Desaturation Validation of INVSENSOR00031: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Desaturation Validation of INVSENSOR00031 (NCT03783780), a na na study. The primary outcome is Oxygen Saturation (SpO2) Accuracy of Sensor by Arms Calculation During Non-motion.

2. Clinical Question

The study's stated purpose was to evaluate the accuracy of a noninvasive oxygen-saturation measurement against laboratory blood-gas reference values. The procedure involved gradually reducing the oxygen concentration inhaled by healthy adult subjects until arterial oxygen concentration was approximately 70%. The registry summary also says that the hand and arm without the arterial line were subjected to motion at various times, followed by a return to room air.

The primary question concerns non-motion measurements, not motion performance. Motion accuracy is identified as a secondary outcome. The record therefore describes more than one measurement condition, but the primary endpoint is specifically the non-motion ARMS measure. It does not report the numerical value of that endpoint, so the degree of agreement cannot be judged from the outcome description alone.

This is a measurement-validation question rather than a treatment-versus-control question. Arterial blood sampling supplies the reference for comparison; it is not an independent participant group. The record does not identify a randomized comparator, and its allocation field is listed as NA. The primary measure was collected over a timeframe of 1-5 hours. The clinical relevance of the measurement should therefore be interpreted within the described study setting and population, rather than assumed to establish performance in other populations or conditions.

3. Trial Design

The registry identifies this as a completed, single-group device study of INVSENSOR00031. The sponsor is Masimo Corporation. The listed enrollment is 21, the phase is NA, allocation is NA, and masking is none. The study start date is 2018-12-11, and primary completion is listed as 2018-12-18. These registry fields describe the design and record status; they do not establish what procedures or analyses were prespecified beyond the descriptions entered in the record.

All subjects who enrolled in the test group and participated in data collection received the device during motion and non-motion. There is no separate control arm described. Instead, the sensor measurement is assessed against arterial oxygen-saturation values obtained from reference blood samples. This paired measurement structure is suitable for asking how sensor readings compare with a reference, but it is not a randomized comparison of two interventions.

The brief summary describes a controlled desaturation procedure in healthy adults: inspired oxygen concentration was slowly reduced until arterial oxygen concentration was approximately 70%, with movement of the hand and arm without the arterial line at various times. Subjects were then returned to breathing room air. The record does not give further procedural detail about the timing or extent of motion, the precise conduct of the desaturation procedure, or how many measurements were collected from each subject.

The primary outcome is non-motion SpO2 accuracy calculated using ARMS, with a timeframe of 1-5 hours. A motion ARMS measure is listed as a secondary outcome. Although the registry says 20 subjects participated, only ten of those 20 underwent testing for the primary non-motion outcome. The reason for this subset and the relationship between that number and the listed enrollment of 21 are not reported in the registry record. The record is marked completed, but completion status does not itself provide an outcome estimate or confirm that every enrolled subject contributed to every outcome.

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: Only ten subjects, out of the 20 who participated in the study, underwent testing for this non-motion outcome measure.

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 Oxygen Saturation (SpO2) Accuracy of Sensor by Arms Calculation During Non-motion, measured over 1-5 hours.

The primary endpoint is the oxygen-saturation accuracy of INVSENSOR00031 during non-motion, expressed as an ARMS value. The sensor provides a noninvasive SpO2 measurement, which the study description compares with SaO2 from an arterial reference blood sample. The endpoint is reported in the unit “% of oxygen saturated hemoglobin” and has a timeframe of 1-5 hours. The registry states that only ten subjects, out of the 20 who participated in the study, underwent testing for this non-motion outcome measure.

In general, an arithmetic root mean square error measure combines the sizes of differences between a measurement method and its reference. It reflects the magnitude of errors, rather than whether readings tend to be higher or lower. Because squaring gives greater influence to larger deviations before averaging and taking the root, the statistic can be affected by relatively large errors. A lower ARMS value would ordinarily indicate closer agreement under the stated calculation, but the record does not report the primary ARMS result, so no direction or magnitude of observed performance can be described.

Using an arterial reference makes the endpoint a comparison of measurement accuracy against a specified reference method. It does not, by itself, indicate how well the sensor performs in every setting or population. The registry does not report the result's uncertainty, the underlying paired observations, or any additional details of the calculation beyond the stated ARMS approach. Motion is addressed separately as a secondary outcome and should not be conflated with the primary non-motion endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation begins by identifying what is being compared. Here, the sensor's noninvasive SpO2 reading is compared with an arterial SaO2 reference value. This is a measurement-agreement question. It differs from a trial that compares outcomes between randomized groups: the record describes one test group, and the arterial reference is a measurement method rather than a second treatment arm.

What does ARMS represent?

In general, root mean square error summarizes the magnitude of deviations between paired measurements and a reference. The deviations are squared, averaged, and then converted back to the original scale by taking a square root. This puts greater weight on larger errors than a simple average of signed differences would. Because errors are not retained as positive or negative in the final magnitude summary, ARMS does not show whether sensor readings tend to overestimate or underestimate the reference. It also does not describe the full distribution of errors. A reported ARMS value would need to be considered alongside the number and range of observations, and ideally other information about agreement, to understand performance in detail. Such numerical results are not reported in the registry record.

The endpoint is titled as oxygen-saturation accuracy during non-motion, with the unit “% of oxygen saturated hemoglobin.” The outcome description says it is calculated by comparing sensor SpO2 with arterial SaO2. The stated timeframe is 1-5 hours. That timeframe identifies when the measurement outcome was assessed in the registry; it does not describe the number of readings or specify how observations were summarized across subjects. Those details are not reported in the registry record.

Why does the reference method matter?

In general, accuracy is assessed relative to a reference considered suitable for the measurement question. The study summary identifies arterial blood samples analyzed by a laboratory blood-gas analyzer as the source of reference values. A comparison with this reference addresses agreement under the study's conditions, rather than agreement with an unobserved true value. The record does not report information about reference-method precision, calibration procedures, or how paired sensor and blood-sample measurements were matched in time.

The study's planned procedure included desaturation and periods of motion, with the arm and hand without the arterial line moved at various times. The primary endpoint concerns non-motion; motion accuracy is listed separately as a secondary outcome. In general, separating conditions can help distinguish measurement performance across different circumstances, but the registry does not report detailed condition-specific results or the protocol's rules for classifying motion and non-motion periods.

What can an uncertainty interval add?

In general, a confidence interval describes the statistical uncertainty around an estimated quantity under the assumptions of the method used. It can show which values are reasonably compatible with the data and analysis, rather than presenting only a point estimate. For a measurement-accuracy estimate, the interval's width would help readers judge precision. No confidence interval or other uncertainty estimate for the primary endpoint is reported in the registry record. Without the point estimate and its uncertainty, the magnitude and precision of the observed non-motion accuracy cannot be evaluated.

The primary outcome was assessed in ten subjects, while the record says 20 subjects participated. In general, estimates based on fewer independent subjects tend to be less precise than estimates based on more subjects, although precision also depends on the variability and structure of the measurements. Multiple readings from a subject are not automatically equivalent to measurements from separate subjects: readings within one subject may be correlated. The registry does not report the number of readings per subject or the analysis method used to account for repeated measurements. Therefore, the effective information behind any result cannot be reconstructed from the participant counts alone.

How should planning calculations be distinguished from results?

The recorded planning assumptions describe an illustration computed from an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. In general, a sample-size calculation asks how much information a future design might need under specified assumptions. It is not a description of the information actually collected, and it is not observed power. Here the stated equal allocation between two arms does not match the registry's single-group design as described. The calculation is therefore best understood as a planning illustration, not a report of the trial design. The observed effect and the numerical planning outputs are not reported in the registry record.

In general, over-dispersion means that observed variability exceeds what a chosen statistical model expects. If present, it can make standard errors too small and uncertainty appear narrower than warranted unless the model accounts for it. The registry record does not report a model for count outcomes or any assessment of over-dispersion, so this cannot be evaluated for this study. More broadly, the record does not report the statistical analysis plan, handling of repeated measurements, or criteria for interpreting accuracy. Statistical concepts can explain what additional reporting would clarify, but they cannot replace the absent estimates and analysis details.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures: the primary non-motion SpO2 accuracy measure and a secondary motion accuracy measure. It does not report a multiplicity adjustment, a hierarchy for formal testing, or an analysis plan specifying how results across outcomes would be interpreted. If multiple outcomes are tested, general statistical practice recognizes that the chance of at least one apparently notable finding can rise when results are considered together. Whether that issue applies to confirmatory testing in this study cannot be determined because the planned testing strategy is not reported in the registry record.

The record reports that only ten subjects, out of 20 participants, underwent testing for the primary non-motion outcome. It does not explain why the primary outcome was available for that subset, whether the remaining subjects lacked measurements, or whether any observations were excluded. Missing-data methods, reasons for missingness, and sensitivity analyses are not reported in the registry record. The enrollment of 21 is also not reconciled with the statement that 20 subjects participated. These counts should be retained as recorded rather than treated as interchangeable populations.

The analysis populations are not reported in the registry record. There is no listed analysis among the registry's outcome information, and the record does not state whether analyses were based on all enrolled subjects, participating subjects, or subjects with evaluable paired measurements. It also does not describe whether measurements were summarized per subject or analyzed at the reading level.

In general, transparent reporting of outcome availability, exclusions, analysis populations, and multiplicity procedures helps readers understand how estimates were selected and how broadly they apply. Those details are not reported in the registry record, so the effect of missingness or multiple outcome assessment on interpretation cannot be established.

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.

Most frequent non-serious adverse event terms

The safety figures list the INVSENSOR00031 group with 21 subjects at risk for serious adverse events and zero subjects affected. The record also lists 21 subjects at risk for other adverse events and one subject affected. Deaths are listed as zero. No serious adverse-event terms are listed in the registry record.

These are counts of subjects, not a description of event severity beyond the registry's serious and other categories. The single subject in the other adverse-event category should not be reclassified based on information that is not reported. The record does not provide the event term, timing, duration, outcome, or an assessment of relationship to the device. It also does not provide detail about how adverse events were collected or the period over which subjects were monitored.

In general, a zero count in a study of this size does not establish that an event cannot occur; it describes what was recorded in this study. Likewise, the other-adverse-event count alone does not establish cause. Because the study has one group and no randomized comparator, these figures do not support a comparative safety estimate. The registry's safety counts should be read as descriptive reporting, not as proof of absence or presence of a causal safety effect.

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

The registry record establishes that a completed single-group study was designed to compare INVSENSOR00031 noninvasive oxygen-saturation readings with arterial reference measurements in healthy adults. Its primary endpoint is an ARMS accuracy measure during non-motion, while motion accuracy is listed as a secondary outcome. These details define the question and measurement approach, but the numerical outcome results are not reported in the registry record.

Accordingly, the record does not allow readers to quantify the sensor's observed non-motion accuracy, determine the precision of an estimate, or assess whether accuracy differed during motion. It also does not report the underlying paired data, a confidence interval, or a listed statistical analysis. The planned desaturation to approximately 70% describes the study procedure; it is not a result showing the distribution of measured oxygen values or performance across a particular range.

The study's population and outcome coverage also matter. The record describes healthy adults, and reports that only ten of the 20 participants underwent testing for the primary non-motion outcome. It does not report why this subset contributed to the endpoint, or reconcile the participation count with enrollment of 21. Generalization to other populations or settings is not established by this record. The study summary's mention of motion and return to room air likewise does not provide separate outcome estimates for those conditions.

The adverse-event listing records zero subjects with serious adverse events, one subject with an event in the other category, and zero deaths. These are descriptive counts in a single group and do not establish causation or comparative safety.

The record therefore supports a description of the validation question and study setup, but not a conclusion about the magnitude or practical significance of observed performance. It neither establishes that the sensor met a performance criterion nor establishes that it failed one: a criterion and numerical result are not reported in the registry record. Any interpretation beyond the stated design, endpoint definition, participant counts, and safety counts would require additional results or methods information.

13. Why This Trial Matters Statistically

Measurement studies raise a statistical question distinct from the usual comparison of treatment outcomes: how closely does a device agree with a reference, and how much uncertainty surrounds that estimate? This study's stated use of arterial blood-gas values as a reference and ARMS as the primary non-motion summary identifies that central issue. A clear distinction between the reference method, the sensor reading, and the calculated error is essential for interpreting device-validation data.

The study also identifies motion as a separate measurement condition. That distinction matters statistically because measurement error may depend on conditions, and a result averaged across conditions could obscure such differences. The registry lists motion as a secondary outcome, but does not report numerical condition-specific findings. Keeping primary non-motion performance distinct from secondary motion performance is therefore important when reading the outcome descriptions.

Participant accounting is another key issue. Enrollment is listed as 21, the record says 20 subjects participated, and only ten of those 20 underwent testing for the primary non-motion outcome. Those are not equivalent analysis populations. The reason for the smaller endpoint subset and the treatment of the other participants are not reported in the registry record. Transparent accounting would help readers assess how much information supports the primary estimate and whether the measured subset represents the wider participant group.

Finally, the record demonstrates why design information and results must be read separately. A completed status and a defined statistical endpoint do not substitute for an estimate, uncertainty interval, or analysis description. Without those results, the study can be understood as a planned comparison of sensor and reference measurements, but its observed accuracy cannot be assessed from this record. That boundary is a central statistical finding about the available reporting, not a judgment about the device.

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 primary limitation for interpreting this record is that numerical results for the primary endpoint are not reported in the registry record. The motion outcome is also listed without a numerical result. No estimate, confidence interval, distribution of paired errors, or analysis output is available in the record, so the observed magnitude and precision of accuracy cannot be assessed.

The primary non-motion endpoint applies to ten subjects, while 20 subjects are described as having participated and enrollment is listed as 21. The record does not explain the difference among these counts or why only a subset underwent the primary outcome testing. It does not report missing-data reasons, exclusions, or the analysis population. This limits assessment of how representative the endpoint subset may be.

The description identifies healthy adults and a controlled desaturation procedure, including oxygen concentration reduced until arterial oxygen concentration was approximately 70%. It does not report detailed results across the observed oxygen-saturation range, or establish performance in populations or circumstances beyond those described. Motion is mentioned in the procedure and as a secondary outcome, but the registry does not give a numerical motion result or detailed classification of motion conditions.

The statistical analysis plan, handling of repeated measurements, uncertainty calculation, and any prespecified performance threshold are not reported in the registry record. The record also does not provide detail on reference-method precision or the matching of sensor readings to arterial samples. These limitations concern what can be interpreted from the registry entry and should not be mistaken for proof that procedures or analyses were absent from the study itself.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Desaturation Validation of INVSENSOR00031 — NCT03783780.
• Registry record last updated 2019-12-17; status last verified 2019-11.
• Results first posted 2019-07-02.
• 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

NCT03783780 is a completed, single-group device study sponsored by Masimo Corporation, with enrollment listed as 21 and no masking. It evaluates noninvasive oxygen-saturation measurements from INVSENSOR00031 against arterial blood-gas reference values in healthy adults undergoing desaturation.

The primary outcome is non-motion SpO2 accuracy calculated using ARMS over a timeframe of 1-5 hours. The record says ten of the 20 participants underwent this primary outcome test. A motion accuracy measure is listed as secondary. Numerical outcome results and a listed analysis are not reported in the registry record. The safety listing reports zero subjects with serious adverse events, one with an event in the other category, and zero deaths.