Evaluating Satisfaction and Quality of Life with Elpenhaler: A Pooled Analysis of Real-Life Studies

Abstract

Purpose

Patient adherence and satisfaction are critical for effective asthma and chronic obstructive pulmonary disease (COPD) management, as adherence strongly influences symptoms and prognosis. Patient-reported outcomes, including treatment satisfaction and quality of life, provide valuable insights into the effectiveness of therapeutic interventions from the perspective of individual patients. This study evaluated patient satisfaction with the Elpenhaler device and its impact on quality of life in routine clinical practice.

Patients and methods

Individual participant data were pooled from seven real-life studies of fixed-dose combinations delivered via the Elpenhaler device (asthma, budesonide/formoterol: BOREAS, SKIRON, COMPLETE, NOTOS; COPD, salmeterol/fluticasone: AEOLOS, AHEAD, RETRIEVE). Patient satisfaction was assessed with the Feeling of Satisfaction with Inhaler (FSI-10) questionnaire in the four studies in which it was administered, and quality of life in asthma with the mini Asthma Quality of Life Questionnaire (mAQLQ) at baseline, 3 and 6 months. The minimum clinically important difference (MCID) was defined as a change of at least 0.5 units in the mAQLQ total score.

Results

In 4333 patients with evaluable FSI-10 score, overall satisfaction was high (mean total score 44.1, standard deviation 5.4), with 94.5% of patients expressing confidence in correct usage. Multivariable analysis showed that age ≥65 years, male sex and lower education level were independently associated with lower satisfaction, whereas never smoking predicted higher satisfaction. In 4212 patients with asthma, mAQLQ total score improved from 4.52 at baseline to 6.06 at 3 months and 6.27 at 6 months; the MCID was achieved by 83.4% and 87.8% of evaluable patients, respectively. Never smoking and lower education level predicted MCID achievement.

Conclusion

Treatment delivered via the Elpenhaler is associated with high patient satisfaction and with significant and sustained improvements in quality of life, supporting its effectiveness in real-world respiratory care. The identification of factors associated with satisfaction and quality of life improvement may help in personalising treatment and optimising care.

Introduction

Asthma and chronic obstructive pulmonary disease (COPD) are chronic inflammatory airway diseases that remain major public health challenges, accounting for substantial morbidity and economic burden worldwide. Asthma affects more than 300 million people globally,¹ while COPD is currently the 3rd leading cause of death, accounting for more than 3 million deaths annually, and the 7th leading cause of disability worldwide.² Although effective treatments are widely available for both diseases, real-world data continue to demonstrate high morbidity and increased healthcare utilisation. Effective disease management depends not only on pharmacological efficacy but also on correct drug administration and on patient adherence.^3,4 Adherence influences nearly every aspect of the disease including symptoms, disease control, exacerbations and prognosis, and is strongly affected by inhaler-related characteristics such as device design and usability, as well as by patient satisfaction.^5,6 The device, particularly, plays a key role in the appropriate inhalation of the drug and the overall adherence to treatment.³

In real-world settings, patient-reported outcomes (PROs), including treatment satisfaction and quality of life (QoL), provide valuable insights on the effectiveness of therapeutic interventions from the perspective of individual patients.⁷ Validated PRO measures such as the Feeling of Satisfaction with the Inhaler (FSI-10) questionnaire and the mini Asthma Quality of Life Questionnaire (mAQLQ) ensure consistency by implementation of standardised questions, response options, and assessment procedures across all participants.

The Elpenhaler is a dry powder inhaler designed to deliver medication efficiently and easily to patients with respiratory diseases. Previous studies have demonstrated the effectiveness of medications delivered via the Elpenhaler device,^8–10 while comparisons with other devices have showed equivalent lung deposition of the delivered drugs.¹¹ Elpenhaler is user-friendly, with lower rates of critical handling errors versus other common inhalers,^12,13 attributes that may lead to increased compliance and adherence.

Despite this evidence, important knowledge gaps remain. First, previous evaluations of satisfaction with the Elpenhaler have been based on individual studies that were designed and powered to describe average satisfaction rather than to identify the patient characteristics that determine it. Second, satisfaction and QoL have generally been reported separately, in different populations, so that the relationship between a well-accepted device and clinically meaningful QoL gains has not been examined within one coherent dataset. Third, and most importantly, no study has systematically explored which demographic and clinical characteristics predict either low satisfaction or failure to achieve a clinically meaningful QoL improvement, information that is required if inhaler choice and inhaler training are to be individualised. The present study addresses these gaps by evaluating patient satisfaction with the Elpenhaler in patients with asthma and COPD, and its impact on asthma quality of life, by integrating individual participant data from multiple real-life studies and exploring clinical characteristics associated with satisfaction and with clinically meaningful improvements in QoL.

Methods

Study Design

This was a pooled analysis of individual participant data from real-life studies evaluating fixed-dose combinations delivered via the Elpenhaler device in patients with asthma and COPD, focusing on two pre-specified domains: patient satisfaction and quality of life. The analysis was performed according to a statistical analysis plan that was finalised before the pooled dataset was assembled and before any analysis was performed. The individual studies were conducted in accordance with the Declaration of Helsinki, were approved by the institutional review boards of the participating institutions and were prospectively registered on ClinicalTrials.gov where applicable; all participants provided written informed consent (Table S1).

Study Selection Criteria

Studies were eligible for the pooled dataset if they met all of the following pre-specified criteria: (i) a real-life (non-interventional, observational) design conducted in Greece; (ii) enrolment of adults (≥18 years) with a physician diagnosis of asthma or COPD; (iii) treatment with a fixed-dose combination delivered via the Elpenhaler device, i.e budesonide/formoterol (Pulmoton Elpenhaler) in asthma or salmeterol/fluticasone propionate (Rolenium Elpenhaler) in COPD; (iv) completion of the study and availability of the individual participant dataset to the sponsor at the time the analysis plan was finalised. Seven studies met these criteria and all seven were included: four asthma studies (BOREAS,¹⁴ SKIRON,⁸ COMPLETE,¹⁵ NOTOS⁹) and three COPD studies (AEOLOS,¹⁰ AHEAD,¹⁶ RETRIEVE¹⁷). No study meeting the selection criteria was excluded, and no additional Elpenhaler study, published or unpublished, was available to the sponsor at the time of the analysis. The design, setting, eligibility criteria, assessment schedule and PRO instruments of each contributing study are summarised in Table S1.

The individual studies were conducted in non-overlapping calendar periods and in different networks of investigational sites, participants were identified by unique study-specific identifiers, and enrolment in a concurrent interventional or non-interventional study was an exclusion criterion in each protocol; accordingly, no patient contributed data to more than one of the pooled studies.

Exposure and Analysis Groups

This was a single-arm analysis. The exposure of interest was treatment with an inhaled fixed-dose combination delivered via the Elpenhaler device; there was no comparator device or control group in any of the contributing studies. No exposure groups were therefore defined. All comparisons presented in this manuscript are between pre-specified subgroups of the same exposed population, defined by contributing study, disease (asthma versus COPD), age group (<65 versus ≥65 years), sex, educational level, and smoking status; these are descriptive and analytical subgroups, not exposure groups.

Assessment Schedule and Contribution of Each Study to the Analyses

Contributing studies scheduled a baseline visit followed by protocol-defined follow-up visits, most commonly at 3 and 6 months, with the exception of AEOLOS, which followed patients for 12 months with visits at 6 and 12 months, and AHEAD, which followed patients for 3 months (Table S1). The FSI-10 was administered in four of the seven studies (AEOLOS, BOREAS, NOTOS and SKIRON); it was not part of the assessment schedule of COMPLETE, AHEAD or RETRIEVE. Because the timing of the FSI-10 assessment differed between protocols, the FSI-10 total score was analysed at the final available post-baseline assessment for each participant, and the contributing study was retained as a fixed effect in all models. The mAQLQ was administered at baseline and at each follow-up visit in the four asthma studies; 6-month mAQLQ data were available from BOREAS, COMPLETE and NOTOS, since SKIRON was a 3-month study. The mAQLQ was not applicable to, and was not administered in, the COPD studies.

AHEAD and RETRIEVE administered neither the FSI-10 nor the mAQLQ and therefore contributed no patient-level satisfaction or QoL data. They were retained in the description of the source population (Table S1) so that the full evidence base from which the analysis populations were drawn is transparent and so that the reader can verify that no eligible study was selectively omitted, but they contributed to none of the analyses reported here. The derivation of the analysis populations is shown in Figure 1.

Figure 1.

Flow of participants through the pooled analysis. Seven real-life studies of fixed-dose combinations delivered via the Elpenhaler device contributed to the source population (N=6,424). The FSI-10 was administered in four studies (AEOLOS, BOREAS, NOTOS, SKIRON) and 4,333 participants had an evaluable FSI-10 total score. The mAQLQ was administered in the four asthma studies and 4,212 participants had an evaluable baseline score. COMPLETE, AHEAD and RETRIEVE did not administer the FSI-10, and the three COPD studies did not administer the mAQLQ; AHEAD and RETRIEVE therefore contributed to the description of the source population only and to none of the analyses.

Abbreviations: FSI-10, Feeling of Satisfaction with Inhaler questionnaire; mAQLQ, mini Asthma Quality of Life Questionnaire; MCID, minimum clinically important difference.

Study Assessments

Patient satisfaction was assessed using the FSI-10 questionnaire, adapted in the Greek language, across all applicable studies.¹⁸ The FSI-10 is a questionnaire that assesses the patients’ satisfaction from an inhaler and the usability of the device.¹⁹ It consists of ten questions, with five possible responses (very, fairly, somewhat, not very, and hardly at all) scored from 5 to 1. The FSI-10 total score ranges from 10 to 50, and higher scores correspond to higher patient satisfaction.

Quality of life was evaluated in patients with asthma using the Greek version of the mAQLQ.²⁰ This 15-item questionnaire has 4 domains (symptoms, environment, emotions, and activities), providing a total score and four subscores from 0 to 7; this is a shorter version of the initial 32-item AQLQ.²¹ The minimal clinically important difference (MCID) is a change of ≥0.5 units.

Study Objectives and Endpoints

1. Patient satisfaction with the Elpenhaler, as assessed with the FSI-10 total score in the overall population of studies in which the instrument was administered. Factors contributing to the FSI-10 score, such as disease, age, and sex, were also investigated.

2. Quality of life improvement achieved by treatment with the Elpenhaler as assessed via the mAQLQ in the asthma population. Specifically, change in mAQLQ total score at 3 months across all asthma studies, and change in mAQLQ total score at 6 months in the BOREAS, COMPLETE, and NOTOS studies. Predictors for achieving the MCID of ≥0.5 units in mAQLQ were also explored.

Statistical Analysis

Descriptive statistics included patient numbers (N), mean, median, minimum and maximum values and standard deviation (SD) for continuous variables, and absolute and relative frequencies for categorical variables. Baseline characteristics were compared across contributing studies using the chi-squared test for categorical variables and analysis of variance (ANOVA) for continuous variables.

The FSI-10 total score was analysed as a continuous variable at the final available assessment time point. Associations between the FSI-10 total score and pre-defined factors were evaluated using linear models and/or ANOVA, with disease type, age group, sex and exacerbation status included as covariates. Overall statistical significance was assessed using F-tests, and for factors with significant effects pairwise comparisons were conducted using Tukey’s test. Individual FSI-10 item responses were compared between subgroups using the chi-squared test. In a secondary analysis, patients were dichotomised at the sample median FSI-10 total score (45 points) and the independent predictors of high satisfaction were explored using logistic regression, so that the estimates would be directly comparable with those obtained for MCID achievement; this cut-off is data-derived and the corresponding results are therefore supportive rather than primary.

Changes in mAQLQ total score over time were analysed using a mixed model for repeated measures (MMRM), with fixed effects for time, study, and time-by-study interaction, and with baseline mAQLQ included as a covariate. The primary contrasts were the estimated mean changes from baseline to 3 and 6 months, with corresponding 95% confidence intervals (CIs). Logistic regression models were used to estimate the probability of achieving the MCID at 3 and 6 months, and results were reported as odds ratios (ORs) with 95% CIs. The candidate predictors specified in the analysis plan were the baseline mAQLQ total score, age group, sex, body mass index (BMI), smoking status, educational level, disease severity, inclusion period and exacerbation status; the variables presented in the Results are those retained in the final models.

Smoking status was modelled as a three-level categorical variable (never, former, current smoker) with current smokers as the reference category. Pack-years were recorded only in current and former smokers, and were missing for a substantial proportion even of these patients, so smoking exposure could be modelled only as a categorical variable. Educational level was modelled as a three-level categorical variable (elementary, high school, bachelor’s degree or higher) with bachelor’s degree or higher as the reference category. To account for potential heterogeneity, disease type, study and inclusion period (seasonality) were included as fixed effects in all primary models (ANOVA, MMRM and logistic regression).

Missing data were below the pre-specified threshold of 5% for all analysed variables; analyses were therefore based on observed data and no imputation was applied. All statistical tests were two-sided with a significance level of 0.05, and the analyses of subgroups and predictors were exploratory. The analysis was done using the language R version 4.5.2.

Results

Study Population

The seven contributing studies enrolled a total of 6424 participants (asthma 4226; COPD 2198), of whom 3962 (61.7%) completed both protocol-defined follow-up visits. Statistically significant differences were observed across all baseline characteristics among the seven study cohorts. Age distribution varied substantially, with asthma studies predominantly including participants younger than 65 years, whereas male sex was in the majority in all COPD cohorts. Detailed baseline characteristics by study are presented in Table S2. The derivation of the two analysis populations from the source population is shown in Figure 1.

FSI-10 Assessment

The FSI-10 questionnaire was administered in the AEOLOS study (COPD) and in the asthma studies BOREAS, SKIRON and NOTOS. Among the 4526 participants enrolled in these four studies, 4333 (95.7%) had an evaluable FSI-10 total score and constituted the satisfaction analysis population; an FSI-10 assessment was recorded at the first post-baseline visit for 3326 participants (76.8%) and at the second post-baseline visit for 3277 participants (75.6%). Table 1 outlines the baseline characteristics of the 4333 patients who provided FSI-10 scores, and Table S2 presents these characteristics by study. The cohort was predominantly younger than 65 years (65.2%), largely driven by patients with asthma, and showed a near-equal distribution between women (51.9%) and men (48.1%).

Table 1.

Baseline Characteristics of Patients Included in the FSI-10 Analysis

Characteristic	N = 4333
Age (years)
<65	2827 (65.2)
≥65	1506 (34.8)
Sex
Female	2250 (51.9)
Male	2083 (48.1)
Race
Asian	5 (0.1)
Black or African American	5 (0.1)
Caucasian	4319 (99.7)
Other	4 (0.1)
Educational level
Elementary	1115 (25.7)
High school	2030 (46.9)
Bachelor’s degree or higher	1188 (27.4)
BMI (kg/m2), mean (SD)	27.9 (5.4)
Smoking status
Current smoker	1092 (25.2)
Former smoker	1101 (25.4)
Never smoked	2140 (49.4)

Notes: Data are n (%) unless otherwise stated. The population comprises all participants of the AEOLOS, BOREAS, NOTOS and SKIRON studies with an evaluable FSI-10 total score. Percentages are calculated on the total of 4333 patients and may not sum to 100 because of rounding.

Abbreviations: BMI, body mass index; FSI-10, Feeling of Satisfaction with Inhaler questionnaire; SD, standard deviation.

Satisfaction in the Overall Population

The mean total FSI-10 score was 44.1 (SD 5.4). Overall, patient satisfaction was consistently high across all ten domains of the FSI-10 questionnaire (Table 2). Most patients reported ease of use, with 88.4% rating it “Very” or “Fairly” easy to learn how to use the device, and 94.5% expressing confidence (“Very” or “Fairly”) in correct usage. High satisfaction was also evident in device handling; 66.2% reported the inhaler’s fit with the lips “Very” comfortable, and 57.5% found it “Very” easy to continue normal activities while using it. Overall satisfaction was high, with 94.3% of patients being either “Very” satisfied (54.5%) or “Fairly” satisfied (39.8%).

Table 2.

Satisfaction with the Inhaler: Results for the FSI-10 in the Overall Population

FSI-10 Item	N = 4,333
Has it been easy to learn how to use the inhaler?
Very	2084 (48.1)
Fairly	1744 (40.3)
Somewhat	408 (9.4)
Not very	71 (1.6)
Hardly at all	26 (0.6)
Was it easy to prepare the inhaler for use?
Very	1786 (41.2)
Fairly	1895 (43.7)
Somewhat	527 (12.2)
Not very	103 (2.4)
Hardly at all	22 (0.5)
Was it easy to use the inhaler?
Very	2214 (51.1)
Fairly	1681 (38.8)
Somewhat	371 (8.6)
Not very	46 (1.1)
Hardly at all	21 (0.5)
Was it easy to keep the inhaler clean and in good working condition?
Very	2502 (57.7)
Fairly	1525 (35.2)
Somewhat	259 (6.0)
Not very	31 (0.7)
Hardly at all	16 (0.4)
Was it easy to continue normal activities with the use of the inhaler?
Very	2490 (57.5)
Fairly	1510 (34.9)
Somewhat	287 (6.6)
Not very	31 (0.7)
Hardly at all	15 (0.4)
Did the inhaler fit your lips comfortably?
Very	2867 (66.2)
Fairly	1255 (29.0)
Somewhat	176 (4.1)
Not very	18 (0.4)
Hardly at all	17 (0.4)
Was using the inhaler easy in terms of size and weight?
Very	2,163 (49.9)
Fairly	1,707 (39.4)
Somewhat	395 (9.1)
Not very	47 (1.1)
Hardly at all	21 (0.5)
Was it easy to carry the inhaler with you?
Very	1,845 (42.6)
Fairly	1,644 (37.9)
Somewhat	640 (14.8)
Not very	164 (3.8)
Hardly at all	40 (0.9)
After you have used the inhaler do you have the feeling that you used it correctly?
Very	2,648 (61.1)
Fairly	1,447 (33.4)
Somewhat	197 (4.6)
Not very	21 (0.5)
Hardly at all	20 (0.5)
Overall, considering your responses to the previous questions, were you satisfied with the inhaler?
Very	2,361 (54.5)
Fairly	1,724 (39.8)
Somewhat	205 (4.7)
Not very	19 (0.4)
Hardly at all	21 (0.5)
FSI-10 total score, mean (SD)	44.1 (5.4)
Median (Q1, Q3)	45.0 (41.0, 49.0)
Observed range	10.0–50.0

Notes: Data are n (%) unless otherwise stated. Response options are presented in descending order of satisfaction and are scored from 5 (very) to 1 (hardly at all); the total score ranges from 10 to 50, with higher scores indicating greater satisfaction. Percentages are calculated on the total of 4333 patients and may not sum to 100 because of rounding. Three patients had no response to the final item and no total score; percentages for that item and the summary statistics for the total score are therefore based on 4330 patients. All observed total scores lay within the instrument range of 10 to 50.

Abbreviations: FSI-10, Feeling of Satisfaction with Inhaler questionnaire; SD, standard deviation. Q1, first quartile; Q3, third quartile.

Satisfaction per Study

Differences in satisfaction were observed across contributing studies for all FSI-10 domains (p<0.001 for each item; Table S3). While overall satisfaction was high across all groups, the BOREAS and NOTOS studies consistently reported the highest proportions of patients responding “Very” satisfied for multiple attributes, including ease of use, lip fit, and device portability. Specifically, NOTOS (62.0%) and BOREAS (57.3%) had the highest percentage of patients reporting they were “Very” satisfied, compared to SKIRON (55.8%) and AEOLOS (42.0%). This pattern was reflected in mean total FSI-10 scores, which were higher for NOTOS [44.8 (5.7)] and BOREAS [44.5 (5.0)], followed by SKIRON [44.2 (5.4)] and AEOLOS [42.8 (5.5)].

Satisfaction by Disease Group

Comparative analysis by disease group demonstrated significantly higher satisfaction among patients with asthma (N=3340) compared to those with COPD (N=993). Patients with asthma consistently reported a higher level of “Very” satisfaction across all domains compared to patients with COPD, who more frequently selected “Fairly” satisfied. For example, a substantially higher proportion of patients with asthma reported being “Very” confident in correct inhaler use (64.3% vs 50.3%) and finding its fit with the lips “Very” comfortable (70.1% vs 53.1%). In the overall satisfaction rating, 58.3% of patients with asthma were “Very” satisfied, compared with 42.0% of patients with COPD (Figure 2 and Table S4). The FSI-10 was administered in only one of the three COPD studies (AEOLOS), so the comparison between disease groups rests on a single COPD cohort and disease and study cannot be separated; Figure 2 should therefore be read as a descriptive summary of the two groups rather than as evidence of a disease-specific effect on satisfaction.

Figure 2.

FSI-10 Total Score by Disease Group. Bars show mean FSI-10 total score; error bars show ±1 SD. Asthma 44.5 (5.3), n=3,340; COPD 42.8 (5.5), n=993; p<0.001 (Supplement Table S4). The FSI-10 was administered in only one of the three COPD studies (AEOLOS), so the COPD estimate derives from a single cohort and the comparison is confounded by study.

Abbreviations: COPD, chronic obstructive pulmonary disease; FSI-10, Feeling of Satisfaction with Inhaler questionnaire; SD, standard deviation.

Satisfaction by Age Group

We observed differences in satisfaction between age groups (p<0.0001 for all measures). Patients younger than 65 years (N=2827) consistently reported higher satisfaction across every domain than those aged ≥65 years (N=1506). A greater proportion of younger patients found the device “Very” easy to learn to use (52.7% vs 39.4%) and reported being “Very” confident they used it correctly (65.6% vs 52.7%). Overall, 58.6% of the group younger than 65 years were “Very” satisfied compared to 46.8% of the group aged ≥65 years, who more frequently selected “Fairly” satisfied. The difference in experience is quantitatively summarised by the higher mean total FSI-10 score in the younger cohort (44.6, SD 5.2 vs 43.1, SD 5.7) (Table S5).

Satisfaction by Sex

Overall satisfaction was high in both sexes, although statistically significant differences were observed in selected domains and in overall satisfaction. Most individual questions did not differ (p>0.05 for 7 of 10 items). Significant differences were noted for ease of keeping the inhaler clean (p=0.004), size and weight (p=0.008), portability (p<0.001), and overall satisfaction (p<0.001).

Female patients (N=2250) reported higher overall satisfaction, with 57.7% being “Very” satisfied compared with 51.1% of male patients (N=2083). Male patients were more likely to report being “Fairly” satisfied (43.5% vs 36.4%) and experienced more difficulty carrying the device (a higher proportion found it “Not Very” or “Somewhat” easy to carry the device). Mean FSI-10 scores were slightly higher among women (44.3, SD 5.4 vs 43.8, SD 5.5) (Figure 3 and Table S6).

Figure 3.

FSI-10 Total Score by Sex. Bars show mean FSI-10 total score; error bars show ±1 SD. Female 44.3 (5.4), n=2,250; Male 43.8 (5.5), n=2,083; p<0.001 (Supplement Table S6).

Determinants of Patient Satisfaction

Based on the sample median FSI-10 total score of 45, patients were categorised as having high satisfaction (≥45; N=2328, 53.8%) or low satisfaction (<45; N=2002, 46.2%); three patients could not be classified because of a missing total score. Univariate analysis identified several patient characteristics significantly associated with higher satisfaction (p<0.001), including age <65 years, female sex, asthma diagnosis, never smoking, and holding a bachelor’s degree. In multivariable analysis, age ≥65 years (OR 0.67; 95% CI 0.58–0.77), male sex (OR 0.80; 95% CI 0.71–0.91), and lower education (elementary: OR 0.68; high school: OR 0.86) were independently associated with lower satisfaction (all p<0.001). Conversely, a history of never smoking was a significant independent predictor of high satisfaction versus current smokers (OR 1.23; 95% CI 1.06–1.43; p=0.023). The analysis of the FSI-10 total score as a continuous variable was concordant for age, sex and education, with lower adjusted mean scores in patients aged ≥65 years (β −1.2; 95% CI −1.5 to −0.78), in men (β −0.35; 95% CI −0.68 to −0.01) and in patients with elementary education compared with a bachelor’s degree (β −0.94; 95% CI −1.4 to −0.46). Smoking status was not a significant predictor in the linear model (p=0.12), and the association of never smoking with satisfaction should therefore be regarded as weak.

Quality of Life (mAQLQ) in Asthma

Baseline Characteristics

Table 3 summarises the baseline characteristics of the 4212 patients from the BOREAS, SKIRON, COMPLETE, and NOTOS studies included in the asthma-specific quality of life analysis. Patients were predominantly Caucasian, younger than 65 years, women, with a mean body mass index within the overweight range.

Table 3.

Baseline Characteristics of Patients Included in the mAQLQ Analysis

Characteristic	N = 4212
Age (years)
<65	3203 (76.0)
≥65	1009 (24.0)
Sex
Female	2541 (60.3)
Male	1671 (39.7)
Race
Asian	6 (0.1)
Black or African American	7 (0.2)
Caucasian	4195 (99.6)
Other	4 (0.1)
Educational level
Elementary	833 (19.8)
High school	2036 (48.3)
Bachelor’s degree or higher	1343 (31.9)
BMI (kg/m2), mean (SD)	27.8 (5.5)
Smoking status
Current smoker	875 (20.8)
Former smoker	761 (18.1)
Never smoked	2576 (61.2)

Notes: Data are n (%) unless otherwise stated. The population comprises all participants of the BOREAS, COMPLETE, NOTOS and SKIRON studies with an evaluable baseline mAQLQ total score. Percentages are calculated on the total of 4212 patients and may not sum to 100 because of rounding.

Abbreviations: BMI, body mass index; mAQLQ, mini Asthma Quality of Life Questionnaire; SD, standard deviation.

Change in QoL at 3 Months

Baseline overall mAQLQ scores were moderate, with a cohort mean of 4.52 (SD 1.09). At three months, all contributing studies demonstrated clinically significant improvement, with the overall mean score increasing to 6.06 (mean change +1.55, SD 1.05). Among the contributing studies, SKIRON consistently reported the most favourable outcomes at 3 months, achieving the greatest improvement (mean change +1.80, SD 1.05; mean score 6.37, SD 0.64) and including the largest proportion of patients selecting the best possible score (“7”) across most individual questions. The NOTOS study demonstrated the smallest improvement (mean change +1.38, SD 1.04) and the lowest mean score at 3 months (5.89, SD 0.79). Table S7 presents the mAQLQ scores for the four contributing studies at baseline and at 3 months. Differences across studies were statistically significant.

Change in QoL at 6 Months

At 6 months (available data from BOREAS, COMPLETE and NOTOS), improvements were sustained and further enhanced across all three studies, with the overall mean mAQLQ score increasing to 6.27 (SD 0.69; mean change +1.77, SD 1.14; p<0.001).

Although all studies presented meaningful improvements in QoL, the COMPLETE study showed the greatest improvement and the highest total score (mean change +2.01, SD 1.14; mean score 6.34, SD 0.62), while BOREAS showed the smallest improvement and the lowest total score (mean change +1.65, SD 1.11; mean score 6.21, SD 0.74). At six months, the distribution of responses across the 15 individual questions was largely positive, with a high frequency of scores of 6 or 7. Table S8 presents the mAQLQ scores for the three contributing studies at baseline and at 6 months. Significant differences between studies (p<0.05) were observed in the responses for most questions, particularly those related to symptom frequency and activity limitation, indicating variability in the patient experience of improvement (Tables S7 and S8). Figure 4 shows the QoL trajectory over time, overall and per study.

Figure 4.

mAQLQ total score trajectory over time, overall and by contributing study. Data are mean values with standard deviation. Assessments were performed at baseline, 3 months and 6 months. SKIRON was a 3-month study and therefore contributes no 6-month data; the overall estimate at 6 months is derived from BOREAS, COMPLETE and NOTOS.

Abbreviations: mAQLQ, mini Asthma Quality of Life Questionnaire; SD, standard deviation.

Minimum Clinically Important Difference (MCID) Achievement

A high proportion of patients achieved clinically meaningful improvement (MCID ≥0.5). At 3 months, 83.4% of patients (3367 of 4036 with evaluable paired assessments) achieved the MCID, increasing to 87.8% (2,515 of 2,864) at 6 months.

Significant differences between the contributing studies (p<0.001) were observed in the proportions of patients achieving an MCID ≥0.5 in mAQLQ. At 3 months, SKIRON had the highest MCID achievement rate (89.2%) followed by COMPLETE (84.5%) and BOREAS (81.9%), while NOTOS had the lowest (78.4%) and contributed the largest proportion of patients (34.5%) who did not experience a clinically meaningful change. At 6 months, COMPLETE had the highest MCID achievement (92.2%), followed by NOTOS (86.8%) and BOREAS (86.3%), the latter contributing the largest proportion of patients (45.6%) who did not achieve a clinically meaningful change by 6 months.

Predictors of QoL Improvement

In the multivariable analysis, significant predictors of MCID achievement at three months were never smoking (never versus current smokers, OR 1.30; 95% CI 1.05–1.59; p=0.020) and lower education level (elementary versus bachelor’s degree, OR 1.48; 95% CI 1.15–1.92; p=0.007). At six months, education level remained significant (p=0.028), as patients with an elementary school education had 63% higher odds of achieving the MCID compared to those with a bachelor’s degree (OR 1.63; 95% CI 1.14–2.38). Notably, sex (p=0.091 at 3 months) and age (p=0.12 at 6 months) were not significant independent predictors.

Discussion

In this pooled analysis of multiple cohorts of patients receiving inhaled treatment with the Elpenhaler device, we comprehensively assessed patient satisfaction in asthma and COPD, as well as quality of life in patients with asthma, in real-world settings. Our findings demonstrated consistently high overall patient satisfaction, while age <65 years, female sex, asthma diagnosis, never smoking, and higher education were identified as predictors of higher satisfaction. Additionally, meaningful and sustained improvements in QoL among patients with asthma were observed, especially in those with lower education levels. This is the first and largest study thoroughly exploring demographic and clinical factors associated with satisfaction and clinically meaningful improvements in QoL, suggesting target groups for enhanced training and support. Overall, our data further support the usability and effectiveness of the Elpenhaler device in everyday practice.

Satisfaction with the inhaler in the overall population was consistently high across all ten domains of the FSI-10 questionnaire, with more than 90% of the patients being either “very” or “fairly” satisfied. Patients reported high usability of the device, particularly regarding ease of learning, confidence in correct use, and ergonomic design. Several types of inhalation devices are currently available and device-related characteristics including size, ease of use, portability, and confidence in correct use contribute to the patient’s perception of an ideal inhaler. Patient satisfaction has been strongly linked to patient adherence, better disease control and improved clinical outcomes.^22,23 In a prospective study on inhalers in 705 Greek patients with COPD, asthma and asthma–COPD overlap,¹³ the FSI-10 questionnaire results at baseline and at 60 days showed that Elpenhaler consistently achieved significantly higher satisfaction compared to Diskus and Turbuhaler at both time points. Similarly, in a multicentre, 4-week observational study evaluating patient satisfaction with three dry powder inhalers including more than 1000 patients with asthma or COPD, Elpenhaler showed superior patient satisfaction versus Diskus and Turbuhaler across most evaluation domains, although all devices were acceptable.²⁴ Interestingly, patients with more severe COPD reported greater satisfaction regardless of the device used.

Satisfaction varied across subgroups. Patients with asthma consistently reported higher satisfaction compared to those with COPD, in agreement with previous findings.²⁵ This result may be explained by variations in patient demographics and disease characteristics, as patients with COPD are older, have more comorbidities and worse overall health status. In the study of Plaza et al,²⁵ patients with asthma were overall more satisfied with their inhaler compared to those with COPD, irrespective of the adherence level. Other factors associated with inhaler satisfaction were younger age, better disease control, and previous training on the use of inhalers.

Older age was also independently associated with lower satisfaction, consistent with previous reports.^25,26 Age-related factors, including cognitive function, visual acuity and fine motor skills may interfere with the correct use of the device, leading to lower satisfaction. A systematic review by Usmani et al showed that age, education status, previous inhaler training, comorbidities and socioeconomic status were factors associated with more handling errors in patients with asthma and COPD.²⁶ These findings suggest that special attention in these populations and careful training may improve usability and related outcomes.

Modest differences between the sexes were observed, favouring women in perceived portability and ease in maintenance, possibly reflecting behavioural differences between men and women. The never-smoking status effect linked with higher satisfaction is not adequately explained and may reflect better lung function and behavioural variations. Finally, higher educational levels may facilitate better understanding of device use and reinforce the perceived importance of treatment adherence.

We observed clinically meaningful improvements in asthma quality of life, with over 80% of patients exceeding the MCID at 3 months and nearly 88% at 6 months. These underscore the effectiveness of treatment delivered via the Elpenhaler in clinical practice. Sustained and increasing improvements in mAQLQ scores over time suggest better adherence, potentially linked to high FSI-10 scores, along with durable treatment effectiveness.

Analysis of predictors for achieving the MCID showed that never smoking and lower educational level were associated with a greater likelihood of clinically meaningful benefit. The negative impact of smoking is consistent with existing literature,^27,28 whereas the finding regarding education deserves careful interpretation. At first sight it appears to contradict the satisfaction analysis, in which lower education predicted lower satisfaction. The two findings are, however, not necessarily inconsistent, because they describe different quantities: the FSI-10 measures the level of satisfaction attained, whereas MCID achievement measures the change in QoL from each patient’s own starting point. Achievement of a fixed absolute improvement of ≥0.5 units is structurally dependent on the baseline score, since patients who begin close to the ceiling of the instrument have limited headroom to demonstrate a change of that magnitude. Patients with lower educational levels are often characterised by lower health literacy and may face greater challenges in self-managing a chronic disease such as asthma, and are therefore expected to enter treatment with lower baseline mAQLQ scores; the greater available headroom contributes mechanically to their higher rate of MCID achievement. The models presented were not adjusted for the baseline mAQLQ total score, which was a pre-specified candidate predictor but was not retained in the final models. The extent to which this ceiling effect accounts for the association between educational level and MCID achievement therefore cannot be quantified in the present analysis, and the finding should be interpreted with corresponding caution. Beyond this statistical explanation, the disproportionately high rate of MCID achievement in this subgroup may also reflect a substantial, readily perceived and objectively measurable clinical benefit, driven by simplified medication use, improved adherence, better symptom control, or the cumulative effect of gaining access to an effective, easy-to-use therapy for the first time. Age and sex were not associated with QoL improvement, suggesting that the Elpenhaler may benefit a broad range of patients.

Our findings have important clinical implications, highlighting the central role of the inhalation device in the treatment of airway diseases in real-world settings. High patient satisfaction, such as that observed with the Elpenhaler, may lead to better adherence and overall effectiveness in everyday clinical practice compared with clinical trials. Additionally, the incorporation of PRO measures, such as the FSI-10 and mAQLQ, along with the identification of characteristics that predict better outcomes, paves the way for a more personalised approach and optimisation of treatment.

Major strengths of our study include the evaluation of heterogeneous real-world data from multiple cohorts and the large sample size, which enhance the generalisability of the results, the use of a pre-specified analysis plan with explicit study selection criteria, and the use of robust multivariable models that allowed the identification of independent predictors for both the FSI-10 score and MCID achievement.

Several limitations should be acknowledged. First, all contributing studies were single-arm and non-interventional, and none included a comparator device or a control group; the observed improvements in QoL therefore cannot be attributed to the device itself, and may reflect regression to the mean, the natural course of disease after a treatment change, seasonal variation, or the effect of participation in a study. Comparative statements about the Elpenhaler in this manuscript rest on the previously published head-to-head studies cited above and not on the present data. Second, the studies differed in design, setting, eligibility criteria, follow-up duration and assessment schedule (Table S1), which introduces heterogeneity; we addressed this by retaining the contributing study and the inclusion period as fixed effects in all models, and the direction of effect was consistent across cohorts, but residual heterogeneity cannot be excluded. Third, the FSI-10 was administered in four of the seven studies and only one of these (AEOLOS) enrolled patients with COPD; consequently, all conclusions regarding satisfaction in COPD rest on a single cohort, and the observed asthma–COPD difference is confounded by study, since disease and study cannot be fully separated in this dataset. Fourth, QoL data were available only for patients with asthma, and no conclusions regarding QoL in COPD can be drawn. Fifth, measures of disease severity and disease control at baseline were not collected uniformly across the contributing protocols; different instruments were used in different studies (ACQ-6, ACQ-7, mMRC, CCQ) and no common measure was available in all cohorts, so baseline severity and control could not be included as covariates. As both are plausibly associated with satisfaction and with QoL gains, residual confounding by disease severity cannot be excluded, and this should be regarded as an important limitation of the predictor analyses. Similarly, pack-years were available only for current and former smokers and were incomplete even among these patients, so smoking exposure could be modelled only categorically. Sixth, the logistic models for MCID achievement were not adjusted for the baseline mAQLQ total score. Because achieving a fixed absolute improvement of ≥0.5 units is structurally easier for patients who begin further from the ceiling of the instrument, the predictors of MCID achievement reported here may partly reflect differences in baseline quality of life between subgroups rather than differences in the benefit derived from treatment. Seventh, the analyses of subgroups and of predictors were exploratory and were not adjusted for multiplicity. Finally, the studies were conducted exclusively in Greece, which may limit generalisability to other healthcare settings.

In conclusion, this large-scale real-world pooled analysis shows that inhaled treatment delivered with the Elpenhaler is related to high patient satisfaction in both asthma and COPD populations. Furthermore, significant and sustained improvements in quality of life were observed in patients with asthma. Patient characteristics that may predict satisfaction or MCID achievement were also identified. These findings further support that a user-friendly device, carefully selected to incorporate patients’ needs and preferences, may enhance a more personalised and effective approach in respiratory medicine.

Funding Statement

The seven contributing studies and the present pooled analysis were sponsored and funded by ELPEN Pharmaceutical Co. Inc., Pikermi, Attica, Greece, the manufacturer of the Elpenhaler device. ELPEN Pharmaceutical Co. Inc. funded the statistical analysis for this manuscript. The sponsor participated in the design of the analysis, the interpretation of the data and the preparation of the manuscript through the co-authors employed by the company (AG, MB, PP). The decision to submit the manuscript for publication was taken by all authors jointly.

Abbreviations

ANOVA, analysis of variance; CI, confidence interval; COPD, chronic obstructive pulmonary disease; FSI-10, Feeling of Satisfaction with Inhaler questionnaire; ICS, inhaled corticosteroid; LABA, long-acting beta-2 agonist; mAQLQ, mini Asthma Quality of Life Questionnaire; MCID, minimum clinically important difference; MMRM, mixed model for repeated measures; OR, odds ratio; PRO, patient-reported outcome; QoL, quality of life; SD, standard deviation.

Disclosure

Alexandros Ginis, Maria Bertoli and Polyanthi Papanastasiou are full-time employees of ELPEN Pharmaceutical Co. Inc., the sponsor of the contributing studies and the manufacturer of the Elpenhaler device. Athena Gogali reports honoraria for presentations and/or consultancy fees from: AstraZeneca, Boehringer Ingelheim, Chiesi, ELPEN, Guidotti, Menarini, Pfizer, and Vicore Pharma. Konstantinos Kostikas reports honoraria for presentations and/or consultancy fees from AstraZeneca, Chiesi, ELPEN, GSK, Guidotti, Menarini, MSD, Pfizer, Sanofi, and Specialty Therapeutics; his department has received funding and/or grants from AstraZeneca, Chiesi, ELPEN, GSK, Menarini, Pfizer. Konstantinos Kostikas reports Grants or contracts from Menarini, Pfizer, GSK, ELPEN, Chiesi, AstraZeneca; Consulting fees from Specialty Therapeutics, Sanofi, MSD, Pfizer, Guidotti, Menarini, GSK, ELPEN, Chiesi, AstraZeneca; Honoraria from Sanofi, Pfizer, Menarini, Guidotti, GSK, ELPEN, Chiesi, AstraZeneca, outside the submitted work. The authors report no other conflicts of interest in this work.