Thiamine as a Renal Protective Agent in Septic Shock: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Thiamine as a Renal Protective Agent in Septic Shock (NCT03550794), a randomized phase 2 study. The primary outcome is Kidney Injury Biomarker.

2. Clinical Question

The clinical question is about a difference between randomized groups, not simply whether creatinine changed within either group. Randomization makes the between-group contrast central to interpreting the trial, although the registry record does not describe the allocation procedure in detail. The primary outcome is described as change in creatinine over time, while the posted arm summaries give means and standard deviations and the sponsor's analysis is labeled a mixed models analysis of final values. The record does not explain the precise relationship between those descriptions.

The registry lists secondary outcomes that include renal replacement therapy, ICU free days, in-hospital mortality, acute renal failure, lactate, delirium, SOFA score and novel renal injury biomarkers. These outcomes address distinct aspects of illness and kidney function. They should not be treated as interchangeable measures of the primary question.

3. Trial Design

This is a randomized, controlled, parallel-group phase 2 study sponsored by Beth Israel Deaconess Medical Center. The registry record lists an enrollment of 95 and a completed status. The study began on 2018-09-04, and primary completion is recorded as 2022-04-05. These dates describe the registry record and do not establish when results were analyzed or published.

Participants were assigned to thiamine or matching placebo. The thiamine regimen was 200mg parenterally twice daily for 3 days, for 6 doses. The placebo was described as 50ml 0.9%NACL, administered on the same twice-daily schedule for 3 days, for 6 administrations. The registry lists masking as quadruple; its brief summary describes the study as double-blind. The record does not explain the difference between those descriptions or identify the masked roles.

The primary outcome is titled “Kidney Injury Biomarker” and specifies creatinine change over time from enrollment to 72-hours, with the unit mg/dL. The listed summary parameter is a mean with standard deviation. The registry also lists 8 secondary outcomes, but detailed arm results are posted for only some of them. The protocol, detailed eligibility criteria, original sample-size justification, allocation-concealment procedure and full statistical analysis plan are not reported in the registry record.

These design details establish the broad comparison, but do not answer how missing observations were handled, which participants entered each analysis, or whether any outcome definitions changed during the study. Those details are important for assessing how closely the posted results reflect the originally planned analysis.

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

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 Kidney Injury Biomarker, measured over Enrollment to 72-hours.

The primary endpoint is described as change in creatinine over time, measured from enrollment to 72-hours. Creatinine is used as a biomarker of kidney injury, so comparing its course between randomized groups addresses the study's central renal-function question. The registry identifies mg/dL as the unit and reports the arm summaries as means with standard deviations.

The primary outcome's title, “Kidney Injury Biomarker,” is broader than its description, which specifically names creatinine. The record does not describe a composite endpoint or another biomarker as part of this primary outcome. It also does not explain the exact calculation used to summarize change over time, or whether all measurements in the stated timeframe contributed to the reported arm means.

General statistical context: a continuous measure such as creatinine can be summarized by a mean, while its standard deviation describes how dispersed individual measurements are around that mean. A mean difference compares the group averages, but its interpretation depends on which measurements and analysis population contributed to each average. For repeated measurements, a model may account for how observations from the same participant relate to one another; the registry's brief method label does not give enough detail to establish precisely how that was done.

The posted primary results therefore offer a group comparison, but the endpoint definition and analysis details remain incomplete in the registry record. The reported numbers should be read with that limitation in mind.

7. Results Posted in the Registry

• Thiamine: mean 2.24 versus 2.79 on control; difference -0.55 (95% CI -1.26 to 0.16), standardized effect d = -0.32; z-test p = 0.131.

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.

Number of Participants Receiving Renal Replacement Therapy

• 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:
  • Thiamine: 14.3% (6/42) versus 21.7% (10/46) on control; risk difference -7.5 percentage points (95% CI -23.4 to 8.5), risk ratio 0.66 (0.26–1.65); unadjusted two-proportion z-test p = 0.365.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Thiamine and the control. The interval gives the range of effects that remain compatible with the data.

In-hospital Mortality

• 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:
  • Thiamine: 35.7% (15/42) versus 54.3% (25/46) on control; risk difference -18.6 percentage points (95% CI -39.1 to 1.8), risk ratio 0.66 (0.40–1.07); unadjusted two-proportion z-test p = 0.080.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Thiamine and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Experiences Acute Renal Failure

• 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:
  • Thiamine: 54.8% (23/42) versus 73.9% (34/46) on control; risk difference -19.2 percentage points (95% CI -38.8 to 0.5), risk ratio 0.74 (0.54–1.02); unadjusted two-proportion z-test p = 0.060.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Thiamine and the control. The interval gives the range of effects that remain compatible with the data.

The listed secondary outcomes span kidney support, organ function, illness course, mortality, delirium and biomarker measures. Detailed results are posted for renal replacement therapy, in-hospital mortality and acute renal failure. Detailed results for the other listed secondary outcomes are not reported in the registry record. This incomplete reporting means the overall secondary-outcome pattern cannot be assessed across every outcome named in the record.

For the three outcomes with detailed results, the posted estimates do not provide a uniformly precise picture. Their reported intervals include values compatible with no group difference, and the unadjusted comparisons do not establish clear separation between groups. The sponsor's reported regression analyses likewise have intervals that include the null value for their respective measures. The particular outcome estimates and their meanings are reported with the corresponding results; they should be interpreted within their stated timeframes and analysis types.

These secondary findings do not settle the primary creatinine comparison. They address different clinical events and are not substitutes for the primary endpoint. Nor does a broadly similar direction across some estimates establish a consistent effect, particularly when the estimates are uncertain and outcomes differ in definition and timing.

General statistical context: when many outcomes are examined, the chance of at least one apparently notable result can increase unless the analysis accounts for multiplicity. The record does not report a multiplicity procedure or a hierarchy for secondary outcomes. Accordingly, unadjusted secondary results carry less weight as confirmatory evidence than a clearly prespecified primary analysis. They may generate questions for further study, but the registry record does not establish that they were designed or analyzed as confirmatory tests.

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

• Number of Participants Receiving Renal Replacement Therapy
• ICU Free Days
• In-hospital Mortality
• Number of Participants Experiences Acute Renal Failure
• Change in Lactate Level
• Number of Participants With Delirium on Day 3
• Change in the Sequential Organ Failure Assessment Score
• Novel Biomarkers of Renal Injury

9. Statistical Interpretation

10. Statistical Methods Explained

General statistical context: a randomized comparison estimates how outcomes differ between groups assigned to different study conditions. The estimate is the size and direction of the observed contrast; a confidence interval describes a range of values compatible with the data and the analysis assumptions. A p-value measures how unusual the data would be under a specified null model. It does not measure the probability that the null hypothesis is true, nor does it quantify the clinical importance of an estimate.

What do the mean and standard deviation describe?

The registry reports a mean and standard deviation for each arm. A mean is the arithmetic average of the recorded values included in that summary. The standard deviation describes how spread out those values are around the mean. It is not a confidence interval for the mean and does not show the uncertainty in the between-group contrast by itself. The posted primary summaries are 2.24 mg/dL with a standard deviation of 1.5 in the thiamine group, and 2.79 mg/dL with a standard deviation of 1.91 in the placebo group.

What is the unadjusted recalculation?

The unadjusted recalculation compares the posted group means directly. Its mean difference is -0.55 mg/dL, and its p-value is 0.131. This is a summary-statistic calculation, not a reconstruction of the trial's full statistical model. It does not account for information that may have been used in the sponsor's analysis beyond the posted arm-level means and standard deviations. General statistical context: unadjusted comparisons are useful for a transparent check of group summaries, while a prespecified model can incorporate the trial's design and repeated measurements when those methods are specified.

What does the mixed models label tell us?

The sponsor reports a “Mixed Models Analysis” of “Mean Difference (Final Values).” General statistical context: mixed models can represent relationships among repeated observations from the same participant, and can estimate group contrasts while accounting for that dependence. The exact model matters: choices about time, treatment, their interaction, covariance structure, baseline values and missing observations can affect the estimate and its uncertainty. The registry record does not report those specifications, so it is not possible to say which were used in this analysis. The label alone should not be taken as evidence that any particular adjustment was made.

How should the interval and p-value be read?

The sponsor reports a 95% confidence interval from -1.18 to 0.04 and p = 0.07. General statistical context: a confidence interval expresses uncertainty around an estimate under the model used; wider intervals indicate less precision. Because this interval includes zero, the data and model are compatible with no difference as well as with effects on either side of zero. The p-value is not a measure of the probability that the observed result occurred by chance, and it does not determine whether a difference matters clinically. The registry record does not report a threshold for interpreting this endpoint or a detailed plan for interpreting the interval and p-value.

Why can the sponsor and unadjusted results differ?

The unadjusted comparison uses arm-level summary statistics; the sponsor's result comes from a mixed models analysis of final values. Differences in model structure, included observations or handling of repeated measurements can produce different standard errors, intervals and p-values. The record does not report enough detail to attribute the difference to any particular modeling choice. There is no reported discrepancy flag, and the sponsor's own analysis should be distinguished from rather than replaced by the recalculation.

What can be inferred from the other analyses?

The posted secondary outcomes include participant counts and sponsor analyses using logistic or Cox regression. General statistical context: logistic regression commonly estimates associations for binary outcomes, while Cox regression analyzes time-to-event data. The registry gives odds ratios for some outcomes and a hazard ratio for in-hospital mortality. These measures have different interpretations and should not be read as if they were mean differences or risk differences. For each analysis, the reported estimate and interval should be interpreted according to its stated measure and timeframe. Detailed model specifications are not reported in the registry record.

11. Confidence Intervals and Precision

The sponsor's primary estimate is a mean difference of -0.57 mg/dL, with a 95% confidence interval from -1.18 to 0.04. The interval includes zero, indicating that the reported data do not pin down the direction of the group difference with high precision. Its endpoints span values on both sides of zero, and the point estimate should not be read as if it were known exactly. The sponsor reports p = 0.07; that value is consistent with an estimate whose interval includes zero under the reported analysis.

General statistical context: a confidence interval is a measure of uncertainty around an estimate, conditional on the study data and the analysis assumptions. A narrower interval indicates greater precision than a wider interval, but precision is not the same as freedom from bias. An interval also does not assign a probability to the true effect lying within its realized endpoints.

The unadjusted recalculation gives a mean difference of -0.55 mg/dL and p = 0.131. It is based on the posted means and standard deviations, rather than the sponsor's mixed models analysis. The two estimates are close, but the reported sponsor interval is tied to its own analysis and should not be substituted with an interval from a different calculation.

The reported group sizes for the primary summary are 42 and 46. The registry does not explain how those analysis denominators relate to the enrollment count of 95, whether every randomized participant had an observed endpoint, or whether missing observations were modeled or excluded. Without those details, the precision reflected by the interval cannot be assessed for sensitivity to missing data. The interval quantifies sampling uncertainty under the reported model; it does not address every uncertainty in study design or reporting.

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 illustrations based on the observed effect, not the trial's original design calculation. For the continuous primary outcome and the posted binary secondary outcomes, they are participants per arm. No time-to-event sample-size figures are included among these calculations; in general, a time-to-event planning display expressed in events would refer to total events rather than participants per arm.

The assumptions for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect: 100%, 75% or 50%. The trial's own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the planned participants per arm for an illustration designed to have an 80% chance of detecting the specified effect under those assumptions, if the assumed effect and analysis conditions hold. The 90% power column uses the corresponding 90% target. These are planning targets, not guarantees about what a completed study will detect. The displayed counts are per arm, not totals across both arms.

Smaller effects require more participants for the same stated power target: planning for 75% or 50% of the observed effect therefore leads to larger per-arm figures than planning for 100% of it. This is not observed power. The illustrations do not establish what power the completed trial had, and they do not show that the observed effect is the effect a future study should expect. Their purpose is to show how sample-size requirements change when the target effect is varied under the stated assumptions.

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 = -0.550; pooled standard deviation σ = 1.727.
2. n = 2 × (1.96 + 0.842)² × 1.727² / -0.550² = 154.7, rounded up to 155 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 <- -0.55                      # observed difference in means
sigma <- 1.72672057771035                     # 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(3550794)
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 = 155 (80% power), 208 (90% power)
 75% of observed effect: n per arm = 276 (80% power), 369 (90% power)
 50% of observed effect: n per arm = 619 (80% power), 829 (90% power)
power.t.test: 155.7 
Simulated power at n = 155 per arm: 0.800 (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 sponsor's primary analysis is reported as a mixed models analysis of mean difference in final values. It gives an estimate of -0.57 mg/dL, a 95% confidence interval from -1.18 to 0.04, and p = 0.07. This is the trial's own reported analysis and should be the main reference for the primary result.

The unadjusted recalculation from the posted arm summaries gives a mean difference of -0.55 mg/dL and p = 0.131. It is a simple comparison of the posted means and standard deviations, not a reproduction of the sponsor's model. Similar point estimates do not imply that the methods are identical; the interval and p-value depend on the analysis and its uncertainty calculation.

The registry record does not report the mixed model's full specification, including its treatment of repeated measurements or missing observations. It is therefore not possible to identify why the p-values differ or to independently reproduce the sponsor's estimate from the posted summary statistics. The record does not flag a discrepancy between the analyses. The sponsor's own result remains distinct from the recalculation, and neither should be presented as a substitute for the other.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 secondary outcomes in addition to the primary outcome. It does not report whether the trial used a formal multiplicity adjustment, a prespecified order for testing outcomes, or a rule for deciding which secondary analyses were confirmatory. Without that information, the secondary p-values should be read as unadjusted results rather than as a coordinated family of tests with a known overall error rate.

General statistical context: testing several outcomes can increase the chance of observing at least one result that appears unusual even when there is no corresponding effect. Multiplicity procedures can control that risk, but they may also affect the interpretation and precision of individual tests. The registry record does not state whether such a procedure was planned or applied.

Missing-data procedures are also not reported in the registry record. The posted primary arm summaries include 42 participants for thiamine and 46 for placebo, while the total enrollment is listed as 95. The record does not explain the relation between those figures, which participants were included in each analysis, whether measurements were missing, or how any missing values were handled. It also does not report whether the analysis followed intention-to-treat, per-protocol or another population definition.

These omissions limit assessment of the analysis population and the possibility that incomplete observations affected the estimate. The mixed models label does not by itself establish how missing measurements were handled. Details about allocation concealment, protocol deviations, the original analysis plan and any interim analyses are also not reported in the registry record. Their absence from the record does not show that a procedure was or was not used; it means its role cannot be evaluated from the registry information.

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.

The adverse-event summary reports 0 serious adverse events among 42 participants at risk in the thiamine group and 0 among 46 participants at risk in the placebo group. It also reports 0 other adverse events in each group, using the same at-risk counts. No serious adverse-event terms are listed.

The registry's adverse-event figures are counts, not estimates that establish equal risk between groups. Zero recorded events in these groups does not establish that the interventions have no risks, and the record does not provide detail on event definitions, surveillance procedures or the duration of safety follow-up beyond the relevant outcome timeframes. The registry also reports 15 and 25 deaths in the respective adverse-event group summaries. Those counts are not, by themselves, evidence that deaths were caused by either study assignment; the record does not attribute them to treatment.

Safety information should therefore be read as a description of what the registry records, not as a complete safety profile. The listed serious-event counts do not resolve uncertainty about uncommon events or events not captured in the registry summary.

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

The primary result is a comparison of creatinine values over time in a randomized thiamine-versus-placebo study. The sponsor reports a mean difference in final values of -0.57 mg/dL, with a 95% confidence interval from -1.18 to 0.04 and p = 0.07. The interval includes zero, so the posted estimate is uncertain about the direction and size of the group difference. The unadjusted recalculation is similar in point estimate but uses a different method and should not replace the sponsor's analysis.

This result does not establish that thiamine caused a change in kidney function, nor does it establish that there was no difference. An uncertain estimate can be compatible with more than one underlying effect. The primary finding should be described with its interval and analysis method rather than reduced to a binary label based only on a p-value.

The trial also reports selected secondary outcomes, but the registry record does not provide detailed results for every listed secondary measure. The available secondary results are uncertain, and the record does not describe a multiplicity adjustment. They therefore do not provide a firm basis for overriding or confirming the primary result.

The findings apply to the trial comparison and the recorded outcomes. The registry record does not provide enough detail about eligibility, analysis populations, missing observations, or model specifications to assess how those features may affect interpretation. The posted results are not a substitute for a complete protocol and analysis report, and they do not establish what would occur in other populations or settings.

18. Why This Trial Matters Statistically

This trial illustrates why an estimate, its uncertainty and the analysis method need to be read together. The primary point estimate is close between the sponsor analysis and the unadjusted recalculation, but the analyses are not identical. The sponsor reports a mixed models analysis, while the recalculation uses only arm-level means and standard deviations. Their different p-values underscore that statistical conclusions depend on the analysis, not only on the observed group averages.

The confidence interval is also central to interpretation. It includes zero, so the primary result does not precisely resolve whether the underlying group difference is in one direction or the other. Reporting the estimate without its interval would make the result appear more definite than the posted analysis supports. Conversely, treating a p-value as a complete measure of evidence would omit both effect size and uncertainty.

The record also highlights the importance of reporting analytic details. A mixed models label does not identify the model specification, and the registry does not state how missing data, analysis populations or multiple outcomes were handled. These details are necessary for independent appraisal and for understanding how results were produced.

Finally, the planning illustrations show how required sample sizes depend on the target effect and desired power under stated assumptions. They are not evidence of the completed study's power or proof that the observed effect is a reliable planning value. Clear separation of trial findings, recalculations and planning illustrations helps prevent these different statistical quantities from being conflated.

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.

Several trial-specific details needed to evaluate the primary result are not reported in the registry record. The primary outcome description refers to change in creatinine over time, while the sponsor's reported analysis is labeled as a mean difference in final values. The record does not explain how those descriptions correspond or specify the timing and modeling of repeated measurements.

The reported mixed models analysis lacks a full model specification. The registry does not state which covariates, time terms, covariance structure or missing-data procedures were used. It also does not define the analysis population or explain how the primary outcome denominators relate to total enrollment. These gaps limit reproducibility and make it difficult to assess sensitivity to analytic choices.

The registry lists 8 secondary outcomes, but detailed results are not reported for all of them. It does not state whether secondary tests were adjusted for multiplicity or organized in a prespecified hierarchy. This limits the confirmatory interpretation of the available secondary findings.

The registry lists quadruple masking, whereas the brief summary calls the study double-blind. The record does not reconcile those descriptions or identify the roles masked. It does not provide the original sample-size justification, detailed eligibility criteria, allocation-concealment procedure, interim-analysis plan or full protocol. These items are not reported in the registry record; no conclusion about whether they were used can be drawn from that absence.

These limitations concern reporting available in the registry record. They do not, by themselves, establish that the trial procedures were inadequate. They do constrain the extent to which the posted analyses can be independently reproduced and interpreted.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Thiamine as a Renal Protective Agent in Septic Shock — NCT03550794.
• Registry record last updated 2023-06-27; status last verified 2023-06.
• Results first posted 2023-06-27.
• 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

NCT03550794 is a completed randomized, parallel-group phase 2 study sponsored by Beth Israel Deaconess Medical Center. The registry lists enrollment of 95, quadruple masking and a comparison of parenteral thiamine with matching placebo in septic shock. The primary outcome is described as change in creatinine from enrollment to 72-hours.

The sponsor's primary analysis reports a mean difference in final values of -0.57 mg/dL, with a 95% confidence interval from -1.18 to 0.04 and p = 0.07. The interval includes zero and reflects uncertainty in the estimated group difference. An unadjusted recalculation from posted arm summaries gives a mean difference of -0.55 mg/dL and p = 0.131; it is methodologically distinct from the sponsor's mixed models analysis.

Detailed secondary results are available for some, but not all, listed outcomes. Analysis-population definitions, missing-data methods, multiplicity procedures and full model specifications are not reported in the registry record. Interpretation should retain those limits and distinguish the sponsor's analysis from the summary-statistic recalculation.