California Transport Cooling Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for California Transport Cooling Trial (NCT01683383), a randomized na study. The primary outcome is Percentage of Temperatures in Target Range During Transport.

2. Clinical Question

The trial addresses a practical transport-care question: whether a temperature-regulating device can help maintain target temperatures while newborns with HIE are being moved to a cooling center. Its primary outcome concerns temperature control during transport, not later survival, disability, or neurodevelopment. The registry describes therapeutic hypothermia as an intervention of interest for HIE and states that the goal was to test the device’s ability to achieve target temperatures compared with current practices.

The record does not establish that the groups were comparable on every baseline feature, nor does it report detailed eligibility criteria or the age of participants when cooling began. Those details are not reported in the registry record. The results should therefore be read as a comparison of the posted transport-temperature outcomes, rather than as evidence about all possible clinical outcomes or all newborns with HIE.

3. Trial Design

The California Transport Cooling Trial (NCT01683383) was a randomized, parallel-group clinical trial with no masking. Stanford University is listed as the sponsor. The registry records enrollment of 101 and a completed status. The recorded start was 2012-09 and primary completion was 2013-10. The phase is listed as NA.

One group received device-regulated cooling, using a cooling blanket connected to the Tecotherm Neo and continuous rectal temperature monitoring with servo-regulation. The comparator group received passive or active cooling according to center practice, with rectal temperature recorded every 15 minutes. The registry reports outcome denominators of 49 for standard cooling and 51 for device-regulated cooling for the primary endpoint. Other outcomes have their own posted denominators, which differ by outcome.

The primary outcome was the percentage of temperatures in the 33°-34°C target range during transport after cooling initiation by the transport team. The stated observation period was transport from the birth hospital to the cooling center, expected to average 4 hours. Secondary outcomes concerned time to target temperature, target-range status at 1 hour, and target-range attainment at any time during transport.

Randomization helps reduce systematic differences between groups, on average, but does not ensure identical groups in a trial of this size. Because masking was absent, staff knew which cooling approach was being used. That matters when considering implementation and measurements, particularly because the comparator followed center practice while the device arm used continuous monitoring and automated regulation. The registry record does not report the randomization procedure, allocation concealment, detailed eligibility criteria, or a protocol for handling deviations from assigned cooling.

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: Infants were excluded if they had a lethal congenital or chromosomal anomaly, the decision had been made to not provide full intensive care, or there was refusal of consent. Of 131 infants screened for eligibility, 30 were excluded, and 101 consented and were randomized.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Total patients in each study arm

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.

Population: Eligible newborns

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Temperatures in Target Range During Transport, measured over Participants will be followed for the duration of neonatal transport from the birth hospital to the cooling center, an expected average of 4 hours.

The primary endpoint measures the share of temperature measurements recorded during transport that fell within the target range of 33°-34°C after the transport team initiated cooling. It is a measure of temperature control over the transport period, rather than a direct measure of a newborn’s later clinical condition. The registry specifies the population as all patients in each study arm and describes transport from the birth hospital to the cooling center as expected to average 4 hours.

Using the percentage of temperatures in range can reflect both whether the target is reached and how consistently temperatures remain within it during observation. This makes the endpoint relevant to the stated aim of testing whether the device can achieve target temperatures during transport. It also differs from an endpoint that records only a single temperature at one time point: a series of measurements can capture temperature control across the transport period.

The registry reports the primary endpoint as a median with an inter-quartile range. The median describes the midpoint of the observed distribution, while the inter-quartile range describes the spread of its middle portion. In standard cooling, the posted median was 0 with an inter-quartile range of 0 to 52; in device-regulated cooling, the median was 73 with an inter-quartile range of 17 to 88. These summaries describe the temperature measurements recorded in each arm. They do not, by themselves, quantify the uncertainty around a between-group difference, and the registry does not report a confidence interval for that difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a randomized, parallel-group comparison with no masking. Randomization is intended to make treatment assignment independent of participants’ prognostic characteristics, reducing systematic confounding on average. It does not guarantee that every prognostic feature is balanced, especially in a trial with a limited number of participants. A parallel design assigns participants to one approach for the comparison, rather than having each participant receive both approaches. Masking was absent, so caregivers and staff could know which cooling approach was in use.

What does the primary summary describe?

The primary outcome is expressed as a percentage of temperatures in the target range, and the registry reports the median and inter-quartile range for each arm. The median is a location summary that is less sensitive than the mean to extreme observations. The inter-quartile range describes the spread from the lower to the upper quartile. These summaries help show the typical observed value and the variability in the data, but they do not provide a confidence interval for the difference between arms.

The reported primary test is described as adjusted Wald estimation, with p < 0.001. The record does not specify the adjustment or provide the corresponding estimate and uncertainty interval. In general, a test p-value measures how incompatible the data are with a specified null model, conditional on the test’s assumptions. It is not an effect-size measure and does not convey the range of differences compatible with the data. Without the contrast and its interval, the registry record does not permit a full assessment of the precision of the primary between-group estimate.

How do the secondary analyses differ?

For time to target temperature, the registry reports means and standard deviations by arm. A mean summarizes the arithmetic average, while a standard deviation describes spread around that average. The recalculated comparison uses the posted group means and standard deviations to form a simple, unadjusted mean difference and confidence interval. This is a summary-statistic re-calculation, not the trial’s own analysis. The sponsor’s reported analysis uses the Wilcoxon (Mann-Whitney) method, which is rank-based and does not test exactly the same mean-difference quantity. The reported p-value is 0.04; the unadjusted re-calculation gives p = 0.005. Because these analyses address different summaries and use different methods, their p-values need not match.

For binary outcomes, a risk difference expresses the difference between the observed proportions in the two arms. A confidence interval gives a range of values compatible with the data under the method’s assumptions. The recalculated binary comparisons use the posted counts and denominators and are unadjusted. They are not adjusted for participant characteristics, site, or other factors. They also do not replace a prespecified analysis by the trial investigators. The registry does not report sponsor analyses for the outcome at 1 hour; for attainment at any time during transport it reports a Wilcoxon (Mann-Whitney) test.

General statistical point: confidence intervals communicate uncertainty about an estimate, not certainty that the true value lies within the interval. Their width depends on the amount and variability of information in the data and on the method used. Recalculated intervals from summary statistics may not reproduce a trial analysis that used participant-level data, covariate adjustment, a different test, or a different handling of missing observations. No such additional details are reported for these comparisons.

Why can over-dispersion matter?

General statistical point: when observations vary more than a model expects, standard errors can be too small and confidence intervals too narrow if the model does not account for that extra variation. For an endpoint based on repeated temperatures, measurements from the same participant are not necessarily independent. The primary record does not describe how within-participant dependence was handled. It is therefore not possible to assess from the registry record whether the analysis accounted for the repeated-measure structure or how that structure affected uncertainty.

What do the planning illustrations mean?

The sample-size illustrations are planning calculations based on observed effects, not observed power and not the trial’s own design justification. The record states assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation; the target effect is a stated fraction of the observed effect. The time-to-event convention does not apply to the listed outcomes; the record says those tables give participants per arm for the other outcomes. Such calculations are sensitive to the effect chosen: planning for a smaller fraction of the observed effect requires more participants per arm than planning for the full observed effect. Because the calculations use effects observed in this trial, they should be treated as illustrations rather than reliable forecasts of a future study’s results. The original sample-size justification is not reported in the registry record.

The record does not describe interim analyses, a prespecified statistical analysis plan, or adjustment for multiple testing. These omissions prevent a complete evaluation of whether the posted analyses followed a planned testing hierarchy or how the investigators addressed repeated looks at the data or multiple outcomes.

10. Analyses Reported by the Sponsor

For the primary endpoint, the registry reports an adjusted Wald estimation analysis with p < 0.001. It does not provide the sponsor’s estimated between-arm contrast or its confidence interval. The arm-level medians and inter-quartile ranges give descriptive information, but they are not a substitute for an estimated contrast with uncertainty. There is no primary recalculated contrast presented alongside the sponsor’s result, so there is no recalculation to compare with the sponsor’s primary analysis.

For time to target temperature, the sponsor reports a Wilcoxon (Mann-Whitney) test with p = 0.04. The simple recalculation uses the posted means and standard deviations to estimate a mean difference and gives p = 0.005. The methods summarize the data differently: a rank-based test is not a direct test of the mean difference. The p-values therefore should not be read as competing estimates of the same quantity. Since the record indicates no discrepancy, the sponsor’s own analysis takes precedence over the recalculation.

For attainment at any time during transport, the sponsor reports p < 0.001 using Wilcoxon (Mann-Whitney), and the unadjusted binary recalculation also gives p < 0.001. The registry does not list a sponsor analysis for the 1-hour endpoint. Recalculations are useful for checking what follows from posted summary statistics, but they cannot reproduce analyses based on participant-level data or undocumented choices. They should remain secondary to the trial’s own reported analysis.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary temperature-range endpoint and three secondary outcomes. It reports p-values for the primary analysis and for two secondary analyses, while no sponsor analysis is listed for the 1-hour outcome. The record does not report whether the investigators adjusted for multiple comparisons, prespecified a hierarchy among outcomes, or designated any secondary outcome as confirmatory. Without that information, the secondary p-values should be read as unadjusted evidence rather than as a family of results with controlled overall error.

Multiplicity matters because examining multiple outcomes can increase the probability of at least one result appearing statistically notable by chance. This general concern does not invalidate the results, but it affects how strongly secondary findings can be used to support a claim. The record also does not report interim analyses or whether any adjustments were made for repeated analyses over time.

Outcome denominators are not the same across the posted results. The primary and transport-wide outcomes list denominators of 49 and 51 by arm, while the 1-hour outcome lists 41 and 39. The registry does not explain the difference. It does not report how missing temperature measurements or unavailable outcome data were handled, whether participants were analyzed according to randomized assignment, or whether any sensitivity analyses addressed missingness. The reasons for missing or unobserved data, and whether they differ by arm, are also not reported in the registry record.

Analysis populations matter because results can change depending on whether all randomized participants are retained in their assigned groups, whether participants with protocol deviations are excluded, or whether only those with complete outcome data are analyzed. The registry identifies the primary population as total patients in each study arm, but does not provide enough detail to establish the exact analysis population or the handling of exclusions. These omissions limit assessment of attrition bias and the robustness of the reported comparisons.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: death (17.6%), pulmonary hemorrhage (2.0%).

The registry reports serious adverse events for 6 of 49 participants in the standard-cooling group and 10 of 51 in the device-regulated cooling group. It also lists other adverse events for 0 of 49 in standard cooling and 2 of 51 in device-regulated cooling. These are posted group counts; they do not establish that the cooling approach caused an event or that a difference between groups is statistically reliable.

The record’s list of top serious-event terms includes Death at 17.6% and Pulmonary hemorrhage at 2.0%. It does not make clear in this summary how those terms map to groups, participants, or event counts. The group-level death fields are not reported, and the registry does not describe the event definitions, severity criteria, follow-up duration for safety, or a comparative statistical analysis. The listed terms should therefore be read as a limited summary, not as a complete account of safety or a group-specific estimate.

Counts of serious or other adverse events are different from the primary temperature-control endpoint. They require context about how events were collected, how many participants were monitored, and whether the same observation period applied to each group. Those details are not reported in the registry record. The figures alone do not support a conclusion about comparative safety.

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

The posted results indicate that temperature-range measurements during transport differed between the randomized groups, with the device-regulated cooling arm showing a higher median percentage in the target range than standard practice. The registry reports p < 0.001 for the primary analysis. The secondary summaries also follow a direction consistent with more effective temperature management in the device arm. These findings concern temperature outcomes during transport.

They do not establish that the device improves survival, prevents disability, or changes neurodevelopment. Those outcomes are not reported as results in the registry record. Nor does the trial compare therapeutic hypothermia with no cooling: both arms received cooling, with the control arm following passive or active cooling according to center practice. The comparison is between a servo-regulated device strategy and standard practice during transport.

The primary summaries do not provide a confidence interval for the between-group difference, and the registry does not report the effect estimate from the primary analysis. A statistically small p-value does not by itself indicate a large or clinically important effect, nor does it establish the precision of the estimate. The spread in the posted outcome distributions and the missing contrast interval should remain part of the interpretation.

The trial also does not establish how well the result generalizes to other transport teams, centers, equipment, or patient populations. The control approach varied according to center practice, and details of sites and eligibility criteria are not reported. The result is best understood as evidence about the recorded transport-temperature outcomes in this randomized comparison, with important limits on clinical and external interpretation.

14. Why This Trial Matters Statistically

This trial illustrates how a clinical question can be framed around the performance of care during a constrained period—in this case, neonatal transport—rather than only around a later clinical outcome. A percentage of temperatures in range captures an aspect of temperature control over time, while time to target and point-in-time attainment address related but distinct features. Interpreting these measures together requires attention to their definitions and the fact that they are related outcomes.

It also highlights why a test result and an effect estimate are not the same thing. The primary p-value indicates that the reported analysis found evidence against its null hypothesis, but the registry does not provide a primary contrast estimate or confidence interval. Descriptive arm summaries, while useful, do not fill that gap. Clear reporting of the contrast and its uncertainty would help readers judge magnitude and precision.

The secondary analyses show another important statistical distinction: a rank-based test and a recalculated mean difference address different summaries of the data. Their p-values need not coincide, and a simple recalculation should not displace the sponsor’s own analysis. For binary outcomes, the estimates and intervals calculated from posted counts are likewise unadjusted checks, not a replacement for the investigators’ analysis plan.

Finally, the differing outcome denominators and the absence of reported details on missing data, multiplicity, and analysis populations demonstrate how registry summaries can leave key interpretive questions open. The statistical lesson is not that the results are unusable, but that the strength and scope of an inference depend on transparent reporting of the estimand, analysis, uncertainty, and data handling.

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.

Several limitations specific to this trial affect interpretation. The primary outcome is based on temperatures recorded repeatedly during transport, but the registry does not explain how repeated measurements within each participant were handled statistically. Such measurements can be correlated, and treating them as independent without an appropriate method could affect uncertainty. The primary analysis is named as adjusted Wald estimation, but the adjustment and its implementation are not described.

The primary record gives arm-level medians and inter-quartile ranges and a p-value, but no estimated between-group contrast or confidence interval. This leaves uncertainty about the magnitude and precision of the primary effect. The absence of masking may also matter for care delivery and measurement processes, particularly because the device arm had continuous temperature monitoring while standard practice recorded temperatures every 15 minutes. The record does not report whether outcome measurement or data handling was standardized across arms beyond those descriptions.

Standard practice allowed passive or active cooling according to center practice. That makes the comparator clinically pragmatic, but the record does not describe how practices varied among centers or how center effects were handled in analysis. Eligibility details, baseline characteristics, allocation concealment, protocol deviations, and analysis by randomized assignment are not reported. These omissions limit assessment of comparability, implementation, and potential bias.

The 1-hour outcome has different denominators from the primary outcome and transport-wide endpoint, and the reason is not stated. Missing-data definitions and handling are not reported. The registry also does not report a multiplicity plan or a prespecified hierarchy for the secondary endpoints. Finally, the recorded transport follow-up and temperature outcomes do not provide information about later clinical outcomes, so the findings cannot establish whether observed temperature differences translated into patient benefit.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: California Transport Cooling Trial — NCT01683383.
• Registry record last updated 2014-12-08; status last verified 2014-12.
• Results first posted 2014-12-08.
• 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

NCT01683383, the California Transport Cooling Trial, was a completed randomized, parallel-group trial with no masking. Stanford University is listed as sponsor, and the record gives enrollment of 101. The study compared servo-regulated device cooling with standard practice during neonatal transport for hypoxic ischemic encephalopathy.

The primary endpoint was the percentage of temperatures in the 33°-34°C target range during transport. The posted median was 0 in the standard-practice arm and 73 in the device arm; the registry reports an adjusted Wald analysis with p < 0.001. Secondary outcomes covered time to target temperature and target-range attainment at 1 hour or at any time during transport. The registry does not report a primary contrast interval, a multiplicity plan, or detailed missing-data and analysis-population procedures.