Newborn Ventilation in the Delivery Room: Could it be Improved With a T-piece Resuscitator?: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Newborn Ventilation in the Delivery Room: Could it be Improved With a T-piece Resuscitator? (NCT00443118), a randomized phase 3 study. The primary outcome is Proportion of Infants With a HR ≥ 100 Bpm at 2 Minutes of Life..

2. Clinical Question

The clinical question concerns the early heart-rate response during newborn resuscitation. The registry identifies three ventilation approaches, so the comparison could involve the T-piece and either of the self-inflating bag approaches. However, the primary endpoint description does not state which specific pairwise contrasts were used for the primary analysis, or whether the analysis tested an overall difference among all arms. Those details are not reported in the registry record.

The record also lists secondary outcomes related to the time to reach a heart rate above 100 beats per minute, oxygen saturation at 2 minutes of life, intubation after failure of mask ventilation, chest compressions or medications, Apgar scores, neonatal encephalopathy, air leaks, and oxygen treatment beyond the delivery room. These outcomes broaden the clinical questions beyond the early heart-rate endpoint. Their results and their role in confirmatory or exploratory inference are not reported in the registry record.

3. Trial Design

The registry describes a multicenter, cross-over, cluster randomized controlled trial in newborn infants requiring assisted ventilation for delivery-room resuscitation. It records the trial as randomized and unmasked. The registry also classifies the model as parallel. The cross-over cluster description and the parallel model classification are not fully aligned; the registry record does not explain how these labels relate to the trial’s implementation.

The listed enrollment is 1032. This is the recorded enrollment, not a statement about how many infants contributed to each endpoint analysis. The population described in the protocol summary is newborn infants at least 26 weeks of gestational age who need positive pressure ventilation. The trial is classified as Phase 3, and its status is completed. The recorded start is 2009-12 and the primary completion is 2012-08.

The intervention arms are Neopuff with PEEP, a self-inflating bag with PEEP, and a self-inflating bag without PEEP. The record names the T-piece resuscitator Neopuff TM and lists both self-inflating bag approaches among the interventions. It does not report how clusters were defined, how allocation was generated or concealed, how cross-over order was assigned, or how periods and possible carry-over effects were handled. These details matter for understanding which units were randomized and how the comparisons were constructed.

The registry identifies intention to treat as the primary endpoint population. It does not report the analysis model, adjustment variables, clustering correction, or handling of observations across cross-over periods. Those omissions limit the detail available for evaluating the statistical design. The sponsor is recorded as Edgardo Szyld.

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: The analysis was performed by intention to treat

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 Proportion of Infants With a HR ≥ 100 Bpm at 2 Minutes of Life., measured over 2 minutes of life.

The primary endpoint is the proportion of infants with a heart rate of at least 100 beats per minute at 2 minutes of life. It is a binary outcome: each infant either meets the stated threshold at the specified assessment time or does not. Reporting it as a proportion describes how often the threshold was reached within an analysis group.

Heart rate is an early measure of a newborn’s response during resuscitation. A threshold-based endpoint gives a clinically interpretable description at a defined time point, rather than tracking every change in heart rate. The specified assessment time also makes the endpoint a snapshot: it does not, by itself, describe the timing or course of a response before or after that point.

The registry states that the primary analysis was performed by intention to treat. This generally means that participants are analyzed according to their assigned group, regardless of treatment received, although the precise implementation and any exceptions are not reported in the registry record. The record does not provide the number or proportion meeting the endpoint in each arm, an effect estimate, a confidence interval, or a p value. Therefore, the primary endpoint definition is available, but its comparative result and precision are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

For a binary endpoint such as reaching a heart-rate threshold by a specified time, a basic descriptive result is the proportion of participants meeting the endpoint in each group. A comparison can then be expressed using an effect measure such as a difference in proportions or a relative measure. The choice of effect measure affects interpretation: an absolute difference describes the separation in percentage points, while a relative measure describes the comparison in relation to one group’s value. The registry record does not report a primary effect estimate, so no such comparison can be interpreted for this trial.

Why does cluster randomization matter?

In cluster randomization, groups of participants rather than individual participants are assigned to study conditions. Participants within a cluster may share caregivers, procedures, or other features and can therefore have correlated outcomes. This correlation reduces the amount of independent information relative to an analysis that treats every observation as unrelated. A statistical analysis generally needs to account for clustering, for example through a model or variance method that reflects the assignment unit and within-cluster similarity. The registry record calls this trial cluster randomized but does not report the cluster definition, the number of clusters, the correlation structure, or the method used to account for clustering.

Ignoring clustering can make uncertainty appear smaller than it is, because the analysis may count correlated observations as if they provided fully independent information. That is why a simple individual-level re-analysis can be misleading for a cluster-randomized study. The registry’s notes indicate that only the sponsor’s own analysis is shown for this design, but the sponsor’s estimates and the analysis details are not reported in the registry record.

What does a cross-over design add?

In a cross-over design, a unit or cluster receives different conditions in different periods, according to an assigned sequence. The comparison may then depend on both the intervention and the period in which it was used. A suitable analysis generally needs to account for this structure and consider whether outcomes could carry over from one period into another. The registry describes the study as cross-over cluster randomized, but does not report the sequences, period definitions, washout arrangements, or analysis method. It is therefore not possible to explain the exact contrast or determine how period effects were handled.

The registry separately classifies the model as parallel. Because the record does not reconcile that classification with its cross-over cluster description, the design structure cannot be reconstructed fully from the available details. This is not just a naming issue: the assignment unit and sequence structure determine which observations are compared and how their dependence should be represented.

How should uncertainty be understood?

A confidence interval is a range of effect values compatible with the observed data under the specified statistical method. It conveys precision as well as direction: a narrow interval indicates more precise estimation, while a wide interval indicates greater uncertainty. A p value addresses compatibility with a null hypothesis under a chosen model and assumptions; it does not measure clinical importance or the probability that a treatment effect is true. Neither confidence intervals nor p values for the primary endpoint are reported in the registry record, so the statistical uncertainty of the primary comparison cannot be evaluated.

For clustered data, uncertainty estimates need to reflect the cluster structure. For cross-over data, they also need to reflect repeated periods or other dependence created by the design. If those dependencies are omitted, standard errors and confidence intervals may not have their intended interpretation. The specific approach used in this trial is not reported in the registry record.

What do the planning assumptions describe?

The registry’s planning illustration is described as using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It is based on the observed effect and considers detection of a stated fraction of that effect. The record specifies that time-to-event planning gives the total number of events, while other planning results are expressed as participants per arm. Such calculations are planning illustrations, not observed results and not observed power. They do not show that the trial had a particular probability of detecting an effect. The original design’s sample-size justification and its assumptions are not reported in the registry record.

These illustrations also require caution because the trial is described as cluster randomized and cross-over, while the listed planning assumptions include equal allocation between two arms. The record does not explain how the illustrations accommodate the cluster structure, the cross-over design, or the three listed intervention arms. They should not be treated as a reconstruction of the trial’s design calculation.

Why can over-dispersion matter?

In general, over-dispersion occurs when the observed variability in data is greater than a statistical model expects. For clustered outcomes, unmodeled similarity within clusters can contribute to underestimating variability. If a model assumes less variation than is present, its uncertainty estimates may be too small. Appropriate modeling or robust variance methods can address some forms of this problem, depending on the design and data. The registry record does not report whether over-dispersion was assessed or how variability was modeled.

These concepts clarify what a complete analysis would need to make transparent: the comparison being estimated, the handling of clustering and cross-over periods, the analysis population, the treatment of missing outcomes, and the uncertainty around estimates. The registry identifies an intention-to-treat primary analysis but does not report these implementation details or its primary results.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("T-Piece", "SIB - Self Inflating Bag With PEEP Valve", "SIB-Without PEEP Valve")
n      <- c(511, 290, 226)
events <- c(479, 255, 210)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  T-Piece  511  479  93.7  91.3  95.7  91.3  95.5  91.3  95.5
ARM  SIB - Self Inflating Bag With PEEP Valve  290  255  87.9  83.6  91.4  83.7  91.2  83.7  91.2
ARM  SIB-Without PEEP Valve  226  210  92.9  88.8  95.9  88.8  95.6  88.8  95.6

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and multiple secondary outcomes, and records 14 outcome measures. When several outcomes or comparisons are tested, the chance of obtaining at least one apparently notable result by chance can increase. Statistical plans may address this through a prespecified hierarchy, adjustment, or clear distinction between confirmatory and exploratory outcomes. The method used for multiplicity, and the intended hierarchy among the listed endpoints, are not reported in the registry record. With no reported outcome estimates or p values, the consequences of multiplicity cannot be assessed for this trial.

The primary analysis population is described as intention to treat. In general, this approach seeks to preserve the comparison created by random assignment by analyzing participants in their assigned groups. The registry does not explain the detailed population rules, such as how withdrawals, eligibility deviations, or unavailable endpoint measurements were handled. Those implementation details are not reported in the registry record.

Missing-data handling is also not reported in the registry record. The record does not state how many primary endpoint measurements were missing, whether missingness differed by group, or whether any imputation or sensitivity analysis was used. Missing outcomes can affect estimates if their occurrence is related to participant characteristics or outcomes, so the handling method is important for judging robustness.

No analysis records are listed in the registry. This should not be interpreted as evidence that analyses were not performed; it means that their methods and results are not available in the registry record. In particular, no multiplicity-adjusted results, missing-data sensitivity analyses, or detailed analysis-population results are reported.

12. 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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: mortality (3.1%).

The serious adverse-event information is listed by group as events and numbers at risk: T-Piece, 11 events among 511 at risk; self-inflating bag with PEEP, 8 among 290 at risk; and self-inflating bag without PEEP, 7 among 226 at risk. These figures describe the registry’s recorded serious-event counts. They do not, on their own, provide a fully adjusted comparison between devices or quantify uncertainty around any difference.

The registry also lists mortality at 3.1% among its top serious-event terms. The denominator for that percentage is not reported in the registry record, and the group-level death counts are not reported. The percentage therefore should not be assigned to a particular arm or treated as a comparison among arms based on the listed information.

Safety comparisons depend on how events were defined, collected, and followed, as well as on the analysis method. Those details, along with confidence intervals or tests comparing groups, are not reported in the registry record. The listed counts are descriptive registry information and should be interpreted with those limits in mind.

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

The record establishes that a randomized trial was conducted in newborn infants requiring assisted ventilation during delivery-room resuscitation, and it identifies the three ventilation approaches and the primary endpoint. It also states that the primary endpoint analysis was intended to use an intention-to-treat population. These details define the clinical question and offer context for interpreting any results when available.

The record does not report the primary endpoint values, a between-group effect estimate, or its uncertainty. It therefore does not establish which approach had a higher or lower proportion of infants reaching the heart-rate threshold at the specified time, nor how precise any comparison was. The absence of a reported result is not evidence that the approaches were equivalent or that one was superior.

The secondary outcome list describes additional aspects of resuscitation and newborn outcomes, but their results are not reported in the registry record. The adverse-event counts offer limited descriptive safety information, without a reported comparative analysis or enough detail to interpret the mortality percentage by group. Those safety entries cannot substitute for a complete primary-outcome report.

General statistical interpretation also depends on whether the analysis reflected the cluster and cross-over features described in the registry. The record’s separate parallel-model classification is not reconciled with those design labels, and the analysis method is not reported. Consequently, the statistical handling of dependence, periods, and uncertainty cannot be evaluated from the registry record.

In short, the registry supports a description of the question, trial classification, and some recorded safety counts. It does not provide the outcome results needed to characterize comparative effectiveness or precision, and it does not support a conclusion about whether any ventilation approach performed better.

14. Why This Trial Matters Statistically

This trial is statistically notable because its registry description combines cluster randomization with a cross-over design. Those features affect the structure of the data: assignment may occur at a cluster level, while conditions may change across periods. Analyses need to reflect the dependence created by both features. Treating all infant observations as independent can understate uncertainty, while ignoring period structure may obscure how the comparison was formed.

The record also contains distinct design labels: it describes the trial as cross-over cluster randomized and classifies the model as parallel. This unresolved difference matters when readers try to identify the assignment unit, treatment sequence, and relevant comparison. Those are not details that can be safely inferred from the endpoint definition or the recorded enrollment.

The endpoint is a binary threshold outcome at a fixed time point. That makes the measure straightforward to describe, but statistical interpretation still requires group-specific results, an effect estimate, and an uncertainty interval. None is reported for the primary endpoint. The list of secondary outcomes further raises the importance of knowing how comparisons were prioritized and whether multiplicity was addressed; that information is also not reported.

The case illustrates a broader principle: a registry can identify a research question without providing enough information to judge its estimate. Design labels, analysis population, missing-data methods, and uncertainty estimates all contribute to interpretation. Here, the primary outcome results and the methods needed to evaluate the clustered cross-over analysis are not reported in the registry record.

15. 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.

Important design and analysis details are not reported in the registry record. The trial is described as cross-over cluster randomized, yet its model is classified as parallel; the record does not explain the relationship between these descriptions. It also does not identify the randomized cluster unit, allocation sequence, period structure, or method for accounting for clustering and repeated periods. These omissions make it difficult to reconstruct the design and assess whether the statistical analysis matched it.

The primary endpoint result is not reported. There are no group-specific endpoint proportions, comparative estimates, confidence intervals, or p values in the registry record. Results for the listed secondary outcomes are also not reported. This prevents assessment of the observed effect, precision, or consistency across outcomes.

The record states that the primary analysis was by intention to treat, but detailed population rules are not reported. Missing-data amounts and handling, multiplicity procedures, and sensitivity analyses are also not reported. The registry lists adverse-event counts, but does not provide a comparative safety analysis, group-level death counts, or the denominator for the listed mortality percentage.

The planning assumptions are described as an illustration based on the observed effect, rather than the trial’s own design justification. The original sample-size rationale is not reported in the registry record, and the relationship between the illustration and the cluster, cross-over, three-arm design is not explained. These limitations mean the available registry information is insufficient for a full appraisal of the trial’s statistical findings.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Newborn Ventilation in the Delivery Room: Could it be Improved With a T-piece Resuscitator? — NCT00443118.
• Registry record last updated 2018-09-28; status last verified 2018-08.
• Results first posted 2015-03-31.
• 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.

19. Record Summary

The registry describes a completed, randomized, unmasked, multicenter trial in newborn infants at least 26 weeks of gestational age who required assisted ventilation for delivery-room resuscitation. The listed enrollment is 1032. The record describes a cross-over cluster randomized controlled trial but also classifies the model as parallel, without explaining the difference.

The three ventilation approaches are Neopuff with PEEP, a self-inflating bag with PEEP, and a self-inflating bag without PEEP. The primary endpoint is the proportion of infants with a heart rate of at least 100 beats per minute at 2 minutes of life, analyzed by intention to treat. Primary and secondary outcome results are not reported in the registry record. Serious adverse-event counts are listed by group, but comparative safety estimates and group-level death counts are not reported.