This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Performance Evaluation of an Advanced Algorithm With CGM in Adults, Adolescents, and Pediatrics (NCT03966911), a na na study. The primary outcome is Percentage of Readings Within 20% Agreement.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03966911 |
| Phase | NA |
| Condition | Type 1 Diabetes Mellitus, Type2 Diabetes Mellitus |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 335 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Medtronic MiniMed, Inc. |
| Results posted | Yes |
2. Clinical Question
The study was designed to evaluate the performance of Guardian™ Sensor (3) with an advanced algorithm during sensor wear in people with diabetes. The intended age range was 2–80 years, and the evaluation lasted 170 hours (7 days). The primary question concerned how often sensor values agreed with a reference measurement within the endpoint’s specified tolerance. The reference was Yellow Springs Instrument (YSI™) for participants outside the youngest age group, while Self-Monitoring Blood Glucose (SMBG) was used instead for participants age 2–6.
Population
People with diabetes age 2–80 years were the intended study population. The primary endpoint population consisted of participants with at least 1 paired sensor-reference point during sensor wear. The registry lists Type 1 Diabetes Mellitus and Type2 Diabetes Mellitus as conditions.
Intervention
Participants wore Guardian™ Sensor (3) connected to a Guardian™ Connect Transmitter. The study description says the sensor was worn over 7 days, participants took part in Frequent Sample Testing (FST), and a zero calibration sensor algorithm was applied to raw sensor data.
Comparator
This was a single-group study, not a randomized comparison. The reference measurement for the endpoint was YSI, except that SMBG was used for participants age 2–6. The reference measurements are not a separate treatment arm.
Primary question
What was the mean of the daily percentages of sensor values within the defined agreement tolerance across participants and FST days during the 7-day wear period? The numerical endpoint result is not reported in the registry record.
The question is about agreement between measurements, not a comparison of clinical outcomes between treatment groups. The record identifies a total enrollment of 335, but it does not report how many participants contributed to the primary endpoint. It also does not report the numerical mean or the confidence-interval limits. As a result, the design and endpoint can be described, but the magnitude and precision of the primary finding cannot be assessed from the registry record.
3. Trial Design
The registry record describes a completed, single-group device study sponsored by Medtronic MiniMed, Inc. It lists phase as NA, allocation as NA, and masking as none. The record gives a start date of 2019-06-14 and a primary completion date of 2020-05-13. Enrollment was 335. These details describe the study’s status and structure; they do not establish how many participants had usable paired measurements for the primary endpoint.
The intervention was Guardian™ Sensor (3) connected to a Guardian™ Connect Transmitter. Participants with diabetes wore the sensor and transmitter over 7 days and participated in Frequent Sample Testing. A zero calibration sensor algorithm was applied to raw sensor data. The study’s stated purpose was to demonstrate sensor performance in subjects age 2–80 years over 170 hours (7 days). The listed conditions were Type 1 Diabetes Mellitus and Type2 Diabetes Mellitus.
Because the study is single-group, it does not provide a randomized control arm or a like-for-like comparison between treatment groups. The reference measurements used to assess agreement are part of the endpoint procedure, not a comparator intervention. The record’s primary analysis population required at least 1 paired sensor-YSI point, or sensor-SMBG point for participants age 2–6, during sensor wear. Whether all enrolled participants met that requirement is not reported in the registry record.
The record lists 1 outcome measure and 4 analyses. It does not describe the purpose of each analysis or establish whether they represent different age groups, endpoint summaries, or other analytic choices. The protocol’s detailed analysis plan, including any prespecified subgroup analyses, is not reported in the registry record. The record also does not report a sample-size justification or explain the enrollment target. Consequently, enrollment alone should not be treated as evidence that a particular precision or statistical power was achieved.
The study is marked completed, but completion does not by itself supply the missing primary endpoint estimate. The registry’s description permits a clear account of the device, study duration, and agreement question, while the numerical result and its uncertainty remain unavailable 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: 335 subjects (188 adult subjects and 147 pediatric subjects) consented and enrolled in the overall study. Of them, 311 subjects (169 adult subjects and 142 pediatric subjects) were eligible.
Overall Study
| Stage | Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) |
|---|---|---|
| STARTED | 169 | 142 |
| COMPLETED | 160 | 126 |
| NOT COMPLETED | 9 | 16 |
Reasons for not completing this period:
| Reason | Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) |
|---|---|---|
| Adverse Event | 1 | 1 |
| Withdrawal by Subject | 8 | 2 |
| Unable to complete study visits due to COVID-19 | 0 | 13 |
Analysis population for the primary outcome, as defined in the registry: Subjects with at least 1 sensor-YSI (or sensor-SMBG for subjects 2-6 years) paired point during the sensor wear
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.
Population: Demographics and other baseline characteristics were documented for eligible subjects (i.e., subjects who passed screening).
| Characteristic | Unit | Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) | Total |
|---|---|---|---|---|
| Participants analysed | 169 | 142 | 311 | |
| Age, Continuous | years | 45.9 (SD 17.8) | 10.9 (SD 4.0) | 29.9 (SD 22.0) |
| Sex: Female, Male: Female | Participants | 91 | 59 | 150 |
| Sex: Female, Male: Male | Participants | 78 | 83 | 161 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 21 | 17 | 38 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 148 | 125 | 273 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 |
| Region of Enrollment: United States | participants | 169 | 142 | 311 |
6. Primary Endpoint
The registry lists the primary outcome as Percentage of Readings Within 20% Agreement, measured over 7 days (170 hours).
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | 91.8% Confidence Interval |
| Unit | Percentage of Readings |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the mean of daily percentages of sensor values that fell within 20% of the YSI reference value. When sensor values were < 80 mg/dL, the allowed difference was within 20 mg/dL. The endpoint was summarized across participants and all Frequent Sample Testing days. For participants age 2–6, SMBG was used instead of YSI. The endpoint population included participants with at least 1 paired sensor-reference point during sensor wear.
This is an agreement measure: it describes how often a sensor reading fell within a defined tolerance of the reference reading. Agreement measures are used to characterize correspondence between two measurement methods. A percentage within tolerance is more directly interpretable than a measure of average difference alone because it expresses the share of readings meeting the stated criterion. It still does not show how far outside the tolerance the remaining readings fell, or whether those readings were systematically higher or lower than the reference.
The record identifies the endpoint as a mean and labels its dispersion “91.8% Confidence Interval.” It does not report the numerical mean or the interval limits. Therefore, the observed level of agreement and the precision of the estimate cannot be described from the registry record. The endpoint definition alone specifies what was intended to be measured, not the result.
Reference method matters when interpreting this measure. The endpoint used YSI for most participants and SMBG for those age 2–6. Since these are distinct reference procedures, the record’s aggregate endpoint combines measurements against the reference assigned by age. The registry record does not report separate numerical endpoint results by reference method.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record provides the definition of the primary agreement endpoint but not its numerical result. It describes a mean daily percentage of sensor readings within the stated tolerance of the reference, summarized across participants and FST days during 170 hours (7 days). Without the mean, there is no reported estimate of the observed agreement level. Without the confidence-interval limits, there is no way to describe the statistical precision of that estimate.
The endpoint’s tolerance is defined as within 20% of the YSI reference value, with an absolute tolerance of 20 mg/dL when sensor values were < 80 mg/dL. The record specifies that SMBG replaced YSI as the reference for participants age 2–6. These details explain what counted as a reading within agreement, but they do not establish how many readings met that criterion. The result is also not a measure of clinical benefit or a comparison with another device or treatment.
The primary endpoint population was not necessarily identical to the total enrolled population. Inclusion in the endpoint population required at least 1 paired sensor-reference point during wear. The registry lists enrollment of 335, but does not report the number contributing to the endpoint, the number of paired points, or the amount of missing measurement data. An endpoint estimate based on a subset could have different precision from one based on all enrolled participants; the record does not provide the information needed to assess that distinction.
The registry labels dispersion as “91.8% Confidence Interval,” but does not give interval limits or further explain the interval calculation. A confidence interval, in general, conveys uncertainty around an estimated quantity under the analysis assumptions. Its width and position relative to the estimate are needed to evaluate precision. The label alone is not enough to determine either.
The record lists 4 analyses, but their definitions and relationship to the primary endpoint are not reported. It is therefore not possible to tell whether the analyses were alternative summaries, subgroup assessments, or another form of analysis. No adjustment for multiple analyses is described in the registry record. That omission does not establish that no adjustment was made; it means the record does not document one.
Safety information is reported separately for adult and pediatric participants. The registry lists serious adverse events affecting 2 of 169 adults and 2 of 142 pediatric participants, and reports no deaths in either group. It also lists other adverse events affecting 37 adults and 32 pediatric participants. These counts are descriptive and do not establish that the device caused the events. The registry record does not provide a control group for comparison or an attribution analysis.
Overall, the available record supports an account of the study question, endpoint rule, study structure, and posted safety counts. It does not support a conclusion about the magnitude or precision of sensor-reference agreement. The numerical endpoint result, interval limits, and detailed analysis methods are not reported in the registry record.
9. Statistical Methods Explained
The primary outcome is a percentage-based agreement measure, summarized as a mean. Its definition combines a tolerance rule for each sensor-reference pair with aggregation across days and participants. The interpretation therefore depends on both the rule that classifies an individual pair as within tolerance and the method used to combine those classifications into a study-level estimate. The registry record describes the classification rule and states that the daily percentages were averaged across participants and FST days. It does not report further computational details.
What does agreement within a tolerance represent?
In general, an agreement-within-tolerance measure asks whether paired readings are sufficiently close according to a prespecified criterion. Each paired sensor and reference value is judged against that criterion. The endpoint counts values within the allowed range as meeting the agreement definition; values outside that range do not meet it. This form of summary is easy to interpret as a proportion of readings meeting a practical threshold, but it does not preserve the full size or direction of every difference.
For this study, the tolerance was within 20% of the YSI reference value, except when sensor values were < 80 mg/dL, where the criterion was within 20 mg/dL. The registry record says SMBG was used instead of YSI for participants age 2–6. The record does not report separate endpoint estimates for these reference methods, so the posted endpoint definition should be understood as applying the assigned reference procedure rather than as a single uniform reference for every participant.
Why summarize daily percentages?
A daily percentage first represents the proportion of paired readings meeting the agreement rule on a given testing day. Averaging daily percentages gives a summary across the FST days and participants described in the endpoint. In general, this can prevent days with more paired readings from automatically contributing more weight than days with fewer readings, depending on the precise averaging procedure. The record identifies the mean of daily percentages, but does not specify the exact weighting across participants and days or how incomplete testing days were handled. Those details are not reported in the registry record.
Repeated readings from the same participant are often statistically related. In general, measurements from one person may share characteristics that make them more alike than measurements from different people. If such dependence is ignored, uncertainty estimates can be too optimistic. Appropriate methods can account for repeated observations, but the registry record does not report whether or how dependence among readings within a participant was handled. This is a question about the analysis method, not evidence that the reported method was inadequate.
What does a confidence interval add?
In general, a confidence interval describes uncertainty around an estimate under a specified statistical procedure and its assumptions. The estimate gives the central result; the interval gives information about the range of values compatible with the data and procedure. Narrower intervals generally indicate greater precision, while wider intervals indicate greater uncertainty. Neither an interval nor its estimate alone determines the clinical importance of a finding.
The registry’s dispersion label is “91.8% Confidence Interval,” but the interval limits are not reported. The record also does not explain the procedure used to calculate the interval. As a result, the level indicated by the label cannot be connected to reported bounds, and the precision cannot be assessed. The primary endpoint’s numerical mean is also absent from the registry record.
Why can variability and over-dispersion matter?
General statistical methods may need to account for variability that is greater than expected under a simple model. For percentage outcomes, participant-to-participant differences, day-to-day changes, and dependence among repeated readings can all affect variability. Over-dispersion refers generally to variation exceeding what a chosen model assumes. If present, it can affect standard errors and confidence intervals. The registry record does not report an over-dispersion assessment or specify a model, so no conclusion about whether this issue arose in the study is possible.
How should the planning illustrations be read?
The registry’s stated planning illustration uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It describes the effect to detect as a stated fraction of the observed effect. These assumptions are for planning illustrations and are not a description of the study design: the trial itself is single-group and has no randomized comparison arm. The illustrations are not observed power. Their numerical outputs and the effect used for them are not reported in the registry record.
In general, sample-size calculations estimate how many participants would be needed under specified assumptions to detect a chosen effect at a chosen error threshold and power target. Such calculations are conditional on those assumptions; they do not prove that an analysis had a particular probability of detecting an effect after the study was completed. No sample-size justification for the trial itself is reported in the registry record. The record does not explain how enrollment of 335 was selected.
10. Multiplicity, Missing Data and Analysis Populations
The registry record lists 1 outcome measure and 4 analyses. It does not describe what the 4 analyses represent, whether they were prespecified, or which results were treated as primary or secondary. It also does not state whether any method was used to account for multiple analyses. When several statistical tests are conducted, general statistical practice recognizes that the chance of at least one apparently notable result can increase if the tests are treated as independent opportunities to find a result. Whether this concern applies, and how it was handled, cannot be determined from the registry record.
The primary endpoint population required at least 1 paired sensor-reference point during sensor wear. This eligibility rule means that the enrollment total of 335 should not automatically be interpreted as the denominator for the endpoint. The number of participants included in the primary analysis and the number excluded for lack of paired measurements are not reported in the registry record. The record also does not report missing-data counts, reasons for missingness, or methods for handling incomplete measurements.
In general, missing data can affect both the amount of information available and the interpretation of an estimate, particularly if the chance of missing a measurement is related to its value or to participant characteristics. This general concern does not establish that missingness biased the study’s result. The registry record lacks the details needed to evaluate missingness patterns or determine whether sensitivity analyses were conducted.
Likewise, the analysis population definitions beyond the stated paired-point requirement are not reported in the registry record. The record does not specify whether there were separate evaluable populations, how deviations from the testing procedure were handled, or whether any participants were excluded from particular analyses. No conclusions about the analysis population or multiplicity procedures should be inferred from the listed outcome and analysis counts alone.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | 2 | 169 | 1.2% |
| Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) | 2 | 142 | 1.4% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | 0 of 169 |
| Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) | 0 of 142 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | 37 of 169 |
| Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) | 32 of 142 |
Most frequent serious adverse event terms
| Event term | Organ system | Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) |
|---|---|---|---|
| Prostatomegaly | Reproductive system and breast disorders | 1 of 169 | 0 of 142 |
| Hypoglycaemia | Metabolism and nutrition disorders | 1 of 169 | 0 of 142 |
| Diabetic ketoacidosis | Metabolism and nutrition disorders | 0 of 169 | 1 of 142 |
| Hyperglycaemia | Metabolism and nutrition disorders | 0 of 169 | 1 of 142 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Adult Subjects (Age 18-80) With Diabetes Wearing Guardian™ Sensor (3) | Pediatric Subjects (Age 2-17) With Diabetes Wearing Guardian™ Sensor (3) |
|---|---|---|---|
| Nasopharyngitis | Infections and infestations | 0 of 169 | 7 of 142 |
| Headache | Nervous system disorders | 4 of 169 | 2 of 142 |
| Hyperglycaemia | Metabolism and nutrition disorders | 1 of 169 | 4 of 142 |
| Upper respiratory tract infection | Infections and infestations | 4 of 169 | 1 of 142 |
| Medical device site haemorrhage | General disorders | 4 of 169 | 0 of 142 |
| Dermatitis contact | Skin and subcutaneous tissue disorders | 1 of 169 | 2 of 142 |
| Medical device site pain | General disorders | 1 of 169 | 2 of 142 |
| Syncope | Nervous system disorders | 3 of 169 | 0 of 142 |
| Contusion | Injury, poisoning and procedural complications | 1 of 169 | 1 of 142 |
| Gastroenteritis viral | Infections and infestations | 0 of 169 | 2 of 142 |
Serious events with the highest arm-level rate: diabetic ketoacidosis (0.7%), hyperglycaemia (0.7%), prostatomegaly (0.6%), hypoglycaemia (0.6%).
The registry reports safety counts separately for adults age 18–80 and pediatric participants age 2–17. Serious adverse events affected 2 of 169 adult participants and 2 of 142 pediatric participants. No deaths were reported in either group. Other adverse events affected 37 adults and 32 pediatric participants. These are descriptive counts for the listed study groups; they are not comparisons against a control group.
The registry lists the top serious terms as Diabetic ketoacidosis (0.7%), Hyperglycaemia (0.7%), Prostatomegaly (0.6%), and Hypoglycaemia (0.6%). The record does not provide further detail in the registry record about event timing, participant overlap, or attribution to the device. The term “serious” describes the registry’s event category; it does not establish that an event was caused by the device or algorithm.
Because this was a single-group study, the safety counts do not show whether event occurrence differed from what would have been observed under another intervention or usual care. The record also does not explain how long participants were followed for safety beyond the stated sensor-wear evaluation. Counts and listed terms should therefore be read as descriptive registry reporting, not as evidence of a causal relationship.
12. What the Result Does, and Does Not, Mean
The record describes an evaluation of Guardian™ Sensor (3) with an advanced algorithm in people with diabetes over 170 hours (7 days). Its primary endpoint was intended to quantify the mean daily percentage of sensor values within a specified agreement tolerance of a reference measurement. The tolerance and reference approach are described, including the use of SMBG instead of YSI for participants age 2–6. These details establish the measurement question, not the observed answer.
The numerical primary endpoint result and confidence-interval limits are not reported in the registry record. Therefore, the record does not support a statement about the achieved percentage of readings within tolerance or the precision of that percentage. It also does not permit assessment of whether the estimate was consistent across participants, days, age groups, or reference methods.
The study’s single-group design does not provide a randomized comparison with another device or treatment. The reference measurement is used to assess agreement and should not be mistaken for a control arm. Consequently, the design cannot, by itself, establish a difference in performance relative to a comparator intervention. The registry record also does not report a clinical outcome comparison.
The safety counts describe events reported among the adult and pediatric study groups, including the serious-event counts and the absence of reported deaths. Without a comparison group and without event attribution details, these counts do not establish that the device caused or prevented events. They should be interpreted as descriptive information from the registry record.
The record also does not give the detailed statistical plan, analysis population counts, handling of missing measurements, or the purpose of each of the 4 analyses. Those omissions limit assessment of how the endpoint estimate was constructed and how robust it may be. The appropriate conclusion is therefore bounded: the registry documents a device-performance study and defines its primary agreement measure, but does not report the numerical primary result needed to characterize observed agreement.
13. Why This Trial Matters Statistically
This trial raises statistical questions that are important for measurement studies. Agreement is not the same as association: two measures can move together while differing enough that many paired readings fall outside a prespecified tolerance. A percentage within tolerance focuses attention on the frequency of acceptable pairwise agreement. Its meaning depends on the tolerance rule, the reference method, and the way repeated measurements are aggregated.
The endpoint uses a relative tolerance for most readings and an absolute tolerance when sensor values were < 80 mg/dL. The record also specifies a different reference method for participants age 2–6. These choices matter because the definition determines which pairs count as meeting the criterion. A single aggregate mean can conceal differences among participants, days, or reference procedures, so subgroup and distributional information can be useful when available. Such numerical details are not reported in the registry record.
The repeated-measure structure is also relevant. Readings are collected over FST days while participants wear the sensor. In general, readings from the same participant may not be statistically independent. Methods for uncertainty estimation should reflect the design and the way measurements cluster, but the registry record does not describe those methods. The absent confidence-interval limits also prevent direct evaluation of precision.
Finally, the design distinguishes device evaluation from comparative effectiveness. A single-group study can describe observed sensor-reference agreement under the study procedures, but a randomized comparator is not part of this record. The registry gives a clear endpoint definition and participant enrollment, yet the numerical endpoint estimate is not reported. This makes the trial statistically informative as an example of how agreement questions are framed, while leaving the central quantitative finding unavailable in the registry record.
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 the posted results is that the registry record does not report the numerical primary endpoint estimate or confidence-interval limits. Although the endpoint is precisely described, its observed magnitude and statistical precision cannot be assessed. The dispersion label “91.8% Confidence Interval” does not substitute for the missing interval bounds.
The study is single-group, with allocation listed as NA and masking as none. It does not include a randomized control arm for a like-for-like comparison. The endpoint uses reference measurements rather than a comparator intervention, so it addresses agreement under the study procedures rather than a treatment-group contrast.
Enrollment was 335, but the number included in the primary endpoint population is not reported in the registry record. That population required at least 1 paired sensor-reference point during wear. The record does not report the total number of paired points, missingness patterns, exclusions, or methods for handling incomplete data. These omissions limit evaluation of the endpoint denominator and possible effects of missing measurements.
The registry lists 4 analyses but does not describe their purpose, whether they were prespecified, or how multiplicity was handled. Detailed statistical methods for summarizing repeated readings, calculating uncertainty, or accounting for within-participant dependence are not reported. The record does not report numerical endpoint results by age group or reference method.
Safety reporting is descriptive. The record gives event counts but no control group or attribution details, and it does not report further information on event timing or participant overlap. These limitations do not establish that the study methods were deficient; they identify what cannot be evaluated from the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Performance Evaluation of an Advanced Algorithm With CGM in Adults, Adolescents, and Pediatrics — NCT03966911.
- Registry record last updated 2021-06-07; status last verified 2021-05.
- Results first posted 2021-06-07.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT03966911 is a completed, single-group device study sponsored by Medtronic MiniMed, Inc. The record lists enrollment of 335, a start date of 2019-06-14, and a primary completion date of 2020-05-13. Participants with diabetes wore Guardian™ Sensor (3) connected to a Guardian™ Connect Transmitter over 7 days and participated in Frequent Sample Testing.
The primary outcome was the mean daily percentage of sensor values within the stated tolerance of a reference measurement over 170 hours (7 days). YSI was the reference except for participants age 2–6, for whom SMBG was used. The numerical endpoint result and confidence-interval limits are not reported in the registry record.
Safety counts are reported separately for adults and pediatric participants, including serious and other adverse events; no deaths were reported in either group. The study has no randomized comparison arm, and the registry does not report detailed analysis methods, missing-data handling, or the purpose of its 4 analyses.