Table 1. Study characteristics.

Study publication (Study number/ Trial registration)	Study designa and location	Key patient eligibility criteria	ITT population, n	Device(s) assessed	Key outcomes assessedb	Device training schedule	Study duration, days
van der Palen et al. 2016 (200301/ NCT02184624) [15]	Randomized, cross-over, multicenter study at 2 centers in the UK and 6 centers in the Netherlands	Inclusion criteria • Aged ≥40 years and with a primary diagnosis of COPD, as defined by ATS/ERS criteria [21] • Naïve to ELLIPTA DPI use and ≥1 other inhaler device Exclusion criteria • History of allergy/ hypersensitivity to lactose/milk protein or magnesium stearate or to any other excipient found in commercially available inhaler devices • Current diagnosis of asthma	567	The following devices were assessed vs. ELLIPTA DPI in 5 separate substudies: • DISKUS • MDI • Turbuhaler • HandiHalerBreezhaler	• Correct use, assessed by a trained HCP using a checklist of inhaler errors (based on the ELLIPTA DPI PIL) after patients had read the PIL (primary endpointc) [14] • Ease-of-use, assessed using an ease-of-use questionnaire (see S1 Table) • AEs	Patients were asked to read the PIL of the device and then asked to perform inhaler use. Errors were recorded by a trained respiratory nurse. If the patient made any errors, the nurse demonstrated correct use of the inhaler and the patient was asked to perform inhaler use again. If the patient continued to make errors, the nurse demonstrated the process again up to a maximum of 3 times.	1
Feldman et al. 2019 (201071/ NCT02586493) [16]	Randomized, single-arm, multicenter study at 17 centers in the US	Inclusion criteria • Aged ≥40 years, with an established diagnosis of COPD as defined by ATS/ERS criteria [21], and receiving maintenance inhaler therapy for COPD Exclusion criteria • Current diagnosis of asthma Use of the DPI within the previous 6 months	267	Placebo ELLIPTA DPI	• Correct use, assessed by a trained HCP using a correct use checklist (based on the ELLIPTA DPI PIL) on Visit 2 (Day 28) after a single attempt without further instruction (patients read the PIL before Visit 1 only) [14] • Ease-of-use, assessed using an ease-of-use questionnaire (primary endpoint) (see S2 Table) • AEs	Comprised 2 study visits and a telephone call. At Visit 1 (Day 1), the screening procedure involved an assessment of correct inhaler use within 3 attempts. Prior to their first attempt, subjects reviewed written instructions for the correct use of the inhaler based on the PIL but did not receive any training. Subjects were permitted to receive training (verbal instruction and a demonstration of correct use) from the HCP in between attempts 1 and 2, and/or attempts 2 and 3, if necessary. Subjects unable to demonstrate the correct use of the placebo inhaler within 3 attempts at Visit 1 were considered screening failures and did not continue in the study.	28
van der Palen et al. 2018 (206215/ NCT02982187) [17]	Randomized, cross-over, multicenter study at 2 centers in the UK and 3 centers in the Netherlands	Inclusion criteria • Aged ≥40 years and current or former smokers with ≥10 pack-years of smoking history, with a documented COPD history, as defined by ATS/ERS criteria [21] • Receiving maintenance therapy with a fixed-dose ICS/LABA combination inhaler either with or without concurrent LAMA therapy during the preceding 4 weeks. Short-acting rescue medications were permitted Exclusion criteria • Current diagnosis of asthma and patients with recent experience (within 2 years) of the ELLIPTA DPI, any capsule inhaler, the DISKUS inhaler, or the Turbuhaler	159	The following devices were assessed vs. ELLIPTA DPI in 2 separate substudies: • DISKUS + HandiHalerTurbuhaler + HandiHaler	• Correct use, assessed by a trained HCP using a checklist of inhaler errors (based on the ELLIPTA DPI PIL) after patients had read the PIL (primary endpointc) [14] • AEs	For each device, patients were asked to read the PIL and were then observed by an HCP for correct use. If there were errors, the HCP provided instructions on correcting the observed errors and the patient attempted correct inhaler use again. The process could be repeated once more if the second attempt was unsuccessful, but no more than 3 attempts were permitted.	1
Kerwin et al. 2020 (206901/ NCT03227445) [18]	Randomized, cross-over, multicenter study at 17 centers in the US	Inclusion criteria Aged ≥40 years and current or former smokers with >10 pack-years of smoking history, with a documented COPD history, as defined by ATS/ERS criteria [21] Exclusion criteria Diagnosis of asthma >1 COPD exacerbation requiring hospitalization in the 12 months prior to randomization Use of the ELLIPTA DPI, DISKUS, or HandiHaler DPI in the 12 months prior to randomization Receipt of current COPD maintenance treatment by any of the study inhalers Receipt of a SABA only for COPD maintenance	239	ELLIPTA DPI DISKUS + HandiHaler	Correct use, defined as the patient making 0 errors from an error checklist based on the ELLIPTA DPI PIL (primary endpoint)14 AEs	At randomization, participants were asked to read the IFU section of the approved prescribing information, and then demonstrated use of the assigned inhaler(s) to an HCP. In the case of errors, the HCP gave the participant verbal advice to ensure correct technique was understood before leaving the clinic. The participant took the inhaler(s) and IFU home for 28 days (Period 1). On returning to the clinic at Visit 2 (Day 28), participants were assessed for correct use of their first assigned inhaler(s) and were then given their next assigned inhaler(s) and IFU for Period 2. Demonstration of correct inhaler use was again required, and HCPs advised if the participant made any errors before taking the appropriate inhaler(s) and IFU home. Participants returned on Visit 3 (Day 56) for correct use of the second assigned inhaler, assessed by the HCP.	56 (only 28 days on ELLIPTA DPI)

^aStudy designs did not meet the eligibility criteria (open-label/non-blinded study design).

^bCorrect use and ease-of-use data also analyzed by age subgroups (40–64, 65–74, and ≥75 years).

^cPrimary endpoint defined as proportion of patients making ≥1 critical inhaler error after reading the PIL.