandiHaler and Breezhaler as a Monotherapy or in Combination, in Adult Patients With Chronic Obstructive Pulmonary Disease (COPD)
• Actual Study Start Date: Jun 8, 2017
• Actual Primary Completion Date: Mar 9, 2018
• Actual Study Completion Date: Mar 9, 2018

Arms and Interventions

Arm	Intervention/Treatment
Subjects using RELVAR ELLIPTA Subjects with a fixed dose combination of inhaled corticosteroids/ long-acting beta agonists (ICS/LABA) via a single DPI of RELVAR ELLIPTA for treatment of COPD will be included.	Device: Relvar ELLIPTA; ELLIPTA inhaler containing Relvar will be used by subjects as their maintenance treatment to control COPD.
Subjects using SYMBICORT TURBUHALER Subjects with a fixed dose combination of ICS/LABA via a single DPI of SYMBICORT TURBUHALER for treatment of COPD will be included.	Device: Symbicort TURBUHALER; TURBUHALER inhaler containing Symbicort will be used by subjects as their maintenance treatment to control COPD.
Subjects using SERETIDE DISKUS Subjects with a fixed dose combination of ICS/LABA via a single DPI of SERETIDE DISKUS for treatment of COPD will be included.	Device: Seretide DISKUS; DISKUS inhaler containing Seretide will be used by subjects as their maintenance treatment to control COPD.
Subjects using SPIRIVA HANDIHALER Subjects with a fixed dose monotherapy of long-acting muscarinic antagonists (LAMA) via a single DPI of SPIRIVA HANDIHALER for treatment of COPD will be included.	Device: Spiriva HANDIHALER; HANDIHALER inhaler containing Spiriva will be used by subjects as their maintenance treatment to control COPD.
Subjects using INCRUSE ELLIPTA or ANORO ELLIPTA Subjects with a fixed dose monotherapy of LAMA via a single DPI of INCRUSE ELLIPTA or subjects taking a fixed dose combination of LAMA/LABA via a single DPI of ANORO ELLIPTA for treatment of COPD will be included.	Device: Incruse ELLIPTA; ELLIPTA inhaler containing Incruse will be used by subjects as their maintenance treatment to control COPD.; Device: Anoro ELLIPTA; ELLIPTA inhaler containing Anoro will be used by subjects as their maintenance treatment to control COPD.
Subjects using SEEBRI BREEZHALER or ULTIBRO BREEZHALER Subjects with a fixed dose monotherapy of LAMA via a single DPI of SEEBRI BREEZHALER or subjects taking a fixed dose combination of LAMA/LABA via a single DPI of ULTIBRO BREEZHALER for treatment of COPD will be included.	Device: BREEZHALER; BREEZHALER inhaler containing either Seebri or Ultibro will be used by subjects as their maintenance treatment to control COPD.
Subjects using RELVAR ELLIPTA with HANDIHALER/ INCRUSE ELLIPTA Subjects with a fixed dose combination of ICS/LABA via RELVAR ELLIPTA along with a fixed dose of LAMA via SPIRIVA HANDIHALER or INCRUSE ELLIPTA will be included.	Device: Relvar ELLIPTA; ELLIPTA inhaler containing Relvar will be used by subjects as their maintenance treatment to control COPD.; Device: Spiriva HANDIHALER; HANDIHALER inhaler containing Spiriva will be used by subjects as their maintenance treatment to control COPD.; Device: Incruse ELLIPTA; ELLIPTA inhaler containing Incruse will be used by subjects as their maintenance treatment to control COPD.
Subjects using TURBUHALER with HANDIHALER/INCRUSE ELLIPTA Subjects with a fixed dose combination of ICS/LABA via SYMBICORT TURBUHALER along with a fixed dose of LAMA via SPIRIVA HANDIHALER or INCRUSE ELLIPTA will be included.	Device: Symbicort TURBUHALER; TURBUHALER inhaler containing Symbicort will be used by subjects as their maintenance treatment to control COPD.; Device: Spiriva HANDIHALER; HANDIHALER inhaler containing Spiriva will be used by subjects as their maintenance treatment to control COPD.; Device: Incruse ELLIPTA; ELLIPTA inhaler containing Incruse will be used by subjects as their maintenance treatment to control COPD.
Subjects using DISKUS with HANDIHALER/INCRUSE ELLIPTA Subjects with a fixed dose combination of ICS/LABA via SERETIDE DISKUS along with a fixed dose of LAMA via SPIRIVA HANDIHALER or INCRUSE ELLIPTA will be included.	Device: Seretide DISKUS; DISKUS inhaler containing Seretide will be used by subjects as their maintenance treatment to control COPD.; Device: Spiriva HANDIHALER; HANDIHALER inhaler containing Spiriva will be used by subjects as their maintenance treatment to control COPD.; Device: Incruse ELLIPTA; ELLIPTA inhaler containing Incruse will be used by subjects as their maintenance treatment to control COPD.

Outcome Measures

Primary Outcome Measures

1. Percentage of Participants Making at Least One Critical Error at Visit 1-Primary Device Comparisons [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the health care practitioner (HCP) in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in patient instruction leaflets (PILs) for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The analysis was performed on the Intent to Treat (ITT) Population which comprised of all enrolled participants who demonstrated use of their primary DPI.

2. Percentage of Participants Making at Least One Critical Error at Visit 1 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and reporting and analysis plan (RAP).

3. Percentage of Participants Making at Least One Critical Error at Visit 1 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

4. Percentage of Participants Making at Least One Critical Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented.

5. Percentage of Participants Making at Least One Critical Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

6. Percentage of Participants Making at Least One Critical Error at Visit 1-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

Secondary Outcome Measures

1. Percentage of Participants Making at Least One Overall Error at Visit 1-Primary Device Comparisons [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants with at least one overall error in the primary DPI is presented.

2. Percentage of Participants Making at Least One Overall Error at Visit 1 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants with at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

3. Percentage of Participants Making at Least One Overall Error at Visit 1 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants with at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

4. Percentage of Participants Making at Least One Overall Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in either one or both devices (where applicable) is presented.

5. Percentage of Participants Making at Least One Overall Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. Errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. Percentage of participants making at least one overall error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

6. Percentage of Participants Making at Least One Overall Error at Visit 1-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI) [Day 1]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. Errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. Percentage of participants making at least one overall error in either one or both devices (where applicable) is presented.The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

7. Percentage of Participants Making at Least One Critical Error at Visit 2-Primary Device Comparisons [Week 6]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented.

8. Percentage of Participants Making at Least One Critical Error at Visit 2 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Week 6]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

9. Percentage of Participants Making at Least One Critical Error at Visit 2 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI [Week 6]

   Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

10. Percentage of Participants Making at Least One Critical Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented.

11. Percentage of Participants Making at Least One Critical Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

12. Percentage of Participants Making at Least One Critical Error at Visit 2-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

13. Percentage of Participants Making at Least One Overall Error at Visit 2-Primary Device Comparisons [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in the primary DPI is presented.

14. Percentage of Participants Making at Least One Overall Error at Visit 2 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps listed in PILs for the respective DPI. The errors were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

15. Percentage of Participants Making at Least One Overall Error at Visit 2 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps listed in PILs for the respective DPI. The errors were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP. .

16. Percentage of Participants Making at Least One Overall Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in either one or both devices (where applicable) is presented.

17. Percentage of Participants Making at Least One Overall Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on correct use of inhalers. Any error made by the participant was recorded by HCP in the checklist. Checklist of instructions for correct use were based on steps for correct use listed in PILs for the respective DPI. Errors made during demonstration by participants were defined as critical, when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per PIL. Overall errors is the combination of critical and non-critical errors. Percentage of participants making at least one overall error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.

18. Percentage of Participants Making at Least One Overall Error at Visit 2-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI) [Week 6]

    Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.

Eligibility Criteria

Criteria

• Ages Eligible for Study: 40 Years and Older
• Sexes Eligible for Study: All
• Accepts Healthy Volunteers: No

Inclusion Criteria:

• Subjects with documented Physician's diagnosis of COPD, and currently receiving maintenance therapy.

• Aged >=40 years of age at inclusion.

• Using one of the maintenance therapies of interest for at least 3 months prior to inclusion on the study.

• Males or females.

• Capable of giving signed informed consent, which includes compliance with the requirements and restrictions listed in the consent form and in the protocol.

Exclusion Criteria:

• Asthma: Subjects with a current diagnosis of asthma. Subjects with a prior history of asthma are eligible if they have a current diagnosis of COPD.

• Drug/alcohol abuse: Subjects with a known or suspected alcohol or drug abuse history at screening (Visit 0) that in the opinion of the investigator could interfere with the subject's proper completion of the protocol requirement.

• Investigational product: Subjects who have received an investigational drug and/or medical device within 30 days of entry into this study (Screening/Visit 1), or within five half-lives of the investigational drug, whichever is longer.

• Investigational product: Subjects who have been trained during participation in any device study in the 6 months prior to entry into this study.

Contacts and Locations

Locations

	Site	City	State	Country	Postal Code
1	GSK Investigational Site	Beek En Donk		Netherlands	5741 CG
2	GSK Investigational Site	Beek		Netherlands	6191JW
3	GSK Investigational Site	Eindhoven		Netherlands	5623 EJ
4	GSK Investigational Site	Hengelo		Netherlands	7555 DL
5	GSK Investigational Site	Hoorn		Netherlands	1624 NP
6	GSK Investigational Site	Kloosterhaar		Netherlands	7694 AC
7	GSK Investigational Site	Nijverdal		Netherlands	7442 LS
8	GSK Investigational Site	Rotterdam		Netherlands	3051 GV
9	GSK Investigational Site	Zutphen		Netherlands	7207 AE
10	GSK Investigational Site	Northwood	Middlesex	United Kingdom	HA6 2RN
11	GSK Investigational Site	London		United Kingdom	EC1M 6BQ
12	GSK Investigational Site	London		United Kingdom	SW17 0QT
13	GSK Investigational Site	Sidcup, Kent		United Kingdom	DA14 6LT

Sponsors and Collaborators

• GlaxoSmithKline

Investigators

• Study Director: GSK Clinical Trials, GlaxoSmithKline

Study Documents (Full-Text)

• Study Protocol - Jan 24, 2017
• Statistical Analysis Plan - Apr 26, 2018

More Information

Publications

• Collier DJ, Wielders P, van der Palen J, Heyes L, Midwinter D, Collison K, Preece A, Barnes N, Sharma R. Critical Error Frequency and the Impact of Training with Inhalers Commonly used for Maintenance Treatment in Chronic Obstructive Pulmonary Disease. Int J Chron Obstruct Pulmon Dis. 2020 Jun 9;15:1301-1313. doi: 10.2147/COPD.S224209. eCollection 2020.

• Responsible Party: GlaxoSmithKline
• ClinicalTrials.gov Identifier: NCT03114969
• Other Study ID Numbers: 204981
• First Posted: Apr 14, 2017
• Last Update Posted: Oct 28, 2020
• Last Verified: Oct 1, 2020

• Individual Participant Data (IPD) Sharing Statement: Yes
• Plan to Share IPD: Yes
• Studies a U.S. FDA-regulated Drug Product: No
• Studies a U.S. FDA-regulated Device Product: No
• Keywords provided by GlaxoSmithKline: Overall error; Error rate; Critical error; COPD; ELLIPTA; Dry powder inhaler
• Additional relevant MeSH terms: Lung Diseases; Pulmonary Disease, Chronic Obstructive; Budesonide, Formoterol Fumarate Drug Combination; Tiotropium Bromide; Fluticasone-Salmeterol Drug Combination

Study Results

Participant Flow

Recruitment Details	This was an open-label study investigating error rates for ELLIPTA dry powder inhaler (DPI), alone or in combination when compared to other DPIs (DISKUS, Turbuhaler, HandiHaler and Breezhaler) and combinations of these, as prescribed to chronic obstructive pulmonary disease (COPD) participants, prior to any retraining in correct use.
Pre-assignment Detail	A total of 461 participants were screened of which 11 failed screening and 450 participants were enrolled. The study was conducted in two countries-Netherlands and the United Kingdom.

Arm/Group Title	Relvar Ellipta	Symbicort Turbuhaler	Seretide Diskus	Spiriva Handihaler	Incruse/Anoro Ellipta	Ultibro/Seebri Breezhaler	Relvar Ellipta+LAMA	Symbicort Turbuhaler+LAMA	Seretide Diskus+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.	Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Period Title: Overall Study
STARTED	50	50	50	50	51	49	50	50	50
COMPLETED	50	50	46	48	51	48	49	50	48
NOT COMPLETED	0	0	4	2	0	1	1	0	2

Baseline Characteristics

Arm/Group Title	Relvar Ellipta	Symbicort Turbuhaler	Seretide Diskus	Spiriva Handihaler	Incruse/Anoro Ellipta	Ultibro/Seebri Breezhaler	Relvar Ellipta+LAMA	Symbicort Turbuhaler+LAMA	Seretide Diskus+LAMA	Total
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.	Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Total of all reporting groups
Overall Participants	50	50	50	50	51	49	50	50	50	450
Age (Years) [Mean (Standard Deviation) ]
Mean (Standard Deviation) [Years]	64.6 (10.24)	65.2 (11.43)	68.7 (10.67)	69.5 (9.95)	66.8 (8.93)	67.7 (10.01)	66.6 (8.44)	67.5 (8.21)	68.1 (9.16)	67.2 (9.75)
Sex: Female, Male (Count of Participants)
Female	29 58%	26 52%	29 58%	21 42%	24 47.1%	19 38.8%	28 56%	18 36%	14 28%	208 46.2%
Male	21 42%	24 48%	21 42%	29 58%	27 52.9%	30 61.2%	22 44%	32 64%	36 72%	242 53.8%
Race/Ethnicity, Customized (Count of Participants)
African American/African Heritage	0 0%	1 2%	0 0%	0 0%	0 0%	1 2%	0 0%	1 2%	0 0%	3 0.7%
Asian-Central/South Asian Heritage	0 0%	0 0%	4 8%	0 0%	3 5.9%	1 2%	2 4%	2 4%	3 6%	15 3.3%
Asian-East Asian Heritage	0 0%	0 0%	0 0%	0 0%	1 2%	0 0%	0 0%	0 0%	0 0%	1 0.2%
Asian-Japanese Heritage	1 2%	0 0%	0 0%	0 0%	0 0%	0 0%	0 0%	0 0%	0 0%	1 0.2%
Asian-South East Asian Heritage	0 0%	0 0%	1 2%	0 0%	0 0%	0 0%	0 0%	0 0%	1 2%	2 0.4%
White-Arabic/North African Heritage	0 0%	1 2%	4 8%	1 2%	0 0%	0 0%	1 2%	0 0%	0 0%	7 1.6%
White-White/Caucasian/European Heritage	49 98%	48 96%	41 82%	49 98%	47 92.2%	47 95.9%	47 94%	47 94%	46 92%	421 93.6%

Outcome Measures

1. Primary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 1-Primary Device Comparisons
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the health care practitioner (HCP) in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in patient instruction leaflets (PILs) for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The analysis was performed on the Intent to Treat (ITT) Population which comprised of all enrolled participants who demonstrated use of their primary DPI.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta	Symbicort Turbuhaler	Seretide Diskus	Spiriva Handihaler	Incruse/Anoro Ellipta	Ultibro/Seebri Breezhaler	Relvar Ellipta+LAMA	Symbicort Turbuhaler+LAMA	Seretide Diskus+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.	Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	50	50	50	51	49	50	50	50
Number [Percentage of participants]	10 20%	40 80%	26 52%	34 68%	10 19.6%	33 67.3%	12 24%	38 76%	26 52%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.005
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.657
	Confidence Interval	(2-Sided) 95%; 1.584 to 13.686
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.114
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.478
	Confidence Interval	(2-Sided) 95%; 0.805 to 7.632
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.026
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.499
	Confidence Interval	(2-Sided) 95%; 1.160 to 10.555
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.012
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.943
	Confidence Interval	(2-Sided) 95%; 1.348 to 11.534
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.746
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.223
	Confidence Interval	(2-Sided) 95%; 0.361 to 4.151
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.562
	Confidence Interval	(2-Sided) 95%; 1.932 to 16.013
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.050
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.979
	Confidence Interval	(2-Sided) 95%; 0.999 to 8.882
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.006
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.418
	Confidence Interval	(2-Sided) 95%; 1.521 to 12.834
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.009
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.165
	Confidence Interval	(2-Sided) 95%; 1.425 to 12.178
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.761
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.209
	Confidence Interval	(2-Sided) 95%; 0.357 to 4.095
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

2. Primary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 1 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and reporting and analysis plan (RAP).
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]	Relvar Ellipta
Arm/Group Description	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.
Measure Participants	150	50
Number [Percentage of participants]	25 50%	10 20%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.100
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.292
	Confidence Interval	(2-Sided) 95%; 0.853 to 6.163
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.051
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.642
	Confidence Interval	(2-Sided) 95%; 0.994 to 7.022
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

3. Primary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 1 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	100	100	100
Number [Percentage of participants]	11 22%	39 78%	26 52%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.995
	Confidence Interval	(2-Sided) 95%; 1.808 to 8.829
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.069
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.181
	Confidence Interval	(2-Sided) 95%; 0.940 to 5.059
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.855
	Confidence Interval	(2-Sided) 95%; 2.339 to 10.080
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.010
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.710
	Confidence Interval	(2-Sided) 95%; 1.274 to 5.764
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

4. Primary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta	Relvar Ellipta+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	50
Number [Percentage of participants]	10 20%	22 44%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.108
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.503
	Confidence Interval	(2-Sided) 95%; 0.818 to 7.659
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.122
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.409
	Confidence Interval	(2-Sided) 95%; 0.792 to 7.331
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

5. Primary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.
Measure Participants	50	150
Number [Percentage of participants]	10 20%	41 82%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.001
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.887
	Confidence Interval	(2-Sided) 95%; 1.851 to 12.903
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.484
	Confidence Interval	(2-Sided) 95%; 2.106 to 14.284
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

6. Primary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 1-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	100	100	100
Number [Percentage of participants]	16 32%	47 94%	36 72%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.941
	Confidence Interval	(2-Sided) 95%; 1.863 to 8.337
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.030
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.391
	Confidence Interval	(2-Sided) 95%; 1.088 to 5.256
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.728
	Confidence Interval	(2-Sided) 95%; 2.388 to 9.364
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.002
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.957
	Confidence Interval	(2-Sided) 95%; 1.473 to 5.936
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

7. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 1-Primary Device Comparisons
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants with at least one overall error in the primary DPI is presented.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta	Symbicort Turbuhaler	Seretide Diskus	Spiriva Handihaler	Incruse/Anoro Ellipta	Ultibro/Seebri Breezhaler	Relvar Ellipta+LAMA	Symbicort Turbuhaler+LAMA	Seretide Diskus+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.	Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	50	50	50	51	49	50	50	50
Number [Percentage of participants]	34 68%	64 128%	60 120%	74 148%	37 72.5%	55 112.2%	34 68%	70 140%	66 132%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.045
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.404
	Confidence Interval	(2-Sided) 95%; 1.018 to 5.676
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.172
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.823
	Confidence Interval	(2-Sided) 95%; 0.770 to 4.319
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.016
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.989
	Confidence Interval	(2-Sided) 95%; 1.229 to 7.272
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.179
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.747
	Confidence Interval	(2-Sided) 95%; 0.775 to 3.937
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.690
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.189
	Confidence Interval	(2-Sided) 95%; 0.509 to 2.775
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.004
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.363
	Confidence Interval	(2-Sided) 95%; 1.480 to 7.639
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.012
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.848
	Confidence Interval	(2-Sided) 95%; 1.264 to 6.420
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.630
	Confidence Interval	(2-Sided) 95%; 1.986 to 10.791
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.081
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.037
	Confidence Interval	(2-Sided) 95%; 0.916 to 4.528
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	1.000
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.000
	Confidence Interval	(2-Sided) 95%; 0.438 to 2.283
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

8. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 1 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants with at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]	Relvar Ellipta
Arm/Group Description	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.
Measure Participants	150	50
Number [Percentage of participants]	57 114%	34 68%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.067
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.935
	Confidence Interval	(2-Sided) 95%; 0.955 to 3.921
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.007
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.523
	Confidence Interval	(2-Sided) 95%; 1.283 to 4.961
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

9. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 1 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants with at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	100	100	100
Number [Percentage of participants]	34 68%	67 134%	63 126%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.008
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.399
	Confidence Interval	(2-Sided) 95%; 1.252 to 4.596
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.082
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.812
	Confidence Interval	(2-Sided) 95%; 0.926 to 3.544
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.840
	Confidence Interval	(2-Sided) 95%; 2.135 to 6.905
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.231
	Confidence Interval	(2-Sided) 95%; 1.810 to 5.766
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

10. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in either one or both devices (where applicable) is presented.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta	Relvar Ellipta+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	50
Number [Percentage of participants]	34 68%	46 92%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.120
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.931
	Confidence Interval	(2-Sided) 95%; 0.842 to 4.428
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.232
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.636
	Confidence Interval	(2-Sided) 95%; 0.730 to 3.664
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

11. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 1-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. Errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. Percentage of participants making at least one overall error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.
Measure Participants	50	150
Number [Percentage of participants]	34 68%	71 142%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.217
	Confidence Interval	(2-Sided) 95%; 2.019 to 8.807
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.410
	Confidence Interval	(2-Sided) 95%; 2.660 to 11.005
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator.

12. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 1-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 1 and any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. Errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. Percentage of participants making at least one overall error in either one or both devices (where applicable) is presented.The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Day 1

Outcome Measure Data

Analysis Population Description

ITT Population

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	100	100	100
Number [Percentage of participants]	40 80%	72 144%	74 148%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.007
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.477
	Confidence Interval	(2-Sided) 95%; 1.274 to 4.815
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.006
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.748
	Confidence Interval	(2-Sided) 95%; 1.328 to 5.683
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.896
	Confidence Interval	(2-Sided) 95%; 2.133 to 7.118
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	<0.001
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.777
	Confidence Interval	(2-Sided) 95%; 2.508 to 9.100
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator.

13. Secondary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 2-Primary Device Comparisons
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta	Symbicort Turbuhaler	Seretide Diskus	Spiriva Handihaler	Incruse/Anoro Ellipta	Ultibro/Seebri Breezhaler	Relvar Ellipta+LAMA	Symbicort Turbuhaler+LAMA	Seretide Diskus+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.	Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	48	45	47	50	47	49	49	46
Number [Percentage of participants]	0 0%	8 16%	11 22%	21 42%	0 0%	17 34.7%	6 12%	6 12%	13 26%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.176
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	7.347
	Confidence Interval	(2-Sided) 95%; 0.408 to 132.143
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.128
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	9.300
	Confidence Interval	(2-Sided) 95%; 0.526 to 164.465
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.038
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	20.454
	Confidence Interval	(2-Sided) 95%; 1.190 to 351.613
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.041
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	18.776
	Confidence Interval	(2-Sided) 95%; 1.131 to 311.669
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.141
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	8.873
	Confidence Interval	(2-Sided) 95%; 0.484 to 162.775
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.126
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	10.215
	Confidence Interval	(2-Sided) 95%; 0.519 to 200.904
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.082
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	13.717
	Confidence Interval	(2-Sided) 95%; 0.715 to 263.125
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.024
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	28.283
	Confidence Interval	(2-Sided) 95%; 1.561 to 512.532
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.038
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	21.736
	Confidence Interval	(2-Sided) 95%; 1.183 to 399.541
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.188
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	7.603
	Confidence Interval	(2-Sided) 95%; 0.371 to 155.763
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

14. Secondary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 2 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]	Relvar Ellipta
Arm/Group Description	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.
Measure Participants	144	50
Number [Percentage of participants]	8 16%	0 0%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.147
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	7.781
	Confidence Interval	(2-Sided) 95%; 0.488 to 124.117
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.120
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	9.786
	Confidence Interval	(2-Sided) 95%; 0.553 to 173.301
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

15. Secondary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 2 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in the primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	99	97	91
Number [Percentage of participants]	3 6%	7 14%	12 24%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.395
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.089
	Confidence Interval	(2-Sided) 95%; 0.383 to 11.389
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.166
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.250
	Confidence Interval	(2-Sided) 95%; 0.612 to 17.244
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.175
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.196
	Confidence Interval	(2-Sided) 95%; 0.596 to 17.140
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.042
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.408
	Confidence Interval	(2-Sided) 95%; 1.059 to 27.610
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

16. Secondary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta	Relvar Ellipta+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	49
Number [Percentage of participants]	0 0%	8 16%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.090
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	12.145
	Confidence Interval	(2-Sided) 95%; 0.680 to 216.818
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.130
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	9.991
	Confidence Interval	(2-Sided) 95%; 0.508 to 196.435
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

17. Secondary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.
Measure Participants	50	144
Number [Percentage of participants]	0 0%	13 26%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.079
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	11.907
	Confidence Interval	(2-Sided) 95%; 0.753 to 188.208
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.063
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	15.060
	Confidence Interval	(2-Sided) 95%; 0.866 to 262.042
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

18. Secondary Outcome

Title	Percentage of Participants Making at Least One Critical Error at Visit 2-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose. The percentage of participants making at least one critical error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one critical error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	99	97	91
Number [Percentage of participants]	4 8%	11 22%	13 26%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.239
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.670
	Confidence Interval	(2-Sided) 95%; 0.521 to 13.690
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.200
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.929
	Confidence Interval	(2-Sided) 95%; 0.567 to 15.125
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.078
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.269
	Confidence Interval	(2-Sided) 95%; 0.848 to 21.489
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.046
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.132
	Confidence Interval	(2-Sided) 95%; 1.031 to 25.550
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

19. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 2-Primary Device Comparisons
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in the primary DPI is presented.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta	Symbicort Turbuhaler	Seretide Diskus	Spiriva Handihaler	Incruse/Anoro Ellipta	Ultibro/Seebri Breezhaler	Relvar Ellipta+LAMA	Symbicort Turbuhaler+LAMA	Seretide Diskus+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.	Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	48	45	47	50	47	49	49	46
Number [Percentage of participants]	12 24%	35 70%	27 54%	36 72%	8 15.7%	32 65.3%	14 28%	14 28%	28 56%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.020
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.483
	Confidence Interval	(2-Sided) 95%; 1.218 to 9.961
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.143
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.284
	Confidence Interval	(2-Sided) 95%; 0.757 to 6.895
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.006
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.268
	Confidence Interval	(2-Sided) 95%; 1.615 to 17.187
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.009
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.655
	Confidence Interval	(2-Sided) 95%; 1.478 to 14.666
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 5

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.679
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.274
	Confidence Interval	(2-Sided) 95%; 0.405 to 4.005
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 6

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.010
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	3.804
	Confidence Interval	(2-Sided) 95%; 1.372 to 10.544
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 7

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.083
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.555
	Confidence Interval	(2-Sided) 95%; 0.886 to 7.369
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 8

Statistical Analysis Overview	Comparison Group Selection	Spiriva Handihaler, Incruse/Anoro Ellipta
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.002
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	5.930
	Confidence Interval	(2-Sided) 95%; 1.890 to 18.611
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 9

Statistical Analysis Overview	Comparison Group Selection	Incruse/Anoro Ellipta, Ultibro/Seebri Breezhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.007
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	4.929
	Confidence Interval	(2-Sided) 95%; 1.556 to 15.612
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Incruse/Anoro Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 10

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Relvar Ellipta+LAMA
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.746
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.208
	Confidence Interval	(2-Sided) 95%; 0.385 to 3.788
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

20. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 2 in Primary DPI (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps listed in PILs for the respective DPI. The errors were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]	Relvar Ellipta
Arm/Group Description	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.
Measure Participants	144	50
Number [Percentage of participants]	19 38%	12 24%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.404
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.502
	Confidence Interval	(2-Sided) 95%; 0.578 to 3.905
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.330
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.591
	Confidence Interval	(2-Sided) 95%; 0.625 to 4.050
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

21. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 2 in Primary DPI (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps listed in PILs for the respective DPI. The errors were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in primary DPI is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP. .
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	99	97	91
Number [Percentage of participants]	13 26%	25 50%	27 54%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.189
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.738
	Confidence Interval	(2-Sided) 95%; 0.761 to 3.968
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.086
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.079
	Confidence Interval	(2-Sided) 95%; 0.901 to 4.799
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.082
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.950
	Confidence Interval	(2-Sided) 95%; 0.918 to 4.142
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.018
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.415
	Confidence Interval	(2-Sided) 95%; 1.161 to 5.024
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

22. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus Relvar Ellipta With Any Other LAMA)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The percentage of participants making at least one overall error in either one or both devices (where applicable) is presented.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta	Relvar Ellipta+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
Measure Participants	50	49
Number [Percentage of participants]	12 24%	16 32%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.472
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.510
	Confidence Interval	(2-Sided) 95%; 0.492 to 4.633
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.553
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.402
	Confidence Interval	(2-Sided) 95%; 0.459 to 4.283
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

23. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 2-Dual Device Comparisons (Relvar Ellipta DPI Versus All ICS/LABA DPIs With a LAMA Second DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on correct use of inhalers. Any error made by the participant was recorded by HCP in the checklist. Checklist of instructions for correct use were based on steps for correct use listed in PILs for the respective DPI. Errors made during demonstration by participants were defined as critical, when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per PIL. Overall errors is the combination of critical and non-critical errors. Percentage of participants making at least one overall error in either one or both devices (where applicable) is presented. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta versus all ICS/LABA+LAMA DPIs; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta	[Relvar Ellipta+LAMA]+[Symb Turb+LAMA]+[Seretide Diskus+LAMA]
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.	All participants from Relvar Ellipta+LAMA, Symbicort Turbuhaler (Symb Turb)+LAMA, and Seretide Diskus+LAMA arms were included.
Measure Participants	50	144
Number [Percentage of participants]	12 24%	29 58%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.052
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.523
	Confidence Interval	(2-Sided) 95%; 0.993 to 6.407
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.029
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.757
	Confidence Interval	(2-Sided) 95%; 1.107 to 6.862
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

24. Secondary Outcome

Title	Percentage of Participants Making at Least One Overall Error at Visit 2-Dual Device Comparisons (Relvar Ellipta With or Without a LAMA DPI) Versus Any Other ICS/LABA DPI With or Without a LAMA DPI)
Description	Participants were asked to demonstrate use of their prescribed DPI at Visit 2 within 6 weeks after they were retrained on the correct use of inhalers. Any error made by the participant was recorded by the HCP in the checklist. Checklist of instructions for correct use were based on the steps for correct use listed in PILs for the respective DPI. The errors made during demonstration by participants were defined as "critical", when the participant received a lesser/no dose and non-critical when the dose may not be affected, but the participant has demonstrated improper use of their DPI, as per the PIL. Overall errors is the combination of critical and non-critical errors. The aim of the analysis was to compare percentage of participants making at least one overall error with Ellipta and Ellipta+LAMA versus Turbuhaler and Turbuhaler+LAMA and Diskus and Diskus+LAMA; hence, the arms were combined as pre-specified in the protocol and RAP.
Time Frame	Week 6

Outcome Measure Data

Analysis Population Description

ITT Population. Only those participants with data available at Visit 2 were analyzed.

Arm/Group Title	Relvar Ellipta+[Relvar Ellipta+LAMA]	Symbicort Turbuhaler+[Symbicort Turbuhaler+LAMA]	Seretide Diskus+[Seretide Diskus+LAMA]
Arm/Group Description	All participants from Relvar Ellipta and Relvar Ellipta+LAMA arms were included.	All participants from Symbicort Turbuhaler and Symbicort Turbuhaler+LAMA arms were included.	All participants from Seretide Diskus and Seretide Diskus+LAMA arms were included.
Measure Participants	99	97	91
Number [Percentage of participants]	14 28%	33 66%	34 68%

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.024
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.434
	Confidence Interval	(2-Sided) 95%; 1.123 to 5.276
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.035
	Comments	Analysis was performed using logistic regression with treatment cohort as fixed effect and adjusting for the covariate of time on current primary DPI.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.389
	Confidence Interval	(2-Sided) 95%; 1.061 to 5.376
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Symbicort Turbuhaler
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.003
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.889
	Confidence Interval	(2-Sided) 95%; 1.434 to 5.819
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	Relvar Ellipta, Seretide Diskus
	Comments
	Type of Statistical Test	Superiority
	Comments
Statistical Test of Hypothesis	p-Value	0.003
	Comments	Sensitivity analysis was performed using logistic regression with treatment cohort as fixed effect. Time on current primary device was removed from the model for this sensitivity analysis as it was confounded with primary device/treatment cohort.
	Method	Regression, Logistic
	Comments
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	2.974
	Confidence Interval	(2-Sided) 95%; 1.461 to 6.055
	Parameter Dispersion	Type: Value:
	Estimation Comments	Odds ratio was calculated considering Relvar Ellipta+[Relvar Ellipta+LAMA] as comparator. Note: Statistical testing with a small number of events may produce unusual results; odds ratios and confidence intervals should be interpreted with caution.

Adverse Events

Time Frame	Serious adverse events (SAEs) and non-SAEs were collected until completion of the study (Up to Week 6)
Adverse Event Reporting Description	SAEs and non-SAEs were collected in the ITT Population.
Arm/Group Title	Relvar Ellipta		Symbicort Turbuhaler		Seretide Diskus		Spiriva Handihaler		Incruse/Anoro Ellipta		Ultibro/Seebri Breezhaler		Relvar Ellipta+LAMA		Symbicort Turbuhaler+LAMA		Seretide Diskus+LAMA
Arm/Group Description	Participants prescribed a fixed dose combination of inhaled corticosteroid/long-acting beta agonist (ICS/LABA) via Relvar Ellipta for treatment of COPD were included. Participants were assessed for errors while using Relvar Ellipta prior to retraining on the correct use of Ellipta DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Relvar Ellipta within 6 weeks at Visit 2.		Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler for treatment of COPD were included. Participants were assessed for errors while using Symbicort Turbuhaler prior to retraining on the correct use of Turbuhaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Symbicort Turbuhaler within 6 weeks at Visit 2.		Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus for treatment of COPD were included. Participants were assessed for errors while using Seretide Diskus prior to retraining on the correct use of Diskus at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Seretide Diskus within 6 weeks at Visit 2.		Participants prescribed a fixed dose of long-acting muscarinic antagonist (LAMA) via Spiriva Handihaler for treatment of COPD were included. Participants were assessed for errors while using Spiriva Handihaler prior to retraining on the correct use of Handihaler at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Spiriva Handihaler within 6 weeks at Visit 2.		Participants prescribed a fixed dose of LAMA via Incruse Ellipta or participants taking a fixed dose combination of LAMA/LABA via Anoro Ellipta for treatment of COPD were included. Participants were assessed for errors while using Incruse/Anoro Ellipta prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Incruse/Anoro Ellipta within 6 weeks at Visit 2.		Participants prescribed a fixed dose of LAMA via Seebri Breezehaler or participants taking a fixed dose combination of LAMA/LABA via a single DPI of Ultibro Breezehaler for treatment of COPD were included. Participants were assessed for errors while using Ultibro/Seebri Breezehaler prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of Ultibro/Seebri Breezehaler within 6 weeks at Visit 2.		Participants prescribed a fixed dose combination of ICS/LABA via Relvar Ellipta along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.		Participants prescribed a fixed dose combination of ICS/LABA via Symbicort Turbuhaler along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.		Participants prescribed a fixed dose combination of ICS/LABA via Seretide Diskus along with a fixed dose of LAMA via Spiriva Handihaler or Incruse Ellipta were included. Participants were assessed for errors while using combination DPI prior to retraining on the correct use of DPI at Visit 1 (Day 1). The participants were retrained on inhaler use following which they were reassessed for errors during use of combination DPI within 6 weeks at Visit 2.
All Cause Mortality
	Relvar Ellipta		Symbicort Turbuhaler		Seretide Diskus		Spiriva Handihaler		Incruse/Anoro Ellipta		Ultibro/Seebri Breezhaler		Relvar Ellipta+LAMA		Symbicort Turbuhaler+LAMA		Seretide Diskus+LAMA
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	0/50 (0%)		0/50 (0%)		0/50 (0%)		0/50 (0%)		0/51 (0%)		0/49 (0%)		0/50 (0%)		0/50 (0%)		0/50 (0%)
Serious Adverse Events
	Relvar Ellipta		Symbicort Turbuhaler		Seretide Diskus		Spiriva Handihaler		Incruse/Anoro Ellipta		Ultibro/Seebri Breezhaler		Relvar Ellipta+LAMA		Symbicort Turbuhaler+LAMA		Seretide Diskus+LAMA
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	0/50 (0%)		0/50 (0%)		0/50 (0%)		0/50 (0%)		1/51 (2%)		1/49 (2%)		0/50 (0%)		1/50 (2%)		0/50 (0%)
General disorders
Chest pain	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	0/51 (0%)	0	1/49 (2%)	1	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0
Infections and infestations
Appendicitis	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	1/51 (2%)	1	0/49 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0
Respiratory, thoracic and mediastinal disorders
Chronic obstructive pulmonary disease	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	0/50 (0%)	0	0/51 (0%)	0	0/49 (0%)	0	0/50 (0%)	0	1/50 (2%)	1	0/50 (0%)	0
Other (Not Including Serious) Adverse Events
	Relvar Ellipta		Symbicort Turbuhaler		Seretide Diskus		Spiriva Handihaler		Incruse/Anoro Ellipta		Ultibro/Seebri Breezhaler		Relvar Ellipta+LAMA		Symbicort Turbuhaler+LAMA		Seretide Diskus+LAMA
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	4/50 (8%)		1/50 (2%)		0/50 (0%)		1/50 (2%)		2/51 (3.9%)		2/49 (4.1%)		3/50 (6%)		3/50 (6%)		5/50 (10%)
Respiratory, thoracic and mediastinal disorders
Chronic obstructive	4/50 (8%)	4	1/50 (2%)	1	0/50 (0%)	0	1/50 (2%)	1	2/51 (3.9%)	2	2/49 (4.1%)	2	3/50 (6%)	3	3/50 (6%)	4	5/50 (10%)	5

Limitations/Caveats

[Not Specified]

More Information

Certain Agreements

Principal Investigators are NOT employed by the organization sponsoring the study.

GSK agreements may vary with individual investigators, but will not prohibit any investigator from publishing. GSK supports the publication of results from all centers of a multi-center trial but requests that reports based on single-site data not precede the primary publication of the entire clinical trial.