Randomized Crossover Study of Neurally Adjusted Ventilatory Assist (NAVA) in Preterm Infants: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Randomized Crossover Study of Neurally Adjusted Ventilatory Assist (NAVA) in Preterm Infants (NCT01389882), a randomized phase 2 study. The primary outcome is Peak Inspiratory Pressure.

2. Clinical Question

3. Trial Design

The registry describes this as a randomized, single-masked, crossover study and lists it as phase 2. It records 26 enrolled participants, the condition as preterm infant, and Seoul National University Hospital as the sponsor. The study status is completed. The registry dates the study start to 2011-03 and primary completion to 2011-08.

In general, a crossover design compares conditions within participants across different treatment periods. That structure can account for stable differences between participants, but interpretation depends on details such as the order of conditions, how periods were scheduled, whether there was a washout, and how period or carryover effects were handled. Those details are not reported in the registry record. Randomization is recorded, but the randomization procedure and the allocation sequence are not described.

The registry records single masking but does not identify who was masked or how masking was maintained. It does not describe the number or identity of study arms in a way that maps the coded analysis groups to NAVA and SIMV+PSV. The primary analysis is described as per protocol and includes 19 patients who completed the study; a reason for the difference between enrollment and the analyzed population is not reported in the registry record.

There is an important design-description caveat. The registry fields call the study a crossover, while an analysis note characterizes the design as cluster-randomized and cautions that a simple individual-level re-analysis would be misleading. The record does not reconcile these descriptions. In keeping with that caution, the posted statistical result is the sponsor’s paired t-test rather than an independently reconstructed analysis. The sample-size illustrations are planning exercises based on the observed effect, not evidence of the trial’s original design rationale or its observed power.

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: 26 patients enrolled; 24 assigned to randomization and 2 excluded (1 extubated prior to the study and 1 developed exclusion criteria)

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis was "per protocol". 19 patients who had completed the study was used for the analysis.

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 Peak Inspiratory Pressure, measured over four hours.

Peak inspiratory pressure is the maximum pressure recorded during inspiration by the ventilator. The registry gives cmH2O as the unit and describes measurement over four hours with each ventilator mode. It treats the outcome as a continuous measure and reports a mean-based analysis with standard deviation as the dispersion measure. The registry record does not give the group-specific means or standard deviations.

As general context, airway pressure is a physiologic measure that can help characterize the pressure used to deliver ventilation. It is relevant to the study’s stated question because the investigators proposed comparing ventilator pressure while maintaining equivalent fractions of inspiratory oxygen and capillary-blood carbon dioxide partial pressure. Those equivalence conditions are part of the study rationale; the reported primary result does not itself establish that oxygen or carbon dioxide values were equivalent.

A mean summarizes the average measurement in a group or condition, while individual observations can vary around that average. In a crossover comparison, the paired analysis focuses on within-participant differences rather than treating each condition as unrelated. The endpoint therefore addresses a short-term ventilator measurement under the study conditions, not longer-term clinical outcomes. The registry’s four-hour timeframe per mode does not establish the duration of any later follow-up, which is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports a paired t-test for the primary comparison. In general, a paired t-test is used when observations are meaningfully linked, such as measurements from the same participant under different conditions. Instead of comparing unrelated group averages, it analyzes the within-pair differences and estimates their mean. This approach is naturally relevant to a crossover design, where a participant may contribute measurements under more than one condition. The registry’s designation of the analysis as paired is explicit, but the record does not provide the underlying paired observations or describe how incomplete pairs were handled.

What does the mean difference represent?

The reported estimate is 1.007 cmH2O for the mean difference in final values. This is the estimated average contrast in the coded order used by the analysis. A positive estimate means the first coded quantity exceeds the second under that subtraction convention; however, the registry does not map the group codes to NAVA and SIMV+PSV or state the subtraction order in plain language. Accordingly, the result cannot be assigned a clinical direction from the record alone. Group-specific means are also not reported in the registry record.

How should the confidence interval be read?

The 95% confidence interval runs from 0.036 to 1.978 cmH2O. As a general statistical explanation, a confidence interval is an uncertainty range derived from the data and the assumptions of the analysis. It gives a sense of the precision of an estimated average contrast, rather than describing the spread of individual measurements or guaranteeing that the true value lies in the interval. The interval is on the positive side of zero in the stated coding, but the unreported group mapping still prevents assigning that direction to a particular mode. The record does not describe any adjustment beyond the paired comparison.

What does the p-value say?

The reported p-value is 0.043. In general, a p-value evaluates how compatible the observed data are with a null model of no mean difference, assuming the test’s model and design conditions are appropriate. It does not express the probability that there is no effect, and it is not a measure of clinical importance. The registry lists a two-sided significance level of 0.05 for the planning illustration, but does not report whether that threshold was specified in the original protocol or analysis plan. A conclusion based only on comparing the reported p-value with a threshold would leave the group-direction uncertainty and design limitations unresolved.

Why does pairing matter?

Pairing can improve efficiency when measurements from the same participant are correlated, because stable participant-level differences are accounted for in the within-participant contrast. That benefit relies on analyzing the data in a way that respects the actual assignment and measurement structure. The registry labels the study crossover and reports a paired test, yet its analysis note also describes the design as cluster-randomized and warns against a simple individual-level re-analysis. The record does not explain that apparent inconsistency. The sponsor’s analysis is therefore the only result posted for this comparison, and no independent re-calculation is presented.

General statistical concerns in a crossover study include treatment order, changes over time, period effects, and carryover from an earlier condition. These may affect a paired contrast if they are present and not adequately addressed. The sequence, washout arrangements, and methods for evaluating period or carryover effects are not reported in the registry record. The posted result should consequently be interpreted as the reported analysis, not as evidence that these issues were absent.

What does the analysis population mean?

The primary-outcome description says the analysis was per protocol and used 19 patients who completed the study. In general, a per-protocol analysis includes participants meeting specified adherence or completion criteria, but the registry does not state the exact criteria used. It also does not report an intention-to-treat analysis, the reasons participants did not contribute to the per-protocol result, or the method used for missing outcome measurements. The reported estimate therefore pertains to the described completed-study population; the impact of exclusions or missingness cannot be assessed from the registry record.

The sample-size material is described as an illustration computed from the observed effect, using a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the target effect expressed as a fraction of the observed effect. Such values are planning illustrations, not observed power and not evidence of the trial’s original sample-size justification. The original justification is not reported in the registry record. The registry also notes that outcome and analysis counts exist, but it does not provide a multiplicity procedure for interpreting the collection of outcomes. Whether the primary comparison was selected in advance or adjusted for other analyses is not reported in the registry record.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("SIMV With PS", "NAVA")
n   <- c(19, 19)
m   <- c(13.45, 12.45)
sp  <- c(3.44, 2.66)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  SIMV With PS  19  13.450  3.440  11.792  15.108
CARM  NAVA  19  12.450  2.660  11.168  13.732

11. Analyses Reported by the Sponsor

The posted primary result is the sponsor’s own paired t-test: a mean difference in final peak inspiratory pressure of 1.007 cmH2O, with a 95% confidence interval from 0.036 to 1.978 and p = 0.043. The record does not provide a separate recalculated estimate for comparison. It also includes an analysis note warning that simple individual-level re-analysis would be misleading and stating that only the sponsor’s analysis is shown. Thus, there is no recalculation in the posted results against which to assess agreement or discrepancy.

The registry does not provide the data needed to reconstruct the paired test, such as the individual paired measurements or a complete set of paired summary statistics. Nor does it map the coded groups to the named ventilator modes in the analysis description. The sponsor’s estimate should therefore be read as the reported result, with its direction expressed only in the coded contrast. The registry record does not report enough detail to independently verify the calculation or explain how the crossover and cluster-randomized design descriptions relate to one another.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint of peak inspiratory pressure and also names secondary outcomes including mean airway pressure, minute ventilation, expiratory tidal volume, dynamic compliance, work of breathing, peak EAdi, fraction of oxygen, and capillary blood pH. It records 12 outcome measures and 12 analyses. The primary result is reported with a confidence interval and p-value, but a multiplicity adjustment or a hierarchy for interpreting the multiple outcomes is not reported in the registry record.

As general context, when several outcomes or comparisons are examined, the chance of at least one apparently unusual result can increase if each is assessed independently. Methods for controlling this concern depend on the prespecified analysis plan and the role assigned to each outcome. The registry does not state whether the secondary outcomes were confirmatory or exploratory, how many comparisons were planned, or whether any correction was used. The p-value for peak inspiratory pressure should therefore be interpreted as the reported result for that endpoint, without assuming a multiplicity strategy that is not documented.

The analysis population is described as per protocol, using 19 patients who completed the study. The reasons for not including all enrolled participants in this analysis, the number with missing primary measurements, and the approach to missing data are not reported in the registry record. The record does not state whether missingness was related to treatment, outcome, or completion status. Without those details, it is not possible to judge whether the analyzed participants represent the enrolled population or how missing observations may have affected the estimate.

The registry does not report an intention-to-treat analysis, prespecified handling of deviations, or a full statistical analysis plan. It also does not report whether there were interim analyses. These absences do not establish that such plans or procedures were not used; they mean that they are not documented in the registry record available for interpretation. The reported per-protocol result should be distinguished from a complete randomized-population analysis.

13. 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 5% reporting threshold.

The registry reports serious adverse event figures for the SN Group and NS Group. For the SN Group, 0 serious events are listed among 13 at risk; for the NS Group, 0 are listed among 11 at risk. The registry also lists 0 other events among 13 at risk in the SN Group and 0 among 11 at risk in the NS Group. The group labels are not mapped to the named ventilator modes in the record.

These figures describe the event counts entered in the registry for the stated groups. They do not establish that no adverse events of any kind occurred: the listed figures distinguish serious events from other events, and the registry does not define the scope or collection procedures in the information available. Deaths are recorded as null for both groups, which means death counts are not reported in the registry record; null should not be interpreted as zero.

As general context, absence of a recorded event in a group does not prove that a risk is absent. Adverse-event interpretation depends on how events were defined, monitored, attributed, and collected, as well as on the amount of exposure and follow-up. Those details, the follow-up period for safety, and any treatment-related assessment are not reported in the registry record. These counts should be read as registry entries, not as a complete safety assessment or a comparison establishing that one mode is safer.

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

The result gives a sponsor-reported estimate of the average difference in final peak inspiratory pressure between two coded conditions in a paired analysis. The estimate is 1.007 cmH2O, with a 95% confidence interval from 0.036 to 1.978 and p = 0.043. The interval and p-value convey statistical uncertainty under the reported test, but the record does not identify which coded group corresponds to NAVA or SIMV+PSV. The numerical sign therefore cannot be translated into a statement about which mode used higher pressure.

The endpoint is a short-term ventilator measurement taken over four hours with each mode. It is not itself a measure of longer-term respiratory health, development, survival, or other patient-centered outcomes. The stated study rationale refers to pressure at equivalent fractions of inspiratory oxygen and capillary-blood carbon dioxide partial pressure, but the registry’s primary result does not demonstrate that those quantities were equivalent in the analyzed data.

In general, statistical evidence about a mean contrast is different from evidence about clinical importance. A small p-value does not determine whether an observed difference matters in practice, and the confidence interval does not describe how individual infants respond. The registry does not report a threshold for a clinically important pressure difference, nor does it provide enough detail to evaluate how the contrast relates to other outcomes.

The result also applies to the stated per-protocol analysis of patients who completed the study, not automatically to everyone enrolled or to other populations. Missing-data handling, reasons for non-completion, crossover sequence details, and the apparent inconsistency between the design descriptions are not reported or reconciled. The figures should thus be read narrowly: they characterize the sponsor’s reported paired comparison under the recorded analysis, rather than establishing an overall treatment benefit, lack of benefit, or safety conclusion.

15. Why This Trial Matters Statistically

This trial is statistically interesting because it uses a randomized crossover design to compare ventilator modes on a continuous physiologic endpoint. In general, a crossover structure can make within-participant contrasts informative by controlling for stable differences among participants. A paired t-test is consistent with analyzing linked measurements, and the registry reports the estimated mean difference, confidence interval, and p-value rather than only a significance label.

At the same time, the study illustrates why design and analysis descriptions need to align. The registry labels the design crossover, while an analysis note refers to cluster randomization and cautions against a simple individual-level re-analysis. The record does not explain this discrepancy or provide enough detail to reconstruct the analysis. Readers must therefore rely on the sponsor’s result and retain uncertainty about the exact design structure behind it.

The study also demonstrates the importance of analysis-population transparency. The registry records 26 enrolled participants but describes a per-protocol primary analysis using 19 patients who completed the study. Without reasons for exclusion or non-completion, and without missing-data methods, it is difficult to judge the reach of the estimate. Clear reporting of these details would help distinguish a within-completer contrast from the effect that might be estimated for the randomized population.

Finally, reporting the group coding, treatment sequence, period handling, and multiplicity plan would make the result easier to interpret. The primary estimate is numerically precise to the reported confidence interval, but its treatment direction and robustness cannot be fully assessed from the registry record. Those limitations make this a useful example of how a statistical result can be reported while leaving key interpretive details unresolved.

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 primary limitation for interpretation is that the coded analysis groups are not mapped to NAVA and SIMV+PSV. Although the mean difference and its confidence interval are reported, the sign cannot be connected to a named mode. Group-specific means and standard deviations are not reported in the registry record, so the contrast cannot be examined alongside the observed distributions.

The registry’s descriptions of study structure are not fully consistent: the design fields specify crossover, while an analysis note calls the design cluster-randomized and warns against a simple individual-level re-analysis. The record does not explain this discrepancy. Treatment sequence, period effects, washout, and carryover handling are not reported in the registry record, limiting assessment of whether the paired test captures the intended comparison.

The primary result uses a per-protocol population described as 19 patients who completed the study, while enrollment is recorded as 26. Reasons for non-completion or exclusion and methods for handling missing measurements are not reported in the registry record. The registry does not provide a corresponding intention-to-treat result or the criteria used to define the per-protocol population.

The registry lists multiple outcomes and analyses, but a multiplicity plan is not reported. It does not state whether the p-value threshold was prespecified for the primary analysis or whether the secondary outcomes were exploratory. It also does not report the original sample-size justification or whether interim analyses were conducted.

Finally, the outcome is a short-duration physiologic measure, and the record does not establish the longer-term clinical significance of the pressure contrast. Safety entries are limited to the listed event counts; deaths are null and therefore not reported. These limits do not negate the reported result, but they constrain the conclusions that can be drawn from it.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Randomized Crossover Study of Neurally Adjusted Ventilatory Assist (NAVA) in Preterm Infants — NCT01389882.
• Registry record last updated 2012-01-09; status last verified 2012-01.
• Results first posted 2011-10-28.
• 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

The registry describes a completed, randomized, single-masked crossover study of NAVA in preterm infants, with SIMV+PSV named as the comparator in the study rationale. It records 26 enrolled participants and a per-protocol primary analysis using 19 patients who completed the study. The primary endpoint is peak inspiratory pressure measured over four hours with each ventilator mode.

The sponsor reports a paired t-test with a mean difference of 1.007 cmH2O, a 95% confidence interval from 0.036 to 1.978, and p = 0.043. The coded groups are not mapped to the ventilator modes, so the direction cannot be assigned. The record also contains an unreconciled design-description caveat, and does not report key details on missing data, treatment sequence, multiplicity, or the original sample-size rationale. Serious and other adverse-event entries are zero in both listed groups; deaths are not reported.