Forced-air Warming in Endovascular Surgery: Testing Effectiveness of Two Different Blanket Models: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Forced-air Warming in Endovascular Surgery: Testing Effectiveness of Two Different Blanket Models (NCT02742818), a randomized na study. The primary outcome is Values of Body Temperature ( in Celsius Degree ) in Two Groups of Patients Undergoing EVAR and LEA That Will Use Two Different Types of Bair Hugger 3M Body Warming Blankets, Namely: 522 Upper Body Blanket and 635 Full Access Underbody Blanket..

2. Clinical Question

The trial asks a comparative question about temperature patterns during surgery, rather than reporting only a single measurement at one time. This matters because the primary outcome was collected repeatedly while patients were undergoing LEA or EVAR. The registry describes surgery duration as a control variable in the planned comparison, and gives a timeframe of up to 210 minutes.

The posted arm summaries are median temperatures of 36.2 for the upper-body group and 36.1 for the underbody group. The corresponding full ranges are 35.1 to 37.3 and 34.3 to 37.1. These summaries describe the recorded temperatures but do not, by themselves, quantify a model-adjusted difference across the repeated measurements. The registry reports a Generalized Estimation Equations analysis with p = 0.190, but does not report an effect estimate or confidence interval for that analysis.

3. Trial Design

This was a completed, randomized, parallel-group trial with no masking. The registry lists an enrollment of 50 and identifies hypothermia as the condition. The study compared two device interventions: an upper-body Bair Hugger blanket and an underbody Bair Hugger blanket. Both arms were described as involving patients undergoing EVAR and LEA. The registry does not report how many participants underwent each type of surgery.

The record lists the study start as 2014-09 and primary completion as 2015-06. It gives the status as completed and lists the phase as not applicable. Nelson Wolosker is named as sponsor. No further information about the randomization procedure, concealment of allocation, or masking is reported in the registry record.

For the primary temperature outcome, the registry describes turning on the assigned blanket after tracheal intubation at the warming unit’s highest setting, 43 celsius degree. An esophageal thermometer was to record body temperature every 15 minutes until tracheal extubation. The stated observation period was across the duration of surgery, up to 210 minutes. The planned description says the temperature values would be compared using variance analysis with repeated measurements and that surgery duration would be used as a control variable.

The posted primary outcome summaries include 25 participants in the upper-body group and 23 in the underbody group. The record says two underbody participants were excluded from analysis: one due to an anaphylactic reaction and one because surgery time was less than 45 minutes. It does not explain whether any other participants had missing measurements or how incomplete temperature series were handled. The number randomized to each group is not separately described in the arm summaries, although the adverse-event record lists 25 participants at risk in each group.

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: In the Underbody blanket group 2 patients were excluded from analysis. One of the patients due to anaphylactic reaction and the other framed exclusion criteria (Surgery time lesser than 45 minutes)

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 Values of Body Temperature ( in Celsius Degree ) in Two Groups of Patients Undergoing EVAR and LEA That Will Use Two Different Types of Bair Hugger 3M Body Warming Blankets, Namely: 522 Upper Body Blanket and 635 Full Access Underbody Blanket., measured over Across duration of surgery (up to 210 minutes).

The primary endpoint is body temperature, measured in celsius degrees with an esophageal thermometer during surgery. The registry describes measurements every 15 minutes from after tracheal intubation until tracheal extubation. It reports median temperature and full range for each group, with a timeframe of up to 210 minutes. The posted medians are 36.2 for the upper-body blanket and 36.1 for the underbody blanket; the full ranges are 35.1 to 37.3 and 34.3 to 37.1, respectively.

Repeated measurements are relevant because temperature can vary during an operation. General statistical context: measuring an outcome at multiple times can describe its course, but those observations from one participant are related rather than independent. A suitable analysis therefore needs to account for within-participant dependence and, when appropriate, differences in observation duration. The registry’s primary-outcome description says surgery duration was to be used as a control variable.

The endpoint is used to compare temperature patterns associated with the two blanket models during the procedures studied. However, the record does not report a numerical between-group effect across time, its confidence interval, or the detailed model results. The medians and ranges are descriptive summaries and do not establish how large or precise any adjusted difference in temperature over time may be.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome was measured repeatedly during surgery, and the registry identifies Generalized Estimation Equations (GEE) as the reported analysis method. The primary-outcome description separately states that repeated-measures variance analysis was intended, with surgery duration as a control variable. These descriptions concern the same broad statistical challenge: comparing groups when observations are taken more than once from each participant and the length of observation may vary.

Why repeated observations need special treatment

General statistical context: measurements from the same participant tend to be related because they share characteristics of that participant and the same operation. Treating every temperature reading as if it came from a different, unrelated person can misrepresent uncertainty. Repeated-measures methods account for the fact that readings within a participant are clustered. A comparison of full temperature patterns can use information about the sequence of readings, rather than reducing each participant’s experience to one isolated value.

The registry says temperatures were to be measured every 15 minutes from after tracheal intubation until extubation, over a period up to 210 minutes. This design can produce different numbers of recorded measurements for participants whose surgeries have different durations. The registry’s description says duration was to be used as a control variable. It does not provide the model equation, the form in which duration was entered, or details about how incomplete measurement sequences were handled.

What GEE contributes

General statistical context: GEE is a way to estimate population-average relationships when observations are grouped, such as repeated measurements within participants. The method accounts for correlation among observations through a working correlation structure and uses uncertainty estimates designed for clustered data. The method’s name alone does not reveal which outcome scale, time terms, covariates, or correlation structure were used. Those details are not reported in the registry record.

The registry reports a GEE p-value of 0.190 but does not state the precise coefficient or contrast tested. It also does not report an effect estimate or confidence interval. Without those results, a reader cannot determine the estimated direction and size of a temperature difference across time or how precisely it was estimated. A p-value does not replace those quantities: it does not convey the clinical or practical magnitude of a contrast.

How the descriptive summaries differ

The arm summaries are medians and full ranges. The upper-body group has a median of 36.2 and a range from 35.1 to 37.3; the underbody group has a median of 36.1 and a range from 34.3 to 37.1. A median describes the middle of the observed values, while a full range gives the lowest and highest observed values. These summaries can convey the distribution of posted temperatures, but they do not account for the repeated-measures structure in the same way as a fitted model. They also do not provide the uncertainty around a between-group estimate.

General statistical context: ranges are sensitive to extreme observations and do not show how values are distributed between the endpoints. Medians and ranges can be useful descriptive summaries, but they are not substitutes for a model-based comparison when the target is a group difference over time. The registry does not give the number of temperature readings contributing to each arm summary or say how readings at different times were combined for those summaries.

Why over-dispersion can matter

General statistical context: over-dispersion means that observed variation is greater than a model’s assumed variation. When a model does not account for that extra variability, its standard errors can be too small and its uncertainty statements can be too optimistic. Whether over-dispersion is relevant depends on the outcome and model specification. The registry does not report a diagnostic assessment, a dispersion estimate, or whether any adjustment for over-dispersion was used. It would therefore be inappropriate to assert that over-dispersion occurred in this trial.

Likewise, the record does not explain whether the GEE analysis included a robust variance estimator, how missing readings were treated, or whether assumptions were checked. These are useful details for evaluating a repeated-measures analysis, but their absence does not establish that a particular method was or was not used.

Reading a p-value and an interval

General statistical context: a confidence interval describes a range of values compatible with the data and the model at a stated confidence level. Its width communicates precision, and its location can help describe the direction and possible size of an effect. The registry does not report a confidence interval for the primary GEE result, so precision cannot be assessed from the reported p-value alone.

The registry’s sample-size assumptions describe a planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation, a normal approximation, and effects expressed as fractions of the observed effect. These are planning assumptions, not a report of observed power, and the registry does not give numerical planning results or the trial’s own sample-size justification. The actual enrollment was 50. No further design rationale or interim-analysis information is reported.

Finally, the methods description and reported analysis label are not identical: the former mentions repeated-measures variance analysis, while the posted analysis names GEE. Both are approaches that can address repeated data, but the registry does not state why the listed analysis differs from the descriptive plan or supply enough details to reconstruct it. That uncertainty reinforces the need to interpret the reported p-value cautiously and to distinguish it from a complete treatment-effect estimate.

10. Analyses Reported by the Sponsor

The registry identifies GEE as the reported analysis and gives its p-value as 0.190. It does not report a corresponding estimated treatment contrast, confidence interval, or sufficient model details to reconstruct that contrast. The arm-level medians and ranges are descriptive summaries; they are not a recalculation of the GEE effect across repeated measurements.

No separate recalculated effect estimate or confidence interval is reported in the registry record. The record also does not describe a discrepancy between a sponsor analysis and a recalculation. Accordingly, there is no competing numerical result to compare with the sponsor’s analysis. The posted GEE result remains the registry’s reported inferential result, but the absence of an estimate and interval limits what can be learned from it.

General statistical context: a recalculation from summary statistics can be useful as a check when the required quantities are available, but it cannot generally reproduce a repeated-measures model from medians and ranges alone. Such a recalculation would also not replace the model specified for the trial. The registry does not provide the information needed to independently reproduce the stated GEE result, and no recalculated result should be inferred from the reported summaries.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists four secondary outcomes: age, gender, type of surgery (LEA and EVAR), and duration of surgery. It reports five outcome measures and five analyses overall, but does not provide a full account of the analyses for those secondary outcomes. The record does not describe an adjustment for multiple comparisons, identify a hierarchy of outcomes, or state which results were considered confirmatory. Without that information, the extent to which multiple testing was addressed is not reported in the registry record.

The primary outcome is repeated body-temperature measurement over the duration of surgery. The registry gives an analysis p-value of 0.190 but does not report an effect estimate, confidence interval, analysis alpha for that endpoint, or a clear definition of the primary contrast. It also does not explain how the difference between the planned repeated-measures variance analysis and the posted GEE analysis arose. The available record therefore does not establish the full testing strategy.

The primary temperature summaries include 25 participants in the upper-body group and 23 in the underbody group. The registry reports that two underbody participants were excluded from analysis: one due to an anaphylactic reaction and one because surgery time was less than 45 minutes. It does not report whether other participants had missing temperatures, how incomplete repeated records were treated, or whether participants were analyzed according to their randomized assignment. No further analysis-population definition is reported.

The exclusion following an anaphylactic reaction is relevant both to outcome analysis and to safety interpretation. The record does not describe whether temperature measurements before the reaction were retained or how the event affected the participant’s contribution to the repeated-measures analysis. Those details should not be inferred from the final group count.

The registry gives assumptions for a sample-size planning illustration, including a two-sided significance level of 0.05, equal allocation, and a normal approximation. It does not give a numerical sample-size justification for the trial or the resulting planning figures. Interim analyses and missing-data procedures are not reported in the registry record.

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 4% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: anaphylatic reaction (4.0%).

The adverse-event record lists 25 participants at risk in each group. It reports zero serious adverse events in the upper-body group and one in the underbody group. The record identifies “Anaphylatic reaction” among the serious adverse-event terms and lists 4.0%. The registry’s primary-outcome description also says one underbody participant was excluded after an anaphylactic reaction. The record does not explicitly describe whether those entries refer to the same event or participant, so they should not be treated as separate events or as a confirmed match.

The adverse-event record also lists zero other adverse events in the upper-body group and one in the underbody group. It does not clarify whether the serious and other adverse-event entries refer to distinct events or participants. Death counts are not reported in the registry record; blank death fields should not be interpreted as evidence of no deaths.

These figures are counts from a trial with 25 participants at risk in each group. They provide a limited description of events recorded in the registry, not a precise estimate of comparative safety. The registry does not report additional detail about event timing, severity grading, assessment of relatedness, or follow-up duration. The listed figures therefore need to be read alongside those reporting gaps.

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

The trial addresses a focused statistical question: how body-temperature measurements during LEA and EVAR compare between an upper-body and an underbody forced-air warming blanket. Its randomized, parallel-group design is relevant to a comparison between the two assigned devices. The registry reports temperature summaries for each arm and identifies GEE as the analysis method.

The posted medians are 36.2 in the upper-body group and 36.1 in the underbody group. The full ranges are 35.1 to 37.3 and 34.3 to 37.1. These descriptive summaries show the range and median of the recorded values reported by arm, but they do not establish the size or precision of an adjusted difference in temperature over time. The registry’s GEE result is p = 0.190, without an effect estimate or confidence interval. That p-value alone does not indicate how large a difference may be or how uncertain the estimated difference is.

The record does not provide enough detail to determine the exact contrast tested, the fitted model specification, or the handling of incomplete repeated measurements. Two underbody participants were excluded from the primary analysis for stated reasons. This creates a difference between the reported analysis-group counts and the adverse-event risk counts, and the record does not fully describe the analysis population.

The results do not establish a conclusion about temperature differences beyond the quantities actually reported. In particular, no unreported effect estimate, confidence interval, or clinical threshold can be inferred from the medians, ranges, or p-value. The registry record also does not provide enough information to assess how the outcome analysis accounted for the repeated readings in full detail.

More broadly, the posted results concern these two blanket models in the described surgical setting. They do not directly describe other devices, procedures, patient groups, or care settings. The record gives no basis for extending the statistical findings beyond the comparison and participants enrolled in this trial.

14. Why This Trial Matters Statistically

This trial illustrates why the analysis of repeated outcomes needs to be distinguished from a comparison of one summary per group. Body temperature was to be measured every 15 minutes during surgery, from after intubation until extubation. Those readings are linked within each participant, and the period observed can depend on surgery duration. A method that accounts for repeated observations is therefore central to the statistical question.

The registry’s primary-outcome description names repeated-measures variance analysis and says duration would be a control variable; the reported analysis is identified as GEE. Both labels point to the importance of accounting for repeated data, but the record does not explain the difference between the planned description and the posted analysis or give model details. This makes the trial useful for recognizing that an analysis label and a p-value do not, by themselves, give a reader enough information to understand the estimated contrast.

The record also shows why effect estimates and confidence intervals matter. It gives arm medians and ranges and a GEE p-value, but no estimated group contrast or interval. The descriptive values convey some aspects of the observations, while the inferential result gives a test-related summary. Neither alone communicates the magnitude and precision of a model-adjusted comparison over time.

Finally, the underbody analysis excludes two participants for stated reasons, including an anaphylactic reaction, while adverse-event counts are reported separately. Understanding how participants and their measurements enter both efficacy and safety summaries is important when reading small randomized trials. The registry leaves some of those details unreported, making transparent reporting of analysis populations, missing readings, and event handling especially relevant.

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 does not report the GEE treatment-effect estimate, its confidence interval, the exact contrast tested, or key model details. It does not state the working correlation structure, the time terms or covariates included, or how missing repeated observations were handled. These omissions prevent independent interpretation of the direction, magnitude, and precision of the primary comparison from the inferential result.

The primary-outcome description says repeated-measures variance analysis was planned, whereas the reported analysis is GEE. The registry does not explain the difference or give enough information to assess whether the analysis followed a prespecified plan. It also does not report the primary analysis significance level or an interim-analysis plan.

The primary temperature analysis includes 25 upper-body and 23 underbody participants. The two underbody exclusions are described, but the record does not explain whether any other participants had incomplete measurements, how their data were handled, or whether the stated summaries use all available readings. It does not report the number of temperature observations contributing to each group summary.

The record does not give the distribution of participants across LEA and EVAR, or detailed outcomes by procedure type. It also does not report further baseline characteristics beyond listing age, gender, and type of surgery as secondary outcomes. The sample-size assumptions describe a planning illustration, but a trial-specific sample-size justification and numerical planning results are not reported.

Safety reporting is also limited: event details, timing, relatedness, follow-up, and death counts are not reported in the registry record. The listed serious and other adverse-event counts do not establish whether the entries refer to distinct events or participants. These limitations are specific to the posted information and do not establish that the unreported procedures were not performed.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Forced-air Warming in Endovascular Surgery: Testing Effectiveness of Two Different Blanket Models — NCT02742818.
• Registry record last updated 2019-12-30; status last verified 2019-12.
• Results first posted 2018-01-23.
• 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, parallel-group trial comparing an upper-body Bair Hugger blanket with an underbody Bair Hugger blanket in patients undergoing LEA or EVAR. Enrollment was 50. Temperature was to be measured with an esophageal thermometer every 15 minutes from after tracheal intubation until extubation, with a stated timeframe up to 210 minutes.

The reported arm summaries are medians of 36.2 and 36.1, with full ranges of 35.1 to 37.3 and 34.3 to 37.1. The registry reports a GEE p-value of 0.190, but no effect estimate or confidence interval. The primary analysis includes 25 upper-body participants and 23 underbody participants; two underbody participants were excluded for stated reasons.

The record lists one serious adverse event in the underbody group and none in the upper-body group. Further details about analysis methods, missing measurements, multiplicity, and safety-event follow-up are not reported in the registry record.