A Study for the Assessment of the Benefits of a Novel Mesh Nebulizer in the Treatment of Patients With Stable COPD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study for the Assessment of the Benefits of a Novel Mesh Nebulizer in the Treatment of Patients With Stable COPD (NCT03933462), a randomized na study. The primary outcome is Device Preference Among the Number of Participants..

2. Clinical Question

The primary question concerns preference, not a direct measure of lung function, medication delivery, or clinical benefit. The questionnaire was an 11-question survey designed to assess device satisfaction. Its first 10 questions used a scale from 0 to 10, with 0 described as worst and 10 as best; the final question asked about device preference at the end of the study.

The registry describes the study as randomized and crossover. In a crossover design, participants receive more than one study condition, so each participant may contribute experience with both devices. The registry record does not report details such as the order-allocation procedure, whether treatment sequence was balanced, or how the analysis accounted for within-participant comparisons. The posted primary result is expressed as counts of participants preferring each device. It therefore addresses the stated preference question, while leaving important design and analysis details unreported.

3. Trial Design

This completed post-market study evaluated InnoSpire Go, a vibrating mesh nebulizer, against a standard jet nebulizer in stable COPD. The registry records randomized allocation, no masking, and a crossover model. The planned experience was 30 days using each device, with preference and quality-of-life changes to be evaluated. The primary endpoint was assessed after 60 days of device use.

The registry records enrollment of 21 participants. The primary-outcome population description says that one participant was removed from the data set because they were on the incorrect standard-of-care device. The posted outcome counts list 17 participants for InnoSpire Go and 17 for the jet nebulizer. The registry record does not explain how these arm counts relate to the enrollment figure or to the crossover structure, so they should be read as the posted counts rather than treated as an independently reconstructed participant flow.

The study began on 2019-06-12, reached primary completion on 2019-10-28, and is marked completed. The phase is recorded as NA. Philips Respironics is identified as the sponsor. One secondary outcome is listed: change in quality of life from baseline on the CRQ-SR. The registry lists nine outcome measures and one analysis.

Randomization can help reduce systematic differences in how participants experience or are assigned devices. Crossover designs can also enable comparisons within the same participant, but the interpretation depends on details such as treatment sequence, washout, and the statistical handling of paired observations. Those details are not reported in the registry record. Because masking is listed as none and the main endpoint is a stated preference, participants’ awareness of the device could be relevant to how the endpoint is reported. The registry record does not describe additional procedures to address that possibility.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: One participant removed from the data set because they were on the incorrect standard of care device.

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 Device Preference Among the Number of Participants., measured over 60 days.

The primary endpoint is device preference among participants after 60 days of device use. It was assessed using the final question of an 11-question nebulizer satisfaction questionnaire. The questionnaire’s first 10 questions use a 0-to-10 response scale, where 0 is worst and 10 is best; the final question asks which device the participant prefers and is asked at the end of the study.

Preference is a participant-reported endpoint. It can capture the overall experience of using a device, including features that may matter to participants but are not represented by a clinical measurement alone. In general, a preference endpoint is useful for describing acceptability or relative satisfaction in the study setting. It does not, by itself, establish that one device improves health, delivers medication more effectively in practice, or changes disease outcomes.

The posted counts are 17 participants preferring InnoSpire Go and 0 preferring the jet nebulizer. The registry’s primary-outcome population description identifies one participant removed for using the incorrect standard-of-care device. The record does not describe how participants with no preference, incomplete questionnaires, or other missing responses were handled, nor does it provide the questionnaire’s full response distribution beyond the reported preference counts. Those omissions limit what can be learned about the range and certainty of participants’ experiences from this endpoint alone.

7. Results Posted in the Registry

• Jet Nebulizer: 0.0% (0/17) versus 100.0% (17/17) on control; risk difference -100.0 percentage points (95% CI -100.0 to -100.0); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary at the participant level: a participant’s final questionnaire response indicates preference for one device or the other. The registry presents counts, with 17 participants preferring InnoSpire Go and 0 preferring the jet nebulizer. For a binary outcome, a proportion describes the share of observed participants with a specified response, while a risk difference compares two such proportions by subtraction. In general, the direction of a risk difference depends on which group is placed first or treated as the reference. The posted comparison labels the jet nebulizer against InnoSpire Go and reports a risk difference of -1.0.

The registry’s reported method is a One Sample Binomial analysis, with a superiority result and p < 0.001. A binomial model is generally used for counts of outcomes in a group when each observation has two possible categories, under assumptions about the probability of the outcome and the independence structure. The method label alone does not reveal the full null hypothesis, how the test was parameterized, or whether crossover pairing was incorporated. Those details are not reported in the registry record. Because the study is described as crossover, it is especially important not to infer that the observations were independent merely from the displayed arm counts.

A p-value evaluates how compatible the observed data are with a specified null model, assuming that model and its assumptions. A small p-value is not an effect size and is not a measure of the probability that the result occurred by chance. It also does not account automatically for every design feature, missing response, or analysis decision. The sponsor’s reported p-value is evidence relative to its stated analysis, but the registry record gives limited detail about the test’s exact specification.

A confidence interval is generally a range produced by a method that, across repeated use under its assumptions, has a stated coverage property. It is intended to express sampling uncertainty around an estimate, not the range of individual participants’ responses. The posted risk-difference interval runs from -1.0 to -1.0, with no width at the displayed precision. That unusual presentation should be reported as it appears, not expanded or replaced with an interval from an unreported method. The registry record does not provide a confidence interval for the sponsor’s One Sample Binomial analysis.

In a crossover study, a participant’s outcomes under the different devices may be correlated. A paired analysis can use that within-participant relationship, while an analysis that treats each device group as independent may not. In general, accounting for pairing can improve the relevance and efficiency of a comparison when the design and data support it. Whether that happened in this study is not reported in the registry record. Details about sequence effects, period effects, carryover, washout, and the handling of incomplete crossover participation are also not reported. These are not details that can be reconstructed from the endpoint counts.

The posted sample-size illustrations use the observed effect as their starting point and describe planning under a two-sided significance level of 0.05, equal allocation, and a normal approximation. They are planning illustrations, not observed power and not evidence that the trial was designed to achieve a particular power. The registry record does not report the trial’s own sample-size justification, target power, or assumptions used before enrollment. An illustration based on the observed effect is necessarily tied to this study’s observed result and should not be read as independent confirmation of it.

Over-dispersion is a general issue when observed variability in count data exceeds what a simple model assumes. It can make standard errors too small and inferential results too optimistic if the model is not suitable. For a binary preference endpoint, the central concerns include the sample size, independence or pairing, and whether the model matches how the data were collected. The registry record does not report an over-dispersion assessment, and the posted result alone does not show whether one was needed.

The study lists nine outcome measures but only one secondary outcome is identified in the record excerpt. The primary endpoint and the secondary quality-of-life measure answer different questions. If many outcomes or analyses are tested, the chance of obtaining at least one small p-value can increase; this is the general multiplicity problem. The registry record does not report a multiplicity strategy, a hierarchy among outcomes, or whether any adjustments were made. It also does not report the complete statistical analysis plan, interim analyses, or detailed analysis populations. These omissions limit independent assessment of how the reported test fits into the full set of planned and conducted analyses.

10. Confidence Intervals and Precision

The primary comparison is reported with an interval from -1.0 to -1.0 for the risk difference. Since the lower and upper limits are the same at the displayed precision, the interval does not communicate a visible range of uncertainty around the estimate. The registry record does not explain the interval method or why the endpoints coincide. The interval should therefore be read as the value recorded, not as evidence that the underlying preference difference is known without uncertainty.

The sponsor’s One Sample Binomial analysis reports p < 0.001 but no confidence interval. A p-value and a confidence interval answer different questions: a p-value summarizes compatibility with a null hypothesis under a specified test, while a confidence interval is intended to describe uncertainty around an estimate under a specified method. A small p-value does not replace an interval, and it does not state how precisely the preference pattern would reproduce in another sample.

The observed counts are based on a limited set of participants: the registry lists 17 for each device group, while enrollment is 21. The record also notes one removal from the primary-outcome data set for use of the incorrect standard-of-care device. It does not explain the relationship between these counts, the crossover structure, or whether all participants completed all endpoint assessments. The available record therefore does not permit a full assessment of the effective information underlying the reported comparison.

Precision is also shaped by the endpoint and design. A preference response reduces a participant’s experience to a categorical choice, and the posted counts do not show the distribution of questionnaire ratings or how strongly participants preferred their selected device. In general, paired crossover information can affect the precision of a comparison, but the analysis details needed to judge that contribution are not reported in the registry record. The result is precise as a description of the posted counts; the degree of precision for a broader population remains uncertain.

11. Analyses Reported by the Sponsor

The sponsor’s reported analysis is a One Sample Binomial superiority analysis with p < 0.001. The registry also posts a comparison for the jet nebulizer relative to InnoSpire Go, with a risk difference of -1.0 and an unadjusted p-value of <0.001. The sponsor’s reported p-value and the posted unadjusted comparison are consistent in their reported threshold and direction. The record does not flag a discrepancy between them.

The sponsor analysis is the trial’s own reported inferential analysis and takes precedence over a simple unadjusted re-calculation when a discrepancy is identified. No discrepancy is indicated in this record. The sponsor’s result should still be interpreted in light of what is documented: it concerns the participant preference endpoint, and the record gives limited detail about the precise test specification and its handling of the crossover design.

The sponsor’s p-value does not establish clinical benefit or resolve uncertainty about wider applicability. Nor does it provide an interval estimate; no confidence interval for that analysis is reported in the registry record. The count pattern remains the most direct description of the posted primary result: 17 preferences for InnoSpire Go and 0 for the jet nebulizer. The inferential analysis supports a statistical comparison under the sponsor’s method, while the available documentation does not permit a complete independent reconstruction of that method.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists nine outcome measures and identifies change in quality of life from baseline on the CRQ-SR as a secondary outcome. It also records one analysis. The primary outcome is device preference after 60 days. The record does not describe whether outcomes were prioritized in a prespecified hierarchy, whether multiple testing was considered, or whether any p-values were adjusted. These details are not reported in the registry record.

Multiplicity matters in general when several outcomes, comparisons, time points, or analytic approaches are tested. Without adjustment or a planned hierarchy, the chance of at least one apparently strong result can rise across a collection of tests. The information recorded does not show how many tests were contemplated in the full analysis plan or whether the single listed analysis represents all analyses conducted. The result should therefore be considered in the context of incomplete reporting rather than assumed to be the only statistical test performed.

The registry’s primary-outcome description says one participant was removed from the data set because they were using the incorrect standard-of-care device. It does not describe the analysis population in greater detail, give a participant flow by crossover sequence, or state how incomplete questionnaires or other missing observations were handled. No imputation method or sensitivity analysis for missing data is reported in the registry record. The posted counts are not enough to determine whether all randomized participants contributed to each outcome.

The record does not report whether the primary analysis followed an intention-to-treat principle, used a per-protocol population, or applied another population definition. It also does not report a detailed statistical analysis plan, interim analyses, or the timing of any analysis decisions. These omissions matter because population definitions and analytic choices can affect both the estimated result and its uncertainty. The preference result is interpretable as the posted count comparison, but the registry record does not provide enough information to establish how missingness, exclusions, or multiplicity were managed across the full study.

13. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: copd exacerbation (9.5%).

The registry reports serious adverse events for 2 participants at risk of 21 in the InnoSpire Go group and 0 participants at risk of 21 in the jet nebulizer group. It lists 0 deaths in each group. COPD exacerbation is the leading serious adverse event term, reported as 9.5%. These figures describe the safety information recorded in the registry and should be kept separate from the primary preference endpoint.

Serious adverse event counts do not by themselves show that a device caused an event. The registry record does not report enough detail to establish event timing relative to device exposure, adjudication, relatedness, or the method used to attribute events in a crossover design. The figures also do not provide a complete comparison of all potential harms; they are the serious-event summary recorded in the registry.

The at-risk counts are listed as 21 for each group, while the primary endpoint lists 17 participants per device group. These counts refer to different posted outcome summaries and should not be combined or used to infer a participant flow. The record does not explain the relationship between the safety populations, crossover exposure, and primary-outcome population. The absence of listed deaths or serious events in a group is not proof that no risk exists; it means no such events are recorded for that group in this summary.

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

The result describes a preference pattern among participants in this completed randomized crossover study of stable COPD: the posted counts are 17 participants preferring InnoSpire Go and 0 preferring the jet nebulizer. The sponsor’s reported superiority analysis gives p < 0.001. This is evidence about the recorded device-preference endpoint under the sponsor’s analysis, not a direct test of clinical effectiveness.

Preference can be meaningful because a device experience may shape whether participants find a system acceptable. In general, however, preference cannot establish that medication reaches the lungs more effectively, that symptoms improve, or that health outcomes change. The registry lists a secondary quality-of-life outcome, but a conclusion about that measure cannot be drawn from the primary preference result. Its posted results are not reported in the registry record beyond the outcome listing.

The result also does not establish what all people with COPD would prefer. The study enrolled a limited group of stable, ambulatory participants already using a jet nebulizer, and the primary counts list 17 participants per device. The record does not fully explain how those counts relate to enrollment or how crossover pairing and missing data were handled. Because the study was unmasked and preference was reported by participants, the result reflects participants’ stated experience in the study setting, with the limitations of a subjective endpoint and the information recorded.

Finally, a small p-value does not eliminate uncertainty. The displayed risk-difference interval has identical limits, and the sponsor analysis has no confidence interval reported. The registry does not provide the analysis detail needed to evaluate all assumptions or the result’s precision for a wider population. The appropriate statistical reading is therefore specific: the recorded participants expressed a strong preference for InnoSpire Go, while broader clinical meaning and generalizability are not established by these preference counts alone.

15. Why This Trial Matters Statistically

This study illustrates how a clinical device comparison can produce a clear preference pattern while still leaving important statistical questions open. The endpoint is easy to understand as a participant choice, and the posted counts show all recorded preferences favoring InnoSpire Go. Yet a preference endpoint is not interchangeable with a clinical outcome, and the strength of the observed pattern does not settle whether it generalizes beyond the participants represented.

The crossover design makes the analysis structure especially relevant. When participants use both devices, responses can be paired within a person. That can be valuable, but the interpretation depends on how sequence and within-participant dependence are handled. The registry describes the crossover model but does not report the analysis details needed to assess those features. The One Sample Binomial method label does not, on its own, establish how pairing was incorporated.

The result also highlights why reporting uncertainty matters alongside a p-value. The reported p-value is below 0.001, but the posted risk-difference interval has the same lower and upper value. The sponsor analysis does not have an interval reported in the registry record. Readers therefore have a strong test result and a straightforward count pattern, but limited documented information about precision.

Finally, this record shows the value of transparent reporting on analysis populations, missing responses, multiplicity, and sample-size planning. Such details help readers distinguish an observed result from the assumptions and decisions used to estimate it. The available result is informative about stated preference in this study; fuller reporting would make its statistical interpretation easier to assess.

16. 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 posted primary counts require caution because the study is described as crossover, yet the outcome is presented as 17 participants for each device group. The registry also reports enrollment of 21 and says one participant was removed from the primary data set for using the incorrect standard-of-care device. It does not explain the relationship among these figures or provide a complete participant flow. The effective analysis population and contribution of each participant cannot be reconstructed from the registry record.

The record does not describe treatment sequence, washout, period effects, carryover, or the model used to account for repeated observations within participants. These are relevant features of a crossover design. The reported One Sample Binomial analysis does not supply enough detail to determine whether the analysis fully reflects the design. The registry record also does not report the detailed statistical analysis plan or the trial’s own sample-size justification.

The endpoint is participant-reported preference, and the study is listed as unmasked. The record does not report how neutral or missing responses were treated, whether questionnaire completion was complete, or whether preference strength was analyzed. The posted counts alone cannot show how participants scored the other questionnaire items. The sole secondary outcome listed is quality-of-life change on the CRQ-SR, but its results are not reported in the registry record.

The comparison’s displayed interval has identical limits, and no confidence interval is reported for the sponsor’s test. Multiplicity handling, interim analyses, and the full set of analyses conducted are not reported. Together, these omissions limit assessment of precision, robustness, and generalizability. They do not negate the observed count pattern, but they restrict how far it can be interpreted beyond the recorded preference responses.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: A Study for the Assessment of the Benefits of a Novel Mesh Nebulizer in the Treatment of Patients With Stable COPD — NCT03933462.
• Registry record last updated 2020-12-17; status last verified 2020-11.
• Results first posted 2020-12-17.
• 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.

20. Record Summary

NCT03933462 is a completed, randomized, unmasked crossover study sponsored by Philips Respironics. It compared InnoSpire Go, a vibrating mesh nebulizer, with a jet nebulizer in stable COPD. Enrollment was 21, and the registry describes 30 days of use for each device.

The primary endpoint was participant preference after 60 days, assessed with the final question of an 11-question nebulizer satisfaction questionnaire. The posted counts are 17 participants preferring InnoSpire Go and 0 preferring the jet nebulizer, with 17 participants listed for each group. The sponsor reports a One Sample Binomial superiority analysis with p < 0.001. A secondary quality-of-life outcome is listed, but its results are not reported in the registry record.

The registry records 2 serious adverse events among 21 at risk for InnoSpire Go and 0 among 21 at risk for the jet nebulizer; no deaths are listed. The available record does not report important details of the crossover analysis, missing-data handling, multiplicity, or precision for the sponsor’s test. The primary result is a preference finding, not evidence by itself of clinical benefit.