Accuracy of Noninvasive Rainbow Adhesive R1 Pulse Oximeter Sensor: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Accuracy of Noninvasive Rainbow Adhesive R1 Pulse Oximeter Sensor (NCT03134313), a na na study. The primary outcome is Accuracy of SpHb on R1-25 by Arms Calculation.

2. Clinical Question

The clinical question is about measurement accuracy rather than a treatment effect. The study description proposes comparing noninvasive readings with blood-sample values. The outcome definition says the sample-based difference is used to calculate bias and precision, which are then combined into an arithmetic root mean square error. This type of summary can describe both systematic offset and variability in the paired measurements.

The record describes a single test group. It does not describe a randomized comparison between different sensors, a no-sensor group, or a treatment-control contrast. The intended inference is therefore limited to the accuracy assessment described for this sensor and study setting. The registry record does not report the outcome estimate, the number of paired measurements, or participant-level results, so the degree of agreement cannot be evaluated from the recorded results.

3. Trial Design

The registry identifies this as a single-group device study of the Rainbow Adhesive Noninvasive R1 Pulse Oximeter Sensor. Enrollment is listed as 25, and the population is described as healthy. The arm description says all subjects were enrolled in the test group and received the sensor. There is no separate control arm in the described design.

Allocation is recorded as not applicable, and masking is listed as none. These entries are consistent with a measurement evaluation in which the sensor reading is compared with a blood-sample reference, rather than a randomized comparison of clinical interventions. The blood-sample measurement serves as the reference for assessing the sensor reading; it is not described as a parallel trial arm.

The study summary describes measuring hemoglobin noninvasively and comparing it with hemoglobin from a volunteer’s blood sample. The primary outcome timeframe is 1-5 hours per subject. The record does not state how many paired readings were collected per participant, how the timing of readings and blood sampling was managed, or whether all recorded readings were included in the endpoint calculation.

The registry lists the study as completed, with a start date of 2008-12-12 and a primary completion date of 2009-06-19. The sponsor is Masimo Corporation. The listed phase is NA, which is not a phase designation for this device evaluation. The record does not report a protocol rationale for the enrollment, a sample-size justification, or details of monitoring during data collection. Completion status and enrollment do not, by themselves, establish that the primary outcome was analyzed or that results were posted.

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

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 SpHb on R1-25 by Arms Calculation, measured over 1-5 hours per subject.

The primary endpoint is titled Accuracy of SpHb on R1-25 by Arms Calculation, with units of g/dL. It is intended to summarize the discrepancy between the sensor’s noninvasive hemoglobin measurement and the hemoglobin value from a blood sample. The registry describes subtracting the blood-sample value from the pulse oximeter value for a number of samples.

The mean of these differences is called bias in the endpoint description. Bias represents the average direction and size of the measurement offset: a nonzero average difference indicates that readings tend to lie on one side of the reference values. The standard deviation of the differences is called precision in the description. It characterizes how much individual differences vary around their average.

The arithmetic root mean square error combines these features by taking the square root of the sum of squared bias and precision. In general, this kind of combined summary is useful because small average bias alone does not guarantee that individual readings are close to their reference values; variability also matters. Conversely, a measure of spread alone does not capture a consistent offset. The combined endpoint is intended to reflect both aspects of agreement in one summary, expressed in the stated measurement units.

The registry record defines the calculation but does not report an endpoint value, its uncertainty, or the underlying paired measurements. It also does not report a clinically acceptable accuracy threshold. Accordingly, the endpoint’s meaning can be described, but its observed magnitude and practical interpretation cannot be determined from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The endpoint uses paired differences. For each comparison described, the sensor value is compared with the blood-sample hemoglobin value. The registry defines the difference as the pulse oximeter value minus the blood-sample value. This direction matters when interpreting bias: a positive average difference would mean sensor values tend to be higher than reference values, while a negative average difference would mean they tend to be lower. The registry record does not report the observed direction or size of the bias.

What does bias describe?

Bias is the average of the paired differences, as stated in the endpoint definition. In general, an average difference helps identify systematic offset. If differences tend to be consistently positive or consistently negative, the average reflects that direction. A bias close to zero would indicate little average offset, but it would not necessarily mean that individual measurements are close to the reference. Positive and negative differences can cancel in an average even when individual differences vary substantially.

What does precision describe?

The endpoint description calls the standard deviation of the paired differences precision. In general, standard deviation measures the spread of observations around their average. For this endpoint, it describes the variability of the sensor-reference differences around the estimated bias. Greater spread means the paired differences are less tightly clustered around their average. The registry does not report this standard deviation, nor does it explain whether any additional modeling was used to account for the way samples were collected.

How does the combined error work?

The arithmetic root mean square error is defined in the record as the square root of the sum of the squares of bias and precision. Squaring makes the contributions nonnegative, and taking the square root returns the summary to the original measurement scale. In general, this combines systematic offset and variation rather than treating either feature as sufficient on its own. A low average bias cannot alone establish close agreement if differences are highly variable. Likewise, a narrow spread around a consistently shifted average would not capture the offset if only variability were considered.

The endpoint is expressed in g/dL. Its units indicate the scale on which the difference summary is reported; they do not establish whether a particular result is acceptable. A threshold or decision criterion for interpreting the value is not reported in the registry record. Nor does the record report an observed value, so the general properties of the calculation cannot be translated into a study-specific statement about accuracy.

Why does the reference measurement matter?

Measurement accuracy is assessed relative to a reference. Here, the study summary identifies hemoglobin from a blood sample, with measurement by a laboratory analyzer described as the comparison basis. In general, an accuracy estimate reflects not only the device being assessed but also the reference method and the conditions under which paired readings are obtained. The registry record does not provide further details about the laboratory analyzer, sample handling, calibration, or procedures for pairing each sensor reading with a blood-sample value.

The description refers to calculating the endpoint across a number of samples, while the timeframe is stated as 1-5 hours per subject. It does not state how many samples were taken from each participant or whether samples from the same participant were treated as statistically related. In general, repeated observations from one participant may be correlated rather than independent. If such dependence is present, analysis methods that treat every observation as independent can understate uncertainty. The registry record does not report how this issue was handled, so no particular analysis approach should be assumed.

What can be learned from the registry entry?

The outcome definition describes a planned calculation, not a posted result. A complete numerical interpretation would ordinarily consider the reported estimate, its uncertainty, the number and structure of measurements, and a stated criterion for acceptable performance. These details are not reported in the registry record. The record also lists no analysis entries. That absence does not prove that no analysis was conducted; it means the registry record does not document an analysis result for interpretation.

The planning assumptions are separate from the endpoint method. The planning illustration uses a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and planning for stated fractions of an observed effect. The record identifies those as planning assumptions, not as the study’s own design. They do not provide the missing observed endpoint result, and they should not be mistaken for observed power or evidence about the sensor’s measurement performance.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome measure and no analysis entries. It does not describe a statistical analysis plan, a hierarchy of outcomes, adjustment for multiple testing, or rules for selecting among analyses. Because no numerical outcome analysis is reported, the record also does not show whether any multiplicity procedure was applied. The absence of such detail should be read as missing reporting, not as evidence that a particular method was or was not used.

Missing-data handling is not reported in the registry record. The entry does not say whether any sensor readings or blood samples were missing, how unusable measurements were defined, whether incomplete participant records were excluded, or how missing pairs would enter the endpoint calculation. Without those details, it is not possible to assess whether the reported analysis, if conducted, represented all intended measurements or a selected subset.

The analysis population is also not described beyond the single test group. The record does not define an analysis set, state whether all enrolled participants contributed to the primary endpoint, or identify exclusions from the accuracy calculation. Enrollment is listed as 25, but that figure alone does not establish how many participants or paired measurements were included in an endpoint analysis.

In general, these details matter because analytic choices can affect the estimate and its uncertainty. For example, exclusions may alter which measurements contribute, while repeated observations from the same participant may require methods that account for their relationship. The record does not provide enough information to determine how such issues were handled. No missing-data assumptions, sensitivity analyses, or analysis-population definitions should be inferred from the endpoint description alone.

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 safety entry lists the R1-25 Sensor group with 25 participants at risk for the serious adverse event summary. It records 0 participants with a serious adverse event and 0 deaths. No serious adverse event terms are listed.

These figures describe the registry’s recorded serious-event counts for this group. They do not provide a complete account of all possible adverse events, their severity, timing, relatedness to the sensor, or methods of collection. The record does not list non-serious adverse-event results in the information available for this summary, so the serious-event figures should not be interpreted as showing that no adverse events of any kind occurred.

In general, an observed count of zero is a description of events recorded in the study, not proof that the risk of an event is absent. The interpretation of safety counts also depends on how events were defined, monitored, and reported. Those procedures are not described in the registry record. The safety entry should therefore be read narrowly: it reports 0 serious adverse events and 0 deaths in the listed group, without establishing a broader safety conclusion.

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

The record establishes the question the study intended to address: whether a noninvasive pulse oximeter sensor’s hemoglobin readings agree with blood-sample hemoglobin values, summarized by an arithmetic root mean square error. It identifies healthy volunteers, a single sensor group, the measurement reference, and a stated timeframe. These details help explain the design and the intended endpoint.

The record does not report the calculated endpoint value. It therefore does not show how large the observed differences were, whether readings tended to be higher or lower than the reference, how much individual differences varied, or how precisely any estimate was measured. Without those results, it is not possible to characterize the observed degree of agreement or to compare it with a performance criterion. No such criterion is reported in the registry record.

This is a measurement-accuracy question, not a randomized test of a clinical treatment. The design described has no separate control arm, and allocation is listed as not applicable. The blood sample is the reference for the sensor measurement, not a comparator group for estimating a treatment effect. Any interpretation should stay within that distinction.

The study is listed as completed, but completion status is not itself an outcome result. The registry also lists enrollment of 25, but does not state how many participants or paired readings contributed to an endpoint analysis. There are no listed analyses. As a result, the entry supports an explanation of the planned measurement comparison and its statistical summary, but not a conclusion about the numerical accuracy observed.

More broadly, an accuracy result is specific to the population, reference procedure, and measurement conditions used. The registry record does not give enough information to assess variation across other populations or settings. Any broader interpretation would go beyond what the recorded information establishes.

13. Why This Trial Matters Statistically

This study illustrates a statistical distinction between evaluating a measurement and comparing clinical outcomes. Its primary question concerns agreement between a device reading and a reference value. The endpoint is constructed from paired differences, so its interpretation depends on both systematic offset and variation across measurements. This framework is useful when a device is judged against a reference rather than against a separate randomized group.

The arithmetic root mean square error also highlights why one summary can combine complementary features. Bias captures the average direction of discrepancy, while the standard deviation of differences captures spread. Considering both avoids treating a small average offset as sufficient evidence of close individual agreement. The registry describes this calculation clearly, even though it does not report the resulting estimate.

The record also shows why a defined endpoint is not the same as a reported result. A completed study, an enrollment figure, and a detailed endpoint description do not reveal the observed accuracy in the absence of the estimate and its uncertainty. Statistical interpretation requires the actual outcome data or reported analysis, along with sufficient information about the analysis population and measurement structure.

Finally, the single-group structure matters. The blood-sample measurement is a within-study reference, not a randomized control arm. Planning assumptions based on equal allocation between two arms are identified as illustrative rather than as the trial’s design. Keeping those roles distinct prevents a planning calculation from being mistaken for a study result or a direct description of the design.

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 of the registry record for statistical interpretation is that it defines the primary endpoint without reporting its value. Bias, precision, and arithmetic root mean square error are not reported, and there is no uncertainty interval or participant-level measurement information. The record therefore cannot establish the observed magnitude or direction of error.

The measurement process is not described in enough detail to assess how readings were paired. The outcome description refers to a number of samples, and the timeframe is 1-5 hours per subject, but the number of readings per participant and the handling of repeated measurements are not reported. The registry also does not describe the laboratory analyzer, sample handling, calibration, or procedures for quality control.

Other information relevant to interpreting an accuracy estimate is also absent. The record does not report a criterion for acceptable performance, missing-data procedures, exclusions, a defined analysis population, or a statistical analysis plan. It does not say whether uncertainty was estimated in a way that accounted for possible within-participant dependence.

The population is described as healthy, and the record does not report participant characteristics or results for other populations or settings. The single-group design does not provide a randomized comparison with another device or method; instead, the blood-sample value is the reference measurement. The planning assumptions described elsewhere are explicitly illustrative and do not replace the missing endpoint result or document the study’s design rationale. These limitations restrict interpretation to the study objective and calculation method recorded, without supporting a numerical conclusion about accuracy.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Accuracy of Noninvasive Rainbow Adhesive R1 Pulse Oximeter Sensor — NCT03134313.
• Registry record last updated 2017-07-02; status last verified 2017-06.
• Results first posted 2017-07-02.
• 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

The registry record describes a completed, single-group device study of the Rainbow Adhesive Noninvasive R1 Pulse Oximeter Sensor in healthy volunteers. Enrollment is listed as 25. The study summary proposes comparing the sensor’s noninvasive hemoglobin measurement with hemoglobin from a blood sample, using an arithmetic root mean square error as the primary accuracy summary. The outcome timeframe is 1-5 hours per subject.

The record lists allocation as not applicable and masking as none. It gives the study start date as 2008-12-12 and primary completion date as 2009-06-19, with Masimo Corporation listed as sponsor. The serious adverse event entry reports 0 affected participants and 0 deaths in the sensor group.

The registry record defines the endpoint calculation but does not report its numerical result, an uncertainty estimate, or any analysis entries. It also does not report missing-data methods or analysis-population details. Thus, the intended statistical comparison is clear, but the observed accuracy cannot be evaluated from the recorded results.