Diabetes Remote Care Management System: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Diabetes Remote Care Management System (NCT01354015), a randomized na study. The primary outcome is Change in HbA1c.

2. Clinical Question

The question is about the observed change in a laboratory measure during this pilot, not a broader judgment about diabetes care. The registry record does not report enough detail to establish why the sponsor’s analysis and the summary-statistic recalculation yield different inferential results. It also does not report the original sample-size justification or a more detailed statistical analysis plan. The results should therefore be read in light of the available outcome summaries, the sponsor’s stated analysis, and the limits of the registry record.

3. Trial Design

This was a randomized, parallel-group trial with no masking. The registry lists Tulane University School of Medicine as the sponsor, enrollment of 98, and a completed status. The recorded study start is 2011-05, and primary completion is recorded as 2014-09. The phase is listed as NA. The brief description calls the study a pilot and describes a plan to enroll 100 injectable-insulin-dependent participants with uncomplicated Type 1 or Type 2 diabetes, with randomization 1:1 to DRMS or standard care. The record’s enrollment figure is 98; the record does not explain the difference between that figure and the planned enrollment in the brief description.

The intervention was called Use of Messaging System or Use of DRMS, and the comparison was Usual Care. DRMS was described as using multiple communication channels, including cell phones, IVR, email, web, and SMS messaging. The listed aims included reminders to fill prescriptions and take medication as prescribed, tracking insulin use and timing, collecting glucose levels, and sending customized information or automated dose-adjustment directions based on provider instructions. These are descriptions of intended system functions, not evidence that each function was used by every participant.

The primary outcome was change in HbA1c from baseline to six months. The record states that only participants who completed the six-month visit were included in that analysis: 44 in the messaging group and 43 in usual care. A secondary HbA1c change outcome was recorded for baseline to three months, with results for 46 and 45 participants. The registry record does not report the reasons for missed visits, the handling of missing outcome values beyond the stated six-month completer analysis, or whether any additional analysis populations were used.

Random assignment can support a group comparison by distributing participant characteristics across groups on average, but the posted six-month analysis is limited to visit completers. The record does not report whether the analysis preserved the original randomized groups for all assigned participants, nor does it describe masking of outcome assessment. The available results therefore describe the recorded group summaries and the reported sponsor analysis, while important details about conduct and analysis remain unreported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

Analysis population for the primary outcome, as defined in the registry: Of the 50 enrolled in the Use of DRMS group, only 44 completed the visit at 6 months, and out of 48 enrolled in the Usual Care group only 43 completed the visit at 6 months. Therefore only those who completed the visit at 6 months were included in this analysis.

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 Change in HbA1c, measured over baseline to 6 months.

The primary endpoint is change in HbA1c from baseline to six months, reported as a mean in percent for each group with a standard deviation. HbA1c is a measure used to summarize blood-glucose exposure over time, so a change from baseline provides a way to describe how the measure moved during follow-up. Reporting change, rather than only a final value, makes the baseline-to-follow-up comparison explicit in the endpoint definition.

In the messaging group, the posted mean change was -0.25, with a standard deviation of 0.2, among 44 six-month visit completers. In usual care, the mean change was -0.4, with a standard deviation of 0.3, among 43 completers. Negative changes indicate decreases from baseline in the direction represented by the reported measure. These are group averages; they do not describe the change for every participant or show how individual values were distributed.

The endpoint is clinically relevant as a measure of glycemic control, but the registry result is not, by itself, a complete assessment of the intervention or of diabetes care. It does not report all possible health outcomes, participant experience, or longer-term results. Nor does the endpoint summary establish why any change occurred. It quantifies the recorded HbA1c changes in the analyzed groups; interpretation of the group comparison also depends on the analysis method and the missing-data context.

7. Results Posted in the Registry

• Use of Messaging System: mean -0.25 versus -0.40 on control; difference 0.15 (95% CI 0.04 to 0.26), standardized effect d = 0.59; z-test p = 0.006.

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.

Change in HbA1c Over 3 Months

• 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:
  • Use of Messaging System: mean -0.50 versus -0.20 on control; difference -0.30 (95% CI -0.40 to -0.20), standardized effect d = -1.17; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Use of Messaging System than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry posts one secondary endpoint: change in HbA1c from baseline to three months. Its group summaries show a larger decrease in the messaging group than in Usual Care at that timepoint. The direction of this descriptive pattern differs from the six-month group means, where the mean decrease is greater in Usual Care. The posted results do not report a direct test comparing the time course between groups, so the two timepoint summaries cannot establish whether the intervention’s association changed over time.

The three-month result is a secondary outcome, and no sponsor-reported analysis is listed for it. Its recalculated comparison is therefore based on the posted summary statistics. It should be interpreted as exploratory evidence rather than as confirmation of the primary result. The registry record does not report a prespecified hierarchy of outcomes or a multiplicity adjustment for secondary analyses.

When multiple outcomes or timepoints are examined, the chance of finding an apparently unusual result can increase unless the analysis plan accounts for that testing. This is a general statistical concern. An unadjusted secondary p-value does not reflect such a correction, and a statistically striking secondary result should not automatically outweigh the primary endpoint or the sponsor’s own analysis. The secondary result can help describe the record’s pattern across follow-up, but it does not resolve the primary discrepancy or establish a durable difference between groups.

The numbers of participants represented at three months and six months differ. The record does not state whether the same participants contributed measurements at both visits, how missing values were treated for the secondary endpoint, or whether any repeated-measures model was used. Accordingly, cross-timepoint comparisons are descriptive and should not be interpreted as if they came from a reported longitudinal analysis.

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

• Change in HbA1c Over 3 Months

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean difference describe?

For a continuous outcome such as HbA1c change, a group mean describes the average recorded change among the participants included in that group’s analysis. A mean difference compares the two group averages using the stated order of the comparison. The sign depends on that order: a positive value does not automatically mean benefit, and a negative value does not automatically mean harm. It must be interpreted alongside the endpoint definition and the group means.

The primary summary reports the mean change and standard deviation for each group. A standard deviation describes how spread out individual measurements are around the group mean. It is not the same as a confidence interval or a standard error. The posted recalculation uses the group summaries to estimate an unadjusted difference and its uncertainty. It does not incorporate participant-level information that is not present in those summaries, nor does it adjust for other variables.

What does an unadjusted p-value tell the reader?

A p-value is a measure of how compatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. A smaller p-value indicates less compatibility with that null model under the stated assumptions; it does not, on its own, establish that the analysis is unbiased or that the finding will replicate.

The recalculated primary p-value is unadjusted. That label means the result is not reported as corrected for multiple outcome or analysis testing. The registry’s sponsor-reported ANOVA p-value is separate and differs from the recalculation. The record does not describe enough about the sponsor analysis to reconstruct it or determine the precise source of the disagreement. The discrepancy is not resolved by noting that one value is smaller: the sponsor’s own analysis takes precedence for the trial’s reported inference.

Why does a standard deviation matter?

The spread of observations affects how precisely group means and their contrast can be estimated. When individual observations vary more, the mean difference is generally less precise if other features of the data are similar. The posted summaries include standard deviations of 0.2 and 0.3 for the primary outcome, and 0.3 and 0.2 for the secondary outcome. The recalculation uses the reported summary information; it does not model individual trajectories or account for unreported features of the outcome distribution.

A standardized effect size expresses a difference relative to a measure of variability, allowing effects measured on the same scale to be compared more readily. The posted primary standardized effect is d = 0.590, and the secondary standardized effect is d = -1.174. Such standardized values depend on the outcome variability used in the calculation. They do not replace the original-unit mean differences and do not directly convey clinical importance.

What can randomization and a completer analysis establish?

Randomization is intended to create groups that are comparable at the time of assignment, on average. It supports a causal comparison when the analysis and follow-up preserve the randomized comparison and other design assumptions are reasonable. In this record, the primary analysis includes those who completed the six-month visit. The record does not report a full analysis of everyone randomized, the reasons participants did not complete visits, or how missing values were handled in any other analysis.

Analyzing only participants with observed follow-up measurements can change the balance created by randomization if visit completion is related to group assignment, participant characteristics, or outcomes. This is a general statistical concern, not evidence that such a bias occurred in this trial. The registry does not report enough information to evaluate it. The secondary result also has a different number of participants represented than the primary result, which means that the two timepoint summaries should not be assumed to represent exactly the same analysis population.

Why does the sponsor analysis take precedence?

The registry lists the sponsor’s ANOVA analysis and separately records the summary-statistic recalculation. They report the same stated mean difference of 0.15 for the primary outcome, but the sponsor’s p-value is 0.254, while the unadjusted recalculation gives p = 0.006. The registry marks the record as discrepant. Under that condition, the sponsor’s analysis is the trial’s own analysis and takes precedence; the recalculation is an additional, simple unadjusted calculation from the posted summaries. The record does not report the methods needed to reconcile their inferential difference.

ANOVA is a general method for comparing means through a model that partitions variation among sources. The precise model, assumptions, and contrast used in the sponsor’s analysis are not reported in the registry record. Therefore, it is not possible to determine whether differences in model specification, endpoint handling, or another analytic choice account for the p-value contrast. Naming a possible explanation would go beyond the recorded information.

11. Confidence Intervals and Precision

A confidence interval describes a range of values compatible with the data and the analysis assumptions at a stated confidence level. Wider intervals indicate less precision than narrower intervals, all else being similar. An interval is not a guarantee that the true effect lies within its bounds, and its interpretation depends on how the interval was constructed.

For the primary summary-statistic recalculation, the posted interval for the mean difference is 0.043 to 0.257. The record does not state the confidence level for that recalculated interval. It therefore should not be labeled with a confidence percentage that is not reported. The sponsor’s analysis is listed with a 95% confidence level, but its confidence-interval bounds are not reported in the registry record. The sponsor’s p-value and the recalculated interval should not be treated as if they necessarily came from the same analysis.

For the secondary three-month comparison, the posted interval is -0.405 to -0.196. As with the recalculated primary interval, the record does not state its confidence level. These intervals reflect uncertainty around the recalculated group contrast under its assumptions; they do not account for every potential source of uncertainty, such as selection into the completer analysis or multiple testing.

Precision is not the same as clinical importance. A narrow interval can still surround an effect of uncertain practical relevance, and statistical compatibility does not resolve the sponsor-versus-recalculation discrepancy. The trial’s own reported analysis remains the sponsor analysis, while the intervals from the recalculations help describe the uncertainty implied by the posted summary statistics.

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 and not total events. They are illustrations calculated from the observed effect, not the trial’s original design assumptions. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each row uses the stated fraction of the observed effect to detect. The original trial’s sample-size justification and design assumptions are not reported in the registry record.

For the primary endpoint, the row planning for 100% of the observed effect gives 46 participants per arm for 80% power and 61 participants per arm for 90% power. Planning for 75% of the observed effect gives 81 participants per arm for 80% power and 108 for 90% power. Planning for 50% of the observed effect gives 181 participants per arm for 80% power and 242 for 90% power. The three-month secondary outcome has its own illustrations: 12 and 16 participants per arm for 80% and 90% power when planning for 100% of its observed effect; 21 and 28 when planning for 75%; and 46 and 61 when planning for 50%.

The 80% power column describes a planning scenario in which a study of the indicated size would have an 80% chance of detecting the specified effect under the stated assumptions. The 90% column uses a higher target chance of detection. Smaller effects require more participants under these assumptions, which is why the planning counts increase as the effect fraction decreases. These figures are planning illustrations, not observed power. They should not be read as a retrospective measure of how likely the completed trial was to detect its result.

A curve associated with these planning calculations should be read as a visual representation of how the required participants per arm vary with the effect size and power target under the same assumptions. It does not add evidence about the trial outcome, and it does not establish that the observed effect is a reliable expectation for a future study. Planning around an observed effect can be sensitive to the uncertainty in that estimate, especially in a pilot.

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 = 0.150; pooled standard deviation σ = 0.254.
2. n = 2 × (1.96 + 0.842)² × 0.254² / 0.150² = 45.1, rounded up to 46 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 <- 0.15                      # observed difference in means
sigma <- 0.254373509534584                     # 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(1354015)
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 = 46 (80% power), 61 (90% power)
 75% of observed effect: n per arm = 81 (80% power), 108 (90% power)
 50% of observed effect: n per arm = 181 (80% power), 242 (90% power)
power.t.test: 46.1 
Simulated power at n = 46 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 registry’s own primary analysis is an ANOVA comparison of final values. It reports a mean difference of 0.15 and p = 0.254. A 95% confidence level is listed, but the interval bounds are not reported in the registry record. This is the sponsor-reported analysis and is the one that takes precedence when interpreting the trial’s reported inference.

The simple recalculation from the posted group summaries also gives a mean difference of 0.15, but its unadjusted p-value is 0.006 and its listed interval is 0.043 to 0.257. The registry flags a discrepancy. The matching stated point estimate does not make the inferential results interchangeable: the p-values differ, and the record does not explain why. The recalculation is not a replacement for the sponsor’s analysis, and its interval should not be attached to the sponsor’s ANOVA result.

The record does not report enough methodological detail to reproduce the ANOVA or identify the source of the difference. It would be speculative to attribute the disagreement to a particular model, data-handling choice, or endpoint definition. Readers should retain the sponsor result as the trial’s reported analysis and regard the recalculation as a separate, unadjusted analysis of posted summaries. The registry’s missing details limit further reconciliation.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary endpoint and one secondary endpoint, both based on HbA1c change at different follow-up times. It does not report a multiplicity adjustment, a formal testing hierarchy, or a plan specifying how conclusions should be drawn across these outcomes. The recalculated p-values are explicitly unadjusted. Without a reported approach to multiple testing, a secondary result should be treated cautiously and should not be used as though it had the same confirmatory role as a prespecified primary analysis.

The six-month primary analysis included participants who completed the visit: 44 in the messaging group and 43 in Usual Care. The record gives larger numbers at risk for adverse-event reporting, and its brief description states 50 enrolled in the messaging group and 48 in Usual Care. These different counts reflect different reported contexts; the adverse-event denominators should not be substituted for the primary endpoint analysis population. The registry does not report reasons for missing six-month visits or whether missingness was related to treatment assignment or outcome.

The record does not report an intention-to-treat analysis, imputation method, sensitivity analysis for missing outcomes, or a comparison of included and excluded participants. It states that six-month completers were included, but does not provide enough information to assess whether that restriction could have affected the group comparison. This is a limitation in the interpretability of the posted summary, not proof that missing data biased the result.

The original sample-size justification, interim analysis plan, detailed statistical analysis plan, and complete sponsor-model specification are not reported in the registry record. The record also does not explain the discrepancy between the sponsor’s primary p-value and the summary-statistic recalculation. These omissions limit the ability to evaluate prespecification, multiplicity control, analysis-population choices, and reproducibility. No conclusion about those unreported procedures should be inferred from the posted results alone.

16. 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.

The registry reports serious adverse events for 0 of 50 participants at risk in the Use of Messaging System group and 0 of 48 at risk in Usual Care. It also reports other adverse events for 0 of 50 and 0 of 48, respectively. These are the recorded counts for the registry’s stated at-risk groups.

The deaths field is not populated for either group, so deaths are not reported in the registry record; a blank or null entry should not be read as a recorded count of zero. The adverse-event figures also do not establish that no event occurred outside the reported categories or reporting period. The record does not provide event descriptions, severity detail beyond the serious-event category, or additional safety-analysis methods. These counts are a limited account of what was posted, not a complete assessment of safety.

The at-risk groups are not the same as the six-month primary analysis groups described in the outcome record. Safety denominators and outcome completers answer different reporting questions and should be kept distinct. Zero reported events in these summaries does not establish the absence of risk, especially when the record does not describe detailed ascertainment or follow-up procedures.

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

The registry records a randomized comparison of a messaging-based remote care system with Usual Care and reports HbA1c change at six months as the primary outcome. Its posted group means describe decreases from baseline in both groups among participants included in the analysis. The sponsor’s analysis reports p = 0.254, while a simple unadjusted recalculation from the summary statistics reports p = 0.006. Because the record flags this discrepancy, the sponsor’s analysis takes precedence as the trial’s own inference.

That precedence does not mean the analysis difference is explained. The record does not provide enough detail to reconcile the two results, and the sponsor’s confidence-interval bounds are not reported. The recalculation can be useful as a transparent description of what follows from the posted means and standard deviations, but it should not replace the sponsor analysis or be treated as definitive evidence of a group difference.

The recorded outcomes concern HbA1c, not every aspect of health, care delivery, or participant experience. The registry does not establish that any observed change was caused by the messaging system, nor does it show that the system’s intended features were used consistently. The six-month primary analysis includes visit completers, and the record does not describe the effects of missing follow-up measurements or provide an analysis of all participants assigned to groups.

The three-month secondary result adds context but does not settle the primary finding. Its descriptive direction differs from the six-month means, and the record does not report a direct comparison of change over time. No multiplicity adjustment is reported. Taken together, the posted data describe group summaries at two times and a discrepancy between the sponsor analysis and a recalculation. They do not justify a broader conclusion about treatment effectiveness, and they do not show that one approach to care should be chosen over another.

18. Why This Trial Matters Statistically

This trial is statistically instructive because it combines randomized allocation with an analysis based on participants who completed a follow-up visit. Randomization is an important design feature, but interpretation also depends on who contributes outcome data and how missing measurements are handled. The registry explicitly identifies the six-month completer analysis while leaving the reasons for noncompletion and any missing-data analyses unreported.

It also illustrates why a point estimate, a p-value, and an analysis method should be read together. The sponsor’s primary analysis and the unadjusted recalculation report the same stated mean difference, yet their p-values differ. The registry flags that discrepancy, so the sponsor analysis takes precedence. The recalculation helps show what can be derived from the posted group summaries, but it cannot explain the sponsor’s result or substitute for details that the record does not contain.

The two HbA1c follow-up times also underline the difference between describing separate group summaries and testing a group-by-time pattern. The three-month and six-month results do not, on their own, establish how the relative change evolved. A direct longitudinal analysis is not reported in the registry record.

Finally, sample-size planning illustrations based on observed effects are not equivalent to the original design rationale or to observed power. They show how planning requirements change with the effect assumed and the target power, under stated assumptions. The original trial’s design justification is not reported. Together, these features make transparent reporting of analysis populations, model details, missing-data methods, and prespecified outcome handling especially important for interpreting a pilot’s numerical findings.

19. Limitations

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

The primary endpoint analysis is restricted to participants who completed the six-month visit. The record does not report why other participants did not complete that visit, whether their characteristics differed from those analyzed, or how missing outcomes were handled in any additional analysis. This limits assessment of whether the completer results represent the randomized groups.

The sponsor-reported ANOVA and the unadjusted summary-statistic recalculation have different p-values, and the registry flags a discrepancy. The sponsor’s analysis takes precedence, but the record does not report enough detail to reproduce it or explain the difference. The sponsor’s confidence-interval bounds are also absent. The recalculated intervals do not have a confidence level stated in the record.

The registry does not report a detailed analysis plan, the sample-size justification for the original trial, any interim analysis plan, a multiplicity adjustment, or a hierarchy for interpreting the secondary endpoint. It does not report a direct longitudinal comparison across the three-month and six-month measurements. The three-month result is secondary and has no sponsor-reported analysis listed.

The intervention description lists several intended communication and monitoring functions, but the registry does not report how consistently participants used them or which functions were delivered to each participant. No masking was reported. The safety record contains event counts but does not populate the deaths fields or provide detailed event descriptions. These limitations constrain what can be concluded from the posted results; they do not establish that bias or harm occurred.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Diabetes Remote Care Management System — NCT01354015.
• Registry record last updated 2017-05-09; status last verified 2017-03.
• Results first posted 2015-01-13.
• 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

The registry describes a completed, randomized, parallel-group pilot sponsored by Tulane University School of Medicine, comparing a diabetes remote care messaging system with Usual Care. Enrollment is listed as 98, and no masking is reported. The primary outcome is mean change in HbA1c from baseline to six months, based on visit completers.

The sponsor’s ANOVA reports a mean difference of 0.15 and p = 0.254. A separate unadjusted recalculation from the posted summaries reports the same mean difference with p = 0.006. The record flags a discrepancy, so the sponsor’s analysis takes precedence. A three-month HbA1c change is also posted as a secondary outcome, without a sponsor-reported analysis. Missing-data methods, the original sample-size justification, and details needed to reconcile the primary analyses are not reported in the registry record.