Perioperative Accuracy of the Raiing Wireless Axillary Thermometer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Perioperative Accuracy of the Raiing Wireless Axillary Thermometer (NCT02756910), a registered phase not stated study. The primary outcome is Accuracy of iThermonitor in Surgical Patients.

2. Clinical Question

The registry describes the device as intended to estimate core temperature from an axillary measurement. The clinical question is therefore about measurement agreement, not whether one treatment improves a health outcome. The investigators’ stated criterion was that most Raiing temperatures be within ±0.5°C of the reference for the thermometer to be considered sufficiently accurate for clinical use. That criterion is a proposed benchmark, not evidence by itself that the device met it.

The posted results state that there were 3,339 sets of iThermonitor and reference measurements from 80 patients, with 42 sets per patient on average. They describe the temperatures as agreeing well overall, but the quantitative patient-level primary summary is not reported in the registry record. The exact proportion of measurements meeting the stated tolerance, its distribution across patients, and its confidence interval are therefore not available to assess against the benchmark.

3. Trial Design

The registry record describes a completed evaluation sponsored by Peking Union Medical College Hospital. Its enrollment was 80. The record lists the study start as 2016-05 and primary completion as 2016-09. The condition is major surgery under general anesthesia.

The listed device intervention was iThermonitor (WT701). The arm description is “Surgery Patients >1.5 Hrs” and says that esophageal, Foley, and iThermonitor monitoring were used for core, bladder, and axillary temperature monitoring, respectively. The primary outcome description specifies esophageal temperature as the reference during surgery. These concurrent measurements support a within-patient comparison of device readings with a reference rather than a randomized treatment comparison.

The registry record does not state an allocation method. It does not describe a randomized comparison, and its record notes that the arms are not a like-for-like control comparison. Thus, the study should not be interpreted as a trial comparing clinical treatments. Its main purpose, as described, was to assess the accuracy and precision of a measurement device in surgical patients.

The outcome timeframe is from 1 to 3 hours after induction of anesthesia. The brief summary also describes an intended evaluation during the initial hour of recovery, but whether recovery-period results were reported is not reported in the registry record. The record describes repeated temperature measurements within patients. Repeated readings from the same patient are correlated, so they cannot automatically be treated as independent observations when estimating uncertainty.

The registry record does not report the protocol, a sample-size justification, or a detailed statistical analysis plan. It also does not report whether any interim analyses were conducted. The posted study status is COMPLETED, but completion status alone does not establish that every planned analysis or outcome was reported.

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

Analysis population for the primary outcome, as defined in the registry: There were a total of 3,339 sets of iThermonitor and reference measurements from 80 patients (42 sets per patient on average). The iThermonitor temperature and the reference esophageal temperature agreed well overall. The iThermonitor temperatures ranged from 34.7 to 38.3°C, while the esophageal core temperatures ranged from 34.6 to 37.9°C

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 Accuracy of iThermonitor in Surgical Patients, measured over From 1 to 3 hours after induction of anesthesia.

The primary outcome description defines accuracy through the proportion of each patient’s iThermonitor measurements that fell within 0.5° C of the esophageal reference. This is a patient-level summary built from repeated paired readings. Summarizing the proportion separately for each patient and then describing those proportions across patients keeps the patient, rather than each individual reading, as the unit for the primary summary.

The stated tolerance gives the agreement measure a direct interpretation: it counts how often the device reading was close to the reference within a prespecified temperature margin. The investigators stated that they would consider the device sufficiently accurate for clinical use if most Raiing temperatures were within ±0.5°C of the reference. The registry record does not report a quantitative estimate of the proportion meeting this criterion, its median, quartiles, or 95% confidence interval.

The outcome metadata also labels the measure as continuous, with a mean and standard deviation, while the detailed description defines the primary measure as a patient-level proportion summarized by its median, quartiles, and 95% confidence interval. That difference in presentation is not resolved in the registry record. The record additionally describes bias and repeated-measures Bland–Altman analysis as secondary measures. Those analyses address agreement and systematic differences, but their numerical results are not reported in the available registry text.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Measurement studies ask whether a device produces readings sufficiently close to a reference method for the intended setting. Here, the registry identifies esophageal temperature as the core-temperature reference during surgery and defines the primary measure using the distance between each iThermonitor reading and that reference. The central statistical issue is agreement, not merely whether the two measurements tend to move in the same direction.

What does the patient-level proportion describe?

For each patient, the stated primary measure is the proportion of iThermonitor measurements within 0.5° C of the reference. This reduces a sequence of readings to a patient-level summary that is easy to relate to the stated tolerance. Summarizing those proportions across patients describes how performance varied across the study participants rather than allowing patients with more readings to dominate simply because they contributed more observations. The registry says the proportions were to be summarized by median, quartiles, and a 95% confidence interval. Those numerical summaries are not reported in the registry record.

A proportion within a tolerance answers a threshold question, but it does not describe every feature of the differences. Readings just inside and just outside the boundary count differently under this rule even if their actual differences are very similar. The proportion also does not show whether errors tended to be positive or negative. For those reasons, the outcome description includes additional analyses of bias and paired differences.

Why use repeated-measures Bland–Altman analysis?

Bland–Altman methods examine differences between paired measurements in relation to their magnitude. Bias describes a systematic tendency for one method to read higher or lower than the reference. The dispersion of differences describes how much individual readings vary around that tendency. In repeated-measures data, each patient contributes multiple pairs, and readings from the same person may be more alike than readings from different people. A repeated-measures approach accounts for this clustering when estimating uncertainty; treating every pair as independent can make uncertainty appear smaller than it is.

The registry says repeated-measures Bland–Altman analysis was planned for bias and 95% confidence intervals. It does not report the resulting bias estimate, its confidence interval, or numerical information about the spread of differences. Without those estimates, the record’s qualitative statement of overall agreement cannot be assessed for systematic offset or variability across readings.

What does correlation add?

The outcome description also proposes regressing patient iThermonitor measurements on patient reference measurements to assess Pearson correlation and estimate bias. Correlation measures how strongly paired values vary together in a linear pattern. It does not measure closeness to equality: a device could track the reference while maintaining a consistent offset, or while showing substantial scatter. Therefore, correlation is complementary to, not a replacement for, summaries of paired differences and the fraction within the stated tolerance. The correlation estimate and its confidence interval are not reported in the registry record.

How should uncertainty be understood?

A confidence interval expresses the statistical uncertainty around an estimate under the analysis assumptions. A narrower interval indicates greater precision, while a wider interval indicates more uncertainty. It does not guarantee that the estimate is correct, and it does not represent the range of readings for individual patients. For repeated measurements, the interval should reflect the structure of the data, including within-patient dependence. The registry specifies 95% confidence intervals for planned summaries, but the relevant numerical intervals are not reported in the registry text.

The registry’s outcome metadata calls the measure continuous and names mean and standard deviation, whereas the detailed primary-outcome description specifies patient-level proportions summarized by median and quartiles. Means and standard deviations describe a distribution using its center and spread; medians and quartiles describe its center and spread using ranks and are less sensitive to extreme observations. The record does not explain this difference in outcome presentation, so readers should avoid assuming which summary was used for the primary reported analysis.

What can planning calculations say?

Any sample-size figures associated with planning are illustrations based on an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They are not the trial’s reported design justification, and sample-size illustrations are not observed power. They should not be interpreted as evidence that the study had a particular probability of detecting an effect. The registry record does not report its own sample-size justification or establish that the planning assumptions describe the study design. In particular, an equal-allocation assumption for a planning illustration does not turn this device evaluation into a randomized two-arm comparison.

Overall, the analysis concepts are appropriate to questions about measurement agreement, but interpretation depends on the estimates, uncertainty intervals, and handling of repeated observations. The registry supplies a qualitative conclusion and some measurement counts and ranges, but the numerical primary and secondary agreement results are not reported in the registry record.

10. Confidence Intervals for Each Arm

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

R code and output

arm <- c("Surgery Patients >1.5 Hrs")
n   <- c(80)
m   <- c(0.14)
sp  <- c(0.26)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Surgery Patients >1.5 Hrs  80  0.140  0.260  0.082  0.198

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 1 outcome measure and 1 analysis. The detailed outcome description nevertheless names several statistical summaries and approaches: the patient-level proportion within the stated tolerance, bias and confidence intervals from repeated-measures Bland–Altman analysis, and correlation and regression-based assessment. The record does not report whether these were treated as separate confirmatory tests, whether any outcomes were designated exploratory, or whether a multiplicity adjustment was planned. The number of listed outcomes alone does not establish how many statistical tests were conducted.

Multiplicity matters when many hypotheses or endpoints are tested because the chance of at least one apparently notable result can increase. Methods that control this risk depend on which results are considered confirmatory and on the analysis plan. The registry record does not report such a plan, so no conclusion about adjustment or control of false-positive findings can be made.

The record reports 3,339 measurement sets from 80 patients, but does not report missing-measurement counts, exclusions, or the number of patients included in each analysis. The phrase “42 sets per patient on average” describes the average number of sets; it does not establish that each patient contributed the same number or that there were no missing readings. The handling of incomplete records, invalid measurements, or patients without usable paired readings is not reported in the registry record.

The analysis population is also not described in detail. The registry does not report whether all enrolled patients were included in each outcome analysis, how eligibility for analysis was determined, or whether any sensitivity analyses were conducted. These omissions matter because missingness or exclusions can affect both the estimated agreement and its uncertainty, particularly if unavailable readings are related to measurement difficulty or patient condition.

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.

The registry’s adverse-event figures list 80 participants at risk for serious adverse events and 0 affected. It also lists 80 at risk for other adverse events and 0 affected, with 0 deaths. No serious-event terms are listed. These are the counts recorded for the adverse-event categories in the registry; they should not be read as proof that no unrecorded or unmeasured events occurred.

The record does not report the adverse-event collection procedures, the observation period used for safety reporting, event definitions, or how completeness of reporting was checked. It also does not provide event-level descriptions beyond the absence of listed serious terms. Without those details, the counts do not establish the intensity of surveillance or the duration over which participants were monitored.

These figures concern the study’s recorded safety reporting, not the measurement agreement endpoint. They do not provide information about the accuracy of the thermometer, and they are not a comparative safety estimate because the registry does not describe a like-for-like randomized control comparison. A zero count in a registry field is best interpreted narrowly as no events recorded in that field, with the limitations of the available reporting information kept in view.

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

The registry record supports a limited description of the study: it evaluated the iThermonitor in perioperative patients and collected paired device and reference measurements. The posted results report 3,339 sets from 80 patients, 42 sets per patient on average, and an iThermonitor temperature range of 34.7 to 38.3°C. The registry characterizes the overall agreement with esophageal temperatures as good. These statements describe the record, but they do not supply the quantitative results needed to judge the size and precision of agreement.

The primary measure was designed to show how often readings fell within 0.5° C of the reference for each patient. The investigators’ stated criterion was that most Raiing temperatures be within ±0.5°C. Because the primary patient-level summaries and confidence interval are not reported in the registry record, the numerical evidence needed to assess that criterion is unavailable. The record also does not report the numerical bias, correlation, or repeated-measures Bland–Altman results described in the outcome text.

This evidence should not be interpreted as demonstrating that the device works or does not work for clinical use. The registry’s qualitative statement is not enough to quantify accuracy, establish performance in every patient, or determine how the device behaves in settings outside the described perioperative context. Nor does a measurement evaluation show that using the device changes patient outcomes. No medical recommendation follows from the reported figures.

The registry record also leaves important analytic details unreported, including missing-data handling, the analysis population, and a detailed statistical plan. These gaps affect how fully the results can be checked and interpreted. The careful conclusion is that the record documents a completed device evaluation and reports an overall favorable qualitative description of agreement, while the numerical endpoint results and their precision are not reported in the registry record.

14. Why This Trial Matters Statistically

Temperature monitoring studies are statistically distinct from trials of treatments. Their central challenge is to establish whether a new measurement method is close enough to a reference across the conditions and patients in which it is intended to be used. A high correlation or a general statement of agreement does not fully answer that question; systematic bias, variability in paired differences, and the fraction of readings within an acceptable tolerance all contribute to interpretation.

This registry describes a clinically demanding measurement context: patients undergoing major surgery under general anesthesia, with the investigators noting thermal perturbations over a range of several °C. It also describes repeated measurements and an intended tolerance of ±0.5°C. Such features make it important to distinguish patient-level performance from the apparent precision that can result from treating numerous readings as independent. The record’s stated use of repeated-measures Bland–Altman analysis recognizes that readings from the same patient are clustered.

The registry further illustrates why complete statistical reporting matters. It describes several relevant measures of agreement, but does not report the numerical estimates and confidence intervals for the primary and secondary analyses in the available text. Without them, readers cannot assess how much readings differed, how consistently the tolerance was met across patients, or how uncertain the estimates are. The qualitative phrase “agreed well overall” cannot replace those quantities.

Finally, the record distinguishes neither a randomized clinical comparison nor a treatment-effect question: it is a device-accuracy evaluation with reference measurements. Interpreting the design on its own terms helps avoid causal claims and keeps attention on the statistical evidence required for measurement agreement.

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 most important limitation for interpretation is incomplete numerical outcome reporting in the registry text. Although the primary measure is described as a patient-level proportion within 0.5° C of the reference, its median, quartiles, and 95% confidence interval are not reported in the registry record. The planned bias, correlation, and repeated-measures Bland–Altman estimates are also not reported. The qualitative statement of good overall agreement does not supply the size or precision of these effects.

The available result text is truncated after beginning to describe the esophageal temperature range. The only complete temperature range in the record is for iThermonitor, 34.7 to 38.3°C. The full reference range and the distribution of paired differences are not reported in the registry text, limiting assessment of agreement across the observed measurement values.

The registry’s metadata and detailed outcome description present the primary measure differently: the metadata labels it continuous with a mean and standard deviation, while the detailed description specifies per-patient proportions summarized with a median, quartiles, and confidence interval. The record does not explain this difference. It also does not report a detailed statistical analysis plan, a sample-size justification, missing-data handling, analysis populations, or whether any interim analyses were conducted.

There is no reported randomized like-for-like control comparison. Allocation is not stated, and the registry itself cautions that the arms are not a like-for-like control comparison. The findings therefore should be read as measurement-performance information in the described perioperative context, not as a comparative treatment result. Whether recovery-period results were reported is not reported in the registry record, despite the brief summary’s mention of an intended initial-hour recovery evaluation.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Perioperative Accuracy of the Raiing Wireless Axillary Thermometer — NCT02756910.
• Registry record last updated 2021-05-14; status last verified 2021-02.
• Results first posted 2021-05-14.
• 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 record describes a completed evaluation of the Raiing iThermonitor (WT701) in patients undergoing major surgery under general anesthesia. It lists an enrollment of 80, sponsorship by Peking Union Medical College Hospital, a study start of 2016-05, and primary completion of 2016-09. Esophageal temperature was identified as the core-temperature reference during surgery.

The primary outcome was defined as the patient-level proportion of device readings within 0.5° C of the reference, with patient-level proportions intended to be summarized across patients. The posted results report 3,339 measurement sets from 80 patients, 42 sets per patient on average, and an iThermonitor range of 34.7 to 38.3°C. The record describes overall agreement as good, but does not report the quantitative primary summary or the numerical bias, correlation, and repeated-measures Bland–Altman results.

The adverse-event record lists 0 serious events, 0 other events, and 0 deaths. Further details about event ascertainment and statistical analysis are not reported in the registry record.