A Pivotal Study to Evaluate Enlite™ Sensor Performance With iPro™2 in China: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Pivotal Study to Evaluate Enlite™ Sensor Performance With iPro™2 in China (NCT03417466), a na na study. The primary outcome is Mean Percentage of Enlite Sensor Values That Are Within 20% Agreement of Gold Standard (Yellow Springs Instrument (YSI) YSI Plasma Glucose Values).

2. Clinical Question

3. Trial Design

This was a completed, single-group device study titled “A Pivotal Study to Evaluate Enlite™ Sensor Performance With iPro™2 in China.” The registry lists the phase as NA, allocation as NA, and masking as none. Those entries describe a study without a randomized, masked, parallel comparator; they do not establish how every operational decision was made.

The intervention was Enlite 1 connected to an iPro2 recorder. The study description specifies sensor use over 144 hours (6 days), with insertion in the abdomen. Participants underwent one YSI frequent sample testing session on Day 1, Day 3-4, or Day 6 of sensor wear. YSI values served as the reference for the paired glucose comparisons.

The registry lists enrollment as 72. It separately describes 66 subjects as having an Enlite sensor inserted for the primary-endpoint population, and says one of those subjects had no paired Enlite and YSI measurement. The record does not explain the difference between the enrollment figure and the number described as having an inserted sensor. The reason for the absent paired measurement is not reported in the registry record.

The study started on 2018-06-28, had primary completion on 2018-09-20, and is listed as completed. The stated purpose was to evaluate sensor performance and safety. Secondary outcomes included mean absolute relative difference (MARD) and Clarke and Consensus Error Grid analyses of paired sensor and YSI plasma glucose values. The registry lists four outcome measures, but the results and analyses for those outcomes are not reported in the registry record.

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.

Before assignment: A total of 72 subjects signed the informed consent form and completed the screening. Five subjects failed the screening. One subject withdrew before wearing the sensor.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Subjects who have Enlite sensor inserted and have at least one paired Enlite sensor and YSI measurement. 66 subjects have Enlite sensor inserted, of which one subject has no paired Enlite sensor and YSI measurement.

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 Mean Percentage of Enlite Sensor Values That Are Within 20% Agreement of Gold Standard (Yellow Springs Instrument (YSI) YSI Plasma Glucose Values), measured over YSI FST days (Day 1, 3-4, or 6 of sensor wear).

The primary endpoint was the mean percentage of primary Enlite Sensor values meeting an agreement criterion against YSI plasma glucose values during frequent sample testing days. The criterion was agreement within 20%; when the YSI value was at or below 80 mg/dL, the description specifies agreement within 20 mg/dL. The unit is percentage of paired readings, and the endpoint is a mean of a continuous measure, with standard deviation identified as its dispersion measure.

Paired measurements are central to this endpoint: each sensor value is assessed against a YSI value taken as the reference. The percentage expresses how often sensor readings fall within the stated agreement limits among the paired readings considered. The mean then summarizes that percentage for the endpoint population. This is a measure of agreement under the study’s specified testing conditions; it is not itself a measure of clinical outcomes or a comparison with another treatment group.

The registry ties the endpoint to YSI frequent sampling on Day 1, Day 3-4, or Day 6 of sensor wear. Its description defines the population as subjects with an inserted Enlite sensor and at least one paired sensor and YSI measurement. It reports that 66 subjects had a sensor inserted, with one lacking a paired measurement. The endpoint estimate, its standard deviation, and any confidence interval are not reported in the registry record, so the degree of observed agreement cannot be evaluated from these entries.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registered primary endpoint is a continuous summary: a mean percentage of sensor values meeting a paired agreement rule. The registry identifies standard deviation as the dispersion measure. Neither the estimate nor its standard deviation is reported in the registry record. The following concepts explain how such measures are generally read; they do not establish which procedures the study team used.

What does a mean percentage summarize?

For a continuous endpoint, a mean describes the average value of the measured quantity across the analysis units. In this study, the quantity is the percentage of Enlite readings within the endpoint’s agreement limits against YSI values. A percentage based on paired readings is not the same thing as a percentage of subjects who met a separate participant-level success criterion. The registry’s endpoint wording refers to sensor values and paired readings; it does not define a participant-level pass-or-fail rule.

When repeated readings come from the same person, those readings may be more alike than readings from different people. General statistical practice therefore considers the level at which an endpoint is calculated and the dependence among repeated observations. The registry identifies the endpoint population and the mean, but does not report the calculation details needed to determine how repeated readings were handled or how the mean was formed.

What does standard deviation describe?

Standard deviation describes how spread out individual values are around their mean. For this endpoint, it would help indicate how much the subject-level or analysis-unit percentages varied, depending on the method used. It is not the same as a confidence interval: standard deviation describes variability in the observations, while a confidence interval is generally used to express uncertainty around an estimated quantity. The registry does not report the standard deviation or a confidence interval for the primary endpoint.

Why does the paired reference matter?

The endpoint compares Enlite readings with paired YSI plasma glucose values. Pairing means that a sensor value is interpreted alongside its corresponding reference measurement rather than as an unrelated observation. The rule distinguishes agreement within 20% from agreement within 20 mg/dL when the YSI value is at or below 80 mg/dL. This defines what counts as an agreeing pair for the endpoint, but it does not tell the reader how many pairs met the rule or the resulting mean percentage.

Agreement and correlation answer different statistical questions. Two measurement methods can track changes together without yielding sufficiently close values, so a correlation alone would not establish that values agree within a specified tolerance. Here, the registered primary endpoint directly describes the fraction of paired sensor values that satisfy its tolerance rule. The record also lists MARD and Clarke and Consensus Error Grid analyses as secondary outcomes, but their numerical results are not reported.

What can a confidence interval add?

In general, a confidence interval gives a range of values compatible with the data and the method used to calculate the interval. A narrower interval indicates greater precision under that method; a wider interval indicates more uncertainty. An interval does not remove possible bias from study design, measurement procedures, or missing observations. Because no interval is reported for the primary endpoint, its precision cannot be assessed from the registry record.

How should the planning assumptions be read?

The registry describes sample-size calculations as planning illustrations based on the observed effect, not as the trial’s own design. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. Those hypothetical two-arm assumptions do not describe this single-group study’s allocation. Any illustrated sample size is a planning quantity, not observed power and not evidence about the study’s precision. No calculated planning sample-size values are reported in the registry record.

What is not specified about the analysis?

The registry lists zero analyses. It does not report a statistical analysis plan, a model for the primary endpoint, a method for calculating uncertainty, or procedures for handling repeated readings. It also does not report whether analyses were adjusted for participant or measurement characteristics. These omissions prevent a detailed assessment of the calculations behind the endpoint. General statistical principles can clarify the meaning of the registered measures, but cannot fill in those missing study-specific methods.

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("Study Arm")
n   <- c(65)
m   <- c(85.74)
sp  <- c(22.26)
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  Study Arm  65  85.740  22.260  80.224  91.256

11. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures, including the primary agreement endpoint and secondary outcomes for MARD, Clarke Error Grid analysis, and Consensus Error Grid analysis. It lists zero analyses, and numerical results for these outcomes are not reported in the registry record. A statistical analysis plan, including any prespecified testing hierarchy or multiplicity adjustment, is also not reported in the registry record.

Multiplicity is a general concern when several outcomes or statistical tests are examined. With more opportunities to declare a result noteworthy, the chance of at least one apparently unusual finding can increase. Researchers may address this through prespecified priorities or adjustment methods, but the registry does not say whether any such approach was used. It would therefore be inappropriate to infer that the listed outcomes were all tested, or that a particular correction was applied.

The primary-endpoint population description says 66 subjects had an Enlite sensor inserted and one had no paired sensor and YSI measurement. The registry does not report why that pair was absent, how missing measurements were handled, whether other observations were incomplete, or whether any imputation was performed. It also does not report a complete analysis-population definition beyond the stated requirement for at least one paired measurement.

Those details matter because missing paired readings can affect which observations contribute to an agreement estimate. General statistical methods for missing data rely on assumptions about why values are absent; the record does not provide enough information to assess those assumptions. The stated counts should be read as registry descriptions, not as a complete account of participant flow or analytic exclusions.

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.

Most frequent non-serious adverse event terms

The registry lists 66 subjects at risk for serious adverse events, zero affected, and zero deaths. It also lists 66 at risk and five affected in the other adverse-event category. The registry record does not provide terms for serious events, while the list of top serious terms is empty. The meaning and scope of the other category are not defined in the available record.

These are event-category counts, not a comparison between randomized groups. The study had one arm, and the record does not provide event descriptions, timing, duration, severity detail for the other category, or information about how events were collected. It also does not describe an attribution assessment. The counts therefore provide a limited descriptive account of the reported safety entries, rather than a complete characterization of safety.

Zero reported serious events should be read as the registry’s recorded count for this study group, not as proof that serious events are impossible or that a device is safe in every setting. Likewise, the five entries in the other category cannot be interpreted beyond their registry classification without further detail. These figures do not establish a comparative risk or support a treatment recommendation.

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

The study addresses measurement performance under a defined set of conditions: Enlite sensors were used with an iPro2, and sensor values were compared with paired YSI plasma glucose values during frequent sampling testing. The primary endpoint specifies the agreement threshold and the testing days. That design makes the endpoint a direct summary of how often paired readings met the stated rule, rather than a measure of benefit on a health outcome.

The record does not report the primary endpoint estimate or its uncertainty. Consequently, it does not show how large the mean agreement percentage was, how variable it was, or how precisely it was estimated. The secondary outcome names are listed, but their results are not reported either. It is not possible to infer an outcome from the study title, purpose, or completion status.

The single-group design also limits what the study can establish comparatively. YSI provides the reference values for the paired agreement endpoint, but it is not a randomized comparator arm. There is no registered like-for-like control group against which to estimate a difference in performance. The absence of masking is another design feature to consider, while the registry does not describe its practical influence on measurements or conduct.

The registry’s safety counts are limited descriptive information: zero serious events and zero deaths are listed among 66 at risk, while five are listed in the other category. The record does not supply the event details needed for a fuller assessment. These entries should not be generalized beyond the study record.

In short, the registry identifies what was measured and under what broad conditions, but does not report the numerical performance findings needed to assess the observed agreement. It cannot establish a comparative advantage, a clinical outcome effect, or a broad safety conclusion. The distinction between a clearly defined endpoint and a reported endpoint result is essential: the former is available in the registry description, while the latter is not reported in the registry record.

14. Why This Trial Matters Statistically

This trial is statistically notable for the way it frames device performance as agreement with a reference measurement. A sensor reading can be compared directly with a paired YSI plasma glucose value, and the primary endpoint defines an explicit tolerance for deciding whether the pair agrees. That makes the measurement target more concrete than a general statement that two methods are associated.

The record also illustrates why design and endpoint interpretation belong together. The study is single-group, with no masked assignment and no randomized control arm. Its primary question is a paired measurement question, not a comparison of treatments. Reading the endpoint as though it estimated a contrast between groups would misrepresent the design.

Repeated measurements and participant-level summaries raise statistical questions that the registry does not resolve. Participants may contribute paired observations, and the record does not explain the aggregation procedure or treatment of within-person dependence. Those details can influence how variability and precision should be assessed. The listed standard deviation would describe spread, while an interval could communicate uncertainty around an estimate; neither value is reported for the primary endpoint.

Finally, the record separates a defined analysis target from evidence about its result. It names the endpoint and secondary measures, but provides no numerical outcome findings. That distinction matters for evidence appraisal: a protocol-level description can show what investigators intended to measure, but cannot by itself establish the size or precision of the observed agreement. The study is therefore useful as an example of the questions a reader should ask about paired measurement studies, even though the registry entries do not provide the performance estimate needed to answer the primary question.

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 main trial-specific limitation for interpreting performance is that the primary endpoint result is not reported in the registry record. The mean percentage, its standard deviation, and any confidence interval are absent. The record therefore does not permit assessment of either the observed agreement level or its precision.

The study has a single-group design, with allocation listed as NA and masking as none. There is no randomized control arm for a comparative estimate. The endpoint’s paired YSI measurements are a reference, not a control group. Thus, the design does not answer a question about differences between assigned groups.

The registry gives enrollment as 72 and separately describes 66 subjects with a sensor inserted for the primary-endpoint population. It notes that one of those subjects had no paired sensor and YSI measurement. The record does not explain the enrollment-to-insertion difference, the absent pair, or how other incomplete data were handled. It does not provide a full participant-flow account or details of the analysis population.

Methods needed to evaluate the endpoint calculation are also not reported: the handling of repeated readings, aggregation across participants, missing-data procedures, and uncertainty estimation are unspecified. The registry lists four outcome measures and zero analyses, without describing multiplicity procedures or providing secondary outcome results. Safety information is similarly limited to counts, with no event descriptions or details for the other category. These omissions restrict interpretation to the study’s stated design and registry-listed counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Pivotal Study to Evaluate Enlite™ Sensor Performance With iPro™2 in China — NCT03417466.
• Registry record last updated 2019-08-20; status last verified 2019-07.
• Results first posted 2019-08-20.
• 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.

19. Record Summary

NCT03417466 is a completed, single-group device study of Enlite Sensor performance with an iPro2 in China. The registry lists enrollment as 72, masking as none, and phase and allocation as NA. Sensor use was planned over 144 hours (6 days), with one YSI frequent sampling session on Day 1, Day 3-4, or Day 6.

The primary endpoint was the mean percentage of Enlite values within the specified agreement limits of paired YSI plasma glucose values. The registry describes 66 subjects with a sensor inserted and notes one without a paired measurement. The numerical primary and secondary outcome results are not reported in the registry record.

Safety entries list zero serious adverse events, five events in the other category, and zero deaths. The record does not provide details of those events or a comparative group.