Randomized Trial Comparing Sitagliptin to Placebo in Closed Loop: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Randomized Trial Comparing Sitagliptin to Placebo in Closed Loop (NCT02328040), a randomized phase 1 study. The primary outcome is Blood Glucose Measures in Subjects Treated With Sitagliptin, Compared to the Placebo.

2. Clinical Question

The registry describes a randomized, controlled comparison and identifies blood glucose as the primary outcome. The clinical question centers on post-meal glucose, because the study summary explains that meals can pose a concern for people using insulin to manage type 1 diabetes. The protocol description says participants received a mixed meal and underwent blood testing for glucose, insulin, and relevant hormones. It also describes monitoring for high glucose in the context of delayed stomach emptying or glucagon suppression. The posted primary result, however, is a mean blood-glucose measure. It does not separately report a post-meal time course, a peak value, or a measure restricted to a specified period after the meal. The registry record does not report how the 50 mg and 100 mg visit-level dosing relates to the two displayed treatment groups. Thus, the group-level result addresses the broad glucose comparison, while leaving several details about the exact estimand and dose-specific questions unresolved.

3. Trial Design

The registry lists this study as a randomized, triple-masked, phase 1 trial sponsored by Albert Einstein College of Medicine. Its title describes a double-blinded, randomized, controlled paired trial, while the registry’s model entry says single group. The intervention listings distinguish placebo with insulin via closed loop from sitagliptin with insulin/novolog via closed loop. Those descriptions establish the intended comparison, but the record does not provide enough detail to reconstruct the assignment and visit sequence or to determine how the paired wording relates to the group-based results.

The registry reports enrollment of 17 participants, with study start in 2014-09 and primary completion in 2016-01. The status is completed. The brief summary describes an initial safety assessment in 10 adults aged 18-30 years with established type 1 diabetes for at least 1 year, followed by recruitment of participants aged 14-19 years after safety criteria were established. It mentions sitagliptin doses of 50 mg and 100 mg and says participants and researchers would not know which dose was taken at a visit. It does not explain the final allocation by age group or dose.

The outcome displays report 15 participants in each treatment group, whereas overall enrollment is 17. The registry record does not explain that difference. It also does not report allocation ratios, sequence generation, concealment procedures, the number or order of treatment visits, washout details, or the handling of repeated measurements. The results should therefore be read as posted group summaries rather than a fully documented account of the trial’s design and 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.

First Intervention (2 Days)

Reasons for not completing this period:

Second Intervention (2 Days)

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 Blood Glucose Measures in Subjects Treated With Sitagliptin, Compared to the Placebo, measured over 18 Months.

The primary endpoint is titled Blood Glucose Measures in Subjects Treated With Sitagliptin, Compared to the Placebo. It is a continuous measure reported as a mean in mg/dl, with standard deviation as its dispersion measure and an 18 Months timeframe. The registry description frames the outcome as better targeted blood-glucose levels in the closed-loop setting in the sitagliptin arm compared with placebo and insulin monotherapy.

Blood glucose is used to represent the glucose outcome that the intervention was intended to affect in this study. The brief summary gives a particular rationale: glucose after a meal can be a concern in type 1 diabetes, and sitagliptin was being tested for a possible role in post-meal glucose regulation. The protocol description includes a mixed meal and blood testing, but the posted primary measure does not specify a particular post-meal interval or distinguish post-meal values from other glucose measurements. It should therefore be interpreted as the registry’s stated mean blood-glucose outcome, not as a separately reported post-meal endpoint or a comprehensive account of glucose control.

A mean summarizes the central level of a continuous measurement across the observations included in its calculation. The standard deviation describes variation among those observations. Neither alone reveals the full distribution, the range of individual responses, or how glucose changed over time. Those details are not reported in the registry record.

7. Results Posted in the Registry

• Sitagliptin/Insulin: mean 155.90 versus 149.70 on control; difference 6.20 (95% CI -24.96 to 37.36), standardized effect d = 0.14; z-test p = 0.697.

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.

Measure of Glucagon Concentration in Subjects Treated With Sitagliptin, Compared to Placebo

• 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:
  • Sitagliptin/Insulin: mean 5.66 versus 6.04 on control; difference -0.38 (95% CI -1.49 to 0.73), standardized effect d = -0.25; z-test p = 0.502.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Sitagliptin/Insulin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists glucagon concentration as the secondary outcome. Its group summaries and unadjusted comparison are presented in the outcome displays. At a broad level, the secondary estimate, like the primary estimate, has uncertainty spanning values on either side of zero. The results therefore do not provide a clearly resolved secondary pattern that would alter the interpretation of the primary glucose comparison.

Glucagon is relevant to the study’s stated rationale because the brief summary identifies it as a hormone that can increase blood glucose and says the study would measure hormones affecting glucose. That biological context explains why the measure appears among the outcomes; it does not establish that the glucagon result mediates or explains the glucose result. The registry does not report a mediation analysis, a relationship between participant-level glucagon and glucose values, or a prespecified framework for combining the endpoints.

Secondary endpoints generally provide supporting or exploratory information rather than replacing the primary endpoint. Their interpretation depends on how many outcomes and analyses were considered, whether hypotheses were prespecified, and whether a multiplicity adjustment was used. The registry reports two outcome measures and no reported analyses, but it does not state whether the secondary outcome was adjusted for multiple comparisons or how confirmatory the analysis was intended to be. An unadjusted result can be more vulnerable to chance findings when several endpoints are examined. This is why the secondary comparison should carry less inferential weight than a well-specified primary analysis and should not be treated as independent confirmation of a treatment effect.

Overall, the secondary information is consistent with caution rather than a stronger conclusion: it does not resolve the uncertainty in the primary result, and the registry record does not supply an analysis plan that would justify combining the two outcomes into one claim.

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

• Measure of Glucagon Concentration in Subjects Treated With Sitagliptin, Compared to Placebo

9. Statistical Interpretation

10. Statistical Methods Explained

Statistical methods help distinguish the observed group summaries from the uncertainty around their comparison. The registry reports continuous outcomes as means with standard deviations. It also reports unadjusted mean differences, confidence intervals, and p-values in the comparison displays. The registry record does not report the trial’s original statistical analysis plan, the exact inferential procedure used by the sponsor, or any reported analysis results. The explanations below describe general statistical concepts and how to read the posted summaries; they should not be mistaken for a reconstruction of an undocumented protocol.

What does a mean difference represent?

In general, a mean difference expresses how far apart two group averages are, using the original measurement units. A positive value in the primary comparison means the sitagliptin/insulin mean is higher than the placebo/insulin mean as the comparison is oriented. The posted estimate is 6.2 mg/dl. That is a summary of the difference in averages, not a statement that every participant had that difference. It does not show the distribution of individual outcomes or, if measurements were paired, the pattern of each participant’s change across conditions.

The standard deviation is a measure of the spread of observations around a mean. It is not the uncertainty of the mean difference itself. Uncertainty in the comparison also depends on the amount of information available and on how observations are related. The posted group standard deviations are 37.8 mg/dl and 48.6 mg/dl, and each group has 15 participants in the outcome display. The registry record does not report whether the primary analysis used participant-level repeated measures, a paired contrast, or another model. A paired analysis would generally use within-participant information, so the relationship between measurements from the same person can matter. That relationship is not reported.

How should the confidence interval be read?

In general, a confidence interval expresses the statistical uncertainty around an estimated quantity under a particular analysis framework. The primary interval runs from -24.96 to 37.36 mg/dl. Because it includes zero and extends in both directions, the posted estimate is compatible with a range of possible differences, including values favoring either group. The interval is not a guarantee that the true value lies inside its endpoints, nor does it mean that all values inside are equally plausible. Its width makes clear that the estimate is not precise enough to pin down a narrow range of group differences.

The registry record does not report the confidence-interval construction method, the model specification, or whether any adjustment was made for the design’s paired wording. The recalculated comparison is an unadjusted calculation from posted summary statistics, not a substitute for a documented sponsor analysis. Since no discrepancy is flagged, there is no stated conflict between the posted outcome summaries and that recalculation; the absence of a discrepancy does not fill in missing analysis details.

What does the p-value say?

In general, a p-value measures how unusual data at least as incompatible with a specified null model would be, if that null model and the analysis assumptions held. The primary unadjusted p-value is 0.697. It does not give the probability that there is no treatment difference, and it does not establish equivalence. It also does not quantify the practical importance of a difference. Interpretation depends on the design, model, outcome definition, and analysis plan, several of which are not reported in the registry record.

The word unadjusted matters. A p-value described as unadjusted has not been adjusted for multiple outcome comparisons in the reported calculation. The registry lists two outcome measures but does not report a multiplicity procedure. When several outcomes are examined, treating each unadjusted result as if it were the sole planned test can make chance findings more likely across the set. This concern is especially relevant when outcomes are secondary or exploratory. The record does not say whether the displayed endpoints were designated in advance as confirmatory or whether additional measures were analyzed.

Why can variability and over-dispersion matter?

In general, greater variability makes a group mean harder to estimate precisely when the number of independent observations is limited. If observed variation is greater than assumed by an analysis model, a model may understate uncertainty; this is one reason analysts assess over-dispersion in settings where the assumed variance structure might not fit. The posted outcomes are continuous, and the registry gives standard deviations. It does not report an over-dispersion assessment, a variance model, or diagnostics. There is no basis in the record to claim that over-dispersion was present or that a particular correction was used.

Repeated measurements can also require methods that account for correlation among measurements from the same participant. The title calls the study paired, while the model entry says single group and the results are presented by treatment group. The registry record does not describe the covariance structure, measurement-level data, or the unit used in the analysis. Those omissions limit the extent to which a reader can judge whether the uncertainty calculation matches the design.

In short, the means and recalculated unadjusted comparisons are useful descriptions of the posted results. They do not reveal the full analysis strategy, and general statistical interpretations must remain conditional on details that the registry record does not report.

11. Confidence Intervals and Precision

The primary estimate is a mean difference of 6.2 mg/dl. Its 95% confidence interval extends from -24.96 to 37.36 mg/dl. That interval is broad relative to the point estimate and includes values on both sides of zero. The data therefore give limited precision about the direction and size of the difference. A point estimate alone would obscure that uncertainty; the interval is essential to understanding what the comparison can support.

In general, a confidence interval describes a range of values compatible with the data under the assumptions of the analysis. It is not a range of individual participant outcomes and does not indicate that every value within it has the same support. It also does not show whether any particular value would be clinically important; the registry record does not report a threshold for a meaningful difference.

The primary group standard deviations, 37.8 mg/dl and 48.6 mg/dl, indicate substantial variation among observations relative to the narrowness a precise estimate would require. Each displayed group includes 15 participants. With a limited amount of information and noticeable variability, uncertainty around a group comparison can remain wide. The registry record does not report the number of usable observations for every measurement, missing-data handling, or whether repeated values were incorporated into the analysis.

The secondary glucagon interval also crosses zero, extending from -1.49 to 0.730 Pmol/L around a mean difference of -0.38 Pmol/L. It likewise leaves uncertainty about direction and magnitude. Neither interval supports a claim of equivalence: demonstrating equivalence would require a defined equivalence margin and an analysis designed for that question, neither of which is reported. The intervals are most useful as a reminder that the displayed point estimates are not precise conclusions.

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 sample-size illustrations are planning calculations based on the observed effect, not a report of the trial’s original design calculation. For these continuous outcomes, the listed values are participants per arm, not a total number of events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation estimates would be needed for an 80% chance of detecting the specified effect under those assumptions, if that effect were real and the assumptions held. The 90% power column gives the corresponding planning count for a 90% chance. These are design targets, not properties measured after the trial. They are not observed power.

The rows represent planning for 100%, 75%, and 50% of the observed effect. As the effect to detect gets smaller, the required number of participants per arm increases. A smaller difference is harder to distinguish from background variability, so the calculation calls for more information. The primary-outcome planning values range from 775 participants per arm for 80% power at 100% of the observed effect to 3,097 participants per arm for 80% power at 50% of the observed effect; the corresponding 90% values are 1,037 and 4,145. These are illustrations tied to the observed estimate and the stated assumptions. They do not establish what the trial’s original target was, and they should not be read as evidence that the completed study had a particular probability of success.

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 = 6.200; pooled standard deviation σ = 43.536.
2. n = 2 × (1.96 + 0.842)² × 43.536² / 6.200² = 774.0, rounded up to 775 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 <- 6.20000000000002                      # observed difference in means
sigma <- 43.5361918408122                     # 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(2328040)
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 = 775 (80% power), 1,037 (90% power)
 75% of observed effect: n per arm = 1,377 (80% power), 1,843 (90% power)
 50% of observed effect: n per arm = 3,097 (80% power), 4,145 (90% power)
power.t.test: 775 
Simulated power at n = 775 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when more than one outcome, time point, subgroup, or analysis is examined. Even if each test is conducted at its own stated threshold, making many comparisons can increase the chance that at least one appears unusual by chance. General statistical practice may address this through a prespecified hierarchy, adjustment procedures, or a clearly exploratory interpretation. The registry record identifies a primary blood-glucose outcome and a secondary glucagon outcome, but it does not report a multiplicity strategy, a hierarchy among additional tests, or the number of analyses conducted. It lists no reported analyses.

The posted p-values are unadjusted. They should not be interpreted as though all outcomes and analyses were considered in a single confirmatory testing plan with controlled error. Without a reported plan, it is not possible to establish whether the displayed results were the only planned comparisons, whether outcomes were selected after examining data, or whether any correction was used in the trial’s own analysis. None of those possibilities should be assumed. The record simply does not report them.

Information on missing data and analysis populations is also limited. The registry reports overall enrollment of 17 and outcome group counts of 15 in each arm, but it does not explain how the counts relate to enrollment. It does not state how many participants completed each assessment, whether any measurements were missing, why any observations might be absent, or how missing values were handled. The record does not identify an intention-to-treat, per-protocol, or safety population for the posted efficacy summaries.

The design description adds another uncertainty. The title calls the study paired, but the model entry says single group and the outcome results are displayed as two treatment groups. The registry record does not report whether the primary analysis accounted for within-participant pairing, how treatment sequence was handled, or how repeated measurements were modeled. These are not minor details when they can affect the estimate and its uncertainty. As a result, the displayed comparison can be read descriptively, but the absence of analysis-population, missing-data, and multiplicity details limits a full appraisal of its inferential validity.

15. 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 5% reporting threshold.

The registry reports zero serious adverse events affecting participants in the placebo/insulin group and zero in the sitagliptin/insulin group. It also reports zero participants affected by other adverse events in each group and zero deaths in both groups. The safety displays list 15 participants at risk in each treatment group.

These figures describe what is recorded for the study and should not be treated as proof that no adverse events occurred beyond the registry’s reporting. A zero affected count means the record lists no affected participants for that category; it does not establish that a treatment is free of risk. The registry does not provide event terms, detailed severity or timing information, exposure duration by group, or an explanation of the relationship between the safety counts and the overall enrollment of 17.

The brief summary describes safety criteria for an initial group, including maintaining glucose below 70 mg/dl for at least 70% of the time and having no episode of severe hypoglycemia. These are protocol-description statements, not adverse-event counts in the posted safety results. The record does not provide results for those specific criteria. The safety numbers should therefore be interpreted narrowly as registry-reported affected-participant counts, rather than as a complete account of all safety monitoring or as a basis for medical advice.

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

The results describe a small phase 1 trial’s posted group summaries. For the primary outcome, the sitagliptin/insulin mean is higher than the placebo/insulin mean, while the estimated difference has a wide confidence interval that crosses zero. The unadjusted p-value does not provide evidence of a sharply resolved group difference under the recalculation, and the interval indicates substantial uncertainty about the direction and magnitude. The secondary glucagon comparison is also uncertain and does not settle the primary question.

These results do not establish that sitagliptin improves or worsens glucose measures in this setting. They also do not establish equivalence or rule out a potentially meaningful difference. The registry does not give a threshold for clinical importance, a dose-specific analysis, a post-meal time course, or the full statistical analysis plan. The displayed averages cannot show how individuals responded, whether responses varied across age groups, or whether glucose changed differently over time.

The study is described as randomized and masked, which are design features intended to reduce some sources of bias. But design labels alone do not resolve uncertainty about the analysis. The paired wording conflicts with a single-group model entry, while outcomes are reported by treatment group. The registry record does not explain the enrollment-to-outcome-count difference or report missing-data handling. Those gaps make it difficult to determine precisely what population and estimand the group summaries represent.

The proper reading is therefore limited: the record presents an estimated comparison and its uncertainty, not a definitive answer to every clinical or mechanistic question raised by the study. It offers no basis for decisions about an individual’s treatment. Larger or more fully documented studies would be needed to make more precise comparisons, but this result alone does not establish what those studies would find.

17. Why This Trial Matters Statistically

This trial is statistically instructive because it shows why an estimated difference should be read together with its uncertainty. The point estimate is straightforward to state, but the confidence interval is wide and crosses zero. A reader who focuses only on the direction of the group means could overstate what the evidence establishes. The interval makes clear that a small trial can leave a broad range of plausible comparisons unresolved.

It also illustrates the distinction between enrollment and the number of participants in outcome displays. The registry reports 17 enrolled, while the outcome summaries list 15 participants in each group. Without an explanation, those figures cannot be reconciled into a clear account of who contributed to each analysis. Reporting flow and analysis populations are important because readers need to know which participants and measurements underlie each estimate.

The record also highlights how a trial’s design description and analysis presentation should align. A paired-trial title suggests a design in which within-participant information could matter, while a single-group model entry and two-arm summary create ambiguity. The registry does not report the analysis details needed to resolve it. Statistical precision depends not just on the number of participants and observed variability but also on the structure of the data and the method used to analyze that structure.

Finally, the planning illustrations put the uncertainty in context without turning it into a post hoc power claim. They show, under stated assumptions, that smaller effects require more participants. The illustrations do not explain the original design or measure the trial’s observed power. Together, these features demonstrate why transparent reporting of design, analysis, missing data, and multiplicity is essential for interpreting clinical-trial statistics.

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.

Several limitations specific to this registry record constrain interpretation. First, the study is described as paired in its title, but its model is listed as single group and its results are presented by treatment group. The record does not explain whether the same participants experienced both conditions, how visits were ordered, or whether the analysis used paired data. Without these details, the displayed group contrast may not fully represent the design stated in the title.

Second, overall enrollment is 17, while each outcome group has 15 participants. The record does not explain why these counts differ, whether the same individuals contributed to both groups, or whether counts represent measurements rather than unique participants. There is also no participant flow description, no accounting of missing outcomes, and no named analysis population. These omissions make the denominator and generalizability of the estimates less clear.

Third, the primary outcome is described broadly as blood-glucose measures, with an 18 Months timeframe. The registry does not specify the exact measurement schedule, aggregation method, or post-meal interval represented by the reported mean. Although the brief summary describes mixed-meal testing, the posted outcome does not provide a separate meal-response trajectory. Dose-specific outcomes for 50 mg and 100 mg are not reported.

Finally, the registry does not report the sponsor’s statistical analysis plan, a multiplicity approach, confidence-interval method, missing-data strategy, or detailed safety-event terms. The displayed comparisons are unadjusted recalculations from summary statistics, not a substitute for those missing methods. These limitations do not invalidate the descriptive summaries, but they prevent a complete assessment of how the trial’s design and analysis support the reported inference.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Randomized Trial Comparing Sitagliptin to Placebo in Closed Loop — NCT02328040.
• Registry record last updated 2018-08-07; status last verified 2018-07.
• Results first posted 2018-08-07.
• 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

NCT02328040 is a completed, randomized, triple-masked phase 1 study sponsored by Albert Einstein College of Medicine. It compares sitagliptin with placebo in a closed-loop insulin setting for type 1 diabetes. The registry reports 17 participants enrolled, a start in 2014-09, and primary completion in 2016-01. The primary outcome is a mean blood-glucose measure in mg/dl over 18 Months, with glucagon concentration listed as a secondary outcome.

The posted primary comparison is imprecise, with a confidence interval crossing zero. The secondary comparison is also uncertain. The registry record does not report the original analysis plan, multiplicity handling, missing-data methods, or an explanation for the outcome counts relative to enrollment. These results should be read as limited group summaries rather than a definitive answer about treatment effects.