Perioperative Treprostinil in Pediatric Patients Undergoing the Fontan Operation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Perioperative Treprostinil in Pediatric Patients Undergoing the Fontan Operation (NCT02498444), a randomized na study. The primary outcome is Chest Tube Duration.

2. Clinical Question

The registry record concerns perioperative treprostinil in pediatric patients with single-ventricle physiology undergoing a non-fenestrated, extracardiac Fontan operation. The clinical question is whether assignment to treprostinil, compared with saline, is associated with a different duration of postoperative chest tube drainage.

This is a question about a postoperative duration outcome, not a direct comparison of every possible measure of recovery. The registry also lists hospital length of stay and postoperative pressure measures as secondary outcomes. The primary endpoint captures how long drainage tubes remain in place, which the study description links to postoperative Fontan pressures and fluid management. That proposed connection is the investigators’ rationale, not proof that any observed difference in drainage duration is caused by a change in pressure or fluid status.

The posted results describe the groups’ chest tube duration and report a model-based comparison. They do not establish how the result generalizes beyond the enrolled trial population or to other Fontan procedures, patient groups, or treatment settings.

3. Trial Design

This was a randomized, parallel-group study with single masking. The registry lists Stanford University as the sponsor and identifies the study as phase NA. It records enrollment of 34 and a status of terminated. The start is listed as 2015-09 and primary completion as 2020-12. The record does not state why the trial was terminated.

The two assigned groups were treprostinil and saline. The posted arm counts are 16.0 for treprostinil and 18.0 for saline. The treprostinil arm received a continuous subcutaneous infusion, with the described escalation and intended subsequent taper. The saline arm received saline by subcutaneous infusion. The registry record does not describe the masking procedure in further detail, so the identity of the party or parties masked is not reported in the registry record.

The primary outcome was chest tube duration, measured as the number of days chest tubes remained in place postoperatively. The outcome timeframe is listed as 2-3 wks, and the registry specifies an intention-to-treat analysis based on study group assignment at randomization. Arm-level results are presented as medians with inter-quartile ranges.

The registry lists four secondary outcomes: length of hospital stay, postoperative Fontan pressure, postoperative atrial pressure, and postoperative transpulmonary gradient. The record lists five outcome measures and 11 analyses. The reported primary comparison is a Poisson regression result identified as a risk ratio. Further details of the analysis specification, including covariates and model implementation, are not reported in the registry record.

Random assignment is intended to support a comparison between assigned groups, but the record’s design label alone does not establish whether allocation was concealed or whether all randomized participants contributed complete outcome data. Those details are not reported in the registry record. The study’s terminated status and enrollment should be considered when judging how much precision the result offers; the reason for termination and any original recruitment target are not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intention to treat analysis based on study group assignment at randomization.

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 Chest Tube Duration, measured over 2-3 wks.

The primary endpoint is chest tube duration: the number of days that chest tubes were in place after surgery. Chest tubes are used to drain postoperative fluid, and the study description gives prolonged drainage as a concern in the perioperative Fontan setting. Duration is therefore a direct measure of how long this particular postoperative device remained in place.

The registry summarizes this endpoint with a median and an inter-quartile range for each arm. The treprostinil group’s median was 8.0 days, with an inter-quartile range from 7.0 to 12.5 days. The saline group’s median was 7.0 days, with an inter-quartile range from 5.0 to 8.0 days. The median describes the center of the observed duration distribution, while the inter-quartile range describes the interval containing the middle half of the observations. These summaries do not show every participant’s duration or the full shape of either distribution.

A duration endpoint can be useful when the clinical question concerns how long a postoperative process continues. It is not interchangeable with hospital length of stay, pressure measurements, or other outcomes. Nor does a difference in tube duration by itself identify why the durations differed. Decisions about when to remove a chest tube and the clinical circumstances surrounding removal can affect the recorded endpoint; the registry record does not describe those decision criteria.

The outcome timeframe is listed as 2-3 wks. The registry’s primary population is described as intention to treat, based on group assignment at randomization. Details about censoring, missing duration measurements, and any adjudication of tube removal are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a duration: days with chest tubes in place after surgery. The registry reports arm-level medians and inter-quartile ranges, and separately lists a Poisson regression analysis with a risk ratio. These summaries and the model result should be kept distinct because they express results in different ways.

What does the median describe?

The median is the central point of an ordered set of observations. It is less affected by unusually long or short observations than the arithmetic mean, which can make it a useful summary when durations are skewed. The reported medians are 8.0 days for treprostinil and 7.0 days for saline. Those figures describe the center of each observed arm distribution; they do not establish a treatment effect on their own.

The inter-quartile range describes the middle portion of the observed values, from the lower quartile to the upper quartile. For treprostinil, the reported range is 7.0 to 12.5 days; for saline, it is 5.0 to 8.0 days. It indicates spread among the observations but does not describe all values, identify outliers, or provide a confidence interval for the median. A confidence interval for a statistic and an inter-quartile range are different quantities: the former expresses uncertainty about an estimate, while the latter summarizes variation in the observed data.

What does the regression result say?

The registry identifies the sponsor’s comparison as Poisson regression and labels its parameter a risk ratio. The reported estimate is 1.29, with a 95% confidence interval from 1.02 to 1.64 and p = 0.03. A ratio compares quantities on a multiplicative scale, rather than expressing an absolute change in days. The registry record does not give enough information about outcome coding, exposure time, covariates, or reference-group convention to describe the ratio more specifically or to reproduce the model.

Poisson regression is commonly used for count outcomes or rates. In general, a Poisson model relates an outcome to predictors under assumptions about its distribution and variance. When the observed variability exceeds what the basic model allows, over-dispersion can make standard errors and confidence intervals too optimistic if it is not addressed. This is a general statistical consideration, not evidence that over-dispersion occurred in this trial. The registry record does not report a dispersion assessment, an alternative variance method, or diagnostic results.

The listed endpoint is a duration, while the listed sponsor model is Poisson regression. The record does not explain how the duration was represented for that model or whether any time-at-risk adjustment was used. It is therefore not appropriate to assume a particular implementation or to recast the reported ratio as a hazard ratio. The registry does not report a hazard ratio for this comparison.

How should the confidence interval and p-value be read?

A confidence interval expresses the uncertainty around an estimate under the statistical model and its assumptions. The reported 95% interval, from 1.02 to 1.64, gives a range of model-compatible values; it is not a range containing 95% of individual participants’ outcomes. The interval’s interpretation depends on the model being appropriate and the analysis being performed as specified.

A p-value measures how unusual a result at least as inconsistent with a null hypothesis would be if that null hypothesis and the model assumptions held. The reported p = 0.03 is not the probability that the null hypothesis is true, and it does not measure the probability that the result is due to chance in a general, assumption-free sense. A p-value also does not express clinical importance. Its interpretation depends on which contrast and analysis were specified, details not fully reported in the registry record.

Why are no reconstructed contrast numbers given?

Arm medians and inter-quartile ranges do not contain enough information to reproduce a Poisson regression estimate, its standard error, or its p-value. Individual-level observations and details of the fitted model would be needed for a direct reproduction. The registry record does not report a recalculated two-arm estimate for the median-time endpoint; it indicates that values were missing or not estimable for that comparison. The sponsor’s reported model result is therefore the inferential comparison available in the record, while the medians remain descriptive summaries.

The registry describes the primary population as intention to treat, based on assignment at randomization. In general, an intention-to-treat analysis preserves participants’ randomized group assignment for analysis, regardless of treatment received. That approach can help retain the benefits of randomization, but its implementation still depends on how missing outcomes and protocol deviations are handled. Those details are not reported in the registry record.

10. Analyses Reported by the Sponsor

The sponsor-reported result is a Poisson regression risk ratio of 1.29, with a 95% confidence interval from 1.02 to 1.64 and p = 0.03. The registry record does not indicate a discrepancy between the sponsor’s analysis and a recalculated result. The arm-level medians and inter-quartile ranges are descriptive summaries, not a substitute for that regression analysis.

Reconstructing the regression result from the posted summaries is not possible: medians and inter-quartile ranges do not retain the participant-level information or model details needed to reproduce a Poisson regression. The registry record does not report a recalculated numerical contrast for the median-time endpoint, noting that values were missing or not estimable for a two-arm comparison. No alternate model estimate should be inferred from the arm summaries.

The ratio’s clinical direction cannot be stated confidently from the record because its contrast coding and reference convention are not reported. The model estimate should be retained in the registry’s stated form and not translated into a difference in days. Its interpretation is also limited by the absence of further details about model specification, dispersion, missing-data handling, and covariate adjustment. The posted sponsor analysis is the available inferential result, while its assumptions and reporting limitations remain relevant to how precisely it can be interpreted.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures, four secondary outcomes, and 11 analyses. These counts indicate that the record includes multiple outcomes and analyses, but they do not reveal which analyses were prespecified as primary, which were exploratory, or how the investigators treated multiple testing. A multiplicity adjustment or a strategy for controlling the chance of false-positive findings is not reported in the registry record.

When several outcomes or analyses are examined, the chance of at least one apparently unusual result can rise if each is assessed separately without accounting for the set of tests. This is a general statistical issue, not a claim that the reported primary result is false. Without the analysis plan and a clear account of which comparisons were planned, the p-value should be read as applying to the reported model comparison, not as a complete summary of evidence across every outcome and analysis.

The primary outcome population is described as intention to treat, based on study group assignment at randomization. This states the intended basis for group classification. It does not explain how participants with unavailable chest tube duration data were handled, whether any values were imputed, or whether exclusions occurred. Missing-data counts and methods are not reported in the registry record. The record also does not report the number of participants contributing data to each outcome beyond the arm-level counts.

The registry’s terminated status is documented, but the reason for termination and its relation to recruitment or outcome availability are not reported in the registry record. Interim analyses, stopping rules, and any changes to the analysis plan are also not reported in the registry record. Without that information, it is not possible to assess whether the reported testing was affected by repeated examination of accumulating data or by changes made during the study.

These omissions do not prove that analyses were changed or that data were missing. They mean that the record does not document the information needed to evaluate those possibilities. The reported estimates can still be described accurately, while their interpretation should remain conditional on analysis and data-handling details that are not available in the registry record.

12. 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 non-serious adverse event terms

The registry reports serious adverse events for treprostinil in 0 of 16 participants at risk and for saline in 0 of 18 participants at risk. Deaths are reported as 0 in each group. The listed other adverse events affected 7 of 16 participants at risk in the treprostinil group and 0 of 18 in the saline group. No serious adverse event terms are listed.

These figures describe events recorded in the registry, not a complete characterization of safety. A zero recorded serious-event count means none were reported in the corresponding group in this record; it does not establish that the risk is absent. The number of participants and the observation conditions limit how much can be concluded about uncommon events. The registry record does not provide event descriptions for the other adverse events, their timing, severity, relatedness, duration, or the criteria used to collect them.

The adverse-event counts should not be treated as directly comparable measures of treatment effect without information about follow-up, ascertainment, and analysis. The record also does not explain whether the listed events represent participants with at least one event or event episodes, beyond the field’s description as affected participants. Safety interpretation is therefore descriptive and restricted to the counts reported.

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

The registry results establish that the recorded arm medians for chest tube duration were 8.0 days for treprostinil and 7.0 days for saline, with the reported inter-quartile ranges. They also establish that the sponsor reported a Poisson regression risk ratio of 1.29, with a 95% confidence interval from 1.02 to 1.64 and p = 0.03. These are the observed summaries and model result in the registry record.

The results do not, on their own, show how many days of chest tube duration differ between groups. The ratio is not a median difference, and the registry does not report sufficient coding information to state which group corresponds to the numerator. The result also does not establish the cause of any observed difference or whether changes in postoperative pressure, fluid management, or another factor account for the outcome.

A confidence interval and p-value summarize uncertainty conditional on an analysis and its assumptions. They do not address every source of uncertainty, including data completeness, endpoint procedures, analysis choices, or the possibility that the trial results differ from results in other settings. The registry does not report enough information to evaluate all these issues.

The record describes a randomized, single-masked, parallel-group study, but the trial was terminated. It does not state why, whether termination was planned, or whether the enrollment target was met. These omissions make it difficult to assess how the trial’s completion affected precision and interpretation. Enrollment of 34 and the reported arm counts should be considered as context, not as evidence that every relevant subgroup or uncommon outcome was adequately represented.

The posted result is specific to perioperative treprostinil versus subcutaneous saline in the study population described in the registry. It should not be generalized automatically to different patients, procedures, dosing approaches, or care settings. The registry record supports a careful description of the reported data; it does not support medical advice or a broad conclusion about treatment beyond this trial.

14. Why This Trial Matters Statistically

This trial raises a useful statistical problem: the clinical endpoint is a duration, the registry summarizes it with medians and inter-quartile ranges, and the sponsor reports a Poisson regression risk ratio. These quantities cannot be substituted for one another. Readers need to distinguish the descriptive center of each group’s outcome distribution from a model-based comparison and to check whether the model’s scale answers the clinical question of interest.

That distinction matters especially when the registry does not explain how the duration was represented in the model or how the ratio was coded. A model estimate can be reported precisely as written while still being difficult to translate into a clinically intuitive statement. The medians are easier to describe, but they do not reproduce the regression analysis or establish its uncertainty. Statistical interpretation therefore requires both accurate reporting and restraint about what each result means.

The study also illustrates why a statistically unusual result is not the same as a fully settled conclusion. The reported p-value and confidence interval describe uncertainty under the sponsor’s model. They do not resolve unanswered questions about multiplicity, missing outcomes, analysis specification, or trial termination. These are not reasons to discard the result; they are limits on what can be inferred from the registry record.

Finally, the registry connects chest tube duration to postoperative pressures and fluid management in its study rationale, while listing those pressure measures and hospital length of stay as secondary outcomes. Keeping primary and secondary outcomes conceptually separate helps prevent an inference about one endpoint from being treated as proof about another. The trial is statistically instructive because it requires careful attention to outcome scale, model interpretation, uncertainty, and the boundary between observed results and proposed explanations.

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.

The registry record does not describe the reason the trial was terminated, whether its enrollment target was reached, or whether termination affected outcome collection. Enrollment is listed as 34, with posted arm counts of 16.0 for treprostinil and 18.0 for saline. A planned sample-size justification is not reported in the registry record, so the enrollment cannot be evaluated against a documented target or power calculation.

Although the primary population is identified as intention to treat, the record does not state how missing primary outcome data were handled, whether all randomized participants had an observed duration, or whether additional analysis populations were used. It also does not report individual observations or enough detail to reproduce the model. The analysis description does not specify covariates, link or offset choices, dispersion handling, or the reference convention for the reported ratio.

The primary endpoint is a postoperative duration, but the record does not explain the criteria for chest tube removal, how the timing was recorded, or how censoring was handled. The stated timeframe is 2-3 wks, but further follow-up details are not reported in the registry record. Such details matter because the meaning and completeness of a duration outcome depend on how it is observed.

The registry identifies the study as single-masked but does not state who was masked or how masking was maintained. Allocation concealment and the specific randomization procedure are not reported in the registry record. Multiplicity procedures, interim analyses, and changes to the statistical plan are also not reported. These are limits on assessment of the evidence, not evidence that a particular procedure was or was not used.

The safety record provides counts but not detailed event descriptions, timing, relatedness, or follow-up conditions. The record therefore supports a narrow reading of the reported safety figures rather than a comprehensive comparison of adverse-event profiles. Generalization is also limited to the pediatric, non-fenestrated, extracardiac Fontan population described; the registry does not establish applicability to other procedures or populations.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Perioperative Treprostinil in Pediatric Patients Undergoing the Fontan Operation — NCT02498444.
• Registry record last updated 2022-03-31; status last verified 2022-03.
• Results first posted 2022-03-31.
• 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

NCT02498444 is a randomized, single-masked, parallel-group trial of perioperative treprostinil versus subcutaneous saline in pediatric patients undergoing a non-fenestrated, extracardiac Fontan operation. The registry lists enrollment of 34, status terminated, and Stanford University as sponsor. The primary outcome is postoperative chest tube duration, measured in days, with an intention-to-treat population based on assignment at randomization.

The reported median duration was 8.0 days for treprostinil and 7.0 days for saline. The sponsor-reported Poisson regression result was a risk ratio of 1.29, with a 95% confidence interval from 1.02 to 1.64 and p = 0.03. The record does not provide enough model detail to translate that ratio into a difference in median days or confidently state its direction by group.

The registry lists four secondary outcomes and reports adverse-event counts, including no serious adverse events or deaths in either arm. Reasons for trial termination, missing-data handling, multiplicity procedures, and further analysis details are not reported in the registry record.