The International Diabetes Closed Loop (iDCL) Trial: Protocol 1: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The International Diabetes Closed Loop (iDCL) Trial: Protocol 1 (NCT02985866), a randomized na study. The primary outcome is Time Below 70 mg/dL.

2. Clinical Question

The registry record describes a parallel-group randomized trial with no masking. The primary comparison is between the Artificial Pancreas, also described as closed-loop control, and sensor-augmented therapy. Its outcome is a continuous measure of time below a glucose threshold, expressed as a percentage of total continuous-glucose-monitor time. That makes the question about average exposure during a defined period, rather than about a count of participants crossing a threshold or a time-to-event outcome.

The trial also listed secondary measures covering other glucose thresholds and coefficient of variation. The record’s available quantitative summaries include a subset of those secondary measures. The remaining secondary results and additional details about participant eligibility are not reported in the registry record. The question answered by the posted primary result is therefore specific: how the groups compared on mean time below 70 mg/dL during the stated post-randomization period, using the reported analysis and the available outcome summaries.

3. Trial Design

The registry record identifies this study as the International Diabetes Closed Loop Trial: Protocol 1, a randomized clinical trial sponsored by the University of Virginia. The trial used a parallel-group design, meaning participants were assigned to one of two intervention groups for the comparison. The record lists 127 enrolled participants and describes the trial as completed. Its dates are listed as starting on 2017-10-27, with primary completion on 2018-09-04.

The Artificial Pancreas group was assigned a study system comprising the inControl Diabetes Management Platform, a study insulin pump, a study continuous glucose monitor, and a study glucometer. The comparator group continued using a personal insulin pump with a study continuous glucose monitor and study glucometer. The trial objective was to assess efficacy and safety of home use of a Control-to-Range closed-loop system. The registry record does not provide further detail about training, monitoring procedures, adherence, or the exact operational rules used by the device.

The trial was unmasked. That means the registry does not describe masking of participants or study personnel for the comparison. The primary outcome was measured by continuous glucose monitoring and concerned percentage of total monitoring time below 70 mg/dL during the final 11 weeks after randomization. The primary outcome is marked as a superiority outcome. The record lists phase as NA; a more detailed phase designation is not reported in the registry record.

For the primary outcome, the posted arm summaries include 64 participants in the Artificial Pancreas group and 61 in the Sensor Augmented Therapy group. These are the numbers attached to the reported outcome summaries, rather than an explanation of how participants were handled in every analysis. The analysis population, exclusions, and participant flow details are not reported in the registry record. No protocol-level sample-size justification or original design assumptions are 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: Participants who used a personal Dexcom G4 or G5 CGM pre-study for at least 21 of the prior 28 days proceeded directly to randomization. For these participants, the data downloaded from the personal CGM provided the 2-week baseline data. All others wore a blinded study CGM for 2 weeks to collect baseline data, followed by 2 weeks of training using an unblinded study CGM. Successful completion of the run-in was followed by randomization.

Overall Study

Reasons for not completing this period:

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 Below 70 mg/dL, measured over Post randomization (final 11 weeks).

The primary endpoint is mean continuous-glucose-monitor time below 70 mg/dL, expressed as a percentage of total monitoring time. It covers the final 11 weeks after randomization. The posted group summaries report a mean of 2.4 for the Artificial Pancreas group and 4.0 for Sensor Augmented Therapy, with standard deviations of 1.7 and 3.4, respectively. The units are percentage of total continuous-glucose-monitor time.

General: a time-below-threshold measure describes the share of monitored time spent under a specified glucose level. It differs from a simple participant count because it summarizes duration across monitoring observations. A group mean gives the average percentage for the participants represented in that summary; the standard deviation describes the spread of participant values around that mean. Neither statistic by itself describes every participant’s experience.

The threshold is used to quantify low-glucose exposure within the trial’s monitoring framework. The endpoint’s percentage scale allows the comparison to be expressed in relation to each participant’s total monitoring time, rather than as an unscaled duration. The record specifies a superiority question and reports the endpoint as a continuous mean. It does not report a clinically important difference threshold or a separate interpretation rule for the endpoint, so the numerical contrast should be read as a statistical comparison of the recorded percentages, not as a complete account of clinical benefit or harm.

7. Results Posted in the Registry

• Sensor Augmented Therapy: mean 4.00 versus 2.40 on control; difference 1.60 (95% CI 0.65 to 2.55), standardized effect d = 0.60; 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.

Time Below 54 mg/dL

• 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 Therapy: mean 1.10 versus 0.60 on control; difference 0.50 (95% CI 0.16 to 0.84), standardized effect d = 0.53; z-test p = 0.003.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Sensor Augmented Therapy than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time Below 60 mg/dL

• 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 Therapy: mean 1.90 versus 1.10 on control; difference 0.80 (95% CI 0.26 to 1.34), standardized effect d = 0.53; z-test p = 0.003.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Sensor Augmented Therapy than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Time in Range 70-180 mg/dL

• 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 Therapy: mean 57.00 versus 64.00 on control; difference -7.00 (95% CI -11.43 to -2.57), standardized effect d = -0.56; z-test p = 0.002.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Sensor Augmented Therapy than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists secondary outcomes covering time below additional glucose thresholds, time in range, time above thresholds, and coefficient of variation. The quantitative secondary summaries available in the record include time below 54 mg/dL, time below 60 mg/dL, and time in range from 70 to 180 mg/dL. Taken together, those available summaries point in a direction broadly consistent with the primary comparison: the Artificial Pancreas group had lower reported time below the specified thresholds and higher reported time in the stated range.

This directional pattern gives context to the primary endpoint, but the secondary measures are related rather than independent confirmations. They describe overlapping aspects of continuous glucose monitoring, and changes in one measure may be associated with changes in another. The registry’s remaining listed secondary measures do not have quantitative results reported in the registry record. A full account of all listed secondary outcomes is therefore not possible from the posted results.

The reported secondary analyses include p-values, but the registry does not state whether they were adjusted for multiple comparisons. General: testing many endpoints increases the opportunity for at least one result to appear statistically notable by chance, even when no real difference exists for any endpoint. Without a reported multiplicity strategy, individual unadjusted secondary p-values carry less weight than a prespecified primary comparison and should be viewed as supportive or exploratory rather than as definitive independent evidence.

The secondary outcomes can help assess whether the primary result sits within a wider pattern of monitoring measures, but they do not establish results for outcomes whose numerical summaries are absent. Nor do they answer questions about follow-up beyond the final 11 weeks, participant subgroups, or clinical outcomes not listed in the registry. The primary endpoint remains the central statistical result specified for this trial.

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

• Time Below 54 mg/dL
• Time Below 60 mg/dL
• Time in Range 70-180 mg/dL
• Time in Range 70-140 mg/dL
• Time Above 250 mg/dL
• Time Above 300 mg/dL
• Coefficient of Variation

9. Statistical Interpretation

10. Statistical Methods Explained

The registry record identifies a randomized, parallel-group comparison and labels the primary endpoint as a mean of a continuous measure. The outcome is percentage of total continuous-glucose-monitor time below 70 mg/dL during the final 11 weeks after randomization. The reported group summaries are means with standard deviations. The record also identifies the primary comparison as a superiority analysis.

What does a mean difference describe?

General: a mean difference subtracts one group mean from another, with the sign depending on the chosen order. The recalculated primary contrast is stated as Sensor Augmented Therapy minus Artificial Pancreas and is 1.6. A positive contrast in that direction means the observed mean was higher in the comparator group. The contrast is expressed on the endpoint’s original scale: percentage of total monitoring time. It is not a relative percentage change.

The arm means are 2.4 and 4.0. Those averages summarize the observed outcome values in each group, while the standard deviations of 1.7 and 3.4 describe how widely participant values varied around their respective group means. A standard deviation is not the same as the standard error. General: the standard deviation characterizes variation among observations; the standard error characterizes uncertainty in an estimated mean and depends on both spread and sample size.

What does the mixed-model result add?

The registry reports a Mixed Models Analysis for the primary comparison and gives a superiority result with a p-value below < 0.001. General: mixed models can be used to represent outcome patterns with multiple sources of variation or repeated measurements, depending on how a study is designed and how the model is specified. The registry record does not give the model formula, included covariates, covariance structure, or detailed assumptions. Those features should not be inferred from the method name alone.

The registry’s model-based estimate and confidence interval are not reported. The simple unadjusted recalculation instead uses the posted arm-level summary statistics. It is useful for understanding the scale and direction of the raw group contrast, but it is not a substitute for the registry’s reported mixed-model analysis. Since the record does not describe the model specification, the two approaches cannot be compared in detail beyond the reported direction and significance information.

How should a p-value be read?

General: a p-value measures how compatible the observed data are with a specified statistical null model, under the assumptions of the analysis. A small p-value is evidence against that null model; it is not the probability that the null hypothesis is true, and it does not measure clinical importance. The unadjusted recalculation reports p <0.001. The registry’s reported mixed-model result is below < 0.001. The p-values come from different analyses, so their numerical thresholds should not be treated as if they were outputs from a single identical calculation.

The superiority label indicates the direction of the stated statistical question: whether the groups differ in the endpoint, rather than whether they meet a non-inferiority or equivalence criterion. The registry record does not report a non-inferiority margin or an equivalence margin, and neither is needed to describe the stated superiority analysis.

What do variability and sample size contribute?

General: the precision of a group comparison depends on the amount of information in the data, including the number of observations and their variability. In this trial’s primary summaries, the outcome counts are 64 and 61, and the standard deviations differ between groups. The available summaries give useful context for the mean difference, but they do not show participant-level distributions, missingness patterns, or whether the same number of monitoring observations contributed for every participant.

The registry reports nine outcome measures and nine analyses. It also lists several secondary glucose measures and coefficient of variation. Examining multiple endpoints creates a multiplicity question: if many statistical tests are conducted, the chance of at least one apparently notable result can increase. General: adjustment procedures can control particular error rates across a family of tests, but the appropriate method depends on the analysis plan. Whether a multiplicity adjustment was used is not reported in the registry record.

The registry’s reported analysis method is Mixed Models Analysis, but other statistical-plan details are not reported in the registry record. These include the analysis population, missing-data method, any covariate adjustments, whether analyses were prespecified in a protocol, and whether interim analyses were conducted. The record therefore supports a description of the endpoint summaries and named analysis method, but not a reconstruction of the complete analysis plan.

11. Confidence Intervals and Precision

General: a confidence interval gives a range of values compatible with the observed data under the assumptions of the calculation. Its width conveys precision: a narrower interval indicates less uncertainty in the estimated contrast, while a wider interval indicates more. A confidence interval is not a range containing every participant’s outcome, and it is not a probability statement about the true value after the data have been observed.

The simple unadjusted recalculation for the primary mean difference produces an interval entirely above zero when the contrast is defined as Sensor Augmented Therapy minus Artificial Pancreas. This is consistent with the direction of the posted group means and the recalculated p-value of <0.001. The interval is not the confidence interval from the registry’s mixed-model analysis: that model-based interval is not reported in the registry record. The recalculated interval should therefore be read as a summary-statistic comparison, not as a model-adjusted estimate.

The trial’s posted primary arm summaries include 64 participants for Artificial Pancreas and 61 for Sensor Augmented Therapy. Their standard deviations are 1.7 and 3.4. Those quantities help explain why an estimate should be considered together with its uncertainty rather than judged by the difference alone. They do not reveal the full distribution of outcomes or the number and pattern of monitoring observations that contributed to each participant’s summary.

General: an interval that excludes zero is consistent with a nonzero difference under the specified calculation, but that fact does not establish how important the difference is in practice. The registry does not provide a clinical-importance threshold for this endpoint. It also does not report enough detail to determine whether the model-based and recalculated intervals would have the same width or use the same assumptions. Precision should therefore be interpreted as precision of the available unadjusted recalculation, with model-specific precision not reported in the registry record.

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 based on the observed effect, not the original trial’s design justification. The registry record does not report the trial’s own sample-size assumptions or justification. The illustrations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as a stated fraction of the observed effect.

For these continuous outcomes, the figures are participants per arm, not total event counts. For the primary endpoint, planning for 100% of the observed effect gives 44 participants per arm for 80% power and 59 participants per arm for 90% power. Planning for 75% of the observed effect gives 78 participants per arm for 80% power and 104 participants per arm for 90% power. Planning for 50% of the observed effect gives 175 participants per arm for 80% power and 234 participants per arm for 90% power.

General: power is the chance that a study will detect an effect of a specified size if that effect is present, under the assumptions used for the calculation. The 80% and 90% columns represent two different target probabilities of detecting the selected effect. A higher target probability generally requires more participants under otherwise fixed assumptions. The smaller-effect planning rows also require more participants per arm than the larger-effect rows, because a smaller difference is harder to distinguish from background variability.

These figures are not observed power. They do not tell whether the completed trial was adequately designed according to its original plan, because the original assumptions are not reported in the registry record. They are conditional illustrations using the stated effect fractions and calculation assumptions, and should not be mistaken for the number actually enrolled or analyzed.

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 = 1.600; pooled standard deviation σ = 2.668.
2. n = 2 × (1.96 + 0.842)² × 2.668² / 1.600² = 43.7, rounded up to 44 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 <- 1.6                      # observed difference in means
sigma <- 2.66819569984717                     # 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(2985866)
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 = 44 (80% power), 59 (90% power)
 75% of observed effect: n per arm = 78 (80% power), 104 (90% power)
 50% of observed effect: n per arm = 175 (80% power), 234 (90% power)
power.t.test: 44.6 
Simulated power at n = 44 per arm: 0.794 (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 reports the sponsor’s primary comparison as a Mixed Models Analysis, marked as a superiority result with a p-value below < 0.001. It does not report the mixed-model estimate or its confidence interval. The arm-level means and the recalculated mean difference are therefore not the sponsor’s reported model estimate; they are an unadjusted comparison derived from the posted summary statistics.

The unadjusted recalculation gives a mean difference of 1.6 when the contrast is stated as Sensor Augmented Therapy minus Artificial Pancreas, with p <0.001 and an interval entirely above zero. The direction agrees with the group means and with the reported superiority result. The registry record does not flag a discrepancy between the recalculation and the reported analysis. The sponsor’s own analysis is the principal reported analysis, while the recalculation provides a transparent, simpler comparison based on group summaries.

Because the mixed-model specification is not reported, it is not possible to determine which adjustments or assumptions account for any difference between the methods, or to reproduce the model-based estimate. The model-based result should not be replaced by the unadjusted calculation; both should be understood according to the information each provides. The registry’s reported model result supplies the sponsor’s analysis conclusion, and the recalculation supplies a separate descriptive estimate on the outcome’s original scale.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists nine outcome measures and nine analyses, including a primary endpoint and secondary glucose outcomes. It identifies the primary comparison as a superiority analysis and names Mixed Models Analysis as the reported method. The record does not describe a family-wise error strategy, false-discovery approach, hierarchical testing plan, or other adjustment for multiple comparisons. Whether the primary endpoint was protected from multiplicity by a prespecified testing hierarchy is not reported in the registry record.

This matters most for interpreting the secondary analyses. General: when investigators examine several outcomes, unadjusted testing can increase the overall chance of false-positive findings across the set. A significant-looking result on one secondary measure is therefore less persuasive when it is not clear whether the outcome was prespecified and whether the analysis accounted for the other tests. The registry record does not report those details, so the secondary p-values should be read cautiously and not as a collection of equally confirmatory findings.

The analysis population is not reported in the registry record. The outcome summaries identify 64 participants in one primary arm and 61 in the other, but the record does not specify whether these are all randomized participants, all participants with usable outcome data, or another analysis set. A participant flow account, reasons for exclusions, and the treatment of withdrawals are not reported in the registry record.

Missing-data handling is also not reported in the registry record. In a continuous-glucose-monitor endpoint, missingness could affect the amount or pattern of monitoring data represented in participant-level summaries, but the record does not describe completeness thresholds, imputation, or sensitivity analyses. The mixed-model name alone does not establish how missing observations were handled in this trial.

The registry does not report interim analyses, stopping rules, or the detailed statistical analysis plan. These omissions limit independent evaluation of prespecification and the conditions behind the reported p-values. They do not erase the posted results, but they constrain how fully the analysis can be audited from the registry record.

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: hypoglycemic seizure (1.5%), dehydration (1.5%).

The adverse-event record lists 65 participants at risk for serious adverse events in the Artificial Pancreas group, with 1 affected, and 62 at risk in Sensor Augmented Therapy, with 0 affected. For other adverse events, the listed numbers at risk are 65 and 62, with 0 and 2 affected, respectively. Deaths are listed as 0 in both groups.

The registry also lists hypoglycemic seizure and dehydration among the serious adverse-event terms, each with a reported 1.5%. The record does not provide additional detail about event timing, severity beyond the serious classification, relatedness to the intervention, recurrence, or the group attribution of those terms. The terms should therefore be read as the adverse-event entries reported, not as a complete account of the circumstances or causes.

General: adverse-event counts describe events or affected participants as defined by the record; they are not automatically measures of comparative risk unless definitions, follow-up, exposure, and analysis methods are clear. The figures are small and should be read as descriptive safety information. The registry record does not report a formal statistical comparison of adverse events, and these counts alone do not establish a difference in safety between the groups.

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

The posted primary data show different group averages for the percentage of monitored time below 70 mg/dL during the final 11 weeks after randomization. The mean was 2.4 for Artificial Pancreas and 4.0 for Sensor Augmented Therapy. The unadjusted mean difference, defined as Sensor Augmented Therapy minus Artificial Pancreas, was 1.6, with an unadjusted p-value of <0.001. The registry’s reported mixed-model analysis is also marked as a superiority result.

This result concerns one continuous-glucose-monitor measure over a specified period. It does not, by itself, establish what happened for every participant, what happened outside that period, or what happened for other outcomes not quantitatively reported. A group mean can conceal substantial variation among participants, and the reported standard deviations indicate that individual results varied around the group averages.

General: statistical evidence of a difference and the clinical importance of that difference are separate questions. The registry does not report a threshold for a clinically important change in this outcome. The p-value is not a measure of the size of the effect, and the observed contrast is not a guarantee that the same result would occur in another population or setting.

The study was randomized, which supports a comparative interpretation of the assigned groups, but it was unmasked. The registry does not report enough detail about the analysis population, missing outcome data, or model specification to resolve all possible sources of uncertainty. The secondary outcomes provide a broadly consistent direction in the numerical summaries available, but multiplicity handling is not reported and quantitative results for some listed secondary outcomes are absent.

Accordingly, the result should be read as evidence of a group difference in the posted primary endpoint under the reported analysis, rather than as a complete statement about overall efficacy, safety, or individual outcomes. The registry record does not establish outcomes beyond those it reports.

18. Why This Trial Matters Statistically

This trial illustrates why a continuous monitoring endpoint requires more than a single average for interpretation. The primary result combines a clear definition of the measurement—percentage of total continuous-glucose-monitor time below 70 mg/dL—with a fixed post-randomization interval, reported group means, measures of spread, and a comparative analysis. Together, these elements let readers distinguish the outcome scale from the statistical evidence about a group difference.

It also shows the value of separating a sponsor-reported model analysis from a simple recalculation based on arm summaries. The registry names a mixed-model method and reports a superiority result, while the posted summary statistics permit an unadjusted mean contrast. Those are related but not interchangeable forms of evidence. The model’s estimate and interval are not reported, so the simpler calculation adds descriptive transparency without reconstructing the sponsor’s model.

The listed secondary endpoints show another important statistical issue: related outcomes can provide context, while multiple tests make isolated secondary findings harder to interpret. Without a reported multiplicity plan, readers should distinguish the primary result from the broader set of secondary analyses.

Finally, precision and design assumptions must not be confused. Planning illustrations can show how sample requirements change under different target effects and power levels, but they do not reveal the original justification for enrollment or the power of the completed study. The registry’s missing details on analysis populations, missing data, and model specification also demonstrate why numerical results are most informative when accompanied by a complete analysis record.

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 record does not report participant eligibility criteria beyond the condition, detailed baseline characteristics, or the demographic composition of the enrolled group. Those omissions limit assessment of how closely the participants represented other populations with type 1 diabetes. The number enrolled is reported as 127, while the primary outcome summaries list 64 and 61 participants; the record does not explain the relationship between those counts or provide participant flow details.

The study was unmasked. The registry does not report procedures intended to reduce the influence of knowing the assigned intervention on participant behavior, monitoring, or outcome handling. The primary endpoint was measured using continuous glucose monitoring, but completeness requirements and the handling of missing monitoring data are not reported. The analysis population and reasons for any unavailable outcome data are also not reported in the registry record.

The registry names Mixed Models Analysis but does not report the model specification, covariates, covariance structure, or model-based estimate and confidence interval. The unadjusted recalculation can describe the arm-level mean contrast but cannot fill in those missing model details. The record also does not report a clinically important difference threshold, a multiplicity adjustment, the statistical analysis plan, interim analyses, or the trial’s original sample-size justification.

The outcome window is the final 11 weeks after randomization, so the posted primary result is limited to that period. Quantitative results for all listed secondary outcomes are not reported in the registry record. The safety entries give counts and event terms, but do not provide detail on event timing, attribution, or a formal between-group safety analysis. These are limitations of the available record and should be kept distinct from the results that are reported.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: The International Diabetes Closed Loop (iDCL) Trial: Protocol 1 — NCT02985866.
• Registry record last updated 2022-12-05; status last verified 2022-11.
• Results first posted 2022-12-05.
• 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

NCT02985866 is a completed, randomized, open-label, parallel-group trial sponsored by the University of Virginia. It enrolled 127 participants with type 1 diabetes mellitus and compared an Artificial Pancreas closed-loop system with sensor-augmented therapy using a personal insulin pump. The primary outcome was mean percentage of continuous-glucose-monitor time below 70 mg/dL during the final 11 weeks after randomization.

The reported primary means were 2.4 for Artificial Pancreas and 4.0 for Sensor Augmented Therapy. The unadjusted recalculated difference was 1.6 in the direction of comparator minus Artificial Pancreas, with p <0.001. The registry reports a mixed-model superiority result but does not provide its estimate or confidence interval. Secondary numerical summaries available in the record generally follow the same direction; multiplicity handling and several analysis details are not reported.