Feasibility and Performance Evaluation for INVSENSOR00024: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Feasibility and Performance Evaluation for INVSENSOR00024 (NCT03570255), a na na study. The primary outcome is Accuracy of Oxygen Saturation (SpO2) Measurement by RMS Calculation.

2. Clinical Question

This is a device performance question rather than a question about comparative treatment effects. The record does not describe a randomized comparison or a control group. Each sensor reading was intended to be assessed against a corresponding laboratory reference measurement, with the discrepancy between the paired values contributing to the accuracy summary. The record describes the evaluation as taking place over up to 8 hours.

Convenience sampling means that recruitment was based on available eligible subjects rather than a probability sample of the broader neonatal population. The registry record does not report recruitment criteria in enough detail to establish how representative the enrolled subjects were. As a result, any performance estimate, if available, would need to be interpreted in light of the setting and the sampled population, as well as the stated exclusions and deviations.

3. Trial Design

The registry record describes a prospective, non-randomized, single-group study of the RD SET Neo SpO2 investigational sensor. The study is listed as completed, with enrollment of 28. It was sponsored by Masimo Corporation. The listed phase and allocation are recorded as NA, and masking is listed as none. A treatment-control comparison is not part of the design.

The study had a feasibility component and a clinical performance component. According to the registry description, subjects enrolled under Protocol Version 1 were part of feasibility work intended to verify sensor design changes. The protocol was amended to Version 2 to enroll subjects for clinical performance evaluation intended for submission to regulatory agencies. The record does not report how many subjects belonged to each protocol version, or how results were separated or combined across the versions.

For the primary endpoint, noninvasive oxygen saturation readings from the investigational pulse oximeter were compared with blood oxygen saturation readings from a laboratory pulse Co-Oximeter. The endpoint timeframe is listed as up to 8 hours. The record does not report the number of readings per subject, the timing or spacing of measurements, or how repeated readings were handled statistically. Those details matter because readings from the same subject may be related rather than statistically independent.

The listed primary outcome is accuracy of SpO2 measurement by RMS calculation. No analysis count is listed, and the numerical outcome result is not reported in the registry record. The record also identifies protocol deviations and data exclusions, but does not report a complete analysis population definition or a flow of subjects and measurements through the endpoint analysis.

The design can describe agreement between a device measurement and a laboratory reference in the observed study setting. Without a separate control group, it cannot estimate a relative effect against another treatment or device group. The record does not report the protocol’s sample-size justification, the planned analysis method beyond the RMS definition, or whether any interim analyses were conducted.

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: Subject 7 \& 11: Protocol deviation Subject 19 \& 26: Outlier data points Subject 5: Removed per ISO standard for SpO2 equal to 100.

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 Oxygen Saturation (SpO2) Measurement by RMS Calculation, measured over up to 8 hours.

The primary endpoint is accuracy of oxygen saturation measurement, expressed using root mean square error. The record defines the comparison as the noninvasive SpO2 measurement from the pulse oximeter minus the oxygen saturation measurement from a laboratory pulse Co-Oximeter for each sample. It describes the average of those differences as bias and their standard deviation as precision. RMS error is then the square root of the sum of the squared bias and squared precision.

In general, an RMS summary combines systematic offset and variability. A consistent tendency for device readings to differ from the reference contributes through bias, while spread in the paired differences contributes through precision. Combining these aspects provides a single measure of the overall magnitude of discrepancies, rather than reporting only the average difference. A small average difference alone would not necessarily imply close agreement if individual differences were widely dispersed.

The registry record gives the endpoint unit as percent of oxygen saturated hemoglobin and its timeframe as up to 8 hours. This endpoint is used to assess measurement agreement against a reference method; it does not by itself establish that the sensor performs similarly in other populations, settings, or measurement conditions.

The actual RMS value, the number of samples included, and an uncertainty interval are not reported in the registry record. Consequently, the definition explains what the endpoint was intended to measure, but the record does not provide the numerical estimate needed to judge the magnitude or precision of observed agreement.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes an RMS error method based on paired measurements: an investigational pulse oximeter reading and a laboratory pulse Co-Oximeter reading for each sample. In general, paired comparisons are appropriate when the scientific question concerns agreement between two methods applied to the same measurement occasion. The analysis focuses on within-pair differences, not on comparing unrelated averages from separate groups.

What do bias and precision represent?

The registry calls the average of the paired differences bias. In general, bias describes the average direction and size of a method’s offset relative to its reference. A positive or negative average can indicate that one method tends to read higher or lower, depending on the subtraction order. Here, the record defines the difference as the pulse oximeter measurement minus the laboratory reference measurement. Precision is described as the standard deviation of those differences. In general, that standard deviation captures how spread out the paired discrepancies are around their average.

Bias and precision describe different features. If discrepancies are consistently similar, their spread may be limited even if their average is not close to zero. Conversely, discrepancies may average near zero while varying considerably across samples. A single summary that reflects both characteristics can be more informative than the average discrepancy alone.

Why use root mean square error?

The stated RMS method takes the square root of the sum of squared bias and squared precision. In general, squaring prevents positive and negative contributions from cancelling and gives greater influence to larger components. Taking the square root returns the combined quantity to the scale of the measurement differences. RMS is therefore a compact summary of overall discrepancy that incorporates both systematic offset and variability.

That compactness also means the RMS value does not show the full pattern of differences. It does not, by itself, reveal whether error is concentrated in certain measurement ranges, whether there are unusual observations, or whether discrepancies vary across subjects. A detailed interpretation can depend on the distribution of paired differences and the circumstances of measurement. The registry record does not report supporting distributional summaries or plots in the outcome description.

What does uncertainty add?

In general, an estimate is more informative when accompanied by a confidence interval or another uncertainty measure. A confidence interval represents a range of values compatible with the data and a specified statistical procedure; it is not a guarantee that the true value lies within that range. The width of an interval depends on factors such as the amount of information in the analysis and the variability of observations. The registry record does not report an interval for RMS error, so its statistical precision cannot be evaluated.

Repeated readings from a subject can be correlated. In general, treating correlated observations as independent can make an analysis appear more precise than warranted. Methods that account for clustering or repeated measurements may be needed, depending on the measurement structure and the intended estimand. The record does not describe how many readings were taken per subject or how within-subject dependence was handled, so the statistical treatment of repeated measurements is not reported in the registry record.

How should exclusions be considered?

The outcome description identifies protocol deviations for subjects 7 and 11, outlier data points for subjects 19 and 26, and removal of subject 5 under the stated ISO standard for SpO2 equal to 100. Exclusions can affect both the analysis population and the resulting estimate. In general, a prespecified, consistently applied exclusion rule is easier to interpret than a decision made after results are examined. The record does not state when these handling decisions were specified, whether the identified subjects were excluded from all analyses, or how the outlier points were treated. It also does not report a sensitivity analysis using alternative handling choices.

For a performance study, it is important to distinguish the enrollment population from the set of valid paired measurements used to calculate the endpoint. The record reports enrollment of 28 but does not give the endpoint analysis denominator or the number of paired samples. These missing details prevent a reader from assessing how much of the enrolled study contributed to the accuracy calculation.

How do the planning assumptions relate to this study?

The registry describes planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. These assumptions describe hypothetical planning calculations, not the design of this single-group study. They do not represent observed power, and planning illustrations cannot substitute for the study’s actual endpoint estimate or its uncertainty. The registry record does not report the numerical planning outputs or a protocol sample-size justification.

More generally, a sample-size calculation depends on a defined target effect and analysis assumptions. It is a planning tool for a future study, not evidence that a completed study had a particular probability of detecting an effect. Because the registry record does not report the endpoint value or the assumptions used in the original design justification, the planning illustrations should not be interpreted as a retrospective assessment of the completed evaluation.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and zero analyses. This does not establish which statistical procedures were prespecified or whether additional analyses were conducted outside the registry entry. A detailed multiplicity plan, including any hierarchy of outcomes or adjustment for multiple testing, is not reported in the registry record. Since only one primary outcome is listed, the record does not describe how possible secondary or exploratory analyses were handled.

The endpoint description identifies several subject and data handling issues: protocol deviations for subjects 7 and 11, outlier data points for subjects 19 and 26, and removal of subject 5 for SpO2 equal to 100 under the stated ISO standard. The record does not provide a complete missing-data account, state whether any expected measurements were unavailable, or explain how incomplete pairs were handled. A formal missing-data method is not reported in the registry record.

The registry also does not define the analysis population for the primary endpoint. It is therefore unclear from the record which enrolled subjects and measurement pairs were included in the RMS calculation, or whether the analysis followed a prespecified population rule. The number of observations contributing to the estimate is not reported.

In general, outcome multiplicity, exclusions, missing observations, and analysis-population definitions can all affect interpretation. A transparent account would state which analyses were planned, which observations were included, and how deviations from the intended measurement process were treated. The registry record does not report those details, so the endpoint methods cannot be fully reconstructed from the entry.

11. 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 reports the RD SET Neo SpO2 group as having 28 subjects at risk for serious adverse events and 0 affected. For other adverse events, it likewise reports 28 at risk and 0 affected. Deaths are recorded as 0. No serious adverse-event terms are listed.

These are registry-recorded counts for this single-group study. They should be read as reported counts, not as proof that an event was impossible or that the sensor has no safety risks. The entry does not provide event narratives, severity classifications beyond the serious-event category, follow-up completeness, or a detailed account of how adverse events were collected. The timeframe for safety reporting is not reported in the registry record.

The at-risk figures are not a measure of exposure duration, and the registry record does not report how much time each subject used the sensor. The absence of listed serious events also does not inform the primary RMS accuracy endpoint. Safety reporting and measurement performance are distinct aspects of the study.

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

The registry record establishes that a prospective, non-randomized, single-group study evaluated the RD SET Neo SpO2 investigational sensor in neonatal subjects. It also describes an RMS-based approach to comparing sensor readings with a laboratory pulse Co-Oximeter reference. Those design details clarify what investigators intended to assess and how the primary measurement summary was defined.

The record does not report the numerical RMS error result, an uncertainty interval, or the endpoint analysis denominator. Consequently, it does not allow a quantitative judgment about the observed degree of agreement or the precision of an estimate. The result should not be inferred from the study’s completion status, enrollment, or description of its regulatory objective.

The design also does not establish comparative effectiveness against another device or treatment. There was no parallel control arm, and no randomized allocation is listed. Agreement with a laboratory reference is a different question from whether the sensor performs better or worse than an alternative in a comparative trial.

The interpretation of any performance estimate would depend on the measurement and analysis details, including repeated observations, the enrolled subjects, and the handling of the reported deviations and exclusions. Those details are not fully described in the registry record. In particular, the entry does not say whether the identified outlier data points or subjects with protocol deviations were included in the endpoint calculation.

The safety entry reports zero affected for serious adverse events and zero deaths, with the at-risk group figures recorded as 28. These counts describe what is entered in the registry; they do not establish the absence of all possible harms. The record does not give a detailed safety collection method or follow-up account.

Thus, the registry record supports a limited description of study purpose, design, endpoint definition, and recorded safety counts. It does not establish a numerical accuracy result, statistical precision, performance in populations beyond the enrolled sample, or a comparative advantage. Those conclusions would require outcome and analysis details that are not reported in the registry record.

13. Why This Trial Matters Statistically

This study is statistically notable because its central question concerns agreement between a measurement device and a reference method, not a treatment contrast. The distinction affects which summaries are relevant. For paired measurement data, a method such as the stated RMS calculation combines average offset and dispersion of the paired differences. A comparison of group means alone would not convey the same information about measurement agreement.

The registry’s definition also illustrates why a single endpoint can depend on multiple statistical components. Bias and precision capture different characteristics of discrepancies, and RMS combines them. Without the numerical estimate and uncertainty, however, the endpoint definition cannot demonstrate what performance was observed. Transparent outcome reporting is essential for a reader to assess both the magnitude and reliability of a performance summary.

The record’s notes about protocol deviations, outlier points, and an ISO-based removal make analysis-population transparency particularly relevant. Exclusion decisions can change the set of paired observations used and may influence an estimate. Reporting the endpoint denominator, the reason for each exclusion, and the sensitivity of results to those decisions would help readers assess robustness. Those details are not reported in the registry record.

The study also underscores the importance of describing repeated-measure structure. If subjects contribute multiple measurements, the observations may be correlated. Appropriate uncertainty calculations need to reflect the way data were collected. The registry record does not report enough measurement-level detail to determine how this issue was addressed.

Finally, a completed study entry with a defined endpoint is not equivalent to a fully interpretable numerical result. The registry lists one primary outcome but does not report its value or analysis. Statistical interpretation depends not only on naming an estimand and method, but also on reporting the estimate, its precision, and the analysis population.

14. 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 main limitation for interpreting this registry entry is that it does not report the numerical primary endpoint result. Neither the RMS estimate nor its uncertainty is available in the registry record. The record therefore cannot show the magnitude or precision of measurement agreement.

The entry reports enrollment of 28, but not the number of subjects or paired samples included in the accuracy analysis. It identifies subjects 7 and 11 as having protocol deviations, subjects 19 and 26 as having outlier data points, and subject 5 as removed under the stated ISO standard for SpO2 equal to 100. The record does not explain how these observations were handled in the endpoint calculation or whether alternative analyses were performed.

The study used convenience sampling in a clinical environment and focused on a neonatal population. The registry record does not report enough detail about eligibility, recruitment, or participant characteristics to assess how well the enrolled subjects represent a broader target population. Generalization beyond the study setting is therefore uncertain.

The record does not specify the number or timing of readings per subject, the approach to correlated repeated measurements, or the endpoint analysis population. It also does not provide a full account of missing measurements, multiplicity procedures, or the protocol’s sample-size justification. These omissions limit reproducibility and prevent a complete assessment of the statistical analysis.

There was no parallel control group, so the study cannot provide a randomized comparison with another device or intervention. Safety information is limited to the recorded adverse-event counts; the safety collection timeframe and event ascertainment details are not reported in the registry record. These limitations concern what can be concluded from the registry entry and do not establish that unreported information was not collected during the study.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Feasibility and Performance Evaluation for INVSENSOR00024 — NCT03570255.
• Registry record last updated 2020-02-20; status last verified 2020-02.
• Results first posted 2020-02-20.
• 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.

18. Record Summary

NCT03570255 is a completed, prospective, non-randomized, single-group study sponsored by Masimo Corporation. The registry lists enrollment of 28 and evaluates the RD SET Neo SpO2 investigational sensor in neonatal subjects using convenience sampling in a clinical environment. The study description includes feasibility work under Protocol Version 1 and clinical performance evaluation under Protocol Version 2.

The primary outcome is accuracy of SpO2 measurement by RMS calculation, comparing noninvasive readings with a laboratory pulse Co-Oximeter reference, over a timeframe of up to 8 hours. The registry explains the RMS method but does not report the numerical endpoint result, an uncertainty measure, or the analysis denominator.

The record notes protocol deviations for subjects 7 and 11, outlier data points for subjects 19 and 26, and removal of subject 5 under the stated ISO standard for SpO2 equal to 100. It records 0 affected for serious adverse events, 0 affected for other adverse events, and 0 deaths. Details of endpoint handling, missing data, and the statistical analysis population are not reported in the registry record.