Evaluating the Non-Inferiority of Airmod to Capnostream™35: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluating the Non-Inferiority of Airmod to Capnostream™35 (NCT05263791), a na na study. The primary outcome is Test of the Difference Between airRR and mancRR.

2. Clinical Question

3. Trial Design

This was a completed, multi-center clinical study of a non-invasive medical device in Taiwan. The registry lists an enrollment of 284, a single-group model, no masking, and allocation as NA. The record explicitly cautions that this is not a randomized comparison and that its arms are not a like-for-like control comparison. The description says Airmod and the reference device were used in the same patient, which makes the primary comparison paired rather than a comparison of independent groups.

The population described in the study summary was people aged 20 years or older who were suitable for intravenous general anesthesia after pre-anaesthesia evaluation. The summary describes recruitment at this site and other clinical centers, but its account of the intended site and center totals is not clear enough to reconcile with the recorded enrollment of 284. The record does not report the number enrolled at each site or the number included in the primary analysis.

The intervention was Airmod software used with the AccurSound Electronic Stethoscope AS-101. The reference was the Medtronic Capnostream™35 Portable Respiratory Monitor, with manual-scored capnography serving as the primary reference measurement. The study also describes recording breathing sounds and investigating respiratory symptoms, accuracy during a less sensitive capnography period, agreement between measurement methods, response time following jaw thrust during apnea, variation in breathing rates, safety, and usability.

The listed start date is 2021-12-16, the primary completion date is 2022-09-27, and the status is completed. The study is listed as phase NA. The registry record does not report a randomized allocation procedure, masking, or a separate concurrent control group. These design features matter when interpreting the comparison: measurements from two devices in the same participant address agreement between measurement methods, but do not constitute a randomized comparison of clinical outcomes.

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

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 Test of the Difference Between airRR and mancRR, measured over up to 4 hours.

The primary endpoint is the mean difference in respiratory-rate measurements, expressed in breaths per minute, for a timeframe of up to 4 hours. The specified subtraction is the manual-scored capnography reference measurement minus the Airmod measurement. A difference near zero would indicate similar average readings, while the direction of a nonzero difference would indicate which method tended to give higher readings. This is a general interpretation of a mean difference, not a statement about the study’s observed result.

Comparing measurements from the same participant can account for the fact that the two readings are paired. The endpoint therefore concerns the average within-pair measurement difference, not a difference between independent groups of participants. A mean difference alone does not describe how widely individual paired readings vary, so agreement measures are relevant complements; the record lists a secondary analysis of agreement and Bland-Altman plots.

The outcome description gives a measurement range of 4 BPM to 35 BPM for Airmod and 0 BPM to 39 BPM for manual-scored capnography. These are described as the minimum and maximum measurement values, not as observed study results for the primary mean difference. Although the study title refers to non-inferiority, the registry record does not report the non-inferiority margin or the primary endpoint estimate and 95% confidence interval. It is therefore not possible from this record to assess whether the primary endpoint met a non-inferiority criterion.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

These explanations describe general statistical concepts and do not imply that a particular method was used unless the registry record says so. The record identifies a mean difference, a 95% confidence interval, and the paired measurement comparison, but does not report a complete statistical analysis plan or the numerical primary result.

What does a mean difference represent?

For each paired observation, one method’s respiratory-rate reading can be compared with the other method’s reading. The primary outcome is defined as the manual-scored capnography reference minus the Airmod measurement, in breaths per minute. The mean of these differences describes the average direction and size of the discrepancy across the analyzed observations. A positive mean would correspond to higher reference readings on average, and a negative mean to higher Airmod readings on average. This is a general explanation of the stated subtraction order; the registry record does not provide the observed mean.

The mean is useful for summarizing systematic average offset, but it does not show whether differences are similar for most participants or whether some readings differ considerably. That is why the record’s listed secondary agreement analysis and Bland-Altman plots address a related but distinct question. The numerical outputs for that analysis are not reported in the registry record.

What does a confidence interval add?

A confidence interval expresses uncertainty around an estimate. In general, a 95% confidence interval is produced by a method designed to cover the true parameter in 95% of repeated samples under the assumptions of that method. It is not a guarantee that a particular interval contains the true value, nor does it show the range of individual participants’ readings. The primary outcome specifies a 95% confidence interval, but its limits are not reported in the registry record. Its precision and compatibility with zero therefore cannot be assessed.

Confidence intervals also need to be interpreted in relation to the study question. For a non-inferiority assessment, an interval is generally compared with a prespecified margin, using a defined direction and analysis plan. The registry record does not report such a margin, so it is not possible to determine what interval would have counted as meeting the study’s stated non-inferiority objective.

Why is pairing important?

The study description says Airmod and the reference device were installed in the same patient. That arrangement produces paired measurements: the readings can be compared within the same participant. In general, paired analyses preserve the link between two measurements made under shared participant conditions. Treating paired observations as independent would discard that relationship and can misrepresent uncertainty.

The record does not state how many observations were collected from each participant, whether measurements were repeated over time, or how repeated observations were incorporated into the primary analysis. It also does not state whether the analysis accounted for center-level differences. The timeframe is listed as up to 4 hours, but that phrase alone does not establish the frequency or timing of measurements.

How is non-inferiority usually assessed?

In general, a non-inferiority design asks whether a new method’s performance is no worse than a comparator by more than a prespecified, acceptable amount. The margin, the outcome’s direction, and the statistical decision rule need to be defined before interpreting results. The study title and objective use non-inferiority language, but the margin and decision rule are not reported in the registry record. The primary endpoint is a difference between measurements; the record does not provide a threshold for deciding how large a difference would be acceptable.

Non-inferiority is not established simply because a difference is small, or because a conventional test fails to find a difference. The estimate and its uncertainty must be assessed against the stated margin using the planned approach. Since neither the numerical estimate nor the margin is reported, no non-inferiority conclusion can be reconstructed from the endpoint description.

What do agreement analyses contribute?

Correlation and agreement are different statistical ideas. In general, two measurement methods can rise and fall together and have strong correlation while still producing systematically different readings. Agreement methods examine the differences themselves. Bland-Altman analysis commonly summarizes the average difference and the spread of paired differences, helping describe whether individual readings are close enough for the intended measurement use. The record lists an analysis of agreement and Bland-Altman plots, but does not report their numerical results or the criteria used to interpret them.

Variation across breath rates is also listed as a secondary question. In general, measurement agreement may differ across the range of values being measured, so an overall average can conceal patterns at lower or higher rates. The record does not report subgroup estimates or the distribution of observed respiratory rates.

What can the enrollment figure tell us?

The registry records enrollment of 284. Enrollment is not necessarily the same as the number analyzed for every endpoint. In general, missing readings, exclusions, or unusable measurements can make the analysis population smaller than the enrolled population. The number included in each analysis, the reasons for any exclusions, and the handling of missing data are not reported in the registry record.

The record includes a general sample-size illustration assumption describing two-sided significance level 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a fraction of the observed effect. It also states that such calculations are planning illustrations based on observed effects, not the trial’s own design. These assumptions do not provide an observed power result or establish how the study’s enrollment was chosen. The registry record does not report the trial’s sample-size justification or its own planned analysis parameters.

10. Multiplicity, Missing Data and Analysis Populations

The record lists 6 outcome measures and 4 analyses. The primary outcome is a mean difference in respiratory-rate readings; the listed secondary questions concern comparison with manual-scored auscultation during a less sensitive capnography period, agreement and Bland-Altman plots, response time after jaw thrust during apnea, variation in breath rates, and safety and usability. The registry record does not report how many statistical tests were conducted for each outcome or whether the analyses were prespecified in a detailed plan.

When a study evaluates multiple outcomes or performs multiple comparisons, the chance of at least one apparently notable result can increase. In general, multiplicity methods may control that risk, or researchers may identify a primary outcome and treat other analyses as exploratory. The record identifies a primary outcome but does not report a multiplicity adjustment or a hierarchy for secondary analyses. No inference about adjustment should be made from the number of listed outcomes alone.

Missing-data methods are not reported in the registry record. It does not state how incomplete paired measurements, missing time points, or participants without usable readings were handled. Nor does it report the analysis population for the primary outcome, the number analyzed, exclusions after enrollment, or sensitivity analyses addressing missingness.

The record describes a single-group, unmasked study in which the devices were used in the same participant. It does not report a detailed protocol-defined analysis population, a statistical analysis plan, or any interim analysis. The absence of these details in the registry record does not establish that no such procedures existed; it means their content cannot be evaluated from this record.

11. 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 5% reporting threshold.

The registry lists 255 participants at risk in the Airmod group. It reports 0 participants with serious adverse events, 0 with other adverse events, and 0 deaths. No serious adverse-event terms are listed. These are reported counts, not evidence about events that may have been unrecorded or outside the registry’s reporting scope.

In general, an at-risk count identifies the number of participants included in the safety accounting, while an affected count gives the number recorded as having an event. The safety figures should not be confused with the enrollment of 284: the registry does not explain why the at-risk count differs from the enrolled count, or whether the populations and follow-up periods differ. It also does not report event definitions, collection procedures, follow-up completeness, or a separate comparison group for safety.

The study is described as a single-group device evaluation, not a randomized comparison. With no separate concurrent control group in the listed design, these counts cannot establish a comparative safety difference. They should be read narrowly as the adverse-event totals reported for the listed Airmod group.

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

The registry record describes an evaluation of respiratory-rate measurements from Airmod against manual-scored capnography readings in the same participants. It identifies a mean paired difference as the primary endpoint, gives the subtraction direction and measurement unit, and specifies a 95% confidence interval over a timeframe of up to 4 hours. Those details make the intended statistical question understandable.

However, the registry record does not report the observed mean difference or confidence interval limits. It also does not report a non-inferiority margin or a decision rule. As a result, the central numerical evidence needed to assess the stated non-inferiority objective is unavailable in this record. The study title alone does not show whether the criterion was met, and the endpoint definition is not a substitute for the result.

Even if an average difference were available, it would not by itself establish close agreement for individual readings. An average can conceal substantial variation among paired measurements. The record lists a secondary agreement analysis, but the numerical findings are not reported in the registry record. No conclusion about the spread of individual differences or performance across different breath rates can be drawn from the information recorded.

The trial design also sets limits on interpretation. This was a single-group, unmasked study, with the devices compared in the same patient rather than through random assignment to separate groups. That design is suited to a paired measurement comparison, but it does not provide a randomized comparison of clinical outcomes. The record does not report enough detail to assess missing measurements, the number analyzed, or the analysis model.

The reported safety totals are zero serious adverse events, zero other adverse events, and zero deaths among 255 listed as at risk in the Airmod group. These figures describe the recorded safety counts and should not be expanded into a claim about unreported events or comparative safety. Overall, the record describes what was intended to be measured, but does not provide the primary numerical results needed to judge measurement differences or the non-inferiority question.

13. Why This Trial Matters Statistically

This study raises a practical statistical issue in device evaluation: a new measurement method should be assessed not only for its average reading, but also for how closely it agrees with a reference across paired observations. The primary endpoint is defined as the mean difference between manual-scored capnography and Airmod. That makes direction explicit and focuses the comparison on the measurement discrepancy rather than on a treatment outcome.

The paired design is relevant because both methods were used in the same participant. In general, paired comparisons can account for participant-specific conditions shared by both readings. But a mean difference does not describe all aspects of agreement, and the record’s listed Bland-Altman analysis addresses that broader statistical question. The registry record does not report its results, so the intended analytical breadth cannot be matched to findings in this record.

The non-inferiority framing also illustrates why statistical criteria need to be explicit. A difference cannot be judged acceptable without a prespecified margin and a defined interpretation of uncertainty. The record does not report the margin or the primary estimate and confidence interval. The title therefore identifies the study’s aim, but the registry information does not allow readers to evaluate whether the aim was achieved.

Finally, registry completeness affects statistical interpretation. Enrollment and safety counts are reported, but the number included in the primary analysis, missing-data procedures, and detailed analysis methods are not. These omissions limit assessment of precision, potential exclusions, and the population represented by the result. The study is statistically important as a measurement-comparison question, while the available record leaves its central numerical answer unreported.

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.

A key limitation of the registry record is that it does not report the primary mean difference or its 95% confidence interval. Without the estimate, its direction and magnitude are unknown; without the interval, its precision cannot be assessed. The record also does not report the non-inferiority margin, so the threshold needed to interpret the stated objective is unavailable.

The record describes Airmod and the reference device being used in the same patient, but does not detail measurement timing, frequency, or how repeated observations were handled. Although the primary timeframe is up to 4 hours, the schedule within that period is not reported. The number of participants or observations analyzed for the primary endpoint is also not reported.

Enrollment is recorded as 284, while safety accounting lists 255 at risk. The record does not explain the difference or indicate whether these counts refer to different analysis populations or follow-up. The study summary discusses planned recruitment at multiple centers, but the site and center totals are not clear enough to reconcile with the recorded enrollment. Site-specific enrollment and results are not reported.

The study is single-group and unmasked, and the registry cautions that it is not a randomized comparison. These features do not invalidate a paired measurement question, but they limit conclusions that depend on randomized group comparisons or blinded assessment. The record also does not report detailed methods for missing data, multiplicity, analysis populations, or adjustment for site effects. The listed secondary outcomes have no numerical results in the registry record, limiting assessment of agreement, response time, rate variation, safety, and usability beyond their stated aims.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Evaluating the Non-Inferiority of Airmod to Capnostream™35 — NCT05263791.
• Registry record last updated 2024-11-19; status last verified 2024-09.
• Results first posted 2024-11-19.
• 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

NCT05263791 is a completed, multi-center evaluation of Airmod respiratory monitoring assistant software used with an AccurSound Electronic Stethoscope AS-101. The reference method was manual-scored capnography using the Medtronic Capnostream™35 Portable Respiratory Monitor. The record lists enrollment of 284, a single-group design, no masking, and allocation as NA. The stated population was people aged 20 years or older suitable for intravenous general anesthesia after pre-anaesthesia evaluation.

The primary outcome is the mean difference between reference and Airmod respiratory-rate measurements, in breaths per minute, with a 95% confidence interval and a timeframe of up to 4 hours. The study title describes a non-inferiority evaluation. The registry record does not report the primary estimate, confidence interval limits, or non-inferiority margin. It lists 6 outcome measures and 4 analyses, but does not provide the numerical findings needed to assess the primary measurement comparison.

The recorded status is completed. The start date is 2021-12-16, and the primary completion date is 2022-09-27. Safety reporting lists 255 participants at risk and zero serious adverse events, other adverse events, or deaths in the Airmod group.