The International Diabetes Closed Loop (iDCL) Trial: Clinical Acceptance of the Artificial Pancreas: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The International Diabetes Closed Loop (iDCL) Trial: Clinical Acceptance of the Artificial Pancreas (NCT03563313), a randomized na study. The primary outcome is Time in Target Range.

2. Clinical Question

The primary endpoint is a continuous measure: each participant’s monitored time in the target range is expressed as a percentage, and the registry reports group means and standard deviations. The registry’s posted group means were 71.0 for CLC and 59.0 for SAP. A simple, unadjusted comparison of those summaries gives a mean difference of -12.0 when expressed as SAP minus CLC. The registry also reports a mixed models analysis with p < 0.001. These are complementary descriptions: the group summaries describe the observed values, while the registry’s analysis reports the result of its stated model. Neither result by itself describes every participant’s experience or establishes how well the findings apply beyond the trial population.

3. Trial Design

The registry describes a randomized, parallel-group trial with no masking. Enrollment was 168, and the listed condition was type 1 diabetes mellitus. The sponsor was the University of Virginia. The study status is completed. The registry records a start date of 2018-06-28 and primary completion on 2019-04-08.

Participants were assigned to one of two device strategies. The CLC arm used t:slim X2 with Control-IQ Technology together with Dexcom G6 continuous glucose monitoring for 6 months. The SAP arm used an insulin pump without automated insulin delivery together with a study Dexcom G6 continuous glucose monitor for 6 months. The registry characterizes the primary outcome timeframe as 26 weeks. The record does not report additional details about how randomization was generated or concealed, or how adherence to the assigned device strategy was assessed.

The primary outcome was mean time in the target range of 70-180 mg/dL, measured by continuous glucose monitoring and expressed as a percentage of time. The listed primary analysis was a mixed models analysis for superiority. The registry also lists secondary outcomes related to glucose levels, time in specified glucose ranges, HbA1c, variability, and standard deviation of continuous glucose monitoring. It records 49 outcome measures, 6 analyses, and 32 posted secondary outcomes. The record does not report the original sample-size justification, the trial’s prespecified power assumptions, or whether interim analyses were planned. Those design details should not be inferred from the later outcome summaries or from planning illustrations based on observed results.

Because masking was listed as none, the trial was open-label as recorded. This describes the masking status, not whether outcome assessment or other trial procedures had safeguards against bias; those specifics are not reported in the registry record. The parallel-group design means the two assigned strategies were compared across separate groups rather than through a crossover. The registry does not report further information about participant flow, protocol deviations, or the analysis population used for each posted 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

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 Time in Target Range, measured over 26 weeks.

The primary endpoint measures the percentage of monitored time when glucose was between 70 and 180 mg/dL, based on continuous glucose monitoring over the 26-week timeframe. A higher mean indicates that, on average, a group spent a greater share of measured time within this defined interval. The endpoint is continuous: it can capture a range of values rather than classifying participants only as having or not having an event. The registry reports the group mean and standard deviation for each arm.

As a general statistical point, a continuous endpoint can retain information about variation across participants and can be useful when the study question concerns a level or duration rather than a binary event. In this trial, the target range gives the measure a specific clinical interpretation: it summarizes the proportion of monitored time within a prespecified glucose interval. It does not, by itself, describe the timing of readings, the pattern of excursions, or each participant’s individual trajectory. Those details are not contained in the posted group means.

The primary outcome description identifies the measurement method, target range, comparison groups, and timeframe. It does not specify in the registry record how much valid monitoring data each participant needed to contribute, how missing sensor readings were handled, or whether the reported endpoint was averaged over particular days. Such measurement and analysis details can affect what a group mean represents, so the posted definition should not be expanded beyond what is recorded.

7. Results Posted in the Registry

• Sensor-Augmented Pump (SAP): mean 59.00 versus 71.00 on control; difference -12.00 (95% CI -16.29 to -7.71), standardized effect d = -0.95; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

CGM Time Above 180

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Sensor-Augmented Pump (SAP): mean 38.00 versus 27.00 on control; difference 11.00 (95% CI 6.49 to 15.51), standardized effect d = 0.84; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Sensor-Augmented Pump (SAP) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

CGM Mean Glucose

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Sensor-Augmented Pump (SAP): mean 170.00 versus 156.00 on control; difference 14.00 (95% CI 6.57 to 21.43), standardized effect d = 0.66; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Sensor-Augmented Pump (SAP) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

HbA1c at 26 Weeks

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Sensor-Augmented Pump (SAP): mean 7.39 versus 7.06 on control; difference 0.33 (95% CI 0.05 to 0.61), standardized effect d = 0.40; z-test p = 0.022.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Sensor-Augmented Pump (SAP) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The three posted secondary comparisons cover time above 180, mean glucose, and HbA1c at 26 weeks. Across those measures, the group means follow a direction consistent with the primary endpoint: the CLC group has lower values than the SAP group. This broad pattern provides context across different aspects of glucose monitoring and HbA1c, but the measures are not interchangeable. Each has its own scale, units, and meaning, and the secondary outcomes do not substitute for the prespecified primary endpoint.

The registry reports mixed models superiority analyses for these three posted secondary comparisons. The record also lists other secondary outcomes, including time below specified glucose thresholds, time in a narrower target range, coefficient of variability, and standard deviation of continuous glucose monitoring. Results for those additional measures are not included in the posted secondary comparisons described in the record data used for this analysis, so their direction or statistical evidence cannot be inferred. The registry lists 32 posted secondary outcomes and 49 outcome measures overall, but those counts do not show how many outcomes were prespecified or how results were selected for posting.

Unadjusted secondary results carry less weight because multiple outcomes create multiple opportunities to observe apparently notable findings by chance. A p-value considered in isolation does not account for how many hypotheses were tested. The registry record does not report a multiplicity adjustment strategy, a hierarchy for testing secondary endpoints, or which secondary results were confirmatory versus exploratory. As a result, the secondary findings are best treated as supportive context rather than independent confirmation of the primary result.

In general, a coherent pattern across related outcomes can help readers understand whether the primary finding aligns with other measures, but consistency alone does not remove uncertainty about analysis choices or multiplicity. The primary endpoint remains the main basis for interpreting the trial’s stated question. Secondary outcomes may motivate further investigation, while their strength as evidence depends on prespecification, analysis methods, and reporting details that are not reported in the registry record.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• CGM Time Above 180
• CGM Mean Glucose
• HbA1c at 26 Weeks
• CGM Time Below 70
• CGM Time Below 54
• CGM Time in Range 70-140 mg/dL
• Coefficient of Variability
• Standard Deviation of CGM

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a continuous measure reported as a mean and standard deviation in each randomized group. The posted summaries are 71.0 with a standard deviation of 12.0 in CLC, and 59.0 with a standard deviation of 14.0 in SAP. The group sizes reported for the endpoint are 112.0 and 56.0, respectively. These values describe the observed endpoint summaries; the standard deviations describe variability among observations, not uncertainty in the group means.

The unadjusted recalculation compares the SAP mean with the CLC mean and reports a mean difference of -12.0. Its 95% confidence interval is -16.29 to -7.712, and the unadjusted p-value is < 0.001. This is a summary-statistic-based comparison, not a reproduction of the registry’s model. The registry’s reported method is a mixed models analysis, with a superiority result and p < 0.001. The model’s estimate, confidence interval, specific model form, and covariates are not reported in the registry record.

General explanation: A mean difference expresses how far apart two group averages are, using the order of subtraction to define its sign. A confidence interval gives a range of values reasonably compatible with the data and the method used to calculate it. A p-value summarizes evidence against a specified null model under the test’s assumptions. These quantities answer related but distinct questions: magnitude, precision, and compatibility with a null model.

Mixed models are a broad family of methods that can account for structured data, such as repeated measurements or clustering, depending on the model specification. The registry names a mixed models analysis but does not provide details sufficient to identify its structure, assumptions, or handling of repeated measurements. It is therefore appropriate to describe the analysis by the name in the record, but not to assume which model terms or data-handling choices it used.

The standard deviation is not the same as a standard error. A standard deviation describes variation in individual measurements around a group mean. A standard error describes the uncertainty of an estimated quantity, such as a mean. The record’s endpoint summaries include standard deviations; the recalculation also reports standard errors of 1.134 for CLC and 1.871 for SAP. These recalculated standard errors are not separate registry estimates and should not be confused with the standard deviations.

General explanation: The width of a confidence interval depends on the amount of information in the data and the variability of the measure, as well as on the statistical method. Greater variability or less information tends to reduce precision, all else being equal. An interval that does not include zero is consistent with evidence of a nonzero difference under the interval’s assumptions, but it does not establish clinical importance or rule out bias, missing-data concerns, or model limitations.

The registry records the primary analysis as a superiority test. In general, a superiority analysis asks whether the data support a difference in a specified direction or whether one group’s outcome is better under the chosen endpoint definition. Here, the record’s analysis entry reports only the analysis type and p-value. The underlying test specification, sidedness, and model-estimated effect are not reported in the registry record. The separate planning illustration lists a two-sided significance level of 0.05, but that planning assumption should not be taken as confirmation of the primary analysis’s full prespecified testing procedure.

Secondary outcome comparisons use the same broad idea of comparing group means, but each endpoint has its own units and interpretation. The posted examples include percentages, mg/dL, and HbA1c percentage. Their scales differ, so raw mean differences cannot be compared as if they were the same quantity. The registry reports mixed models analyses and p-values for the three posted secondary outcomes, while adjusted estimates and confidence intervals from those models are not reported. The recalculated comparisons are unadjusted summaries and do not account for multiple testing.

General explanation: Over-dispersion refers to variability greater than a statistical model expects. When present, it can make uncertainty appear smaller than it should if the model does not accommodate that extra variation. It is a general concern in some data types and models; the registry does not report an over-dispersion assessment for this continuous endpoint, so it should not be asserted as a feature of this trial.

All recalculated estimates should be read as descriptive, unadjusted comparisons from the posted group summaries. They do not replace the sponsor’s reported mixed models analysis. Since the record does not include a detailed analysis plan or model output, the recalculations can provide a transparent comparison of the means and their precision, but cannot verify the complete analysis procedure.

11. Confidence Intervals and Precision

The recalculated primary mean difference is -12.0, with a 95% confidence interval from -16.29 to -7.712, expressed as SAP minus CLC in percentage of time. The interval lies on the negative side of zero, consistent with a lower posted mean in SAP than in CLC under this unadjusted calculation. Its endpoints indicate a range of plausible values for the difference under the calculation’s assumptions, rather than a range containing 95% of individual participant outcomes.

As a general statistical explanation, a confidence interval reflects both an estimated effect and its uncertainty. A narrower interval indicates greater precision for that estimate, while a wider interval indicates less precision. The interval does not account automatically for every source of uncertainty, such as bias from unmasking, missing outcome data, model misspecification, or selective reporting. The registry does not report the model-based primary confidence interval, so the interval above applies only to the recalculation from the group summaries.

The registry reports a mixed models p-value of < 0.001 for the primary comparison, but does not report its model-estimated difference or confidence interval. The recalculated interval and the model p-value are therefore not two parts of a single reported model result. They arise from different summaries, and the model’s full specification is not reported in the registry record.

For the posted secondary comparisons, the recalculated confidence intervals are 6.486 to 15.51 for time above 180, 6.567 to 21.43 for mean glucose, and 0.048 to 0.612 for HbA1c. Their units and directions differ by outcome. These unadjusted intervals describe the recalculations, not the registry’s mixed models estimates. Because the record does not report multiplicity adjustment, the intervals should not be read as a family of jointly controlled confidence statements.

Precision is not the same as certainty about generalizability or importance. Even a relatively precise estimate can be affected by the analysis population, missing-data methods, or measurement procedures, details of which are not reported in the registry record. The interval is best read as a measure of sampling uncertainty for the stated unadjusted comparison, within the limits of the posted summaries.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations expressed as participants per arm, not total event counts. They use the observed effect as a reference and show planning for 100%, 75%, and 50% of that observed effect. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The figures are not the trial’s original design calculations; the original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the estimated participants per arm for a planning scenario with an 80% chance of detecting the stated effect, assuming the listed statistical conditions hold. The 90% power column gives the corresponding planning count for a 90% chance. In general, power is the probability that a specified analysis will detect an effect of a chosen size if that effect is present and the assumptions are valid. The values describe hypothetical future planning under the stated assumptions, not outcomes observed in this trial.

Each row changes the effect targeted for detection. Planning for 75% or 50% of the observed effect means planning around a smaller difference than the observed comparison. Smaller effects are harder to distinguish from random variation, so they generally require more participants for the same target power. The listed figures should not be read as proof that the trial was adequately or inadequately sized, because the trial’s own planned effect, allocation assumptions, and design calculations are not reported in the registry record.

These calculations are not observed power. They are illustrations based on the observed effect and specified assumptions, and they do not replace a prospective sample-size justification. The trial’s reported group sizes were unequal, whereas the planning illustration assumes equal allocation; this is another reason to treat the figures as planning aids rather than a reconstruction of the trial design.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -12.000; pooled standard deviation σ = 12.698.
2. n = 2 × (1.96 + 0.842)² × 12.698² / -12.000² = 17.6, rounded up to 18 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -12                      # observed difference in means
sigma <- 12.697594877088                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3563313)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 18 (80% power), 24 (90% power)
 75% of observed effect: n per arm = 32 (80% power), 42 (90% power)
 50% of observed effect: n per arm = 71 (80% power), 95 (90% power)
power.t.test: 18.6 
Simulated power at n = 18 per arm: 0.788 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry’s own primary analysis is a mixed models superiority analysis, with p < 0.001. It does not report the model-estimated difference or a model-based confidence interval in the analysis entry. The unadjusted recalculation from the posted group means gives a mean difference of -12.0 for SAP minus CLC, with a 95% confidence interval from -16.29 to -7.712 and an unadjusted p-value of < 0.001.

These results agree in direction and in the reported threshold for statistical evidence: the group means are lower for SAP, and both the registry analysis and the recalculation report p-values below 0.001. They are not identical analyses. The recalculation uses the posted summary statistics and does not reproduce the mixed models procedure, whose full specification is not reported in the registry record. The recalculated interval therefore should not be attributed to the registry’s model.

The discrepancy indicator is false, so the posted values do not flag a discrepancy between the registry analysis and the recalculation. The sponsor’s analysis remains the registry’s stated inferential analysis; the recalculation serves as a transparent, unadjusted check of the posted summaries. The registry reports a mixed models analysis for each of the three posted secondary outcomes as well, but their model estimates and confidence intervals are not reported.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a mixed models superiority analysis for the primary endpoint and for the three posted secondary comparisons. It does not report whether the primary analysis was part of a broader testing hierarchy, whether secondary outcomes were tested confirmatorily or exploratorily, or whether any adjustment for multiple comparisons was applied. Without those details, the secondary p-values should be interpreted as unadjusted evidence for their individual comparisons, not as a family of error-controlled conclusions.

The registry record does not report how missing continuous glucose monitoring readings were handled, what threshold of valid monitoring data was required, or how missing endpoint values were addressed. It also does not report whether missingness was assessed by randomized group or whether sensitivity analyses were conducted. These details matter because the endpoint depends on recorded sensor data, and incomplete measurement can affect which observations contribute to a group summary.

The analysis population is not specified in the primary endpoint description. The record does not state whether analyses followed an intention-to-treat principle, whether participants were analyzed according to assigned group regardless of device use, or whether a different population was used. It does not report participant flow, exclusions from analysis, protocol deviations, or the handling of participants who discontinued or changed treatment. These matters cannot be reconstructed from the posted group counts alone.

The registry names a mixed models analysis but does not provide the model equation, covariates, correlation structure, degrees-of-freedom method, or assumptions. The reported sample-size planning illustration is based on an observed effect and is not the original trial design justification. Information on interim analyses and stopping rules is not reported in the registry record. Consequently, the available record supports describing the reported result and its summary-statistic recalculation, but not a full audit of the prespecified statistical plan or the analysis implementation.

These reporting gaps do not establish that a particular method was or was not used. They mark limits on what can be assessed from the registry record. A careful interpretation distinguishes the documented mixed models result from missing information about multiplicity, missing data, and analysis populations.

16. 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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: diabetic ketoacidosis (0.9%), concussion (0.9%), otitis (0.9%), cardiac bypass surgery (0.9%).

The registry reports serious adverse events among 4 participants at risk in the CLC group and 0 participants at risk in the SAP group. It reports other adverse events among 12 participants at risk in CLC and 2 participants at risk in SAP. No deaths are reported in either group. These are counts of affected participants as recorded in the adverse-event summary, not a complete description of event timing, duration, severity beyond the serious classification, or causal relationship to an intervention.

The registry lists Diabetic Ketoacidosis, Concussion, Otitis, and Cardiac Bypass Surgery among the top serious-event terms, each with a reported value of 0.9%. The record does not identify the group assignment for those terms in the listed information, so they should not be attributed to one arm. The term list also does not establish that an event was caused by the device or study strategy.

As a general statistical point, safety counts require context about how events were defined, collected, adjudicated, and followed over time. Those details, along with comparisons adjusted for exposure time, are not reported in the registry record. The counts describe the recorded events in this trial and should not be read as proof of equivalence or a comprehensive estimate of risk.

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

The primary result describes a difference in the mean percentage of continuous glucose monitoring time within the 70-180 mg/dL range at 26 weeks between the randomized groups. The CLC group’s posted mean was 71.0, compared with 59.0 in the SAP group. The registry reports a mixed models superiority result with p < 0.001, while the unadjusted recalculation gives a mean difference of -12.0 for SAP minus CLC and a 95% confidence interval from -16.29 to -7.712. Together, these numbers describe the observed group comparison and the precision of the recalculated estimate.

The result does not establish how every participant’s measurements changed, how individual participants experienced the assigned strategies, or whether the observed difference meets a clinically important threshold. The registry does not report such a threshold. A group average can conceal variation, and the reported standard deviations indicate that individual measurements varied around their group means.

The findings do not by themselves establish effects in populations beyond the trial participants, or under different care conditions, devices, monitoring practices, or follow-up periods. The record identifies a randomized, parallel-group trial with no masking, but detailed information about participant selection, adherence, missing data, and analysis populations is not reported. These factors limit judgments about generalizability and the robustness of the analysis.

The posted secondary comparisons are directionally consistent with the primary comparison, but multiplicity handling is not reported. They therefore add descriptive context rather than unrestricted confirmation. The safety summary reports serious and other adverse-event counts and no deaths; it is not a complete causal assessment of device-related risk. No treatment recommendation follows from these registry statistics. They are a record of measured outcomes and analyses in this particular trial.

Statistical significance is not equivalent to clinical importance, certainty about mechanism, or proof that an outcome will be reproduced. The primary p-value reflects the registry’s stated model analysis, while the recalculated interval reflects unadjusted comparison of group summaries. Both need to be interpreted with their respective methods and the unreported analysis details in mind.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares two assigned device strategies through a randomized, parallel-group design and uses continuous glucose monitoring to quantify time in a defined target range. The primary outcome is expressed on an interpretable scale, and the registry reports both group means and measures of variability. The mixed models analysis gives a stated inferential result, while the summary-statistic recalculation makes the observed group difference and its uncertainty explicit.

The endpoint also illustrates why effect size, precision, and p-value should be read together. The group means describe the magnitude and direction of the observed contrast; the recalculated confidence interval describes uncertainty around that unadjusted contrast; and the registry’s p-value reflects its own mixed models analysis. Each contributes different information, and none alone answers every question about bias, clinical importance, or generalizability.

The trial also offers a useful reminder that registry reporting can leave key statistical details unresolved. The mixed models method is named, but its specification, model-based estimate, and model confidence interval are not reported. Multiplicity, missing-data handling, original sample-size rationale, interim analyses, and analysis populations are not reported either. These omissions do not invalidate the posted result, but they constrain independent evaluation and prevent a complete reconstruction of the planned analysis.

Finally, the secondary findings show why a pattern across related outcomes can be informative without being decisive. Their direction is consistent with the primary comparison, yet unreported multiplicity procedures limit the strength of confirmatory claims. Careful statistical reading preserves the distinction between the primary result, supporting secondary observations, and uncertainty that the record does not resolve.

19. 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 registry’s primary analysis entry reports a mixed models method and p-value but does not report the model-estimated difference, confidence interval, covariates, or model specification. The recalculated primary estimate uses group summary statistics and is not a reproduction of the mixed models analysis. This limits direct comparison of the model’s magnitude and precision with the recalculation.

The record does not report the original sample-size justification, analysis population, missing-data procedures, requirements for valid continuous glucose monitoring data, or participant flow. These omissions make it difficult to assess whether the posted group counts represent all randomized participants, how incomplete sensor data were handled, or whether exclusions could have affected the endpoint summaries. Details about adherence, device use, and protocol deviations are also not reported.

The trial was recorded as having no masking. This may be relevant to behavior, care, and assessment, but the registry does not provide enough detail to determine how such concerns were addressed. The primary endpoint is based on sensor readings, but specific measurement and data-processing procedures are not reported in the record used for these results.

The record does not report multiplicity adjustment or a testing hierarchy for the secondary outcomes. The posted secondary comparisons should therefore be interpreted cautiously. Only three secondary comparisons are described in the posted results used for these analyses, while the registry lists additional secondary outcomes; the available information does not support conclusions about all listed outcomes. The registry also does not report a threshold for a clinically important difference, limiting interpretation of the practical importance of the primary estimate.

These limitations are specific to what the registry record reports. They do not establish that unreported procedures were absent, only that their details cannot be assessed from the record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: The International Diabetes Closed Loop (iDCL) Trial: Clinical Acceptance of the Artificial Pancreas — NCT03563313.
• Registry record last updated 2020-04-28; status last verified 2020-04.
• Results first posted 2020-04-28.
• 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.

23. Record Summary

NCT03563313 is a completed, randomized, open-label, parallel-group trial sponsored by the University of Virginia, with enrollment of 168 participants and type 1 diabetes mellitus listed as the condition. The primary outcome was continuous glucose monitoring time in the 70-180 mg/dL target range at 26 weeks, expressed as a percentage.

The posted primary group means were 71.0 for CLC and 59.0 for SAP. The registry reports a mixed models superiority analysis with p < 0.001. The unadjusted recalculation gives a mean difference of -12.0 for SAP minus CLC, with a 95% confidence interval from -16.29 to -7.712. The registry also posts secondary comparisons for time above 180, mean glucose, and HbA1c at 26 weeks.

Important details—including the original sample-size rationale, missing-data methods, multiplicity approach, and analysis population—are not reported in the registry record. The estimates describe the posted trial results and should be interpreted with those reporting limits in mind.