The Summer Camp Study: 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 The Summer Camp Study: Blood Glucose Control With a Bi-Hormonal Bionic Endocrine Pancreas (NCT01833988), a randomized na study. The primary outcome is Difference in Average Blood Glucose (BG) Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) Periods as Determined From All Scheduled HemoCue Measurements With Mean Evenly Weighted Across the Daytime and Nighttime Hours..

2. Clinical Question

The central statistical question is whether average glucose measurements differed between the two study conditions. The posted estimate is a comparison of condition-period averages, not a measure of how every participant responded. Because the design is crossover, participants were assigned to conditions in a study that compared periods rather than simply assigning one permanent condition to each person. The registry record does not report the sequence of conditions, period-specific completion, or the analysis method used to account for repeated measurements within participants.

The primary outcome uses scheduled HemoCue measurements during a one-week timeframe. Other listed outcomes address related measures, including continuous glucose monitor readings, threshold-based summaries, and hypoglycemic events. These outcomes provide different views of glucose during the study, but the primary question remains the average blood glucose contrast between the two conditions. The record does not report a threshold for deciding whether a difference is clinically important.

3. Trial Design

This was a randomized, unmasked crossover study sponsored by Massachusetts General Hospital. The registry lists 32 participants enrolled and identifies type 1 diabetes as the condition. The study title and summary place the research at the Clara Barton diabetes camps and describe young people ages 12–20. The record lists the study as completed, with a start in 2013-04 and primary completion in 2013-08.

The two conditions were a bi-hormonal bionic pancreas and usual care. The bionic pancreas is described as a device that automatically delivers both insulin and glucagon. Usual care is also identified in the outcome descriptions as the open-loop insulin pump condition. The allocation is listed as randomized, masking as none, and the model as crossover. In a crossover design, participants are assigned to experience study conditions in different periods; that structure can allow comparisons within the same participant and can reduce the influence of stable differences between people. This is a general feature of crossover designs, not a claim about the specific analysis used in this trial.

The primary outcome timeframe is one week. The record does not report the crossover sequence, how long each condition period lasted beyond the primary outcome timeframe, any washout period, or how carryover between periods was assessed. It also does not report how many participants completed both conditions or the statistical model used to handle the repeated observations inherent in a crossover study. Those details matter when interpreting the precision and independence of the condition comparison.

The record lists phase as not applicable. It does not report a statistical analysis plan, a sample-size justification for the original trial, or interim analyses. The displayed estimates can therefore describe the posted comparisons, but they do not establish which analyses were planned in advance or how the study's original enrollment target was selected.

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.

Population: Cross over design, 16 participants in each arm and then switched to second arm after a 1 day washout period.

6. Primary Endpoint

The registry lists the primary outcome as Difference in Average Blood Glucose (BG) Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) Periods as Determined From All Scheduled HemoCue Measurements With Mean Evenly Weighted Across the Daytime and Nighttime Hours., measured over 1 week.

The primary endpoint is the difference in average blood glucose between the closed-loop bionic pancreas period and the open-loop insulin pump period. It uses all scheduled HemoCue measurements and defines the mean with daytime and nighttime hours weighted evenly. The endpoint's unit is mg/dL, and its timeframe is one week.

An average is a continuous summary: it combines measured glucose values into a single value for each condition. Comparing the averages gives a concise account of glucose levels over the measurement period. Even weighting daytime and nighttime hours is intended to keep the daily and nighttime portions from being represented solely according to how many scheduled readings fall in each. That is a feature of the endpoint definition, not evidence about the number or timing of measurements.

As a general statistical matter, a continuous average can make a broad comparison easier to interpret than a list of individual readings. It also compresses information: two periods with the same average can still differ in variability, time spent at particular glucose levels, or occurrence of low readings. The registry lists secondary outcomes addressing some of these other dimensions.

The posted condition averages are 138 mg/dL for the bionic pancreas and 157 mg/dL for control. The estimated contrast is expressed as bionic pancreas minus control, so its negative direction indicates a lower posted average in the bionic pancreas condition. The registry record does not report a clinical threshold for interpreting the size of this average difference.

7. Results Posted in the Registry

• Bionic Pancreas: mean 138.00 versus 157.00 on control; difference -19.00 (95% CI -30.24 to -7.76), standardized effect d = -0.83; 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.

Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Number of Subjects With Mean BG < 154 mg/dl

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Bi-hormonal Bionic Pancreas: 90.6% (29/32) versus 43.8% (14/32) on control; risk difference 46.9 percentage points (95% CI 26.9 to 66.8), risk ratio 2.07 (1.38–3.12); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Bi-hormonal Bionic Pancreas than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Difference in the Percentage of Study Days With Mean CGM BG </= 154 mg/dl Over the Duration of the Closed-loop Period vs. the Usual Care Period

• 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:
  • Bi-hormonal Bionic Pancreas: mean 75.00 versus 49.30 on control; difference 25.70 (95% CI 11.11 to 40.29), standardized effect d = 0.86; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Bi-hormonal Bionic Pancreas than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Number of Severe Hypoglycemic Episodes and Nadir BG During Exercise

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Bi-hormonal Bionic Pancreas: 0.0% (0/32) versus 3.1% (1/32) on control; risk difference -3.1 percentage points (95% CI -9.2 to 2.9); unadjusted two-proportion z-test p = 0.313.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Bi-hormonal Bionic Pancreas and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes span several related aspects of glucose measurements: average glucose by different measurement sources or time periods, the number of participants meeting a threshold, the percentage of days meeting a threshold, time within glucose ranges, hypoglycemic events, and exercise-related measures. Together, these outcomes can add context to an average by describing thresholds, variability across time, or specific events. They are not all measuring the same quantity, so agreement across them would not make their estimates interchangeable.

Some posted secondary comparisons are directionally consistent with the primary average comparison, including summaries based on meeting a glucose threshold and the percentage of days at or below a threshold. The exercise-related severe hypoglycemic episode comparison is less conclusive: its interval includes no difference, and its unadjusted p-value is 0.313. These patterns can be considered alongside the primary outcome, but they do not replace it or establish a single overall treatment effect.

The registry lists 28 outcome measures and 25 posted secondary outcomes. The breadth of outcomes raises a multiplicity issue: when many comparisons are made, some may appear statistically notable by chance even if none of the underlying effects is real. The registry record does not report a correction for multiple comparisons, a hierarchy specifying which outcomes were confirmatory, or a plan identifying how secondary findings should be interpreted.

For that reason, unadjusted secondary results carry less weight as confirmatory evidence than a prespecified primary comparison. They can be useful for describing the data and identifying patterns for future study, but their p-values should not be read as if each outcome were the only comparison undertaken. The record also does not report the analysis methods used for the secondary outcomes, including how repeated measurements within participants were handled.

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

• Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Average BG as Determined From All HemoCue Measurements Taken During the Day/Nighttime Including All Extra Measurements.
• Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Number of Subjects With Mean BG < 154 mg/dl
• Difference in the Percentage of Study Days With Mean CGM BG </= 154 mg/dl Over the Duration of the Closed-loop Period vs. the Usual Care Period
• Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Number of Hypoglycemic Events (BG <70mg/dl) as Determined From HemoCue Measurements
• Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Fraction of Time Spent Within CGMG (Continuous Glucose Monitor) Ranges (< 70 mg/dl, 70-120 mg/dl, 70-180 mg/dl, > 180 mg/dl, > 250 mg/dl)
• Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Mean BG During Exercise
• Difference Between Closed-loop (Bionic Pancreas Arm) and Open-loop (Insulin Pump Arm) in Number of Severe Hypoglycemic Episodes and Nadir BG During Exercise
• Difference in Mean CGMG on Day 1 vs. Remaining Days (Days 2-5) Between Closed Loop (Bionic Pancreas Arm) and Usual Care (Insulin Pump Arm)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a continuous measure: average blood glucose in mg/dL. A mean summarizes measured values by adding their contributions and expressing their central level. The registry specifies that scheduled HemoCue measurements are used and that daytime and nighttime hours are evenly weighted. It does not describe the exact calculation steps beyond that definition.

What does a mean difference represent?

A mean difference places the two condition averages on the original outcome scale. Here, the contrast is bionic pancreas minus control, and the posted estimate is −19 mg/dL. A negative value means the first condition's average is lower than the second condition's average. The contrast is an estimate from this study, not a guarantee that every participant had a lower value in the bionic pancreas period. Individual responses can vary even when a group-level average difference is present.

Why does the crossover design matter?

In general, a crossover study compares outcomes from different conditions experienced by the same participants. A paired analysis can use each participant as their own reference, which may account for stable personal characteristics that differ between participants. However, measurements from the same person are related rather than independent. The analysis should reflect that structure, and the order or timing of conditions may also matter if there are period effects or carryover. The registry identifies the study as crossover but does not report the specific model, sequence, period adjustment, or carryover assessment. It is therefore not possible to establish from the registry record how those features were incorporated.

The posted arm summaries list 32 participants for each condition. In a crossover trial, these are condition-specific summary counts and do not imply two separate sets of 32 people. The registry does not report whether all participants posted measurements for both conditions or how missing period data were handled.

What do standard deviations describe?

The registry reports standard deviations of 18 mg/dL for the bionic pancreas condition and 27 mg/dL for control. In general, a standard deviation describes how spread out individual observations are around a mean. It is not the same as the uncertainty in the mean or the confidence interval for a difference. The relation between those quantities depends on the analysis and, in a crossover study, on the within-person relationship between measurements. That relationship is not reported in the registry record.

How should the confidence interval and p-value be read?

A confidence interval expresses uncertainty around an estimated effect under a specified statistical procedure. The primary interval runs from −30.24 to −7.757 mg/dL. Its endpoints are both below zero, the value corresponding to no mean difference. The interval gives a range of effect estimates compatible with the data and procedure; it is not a range containing the true value with certainty. The registry record does not identify the calculation method or model used for this interval.

The primary p-value is reported as < 0.001 and is unadjusted. A p-value is a measure of how unusual data at least as inconsistent with a null model would be if that model and the analysis assumptions held. It does not give the probability that the observed effect arose by chance, and it does not quantify clinical importance. General explanations of p-values and intervals do not substitute for information about the exact analysis used.

Why can over-dispersion matter?

For a general statistical explanation, over-dispersion occurs when observed variability is greater than a model assumes. If a model understates variability, standard errors can be too small and intervals too narrow, potentially making statistical evidence appear stronger than warranted. This is a general concern for analyses of counts or proportions, not a finding reported for this trial. The record does not identify a model for the binary secondary outcomes or report a dispersion assessment.

What is the role of the secondary outcomes?

Secondary outcomes use different summaries, including counts, proportions, and continuous measures. A risk difference, for example, compares the proportions experiencing a defined outcome across conditions; a mean difference compares average values on a continuous scale. Their interpretation depends on the outcome definition and the method used to account for participants' repeated condition periods. The registry posts selected comparisons but does not report a detailed analysis plan or multiplicity correction. Their p-values should therefore be regarded as unadjusted results rather than as a family of confirmed independent findings.

The record reports no sponsor analysis details beyond the posted summaries and comparisons. The statistical model, analysis population, handling of missing data, assumptions for intervals and tests, and any planned hierarchy of outcomes are not reported in the registry record. These omissions limit how fully the calculations can be evaluated from the record alone.

11. Confidence Intervals and Precision

The primary estimate is −19 mg/dL, and the posted 95% confidence interval extends from −30.24 to −7.757 mg/dL. The interval represents uncertainty around the estimated difference, conditional on the method and assumptions used to calculate it. Its endpoints are below zero, so the interval does not include an exactly null difference. The registry record does not report the interval method or the crossover model underlying it.

The interval allows a range of plausible negative differences rather than identifying a single exact effect. Its width is relevant: greater width indicates less precision, while a narrower interval indicates more precision under the same procedure. The stated interval does not measure how much participants' glucose values varied individually; the condition standard deviations describe that separate feature of the data.

Precision depends on more than the number of participants. In a crossover study, within-person similarity across conditions can affect the uncertainty of the paired contrast. Missing observations, period effects, and the statistical model can also matter. The record lists 32 participants for each condition but does not report the correlation between repeated measurements, completeness of paired data, or the method used to account for these issues.

The secondary estimates have their own uncertainty and should be read in relation to their outcome definitions. A confidence interval that includes zero does not establish that the conditions are equivalent; it indicates that the interval includes no difference as well as other values. Conversely, an interval excluding zero does not establish clinical importance. The registry record does not report a threshold defining a clinically meaningful difference for the primary or secondary outcomes.

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 planning figures are participants per arm, not total participants or total events. They are planning illustrations based on the observed primary effect, rather than the original trial's documented design assumptions. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The registry record does not report the assumptions used to choose the original enrollment target.

For 100% of the observed effect, the illustration gives 23 participants per arm for 80% power and 31 participants per arm for 90% power. For 75% of the observed effect, it gives 41 participants per arm for 80% power and 55 participants per arm for 90% power. For 50% of the observed effect, it gives 92 participants per arm for 80% power and 123 participants per arm for 90% power.

In plain terms, 80% power means that, under the planning assumptions and the specified effect, a study of that size would have an 80% chance of detecting the effect at the stated significance level. The 90% column describes a higher chance of detection under those same assumptions, which generally calls for a larger planned sample. Smaller effects are harder to distinguish from random variation, so planning for smaller effects requires more participants.

These counts are not observed power. They are illustrative sample-size calculations based on the observed effect and do not show the probability that the completed trial was successful. The trial is crossover, while the stated planning assumptions describe equal allocation between two arms; the registry does not report whether these illustrations reflect the original crossover 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 = -19.000; pooled standard deviation σ = 22.946.
2. n = 2 × (1.96 + 0.842)² × 22.946² / -19.000² = 22.9, rounded up to 23 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 <- -19                      # observed difference in means
sigma <- 22.9455878111675                     # 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(1833988)
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 = 23 (80% power), 31 (90% power)
 75% of observed effect: n per arm = 41 (80% power), 55 (90% power)
 50% of observed effect: n per arm = 92 (80% power), 123 (90% power)
power.t.test: 23.9 
Simulated power at n = 23 per arm: 0.782 (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. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and numerous secondary outcomes, but a multiplicity strategy is not reported in the registry record. It is not stated whether the trial used an adjustment for multiple testing, a prespecified hierarchy, or another approach to control the chance of false-positive conclusions across several comparisons. The posted p-values for secondary comparisons are unadjusted. Consequently, they should not be treated as though they arose from a single isolated hypothesis test.

Missing-data methods are also not reported in the registry record. The record does not state how incomplete HemoCue or continuous glucose monitor measurements were handled, whether participants with incomplete condition periods were included, or whether any imputation was used. It does not identify the analysis population, such as whether analyses followed all randomized participants or only those with measurements in both conditions.

These omissions are particularly relevant to a crossover study. The record does not report how the analysis treated participants who may have missing observations in one period, whether the condition comparison was paired, or how period and sequence effects were addressed. It also does not report whether carryover was assessed. These are general considerations for crossover analyses; the registry does not establish that any particular issue occurred.

The record does not report the trial's statistical analysis plan, interim analyses, or the sample-size justification used when the study was designed. It therefore cannot establish which outcomes or analyses were prespecified or whether the posted analysis follows the original plan. These uncertainties do not change the numerical summaries, but they limit assessment of the strength and confirmatory status of the statistical evidence.

15. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The posted serious adverse event figures list 0 affected participants among 32 at risk in the bionic pancreas condition and 0 among 32 at risk in control. This means no serious adverse events were recorded in either condition in the posted summary. It does not establish that the risk of a serious event is absent; a study with 32 participants per condition can provide limited information about uncommon events.

Other, non-serious adverse events are listed for 6 participants in the bionic pancreas condition and 4 in control, with 32 at risk in each. These counts do not describe event types, severity, timing, or whether an individual had more than one event. The registry record does not report those details in the posted safety summary.

Deaths are not reported in the registry record. The serious-event figures are safety summaries, not a statistical test of whether the conditions differ. They should be interpreted as reported counts within this study, without inferring broader safety conclusions from the absence of recorded serious events.

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

The posted primary result estimates a lower average blood glucose in the bionic pancreas condition than in control during the study's one-week primary timeframe. The mean difference is −19 mg/dL, and the confidence interval is entirely below zero. This describes the direction and statistical precision of the posted comparison; it is not, by itself, a conclusion about clinical benefit or about what will happen for an individual participant.

The registry record does not report a threshold for clinical importance. Statistical evidence against a zero difference and clinical importance are distinct questions: the first concerns compatibility with a null model, while the second requires a judgment about whether the magnitude matters in the relevant setting. No such clinical threshold or decision rule is reported.

The study concerns young people with type 1 diabetes in a diabetes camp environment. Its results should not automatically be generalized to other ages, settings, or patterns of device use. The record does not report follow-up beyond the study periods or longer-term outcomes, and the primary endpoint covers one week. The registry also identifies the trial as unmasked, so participants and study staff were not blinded according to the listed masking status. The record does not describe whether or how that could have influenced measurement or study conduct.

The crossover structure means the condition summaries are not equivalent to results from two unrelated groups. Interpretation depends on how repeated measurements, period order, missing observations, and possible carryover were handled. Those analytical details are not reported in the registry record. The posted secondary outcomes add descriptive context but are subject to multiple-comparison concerns because an adjustment plan is not reported.

Accordingly, the statistical evidence supports describing a negative posted average difference under the reported comparison, while leaving important questions about analysis details, clinical relevance, longer-term outcomes, and generalizability unresolved. No treatment recommendation follows from these statistics alone.

17. Why This Trial Matters Statistically

This trial illustrates why design and analysis details matter alongside an estimated mean difference. The registry identifies a randomized crossover design, in which the same participants may contribute measurements under different conditions. That structure can be informative because within-participant comparisons may reduce the influence of stable differences between people. It also requires suitable treatment of repeated observations, period effects, and possible carryover. The record does not report how those issues were addressed, so readers can recognize the design's potential while noting the limits of the available statistical description.

The primary endpoint is an average based on scheduled measurements, with daytime and nighttime hours weighted evenly. This makes the outcome definition part of the interpretation: an average is a useful summary, but it does not show the full distribution of readings or describe every episode. Secondary measures broaden the view by considering thresholds, ranges, and hypoglycemic events, while also creating the need to consider multiplicity.

The reported confidence interval adds information beyond the point estimate by showing uncertainty around the mean difference. Yet statistical precision cannot answer whether the observed difference is clinically important, and the registry record reports no clinical threshold for that judgment. Nor can a one-week outcome answer longer-term questions. These distinctions are central to interpreting clinical trial results responsibly.

Finally, planning illustrations based on an observed effect are not a substitute for the original sample-size rationale. Separating the trial's observed results from later planning calculations helps avoid treating a small study's effect estimate as a guaranteed target for future research. The record does not report the original design assumptions or a formal sample-size justification.

18. 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 trial enrolled 32 participants and took place in a diabetes camp context among young people with type 1 diabetes. The study's size and setting limit how broadly the posted averages can be applied. The registry does not report a broader set of sites or populations, and the findings should not be presumed to describe other clinical settings or age groups.

The primary outcome covers a one-week timeframe. The registry does not report longer-term glucose outcomes or follow-up beyond the study periods. It also does not report the sequence of crossover conditions, any washout period, or how carryover and period effects were assessed. These details are important to evaluating a crossover comparison, but their absence does not establish that a particular design problem occurred.

The study is listed as unmasked. This is a design feature relevant to interpretation, although the registry does not describe its impact on measurement or conduct. The primary outcome relies on scheduled HemoCue measurements, while several secondary outcomes use continuous glucose monitor summaries or event counts. The record does not provide enough detail to assess missing readings, measurement completeness, or how different measurement sources were reconciled.

The statistical analysis plan, analysis population, missing-data procedures, and method for accounting for repeated condition measurements are not reported in the registry record. Neither are the original sample-size justification and multiplicity strategy. The secondary outcomes cover many related measures, and their posted unadjusted results therefore require caution. These reporting gaps constrain statistical appraisal even though the main estimate and confidence interval are posted.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: The Summer Camp Study: Blood Glucose Control With a Bi-Hormonal Bionic Endocrine Pancreas — NCT01833988.
• Registry record last updated 2017-09-08; status last verified 2017-08.
• Results first posted 2017-08-11.
• 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.

22. Record Summary

The registry record describes a completed, randomized, unmasked crossover study sponsored by Massachusetts General Hospital, with 32 participants enrolled. The study concerned young people with type 1 diabetes in the Clara Barton diabetes camp setting and compared a bi-hormonal bionic pancreas with usual care.

The primary outcome was the difference in average blood glucose from scheduled HemoCue measurements, with daytime and nighttime hours evenly weighted over one week. The posted averages were 138 mg/dL for the bionic pancreas condition and 157 mg/dL for control. The mean difference was −19 mg/dL, with a 95% confidence interval from −30.24 to −7.757 and an unadjusted p-value of < 0.001.

The record lists multiple secondary outcomes and posts selected comparisons. It does not report the statistical analysis plan, multiplicity adjustment, missing-data approach, or the specific method used to account for the crossover design. The clinical importance of the primary difference is not defined in the registry record.