A Study Evaluating the Effect of Albiglutide on Gallbladder Emptying in Healthy Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Evaluating the Effect of Albiglutide on Gallbladder Emptying in Healthy Subjects (NCT02496221), a randomized phase 4 study. The primary outcome is Maximum Absolute Value of Gallbladder Ejection Fraction (Emax GEF) During Cholecystokinin (CCK) Infusion, as a Measure of Maximum Effect.

2. Clinical Question

3. Trial Design

This was a randomized, double-blind, single-dose, placebo-controlled, two-way crossover study, with quadruple masking recorded in the registry. It was registered as a Phase 4 study and was marked completed. The registry lists a planned enrollment of 20 subjects. The study began on 2015-06-11, and primary completion was recorded as 2015-10-13.

Each participant was intended to receive both albiglutide 50 mg and placebo, one in each treatment period. The order was assigned by randomization. The treatment periods were separated by a washout period of at least 42 days. The summary describes fasting overnight before dosing and cholecystokinin infusion. Gallbladder volume was assessed by ultrasound, with measurements before, during, and after infusion. The primary endpoint was assessed on Day 4 in each treatment period.

The primary analysis population was narrower than the study’s all-subjects population: it included participants with the specified gallbladder ultrasound assessments in both periods. There were 17 participants contributing to each treatment arm’s posted primary endpoint summary. The registry does not establish from those arm summaries how many complete participant pairs were included in the reported mixed-model analysis.

The registry describes this study as a post-marketing commitment to the United States Food and Drug Administration. It also lists secondary safety and physiological outcomes, including blood pressure, heart rate, electrocardiogram parameters, laboratory abnormalities, adverse events, and urinalysis. Detailed original design assumptions, including the trial’s own sample-size justification, 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.

Before assignment: Eligible participants were randomized to one of the two treatments in Treatment Period 1 followed by a wash-out period and crossed over to other treatment in Treatment Period 2.

Treatment Period 1 - 4 Days

Treatment Period 2 - 4 Days

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Evaluable Subjects: Participants in the 'All Subjects' population who had gallbladder ultrasonography assessment pre-treatment and post-Baseline (during CCK infusion) for both periods. 'All Subjects' population comprised participants who received at least one dose of Investigational product.

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 Maximum Absolute Value of Gallbladder Ejection Fraction (Emax GEF) During Cholecystokinin (CCK) Infusion, as a Measure of Maximum Effect, measured over Day 4 in each treatment period.

The primary endpoint was the maximum absolute value of gallbladder ejection fraction, or Emax GEF, during cholecystokinin infusion. It was measured as a percentage of gallbladder ejection fraction on Day 4 in each treatment period. Ejection fraction represents the reduction in gallbladder volume from baseline, expressed relative to baseline volume. Baseline was defined as the average of three gallbladder volume measurements before cholecystokinin infusion in that period.

Ultrasound measurements were scheduled before infusion, at regular intervals during the 50-minute infusion, and after infusion. Taking the maximum ejection fraction during infusion focuses the endpoint on the greatest observed emptying response during the stimulus, rather than on a single arbitrarily chosen measurement time. This is a general rationale for a maximum-response endpoint; the registry does not report a separate justification for selecting it.

The endpoint is a continuous measure, and the posted arm summaries are least-squares means with standard errors. Its purpose in this study was to quantify gallbladder emptying in response to cholecystokinin after albiglutide or placebo. It is a physiological outcome, not a direct measure of symptoms or a clinical event. The endpoint therefore addresses the study’s stated physiological question and does not by itself establish broader clinical consequences.

7. Results Posted in the Registry

• Albiglutide 50 mg: mean 38.43 versus 47.44 on control; difference -9.00 (95% CI -20.12 to 2.11), standardized effect d = -0.54; z-test p = 0.112.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is continuous: it records a percentage rather than assigning each participant to a category. The registry reports least-squares means and standard errors for the arms. A least-squares mean is a model-based estimate of an arm’s mean outcome, potentially adjusted for terms included in the statistical model. The registry does not list all model terms, so the precise adjustments behind those means are not reported in the registry record.

What does the crossover design contribute?

In a crossover design, participants receive each study treatment in different periods. This allows a comparison to use information from the same participant under both conditions. General statistical reasoning is that within-participant comparisons can account for stable differences between participants, because each participant serves as their own comparison. The value of that approach depends on the design and analysis being suitable, including how period and treatment order are handled. The registry identifies a two-way crossover and reports a mixed-model analysis, but does not report the model specification or the handling of treatment sequence and period.

The planned washout between treatment periods was a minimum of 42 days. A washout is generally intended to reduce the influence of an earlier treatment period on a later one. That is a general description of the purpose of washout; the registry does not report the evidence or assumptions used to select this duration, nor does it report an assessment of carryover effects.

How should the standard error be understood?

A standard error describes the statistical uncertainty in an estimated mean. It is not the same as the spread of individual participant measurements. The posted least-squares means each have a standard error of 4.01. The standard error helps convey how precisely the corresponding mean is estimated, while individual participant outcomes may vary more or less widely. The registry does not post the individual values needed to describe that distribution directly.

The sponsor reported a mixed-model analysis of the mean difference in final values. A mixed model is a general approach that can represent different sources of variation and, when specified appropriately, the dependence between repeated observations from the same participant. The registry’s short analysis description does not identify the model’s covariance structure, fixed or random effects, missing-data assumptions, or other implementation details. Those elements should not be inferred from the method label alone.

What do the confidence interval and p-value express?

A confidence interval gives a range of values compatible with the observed data under the assumptions of the analysis. It conveys uncertainty about the effect estimate, rather than giving the probability that the true effect lies in the particular interval. The sponsor’s reported 95% confidence interval extends from -20.578 to 2.574. Because this range includes zero, the reported data do not rule out no difference at the corresponding two-sided level under the model’s assumptions.

A p-value measures how unusual the observed data, or more extreme data, would be under a specified null hypothesis and analysis model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The sponsor reported p = 0.123. The registry does not report whether the primary test was adjusted for any multiplicity arising from the endpoint or other analyses.

The recalculated estimate and interval are simple, unadjusted re-calculations from the posted summary statistics. They are useful as a check on the scale and direction of the arm contrast, but they do not reproduce the sponsor’s mixed model. The pooled standard deviation used in that calculation is approximate, and summary statistics cannot recover participant-level pairing or model adjustments. Consequently, the recalculated inferential quantities need not exactly match the sponsor’s own result.

What is not specified?

The registry reports 15 outcome measures, 7 analyses, and 6 posted secondary outcomes, but those counts do not establish a prespecified testing hierarchy. The original statistical analysis plan, detailed handling of missing observations, interim analyses, and trial-specific sample-size justification are not reported in the registry record. These omissions limit assessment of whether the reported analysis followed a prespecified plan and how results across outcomes were interpreted.

In general, over-dispersion means that variability is greater than a model assumes. When present, ignoring it can make standard errors too small and confidence intervals too narrow. The posted endpoint is continuous, and no over-dispersion assessment is reported; this general concept should not be taken to mean that over-dispersion was present in this study.

10. Confidence Intervals and Precision

The sponsor’s mean-difference estimate was -9.002, with a reported 95% confidence interval from -20.578 to 2.574. This interval is wide enough to include zero and values on either side of it. The result therefore leaves uncertainty about whether the treatment contrast is negative, near zero, or positive within the interval. The interval describes statistical precision under the sponsor’s model; it is not a guarantee that the true effect lies within those bounds.

The standard errors for the posted least-squares means were 4.01 in both arms. Those values describe uncertainty in the arm means, not the variability among individual participants. Because this was a crossover study, precision of the treatment contrast also depends on the within-participant relationship between measurements. That relationship is not available in the posted arm summaries.

The unadjusted recalculation gives an estimate of -9.002 and an interval from -20.12 to 2.113. Its interval also includes zero. The small differences between the sponsor-reported and recalculated interval limits reflect differences in analysis approach and available information, rather than a contradiction about the overall direction of the evidence. The sponsor’s model-based analysis remains the reported analysis.

The endpoint analysis included 17 participants in each treatment arm. A relatively limited evaluable population can leave estimates imprecise, although the registry record does not report a formal assessment of the contribution of sample size to uncertainty. The confidence interval is the most direct posted summary of uncertainty in the treatment contrast. It cautions against treating the point estimate as an exact or definitive measure of the effect.

11. 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 event counts. They use a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the planned number of participants per arm that would, under those assumptions, yield an 80% chance of detecting the selected effect if that effect were the true one. The 90% power column gives the corresponding planning number for a 90% chance. For the full observed effect, the illustration lists 53 participants per arm for 80% power and 71 per arm for 90% power. For planning at 75% of the observed effect, it lists 95 and 127 participants per arm, respectively. At 50% of the observed effect, it lists 212 and 284 participants per arm.

The pattern reflects a general planning principle: smaller effects are harder to distinguish from random variation, so detecting them with the same power generally requires more participants. The curve should be read as a visual representation of the planning relationship, not as a description of the trial’s actual enrollment or observed result. These figures are planning illustrations based on the observed effect, not observed power.

The assumptions are simplified and are not a reconstruction of the trial’s crossover design or sponsor model. In particular, the calculation describes equal allocation between two arms and a normal approximation. Its figures should therefore be treated as illustrations, not as recommendations or as evidence that the original trial was designed using those numbers.

12. 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 = -9.002; pooled standard deviation σ = 16.534.
2. n = 2 × (1.96 + 0.842)² × 16.534² / -9.002² = 53.0, rounded up to 53 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 <- -9.0021                      # observed difference in means
sigma <- 16.5335298655581                     # 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(2496221)
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 = 53 (80% power), 71 (90% power)
 75% of observed effect: n per arm = 95 (80% power), 127 (90% power)
 50% of observed effect: n per arm = 212 (80% power), 284 (90% power)
power.t.test: 53.9 
Simulated power at n = 53 per arm: 0.791 (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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor’s mixed-model analysis as a mean difference in final values of -9.002, with a 95% confidence interval from -20.578 to 2.574 and p = 0.123. This is the trial’s reported inferential result. The sponsor analysis is based on a mixed model, although the registry does not provide enough detail to reproduce the complete model.

The recalculated estimate is -9.002, with an interval from -20.12 to 2.113 and an unadjusted p-value of 0.112. It is a simple, unadjusted calculation from posted summary statistics and uses an approximate pooled standard deviation. It does not reproduce the sponsor’s model or recover participant-level crossover information.

Both analyses indicate a negative estimated contrast for albiglutide relative to placebo, and both intervals include zero. Their differences in interval limits and p-values are expected when a summary-statistic re-calculation is compared with a model-based analysis. The sponsor’s analysis should be used as the primary reported result; the recalculation is a supplementary check, not a competing analysis.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 15 outcome measures, 7 analyses, and 6 posted secondary outcomes. The secondary outcomes include changes from baseline in systolic and diastolic blood pressure, heart rate, and electrocardiogram parameters; clinical chemistry and hematology abnormalities; adverse events; and urinalysis findings. The presence of multiple outcomes raises a general statistical issue: when many hypotheses are tested, the chance of at least one apparently unusual result can increase.

The registry record does not report a multiplicity adjustment, a hierarchy of hypotheses, or a strategy identifying which secondary results were confirmatory and which were exploratory. It also does not report the full statistical analysis plan. The primary p-value should therefore be read as the reported result for that endpoint, without assuming how testing across the full set of outcomes was controlled.

The primary endpoint population is defined as evaluable subjects who had gallbladder ultrasound assessment before treatment and post-baseline during infusion in both periods. The all-subjects population comprised participants who received at least one dose of investigational product. This makes clear that the endpoint analysis required measurements from both periods, but the registry does not give a detailed accounting of missing measurements, exclusions, or the number of participants omitted for each reason.

The registry does not report the missing-data method, assumptions about why data were missing, or sensitivity analyses for missing observations. It also does not report whether all randomized participants were included in any analysis according to their assigned treatment order. Without those details, it is not possible to assess how robust the endpoint result is to alternative missing-data assumptions or to verify the full analysis population.

Interim analyses and their role in the statistical plan are not reported in the registry record. These gaps do not establish that such procedures were absent; they mean that they cannot be characterized from the registry entry.

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 5% reporting threshold.

Most frequent non-serious adverse event terms

The safety summary reports zero serious adverse events among 18 participants at risk in the placebo group and zero among 20 participants at risk in the albiglutide 50 mg group. These are counts within the safety groups reported for each treatment. The registry lists no serious adverse-event terms. The absence of a listed serious event in these groups is a description of the posted data, not proof that serious risks are impossible or that the treatments have equivalent safety.

The summary also lists 8 participants with other adverse events among 18 at risk in the placebo group and 9 among 20 at risk in the albiglutide group. These figures are separate from the serious-event counts. The registry does not provide event-level descriptions in the serious-event terms list, and the death counts are not reported in the registry record.

Safety counts should be interpreted in the context of the study’s size, single-dose exposure, and follow-up as described in the registry. The record does not provide enough detail in this summary to assess the nature, timing, severity, or relatedness of each other adverse event. These descriptive counts do not support a broad comparison of safety between treatments.

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

The study addresses a focused physiological question: whether a single dose of albiglutide changes maximum gallbladder ejection fraction during cholecystokinin infusion in fasting healthy subjects. The sponsor-reported estimate was -9.002 for albiglutide relative to placebo. The 95% confidence interval, from -20.578 to 2.574, includes zero, and the reported p-value was 0.123. These results leave uncertainty about the magnitude and direction of the treatment contrast.

The result does not establish that the treatments have identical effects. A confidence interval that includes zero indicates that the analysis does not distinguish the observed data from no difference at the stated confidence level; it does not prove absence of an effect. Nor does the result establish that albiglutide caused a particular clinical outcome. The endpoint measures gallbladder ejection fraction, a physiological response, rather than symptoms, disease progression, or another clinical outcome.

The study was a small, single-dose crossover investigation with 20 planned subjects and 17 participants per arm in the posted primary endpoint summaries. The reported result pertains to the defined evaluable population and the measurement conditions in the registry. It should not automatically be extended to other populations, dosing schedules, or clinical settings.

The registry also includes safety outcomes, but the serious-event summary is limited to the reported counts, with no serious events listed in either treatment group. That does not establish that rare or longer-term adverse outcomes are absent. The study’s duration and scope do not answer every question that might arise about treatment effects.

In short, the posted evidence estimates a lower maximum ejection fraction under albiglutide than placebo, but the confidence interval includes no difference and spans effects in both directions. The data are not sufficiently precise to make a definitive claim about the size or direction of the physiological contrast, and they do not answer broader questions about clinical benefit or harm.

17. Why This Trial Matters Statistically

This trial illustrates how a crossover design can be used to study a short-term physiological response. Each participant was intended to receive both albiglutide and placebo in separate periods, allowing the treatment comparison to draw on measurements under both conditions. The design makes the handling of period effects, treatment sequence, and within-participant correlation important parts of the analysis.

The registry identifies a mixed-model analysis, while also reporting arm-level least-squares means and standard errors. Comparing those summaries with a simple recalculation demonstrates why posted summary statistics may not reproduce a model-based result exactly. The sponsor’s estimate and the unadjusted calculation are close in direction and magnitude, but their interval limits and p-values differ. This distinction is useful when readers encounter independently recalculated statistics alongside a trial’s reported analysis.

The result also shows why an estimate should be read together with its confidence interval. A point estimate alone can appear more definite than the data justify. Here, the sponsor-reported interval includes zero and extends to values on either side, so the uncertainty matters as much as the estimated contrast.

Finally, the record distinguishes the planned enrollment from the evaluable endpoint population and documents a range of secondary outcomes. Those details make the analysis population, missing-data handling, and multiplicity strategy relevant to interpretation. Several of these methods are not reported in the registry record, so the result is best understood as a focused estimate with limitations in both precision and available methodological detail.

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 primary endpoint was assessed in an evaluable population requiring gallbladder ultrasound measurements in both treatment periods. The registry reports 17 participants per arm for the endpoint summaries, compared with planned enrollment of 20 subjects, but does not explain the difference or describe participant-level exclusions. This limits assessment of whether missing measurements or exclusions could have influenced the result.

The registry identifies a mixed-model analysis but does not report the model specification, covariance structure, treatment-sequence or period terms, or missing-data assumptions. Those details are especially relevant for a crossover design, where observations from the same participant are related. The simple recalculation is not a substitute for those details and should not be treated as a full reproduction of the sponsor’s analysis.

The original sample-size justification is not reported in the registry record. The planning illustration uses a normal approximation and assumptions based on the observed effect, rather than the trial’s documented original design. It cannot establish whether the study was intended to detect a particular clinically meaningful difference.

The primary endpoint is a physiological measurement collected during a specified infusion and study schedule. The registry does not report a clinical threshold for interpreting the size of the difference, and the endpoint does not directly measure patient symptoms or clinical events. Generalization beyond the study’s healthy-subject description, single-dose exposure, and measurement conditions is not established by this result.

The registry lists multiple outcomes but does not report a multiplicity strategy or a full analysis plan. The safety summary provides counts but limited event-level detail, and death counts are not reported in the registry record. These limitations constrain what can be concluded about the robustness, broader clinical meaning, and safety context of the posted findings.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study Evaluating the Effect of Albiglutide on Gallbladder Emptying in Healthy Subjects — NCT02496221.
• Registry record last updated 2018-04-05; status last verified 2018-03.
• Results first posted 2016-07-21.
• 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.

22. Record Summary

The registry record describes a completed Phase 4 randomized, double-blind, placebo-controlled, two-way crossover study of a single dose of albiglutide 50 mg and placebo in fasting healthy subjects. Cholecystokinin infusion and gallbladder ultrasound were used to assess gallbladder emptying. Planned enrollment was 20 subjects, and the primary endpoint was maximum gallbladder ejection fraction during infusion on Day 4 of each period.

The sponsor’s mixed-model analysis reported a mean difference of -9.002, with a 95% confidence interval from -20.578 to 2.574 and p = 0.123. The interval includes zero, leaving uncertainty about the treatment contrast. The registry also reports no serious adverse events in either treatment group’s safety summary. Detailed model specifications, missing-data procedures, multiplicity handling, and the original sample-size rationale are not reported in the registry record.