Study of Inhaler Device Attributes Investigating Critical and Overall Errors, Ease of Use, and Preference Between a Number of Inhaler Devices: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Inhaler Device Attributes Investigating Critical and Overall Errors, Ease of Use, and Preference Between a Number of Inhaler Devices (NCT02184624), a randomized phase 4 study. The primary outcome is Percentage of Participants Who Made at Least One Critical Error After Reading the Patient Information Leaflet.

2. Clinical Question

The study asks whether adults with chronic obstructive pulmonary disease make critical errors when using ELLIPTA and other inhaler devices after reading the patient information leaflet, and how they rate ease of use and preference. The devices compared with ELLIPTA were DISKUS/ACCUHALER, MDI, TURBUHALER, HANDIHALER, and BREEZEHALER. Each comparison was conducted within a separate sub-study, with the order of device use randomized.

The crossover arrangement means participants in a sub-study demonstrated technique with both devices, with the order randomized. The primary endpoint was reported for each inhaler regardless of sequence. The registry record describes the question and endpoint definition, but the primary outcome results are not reported in the registry record. Consequently, the record does not establish the size or precision of any difference between devices.

3. Trial Design

This was a randomized, open-label, placebo, crossover, multicentre study, described as taking place at a single visit. It was registered as a Phase 4 study and is marked completed. The registry record gives enrollment as 569. The brief summary describes a plan to screen at least 600 people and randomize 570 to five sub-studies; those figures are planned numbers, not the enrollment figure recorded for the completed study.

The five sub-studies each compared ELLIPTA with a different device: DISKUS/ACCUHALER, MDI, TURBUHALER, HANDIHALER, or BREEZEHALER. Within each sub-study, participants were randomized to the order in which they used the two devices. Thus the comparisons were organized by device pair, rather than as a single comparison in which every participant used every listed inhaler. The registry record says that eligibility for a sub-study depended on experience with the other inhaler and that participants had to be naïve to ELLIPTA and the relevant comparator. It adds a specific device-experience requirement for participants using BREEZEHALER or HANDIHALER.

Open-label means participants and study personnel were not masked to the device being used. The crossover design allows each participant’s handling of the paired devices to be assessed within the same study visit, but it does not remove possible effects of order, learning, or the instructions received. Randomization of sequence helps distribute order effects; details of how sequence allocation was generated or concealed are not reported in the registry record.

The study assessed errors after reading the patient information leaflet and also listed outcomes after instruction from a health care professional. It included measures of preference and ease of use. The registry record describes the study as using inactive treatment through the devices, so its central focus was device use and reported experience, not comparative drug efficacy. Further operational details, including how device demonstrations were standardized, are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: A total of 575 participants were enrolled in the study out of them 8 failed in screening and remaining 567 were randomized and included in intent to treat population. Participants were allocated to a particular sub-study depending on their experience of using the other inhaler (i.e., depending on which other inhaler they were naïve to using).

Overall Study

Analysis population for the primary outcome, as defined in the registry: Modified Intent-to-Treat Population comprised of all participants in the Intent-to-Treat Population who completed demonstration of using both study inhaler devices after reading patient information leaflet.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Who Made at Least One Critical Error After Reading the Patient Information Leaflet, measured over Day 1.

The primary endpoint is the percentage of participants who made at least one critical error after reading the patient information leaflet, assessed on Day 1. It is a binary participant-level outcome: a participant either met the definition of at least one critical error or did not. The endpoint is reported as a percentage for each inhaler, regardless of the order in which the devices were used.

The registry defines a critical error as an error most likely to result in no or only minimal medication being inhaled. Listed examples include failing to open the cover, not pushing the lever back, shaking the device after dose preparation, exhaling directly into the mouthpiece, or failing to seal the lips around the mouthpiece during inhalation. This definition makes the outcome clinically relevant to the mechanics of device use: it focuses on errors that could compromise delivery rather than treating every handling mistake as equally consequential.

As a general statistical matter, a binary endpoint is straightforward to describe as a proportion, but the estimate depends on the participants included in the analysis and on consistent assessment of the event. The endpoint is not a direct measure of medication exposure or a clinical health outcome. The registry record identifies the analysis population as participants in the Intent-to-Treat Population who completed a demonstration with both devices after reading the leaflet. It does not report the primary outcome percentages, their uncertainty intervals, or a numerical comparison between devices.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies the primary outcome as a percentage of participants with at least one critical error. This is a binary outcome at the participant level, even though the underlying assessment involves observing device-handling steps. The event definition groups participants according to whether any listed critical error occurred. A percentage communicates the frequency of that event in the analyzed population; it does not describe the number or severity of errors among participants who had an event.

What does a percentage estimate?

In general, a sample percentage estimates the proportion of a broader population that would meet the same event definition under comparable conditions. The estimate is subject to sampling variation: another group of eligible participants could yield a different percentage even if the study procedures were unchanged. A confidence interval is commonly used to convey this uncertainty. It gives a range of values compatible with the data and statistical assumptions, rather than guaranteeing that the underlying population value lies within that range. The primary percentages and confidence intervals are not reported in the registry record.

The analysis population matters for interpreting a percentage. The registry specifies a modified Intent-to-Treat Population consisting of participants in the Intent-to-Treat Population who completed demonstrations using both study devices after reading the leaflet. Requiring completion of both demonstrations defines who contributes to the primary analysis. If completion is related to ability to use a device or to the likelihood of making an error, estimates among completers could differ from estimates among everyone randomized. The registry record does not report how many participants were excluded from this population, why any demonstrations were incomplete, or how missing endpoint data were handled.

How does the crossover structure matter?

In a crossover study, participants experience the paired conditions, and the order is randomized. This creates a within-participant comparison: a participant’s performance with one device can be considered alongside that participant’s performance with the other device. Such comparisons can account for stable differences between people, such as prior dexterity or familiarity, when the statistical analysis uses the pairing appropriately. The registry record says that the primary percentage was reported for each inhaler regardless of sequence. It does not describe the model or test used to compare paired outcomes, or whether sequence was included in an analysis.

Order can matter even when sequence is randomized. Practice with the first device, learning from the first demonstration, fatigue, or carryover of instructions could influence performance with the next device. Randomizing order helps prevent systematic assignment of one device to be used first, but does not guarantee that order effects are absent. Because the study was open-label, participants and personnel knew which device was being used. The registry record does not report masking, and it gives no numerical assessment of order effects.

Why do the definitions and instructions matter?

A binary endpoint depends on how an event is defined and assessed. Here, the critical-error definition focuses on errors considered most likely to result in no or only minimal medication being inhaled. The listed examples concern specific handling and inhalation steps. Consistent training of assessors and standardized demonstrations would support comparability, but details of assessor training, adjudication, and measurement reliability are not reported in the registry record.

The primary endpoint follows reading the patient information leaflet. Other outcomes address errors after the first instruction from a health care professional and the instructions needed to demonstrate adequate inhalation technique. Those outcomes reflect performance after different levels or forms of support. In general, outcomes measured under different instruction conditions should be interpreted as distinct questions, rather than pooled or treated as repeated readings of the same endpoint without an explicit analysis plan.

What is known about planning and multiplicity?

The registry record lists seven outcome measures and 35 analyses. It does not report the statistical plan for controlling the chance of false-positive findings across multiple outcomes or comparisons. When many analyses are conducted, some apparently notable results can arise by chance; a prespecified hierarchy or multiplicity adjustment can help address that issue. Whether such procedures were used is not reported in the registry record.

A sample-size illustration based on an observed effect is a planning exercise, not observed power. The stated assumptions for that illustration are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect specified as a fraction of the observed effect. The illustration does not establish that the original trial was designed using those assumptions. The trial’s own sample-size justification and statistical plan are not reported in the registry record. No sample-size illustration values are discussed in the registry record details described for this analysis, so the illustration cannot be interpreted as a trial result.

More generally, the number of participants alone does not determine precision. Precision also depends on the event frequency, the analyzed population, the pairing, and the statistical method. Without outcome estimates and analysis details, the precision of the device comparisons cannot be evaluated from the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists seven outcome measures and 35 analyses. That means the study record encompasses multiple opportunities to examine device-use outcomes, instruction conditions, and questionnaires. When many outcomes or comparisons are tested, the chance of at least one apparently unusual result can increase. Statistical approaches such as a prespecified testing hierarchy or adjustment for multiple comparisons can address this concern, but the multiplicity strategy is not reported in the registry record. It is also not reported which analyses were prespecified as confirmatory and which, if any, were exploratory.

The record does not report the primary outcome percentages, comparison estimates, confidence intervals, or p-values. As a result, statistical significance and precision cannot be evaluated. It is also not possible to determine whether reported analyses, if available elsewhere, were adjusted for the crossover structure or for multiple comparisons.

The primary endpoint population is defined as participants in the Intent-to-Treat Population who completed demonstrations with both devices after reading the leaflet. This restriction makes completion relevant to interpretation. Missing-data counts, reasons for missingness, and the method used to handle missing outcomes are not reported in the registry record. It is therefore unknown whether participants with incomplete demonstrations differed from those included in the primary analysis.

The registry record does not describe an imputation method, sensitivity analysis, or other assessment of how missing observations might affect the findings. It also does not define the Intent-to-Treat Population in detail beyond its use in describing the modified population. The analysis population, missing-data process, and multiplicity approach are necessary context for interpreting estimates, but cannot be reconstructed from the registry record.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

For each of the five sub-studies, the registry record lists participants at risk for serious adverse events and other adverse events in both device groups, with zero events recorded in each category. The listed at-risk counts are 171 for ELLIPTA and DISKUS/ACCUHALER, 80 for ELLIPTA and MDI, 100 for ELLIPTA and TURBUHALER, 118 for ELLIPTA and HANDIHALER, and 98 for ELLIPTA and BREEZEHALER. These are subgroup entries and should be read as recorded for the respective device comparison, not added together to create an overall count.

Zero recorded events means that no events of those categories appear in the registry entries for these groups. It does not prove that the devices have no risks, and it cannot rule out events that are rare or that were not captured or reported. The record does not list serious adverse-event terms. Deaths are not reported in the registry record.

These safety entries are separate from the primary endpoint. Critical errors measure device-handling performance under the study assessment; serious and other adverse events are safety categories. Neither should be used as a substitute for the other when interpreting the study.

12. What the Result Does, and Does Not, Mean

The study’s primary endpoint is intended to quantify how often participants made at least one critical device-use error after reading the patient information leaflet. The definition focuses on errors considered likely to result in no or only minimal medication being inhaled. The randomized crossover sequence and paired device demonstrations provide a framework for comparing handling performance within each device pair. The registry record explains that framework and identifies the endpoint population.

However, the primary endpoint results are not reported in the registry record. The percentages for the devices, the size and direction of any differences, and the uncertainty around those differences are therefore unavailable. The record cannot support a conclusion that one device had fewer critical errors, that devices performed similarly, or that a difference was statistically or practically important.

The outcomes also include errors after health care professional instruction, instruction needed to demonstrate adequate technique, ease of use, and preference. These are relevant but distinct questions. Leaflet-based performance reflects use after reading written information; performance following professional instruction reflects a different condition. Preference and ease-of-use questionnaire responses represent participants’ reported experience. The results for these outcomes are not reported in the registry record.

The trial was open-label and conducted at a single visit. Participants and study personnel knew which device was being used, and the record does not provide evidence about performance after extended or repeated use. The study also focused on technique and device experience, not on comparative clinical outcomes from active medication treatment. It should not be interpreted as evidence of differences in disease control or other health outcomes.

The adverse-event entries record zero serious and zero other adverse events for the listed groups, while deaths are not reported in the registry record. Those records are limited safety observations and do not show that adverse events cannot occur. Overall, the registry establishes the study question, design, endpoint definition, and analysis population, but not the numerical findings needed to judge comparative performance.

13. Why This Trial Matters Statistically

This trial is statistically interesting because its central comparison is a randomized crossover assessment of device technique. In general, crossover designs can use within-participant information: each participant’s performance with a device is considered alongside performance with its paired comparator. This can reduce the influence of stable differences between participants when the analysis properly accounts for the paired structure. The registry record describes randomized device order, which is relevant because experience with the first demonstration could affect performance with the next.

The endpoint also illustrates how an operational definition shapes a statistical result. Counting whether a participant made at least one critical error turns multiple possible handling mistakes into a binary outcome. That makes the result easy to express as a percentage, while losing detail about the specific error, its frequency, or whether multiple errors occurred. The registry defines the errors of interest and why they are considered critical, allowing readers to understand what the percentage is intended to represent.

The study compares several device pairs through five sub-studies and lists multiple secondary outcomes. That creates an analysis landscape in which the choice of comparisons, outcome definitions, and multiplicity strategy matters. The registry record reports seven outcome measures and 35 analyses, but does not report how multiplicity was addressed or how the paired results were modeled. Those omissions limit statistical interpretation even apart from the absence of the primary endpoint values.

The registry also makes a distinction between performance after leaflet reading and performance after professional instruction, alongside subjective assessments of ease and preference. Keeping these outcomes separate matters: they reflect different measurement conditions and constructs. The record’s value is therefore in clarifying what was intended to be measured and how the study was organized. Without posted outcome estimates and analysis details, it cannot establish the magnitude or precision of device differences.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The most consequential limitation for interpreting the registry record is that the primary endpoint results are not reported. There are no device-specific percentages, effect estimates, confidence intervals, or p-values in the record. Thus the intended comparison cannot be evaluated for magnitude, direction, or precision.

The design was open-label. Knowledge of the device could influence participant behavior or assessments of ease and preference, although the registry record does not quantify such influence. The primary endpoint is based on observed technique after reading a patient information leaflet, but details about assessor training, assessment consistency, and adjudication are not reported in the registry record.

The crossover design makes sequence and learning relevant. Randomization of order is described, but the registry record does not report the method of sequence generation or concealment, whether order effects were assessed, or which statistical model was used for paired outcomes. The one-visit setting also means the record does not establish how performance changes with repeated use over time.

Analysis-population details are limited. The primary population excludes people in the Intent-to-Treat Population who did not complete demonstrations using both devices after reading the leaflet. The number excluded, reasons for incomplete demonstrations, missing-data handling, and sensitivity analyses are not reported in the registry record. The implications of completion-based inclusion therefore cannot be assessed.

The study used five device-specific sub-studies, and the registry lists seven outcomes and 35 analyses. The approach to multiplicity and the distinction between confirmatory and exploratory analyses are not reported. Finally, the study focused on device technique, ease, and preference; it does not provide a basis for conclusions about long-term clinical outcomes or comparative effectiveness of active treatments.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Study of Inhaler Device Attributes Investigating Critical and Overall Errors, Ease of Use, and Preference Between a Number of Inhaler Devices — NCT02184624.
• Registry record last updated 2017-05-23; status last verified 2017-04.
• Results first posted 2017-03-20.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT02184624 is a completed, randomized, open-label, crossover Phase 4 study of inhaler-device attributes in adults with chronic obstructive pulmonary disease. The registry records enrollment as 569 and describes a single-visit assessment conducted in the UK and the Netherlands. Participants were assigned to one of five device-pair sub-studies, with randomized order of ELLIPTA and the relevant comparator.

The primary endpoint was the percentage of participants making at least one critical error after reading the patient information leaflet, assessed on Day 1 in a modified Intent-to-Treat Population of participants who completed both device demonstrations. Other listed outcomes concern errors after professional instruction, instruction needed for adequate technique, ease of use, and preference. The primary endpoint results and the statistical analysis plan are not reported in the registry record. Safety entries show zero serious and zero other adverse events in each listed group; deaths are not reported in the registry record.