Outpatient Automated Blood Glucose Control With a Bi-hormonal Bionic Endocrine Pancreas: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Outpatient Automated Blood Glucose Control With a Bi-hormonal Bionic Endocrine Pancreas (NCT01762059), a randomized na study. The primary outcome is Average Blood Glucose (Co-primary Outcome).

2. Clinical Question

3. Trial Design

This was a randomized, unmasked crossover study sponsored by Massachusetts General Hospital. It enrolled 32 participants and is listed as completed. The study began in 2013-01, with primary completion in 2013-10. The phase is listed as NA. The registry title describes a feasibility study of outpatient automated blood glucose control using a bi-hormonal bionic endocrine pancreas.

A crossover design generally involves participants receiving more than one study condition in different periods, allowing comparisons that may use each participant as their own reference. The registry record classifies this trial as crossover, but does not report the sequence assignment, the order in which participants received the conditions, or whether there was a washout period. Randomization is recorded, but its specific implementation is not described. Masking is listed as none.

The intervention was closed-loop control with the bionic pancreas. The usual-care description specifies insulin pump therapy according to usual practice for 5 days, with participants sleeping at home and keeping their usual daytime schedule. Diet and exercise were not restricted in that description, and participants wore a blinded continuous glucose monitor. The primary outcome timeframe is also stated as 5 days of closed-loop control. The record does not give a corresponding primary-outcome timeframe for usual care.

The registry lists 21 outcome measures, including average blood glucose, low-glucose measurements, hypoglycemic events, and glucose nadir during exercise. It reports 0 analyses. The statistical analysis plan, allocation details, period sequence, and reasons for choosing the enrollment target are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: This outcome measure was only assessed for the closed loop control period, and was not assessed during the usual care arm.

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 Average Blood Glucose (Co-primary Outcome), measured over 5 days of closed-loop control.

The listed co-primary outcome is average blood glucose, measured in mg/dL, during the closed-loop control period. The record defines the measure using HemoCue capillary measurements taken during daytime and GlucoScout venous measurements taken during nighttime. Its timeframe is 5 days of closed-loop control. The record explicitly states that this outcome was assessed only during closed-loop control and not during usual care.

An average is a summary of the overall level of a continuous measurement. In general, it can help describe whether glucose readings tend to cluster at a higher or lower level over the assessment period. It does not, by itself, describe the range or variability of readings, the duration of particular glucose levels, or the timing of changes. The registry lists separate secondary outcomes concerning low readings and hypoglycemic events, which address different aspects of glucose measurements.

The endpoint is relevant to the stated objective of examining glycemic control, but the record does not post the average, its standard deviation, or an estimate of a difference between study conditions. Because the primary measure was not assessed during usual care, the registry entry does not support a direct usual-care comparison for this endpoint. The endpoint definition and planned timeframe therefore describe what was measured, but do not establish what the measurements showed.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary outcome summarize?

The primary outcome is a continuous measurement summarized by a mean. The registry specifies average blood glucose in mg/dL, using capillary measurements during daytime and venous measurements during nighttime. A mean condenses multiple readings into a single measure of central level. That compression can make a complex series easier to describe, but it also discards information about the pattern of readings. Two periods can have similar means while differing in variability, timing, or the frequency of low readings. The registry lists separate secondary measures that address some of those other features.

The standard deviation is the listed measure of dispersion. In general, it describes how spread out observations are around their mean. It is not the same as the uncertainty in the estimated mean: uncertainty also depends on the amount and structure of the data and on the analysis method. The record does not report the primary outcome’s mean or standard deviation, so neither the level nor the observed spread can be interpreted from the posted results.

What would a comparative analysis need to address?

Randomized allocation can support a fair comparison when assignment is implemented as intended and the analysis follows an appropriate plan. The record identifies randomization and a crossover model, but does not describe the allocation procedure or the sequence participants followed. In general, crossover observations from the same person are related. An analysis would commonly account for that within-person relationship rather than treating measurements from the same participant as independent. The statistical method actually used is not reported in the registry record.

Period and sequence effects can also matter in crossover studies. A participant’s measurement in a later period may be influenced by the period itself or by the preceding condition. General crossover analyses can consider such features when relevant, but the registry record does not say whether they were assessed or how. It also does not report a washout period, the length of each usual-care period, or the sequence assignment. These omissions prevent a detailed appraisal of how the design was translated into an analysis.

The main endpoint poses an additional limitation for comparison: the registry says average blood glucose was assessed only during closed-loop control, not during usual care. A crossover label does not make a direct comparison available when the endpoint is not recorded in both conditions. The registry does not give a substitute comparison or explain whether the outcome definition changed for any analysis.

How should uncertainty and significance be understood?

In general, a confidence interval communicates the uncertainty around an estimated effect under the assumptions of a particular analysis. A narrow interval indicates greater precision than a wide interval, though both the estimate and interval must be considered in context. A p-value addresses compatibility between data and a specified null model; it is not a measure of clinical importance, and it does not give the probability that a hypothesis is true. The registry record posts no confidence interval, p-value, or primary comparative estimate. No conclusion about statistical significance or precision is therefore available from this entry.

The registry lists 21 outcome measures. When many outcomes or comparisons are examined, multiplicity can increase the chance that at least one result appears notable by chance. General approaches include identifying a primary hypothesis in advance, adjusting for multiple comparisons, or interpreting secondary findings as exploratory. The record does not report how multiplicity was handled, whether outcomes were prespecified in the analysis plan, or which outcomes were included in any inferential analysis.

What do the planning illustrations represent?

Any sample-size figures associated with planning are illustrations, not observed power. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. In general, such calculations show how a hypothetical target effect and selected assumptions relate to planning requirements. They do not describe what the completed study achieved, and they do not establish the probability that the study detected an effect. The registry does not report the trial’s own sample-size justification, and the listed primary endpoint lacks a usual-care measurement from which a direct observed comparison could be obtained.

The registry’s enrollment of 32 participants is a study characteristic, not a substitute for a reported estimate or uncertainty interval. The analysis population, exclusions, handling of missing readings, and statistical model are not reported in the registry record. Without those details and the outcome estimates, readers cannot assess how much information contributed to an analysis or whether the analysis reflected the crossover structure.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Bionic Pancreas")
n   <- c(20)
m   <- c(138)
sp  <- c(14)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Bionic Pancreas  20  138.000  14.000  131.448  144.552

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 21 outcome measures, covering average glucose, low-glucose readings, threshold-based summaries, hypoglycemic events, and exercise-related nadir glucose. A list of outcomes does not establish which analyses were performed or which were treated as confirmatory. The record reports 0 analyses. It does not describe a prespecified hierarchy among outcomes or a method for controlling the chance of false-positive findings when multiple outcomes are considered.

In general, multiplicity matters because examining many outcomes can make isolated findings difficult to distinguish from chance variation. A statistical plan may address this by prioritizing a primary outcome, adjusting inferential thresholds, or treating secondary analyses as exploratory. Which approach is appropriate depends on the study’s objectives and plan. The statistical plan and multiplicity approach are not reported in the registry record, so no adjustment or hierarchy should be assumed.

Missing-data handling is also not reported. The record does not state how incomplete glucose readings, missing periods, participant withdrawals, or other absent measurements were handled. It does not describe whether an analysis used all randomized participants, only participants who completed both conditions, or another analysis population. The registry’s 32 enrolled participants does not itself identify the number contributing to any outcome analysis.

These details are particularly relevant to a crossover study, where analyses may depend on measurements from more than one condition for the same participant. The record does not report how incomplete crossover participation or within-person dependence was handled. The outcome measure was assessed only during closed-loop control, further limiting direct interpretation against usual care for that endpoint. These are reporting gaps, not evidence that a particular method was or was not used.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry records 20 participants at risk in the bionic-pancreas closed-loop group. It records 0 participants affected by serious adverse events and 8 affected by other adverse events. These are participant counts as entered in the record; they are not a description of the number of distinct event episodes, their severity, or their duration.

No serious adverse-event terms are listed. Deaths are not reported in the registry record, which is different from a stated count of zero. The entry does not provide corresponding safety counts for usual care, so it does not support a comparison of adverse events between conditions. It also does not report event timing, whether events were considered related to the device, or how the counts were collected.

These figures should be read as limited registry reporting, not as a complete safety profile. The at-risk count is specific to the displayed closed-loop group and should not be assumed to represent all enrolled participants or every analysis population. With no event descriptions or comparator counts, the posted figures cannot establish how the event experience differed across study conditions.

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

The record documents a randomized crossover feasibility study of outpatient automated glucose control using a bi-hormonal bionic pancreas. It identifies the intervention, usual-care comparator, enrollment of 32 participants, and an average blood glucose endpoint assessed over 5 days of closed-loop control. It also lists several secondary outcomes and limited adverse-event counts.

What it does not show is the numerical result for the primary endpoint. There is no posted mean, standard deviation, comparative estimate, confidence interval, or p-value. The registry further states that the primary outcome was not assessed during usual care. As a result, the entry does not quantify a difference between the intervention and comparator on that measure, and it does not establish whether the intervention improved glycemic control.

The secondary-outcome descriptions likewise do not establish results. Although the registry names measures of low glucose, threshold-based averages, hypoglycemic events, and exercise-related nadir glucose, it does not report their values. The listed count of 0 analyses means that this registry record does not display statistical analyses for those measures.

The safety figures indicate 0 participants affected by serious adverse events and 8 affected by other adverse events among 20 at risk in the closed-loop group. They are not accompanied by event descriptions or usual-care counts, and deaths are not reported. They therefore cannot answer comparative safety questions.

In general, interpreting a clinical trial requires both a clear design and results that quantify effects and uncertainty. Design labels and outcome definitions explain what was intended; they do not replace estimates from observed data. The registry entry supports a description of the study and its reporting limitations, but not a conclusion about the magnitude, precision, or clinical importance of an intervention effect.

14. Why This Trial Matters Statistically

This study is statistically notable for combining randomization with a crossover design in an outpatient setting. In general, crossover designs can use within-participant comparisons, which may reduce the influence of stable differences between participants. That potential advantage depends on an appropriate sequence and analysis, and on complete enough observations across conditions. The registry identifies the design but does not report the sequence, washout arrangements, or analysis method needed to evaluate how those features were handled.

The outcome definition also highlights a basic principle of comparative research: a comparison requires measurements that permit both conditions to be contrasted. The registry names average blood glucose as a co-primary outcome, but says it was assessed only during closed-loop control and not during usual care. This prevents the posted primary endpoint from serving as a direct comparison between the conditions, regardless of the crossover label.

The record further illustrates the difference between a protocol question and a results report. It lists a clinically relevant objective and a broad set of secondary outcomes, but reports 0 analyses and no numerical primary results. Without effect estimates and measures of uncertainty, statistical interpretation cannot move from describing the planned question to evaluating the observed comparison.

Finally, the safety entry demonstrates why counts need context. The affected-participant counts are recorded for the closed-loop group, but no usual-care comparison or event descriptions are given. These limits do not negate the value of the design information; they define what can responsibly be inferred from the registry record. The central statistical lesson is that design labels, outcome lists, and event counts need reported analyses and clear denominators to support interpretation.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary outcome has a substantial reporting limitation: average blood glucose was assessed during closed-loop control only and was not assessed during usual care. The registry does not report a direct comparative estimate for this endpoint. It also does not provide the mean or standard deviation, despite identifying standard deviation as the dispersion measure.

The entry lists 21 outcome measures but reports 0 analyses. No numerical results are available for the primary or secondary outcomes. The statistical analysis plan, multiplicity approach, missing-data methods, and analysis population are not reported in the registry record. These omissions prevent assessment of which participants and measurements contributed to any analysis, and how uncertainty or multiple outcomes were handled.

Although the study is classified as randomized and crossover, the record does not describe the allocation procedure, treatment sequence, period effects, or washout arrangements. It also does not state the analysis method used to account for repeated measurements within participants. These are important details for understanding a crossover study, but their absence from the registry does not establish that they were omitted from the study itself.

The device description is incomplete in the registry entry, and the record does not give detailed participant characteristics or eligibility criteria. Safety reporting is limited to the closed-loop group counts; event types, timing, relatedness, usual-care counts, and a death count are not reported. The registry lists 32 enrolled participants, but it does not state how many contributed to each outcome. Together, these limits restrict interpretation to the design and reporting details recorded, rather than a quantitative assessment of efficacy or comparative safety.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Outpatient Automated Blood Glucose Control With a Bi-hormonal Bionic Endocrine Pancreas — NCT01762059.
• Registry record last updated 2017-07-21; status last verified 2017-06.
• Results first posted 2017-07-21.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry describes a completed randomized, unmasked crossover feasibility study sponsored by Massachusetts General Hospital. It enrolled 32 participants with type 1 diabetes and compared a bi-hormonal bionic pancreas with usual care. The primary outcome was average blood glucose during 5 days of closed-loop control, assessed using daytime HemoCue capillary and nighttime GlucoScout venous measurements. The record states that this outcome was not assessed during usual care.

There are 21 listed outcome measures and 0 reported analyses. Numerical primary and secondary results, the statistical analysis plan, missing-data methods, and analysis populations are not reported in the registry record. In the closed-loop group, 20 participants were at risk, 0 were affected by serious adverse events, and 8 were affected by other adverse events. Serious-event terms are not listed, and deaths are not reported. The posted information describes the study but does not provide a comparative primary-outcome estimate.