Study design and patients
This was a prospective, open-label and industry-independent, cross-sectional study of asthma and COPD patients. Study participants were recruited from the Department of Pneumonology of the Otto-von-Guericke University Magdeburg, Germany. Both inpatients and ambulant patients were enrolled in the study. Inclusion criteria comprised: a diagnosed obstructive airway disease (asthma or COPD), an age between 18 and 100 years and a signed declaration of consent to participate in the study. The lack of written consent was an exclusion criterion.
All patients were interviewed. We examined the inhaler technique with the patient's own device and queried preferred inhaler attributes. Then we tested the following ten devices that did not contain any drug (placebo inhalers) in random order:
- Dry-powder inhalers (DPI): Breezhaler®, Diskus®, Elpenhaler®, Genuair®, Nexthaler®, Forspiro®, Spiromax®, Turbohaler®
- Pressurized metered-dose inhaler (pMDI): A customary pMDI (Flutiform® pMDI) was used
- Soft mist inhaler (SMI): Respimat®
All selected devices represented approved products in Germany with a high market share. After demonstrating the use of the placebo inhalers, patients were asked to evaluate the handling and other properties of each device and to name the device that they would most or least prefer for daily use.
The study was conducted in accordance with the Declaration of Helsinki [19], and the study protocol was approved by the local Ethics Committee of the Otto-von-Guericke University Magdeburg, Germany (reference number 51/16).
Data collection, check of inhaler technique and preferred inhaler attributes
Data was collected using a structured questionnaire that was filled in by the patient. The questionnaire was developed on the basis of the Patient Satisfaction and Preference Questionnaire (PASAPQ) and the Feeling of Satisfaction with Inhaler (FSI-10) questionnaire [20, 21]. The PASAPQ is a multi-item measure of respiratory inhalation device satisfaction, designed and validated in asthma and COPD patients [20]. The FSI-10 questionnaire evaluates the patient's opinion regarding the ease or difficulty of using an inhaler device [21]. In order to check the comprehensibility of the questions and the practicability of the structured questionnaire used in this study, it was tested in advance in 10 patients.
After study participants had been recruited, the patient's characteristics (gender, age, presence of asthma or COPD, disease duration) were collected first. Second, the patients specified which inhalers they currently or formerly used. Subsequently, all participants were asked to demonstrate the inhaler technique with their prescribed inhaler device for daily use. The investigator used checklists based on the recommendations of the German Respiratory League [22] to determine whether the patient's use of their own device was correct or incorrect.
The investigator next asked which of the following predefined inhaler properties were particularly relevant to the patient: easy handling, short inhalation time, low inhalation resistance, discreet handling, availability of a dose counter, design and colour. Ratings were reported on a Likert scale, ranging from 1 (very important) to 4 (unimportant).
Inhaler preferences and error rates for 10 placebo devices
The investigator then demonstrated how to use each of the ten placebo inhaler devices. The latter were presented in a random order based on a pre-generated randomization list. The handling for each individual device was demonstrated on the basis of standardized checklists for correct use, which had been developed by a panel of German expert pulmonologists [22]. For each device, they included three major steps of inhalation: 1) inhalation preparation, 2) inhalation routine, and 3) closure of inhalation. After each demonstration, patients were asked to repeat the procedure shown for the device. If the inhaler technique was not yet correct, the investigator repeated the instructions for the respective device until the patient demonstrated error-free handling of the placebo inhaler. The number of attempts the patient needed to use the inhaler correctly was noted. To assess the correct inhaler technique, standardized checklists were used again [22].
After completion of the device training, all patients were asked for each of the ten inhalers to assess device-related handling features using a 10-domain questionnaire. For this purpose, the following attributes were to be evaluated on a Likert scale, with response options from 5 ("applies perfectly") to 1 ("not correct at all"): "I like the design of the device", "It was easy to learn how to use the device", "It was easy to prepare the device", "The device is comfortable when held in my hand", "The mouthpiece was comfortable when using the device", "The device was easy to use", "I felt that I had used the device correctly", "I think it is possible to use the device easily and correctly in emergency situations", "I am satisfied with the device", and "Overall rating". Finally, the participants were asked to name the device they would most and least prefer for everyday use, given comparable efficacy.
All interviews, device demonstrations and assessments were conducted by the same person to avoid misjudgement by different investigators.
Statistical analysis
The program Microsoft Excel 2016 MSO (version 16.0.7329.1017), as well as WinSTAT for Microsoft Excel (version 2012.1.0.94) served as a basis for the statistical evaluation. Frequencies, mean values and standard deviations were calculated with WinSTAT.
The p-values for the number of attempts required until a device was used correctly were assessed using the Friedman test, followed by pairwise comparison using the Dunn test and a post-hoc correction according to Bonferroni. The same procedure was applied to evaluate device properties and handling features. Statistical significance was set at p < 0.05, whilst statistical trend was set at p < 0.10.
The significance of all other values was calculated using the Chi-square test and subsequent post-hoc Bonferroni correction. If the requirements for this test were not met, an exact Fisher's test was performed.