Short-term Study of Combination Treatment of Escitalopram and Gaboxadol in Major Depressive Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Short-term Study of Combination Treatment of Escitalopram and Gaboxadol in Major Depressive Disorder (NCT00807248), a randomized phase 2 study. The primary outcome is Montgomery and Åsberg Depression Rating Scale (MADRS).

2. Clinical Question

The registry describes a randomized, double-blind, parallel-group study. Random allocation is intended to make groups comparable on average, while masking can reduce the influence of treatment knowledge on participant responses and outcome assessment. These are general design considerations; the registry record does not provide further details about how allocation or masking was implemented.

The clinical question therefore has two related parts: how each active regimen’s change compares with placebo, and whether the gaboxadol combinations differ from escitalopram alone. These are not interchangeable comparisons. Evidence of a difference between an active group and placebo does not by itself establish an added benefit from combining treatments; the direct comparison of the combination against escitalopram alone is needed for that question.

3. Trial Design

The registry identifies NCT00807248 as a completed Phase 2 randomized, double-blind, parallel-group, placebo-controlled trial. Its title describes fixed-dose escitalopram combined with two fixed doses of gaboxadol, compared with escitalopram in major depressive disorder. The registry lists enrollment of 490. The study start is recorded as 2008-11 and primary completion as 2009-12.

Four groups appear in the posted primary outcome data: placebo; escitalopram 20 mg with placebo; escitalopram 20 mg with gaboxadol 5 mg; and escitalopram 20 mg with gaboxadol 10 mg. The brief description specifies a comparison after 8 weeks of treatment. All listed oral regimens are once daily. The registry lists H. Lundbeck A/S as sponsor.

The primary outcome is MADRS mean change from baseline to Week 8. The registry describes the primary analysis population as the Full-analysis Set, with Last Observation Carried Forward and Analysis of Covariance. The arm results are reported as means with standard errors. Separate sponsor analyses are described as ANCOVA comparisons of final values. This distinction in outcome parameter and analysis matters when reading the recalculated contrasts alongside the sponsor’s estimates.

The registry lists eight secondary outcome measures, including depression, anxiety, insomnia, disability, and global clinical ratings. The posted secondary results include MADRS, the Hospital Anxiety and Depression Scale, and the Insomnia Severity Index. Results for other listed secondary measures are not included in the registry record described for these posted outcomes, so their numerical findings are not reported in the registry record.

Further details on the recruitment setting, eligibility criteria, allocation concealment, adherence, and the original sample-size justification are not reported in the registry record. Nor are detailed protocol or statistical-analysis-plan provisions available in the record details. The design labels establish the broad structure, but do not answer those implementation questions.

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:

Analysis population for the primary outcome, as defined in the registry: Mean change from baseline to Week 8: Full-analysis Set (FAS), Last Observation Carried Forward (LOCF), Analysis of Covariance (ANCOVA)

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 Montgomery and Åsberg Depression Rating Scale (MADRS), measured over Baseline to 8 weeks.

The primary endpoint is mean change in MADRS score from baseline to Week 8. MADRS is a 10-item rating scale designed to assess the severity of symptoms in depressive illness and to be sensitive to treatment effects. Each symptom is rated on a 7-point scale from 0, no symptom, to 6, severe symptom. Definitions of severity are provided at 2-point intervals, and the total score ranges from 0 to 60.

Change from baseline expresses how the average recorded score shifted during the treatment period. In these results, the reported mean changes are negative, indicating that the Week 8 score was lower on average than baseline for each group. A between-group contrast asks how much the average change differs between two assigned groups. That contrast is more informative for a treatment comparison than looking at change in one group alone, because changes can also occur in a placebo group.

MADRS is used as the primary outcome because it gives a structured, quantitative assessment of depressive symptom severity. As a continuous measure, it retains more gradation than an outcome that only sorts participants into broad categories. General statistical interpretation still depends on the quality of measurement, missing-data handling, and the analysis model. The registry identifies an ANCOVA analysis of the Full-analysis Set using Last Observation Carried Forward, but further details of endpoint assessment and missing-data procedures are not reported in the registry record.

The posted change scores and sponsor comparisons should be read according to their stated parameter. The arm summaries concern mean change, while the sponsor’s reported analyses are labeled mean differences in final values. Those estimates are related descriptions of the outcome but are not the same reported quantity.

7. Results Posted in the Registry

• Escitalopram 20 mg and Placebo (Orally, Once Daily): mean -19.00 versus -13.40 on control; difference -5.60 (95% CI -8.39 to -2.81), standardized effect d = -0.55; z-test p = <0.001.
• Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily): mean -18.50 versus -13.40 on control; difference -5.10 (95% CI -7.89 to -2.31), standardized effect d = -0.50; z-test p = <0.001.
• Escitalopram 20 mg and Gaboxadol 10 mg (Orally, Once Daily): mean -19.40 versus -13.40 on control; difference -6.00 (95% CI -8.79 to -3.21), standardized effect d = -0.59; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

MADRS

• 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 errors.
• Result:
  • Escitalopram 20 mg and Placebo (Orally, Once Daily): mean -19.00 versus -13.40 on control; difference -5.60 (95% CI -8.39 to -2.81), standardized effect d = -0.55; z-test p = <0.001.
  • Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily): mean -18.50 versus -13.40 on control; difference -5.10 (95% CI -7.89 to -2.31), standardized effect d = -0.50; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram 20 mg and Placebo (Orally, Once Daily) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Hospital Anxiety and Depression Scale (HADS)

• 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 errors.
• Result:
  • Escitalopram 20 mg and Placebo (Orally, Once Daily): mean -14.70 versus -9.70 on control; difference -5.00 (95% CI -7.51 to -2.49), standardized effect d = -0.55; z-test p = <0.001.
  • Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily): mean -14.10 versus -9.70 on control; difference -4.40 (95% CI -6.91 to -1.89), standardized effect d = -0.48; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram 20 mg and Placebo (Orally, Once Daily) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Insomnia Severity Index (ISI)

• 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 errors.
• Result:
  • Escitalopram 20 mg and Placebo (Orally, Once Daily): mean -10.00 versus -6.90 on control; difference -3.10 (95% CI -4.79 to -1.41), standardized effect d = -0.53; z-test p = <0.001.
  • Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily): mean -9.60 versus -6.90 on control; difference -2.70 (95% CI -4.39 to -1.01), standardized effect d = -0.46; z-test p = 0.002.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram 20 mg and Placebo (Orally, Once Daily) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram 20 mg and Gaboxadol 5 mg (Orally, Once Daily) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcome summaries show a broadly similar descriptive pattern to the primary outcome: average scores changed from baseline in a negative direction across groups on HADS and ISI, and the active groups’ reported changes were more negative than the placebo group’s summaries. The posted recalculations also give active-versus-placebo contrasts for these measures. Sponsor ANCOVA comparisons are reported for the placebo comparisons involving escitalopram alone and the gaboxadol 5 mg combination. These secondary measures complement the MADRS results by covering anxiety and depression symptoms and insomnia severity, but they do not answer exactly the same measurement question.

The overall pattern can be considered alongside MADRS, but secondary results do not replace the primary endpoint. In particular, the registry lists additional measures of disability and global clinical ratings, while numerical findings for those measures are not reported in the registry record details described for these results. A complete interpretation of the full set of listed secondary outcomes is therefore not possible from the posted numerical results.

Multiple outcomes and comparisons create a multiplicity concern: as more hypotheses are tested, the chance of obtaining an apparently unusual result by chance can increase. Unless a testing hierarchy or adjustment method is specified, nominal p-values for secondary contrasts should not be treated as if they came from a single prespecified test. The registry record does not report a multiplicity adjustment or a hierarchy for secondary outcomes. The direction of the descriptive summaries can be noted, but isolated unadjusted findings carry less evidential weight than results from a clearly specified primary analysis.

These measures also differ in scale and construct, so a numerical contrast on one instrument cannot be directly translated into a contrast on another without additional validation and interpretation. The record does not report thresholds for clinically meaningful change on these secondary measures. Their role is therefore supportive and exploratory in this statistical reading, rather than definitive evidence of a broader or lasting outcome.

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

• MADRS
• Hospital Anxiety and Depression Scale (HADS)
• Insomnia Severity Index (ISI)
• Sheehan Disability Scale (SDS): Family Subscale
• SDS: Work Subscale
• SDS: Social Subscale
• Clinical Global Impression - Severity of Illness (CGI-S)
• Clinical Global Impression - Global Improvement (CGI-I)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary measure is continuous: MADRS scores and their changes can take a range of values. The registry reports mean changes and standard errors for each group. A mean describes the group’s average score change; it does not show the range of individual responses or imply that every participant changed by that amount. A standard error describes the precision of an estimated mean, rather than the spread of individual scores. General statistical interpretation should keep those quantities distinct.

What does ANCOVA contribute?

Analysis of Covariance (ANCOVA) estimates a group contrast while accounting for covariates included in the model. In a baseline-and-follow-up outcome setting, adjustment for baseline measurements can improve precision when baseline scores are related to follow-up scores. That is a general explanation of ANCOVA, not a claim about every covariate used in this trial: the registry record does not report the full model specification, covariates beyond the outcome description, or diagnostics. The primary analysis is described as ANCOVA on the Full-analysis Set with Last Observation Carried Forward. The sponsor’s posted comparisons are also labeled ANCOVA, but are reported as mean differences in final values.

The sponsor’s ANCOVA estimates should be distinguished from recalculated comparisons. The recalculated values are unadjusted re-calculations based on the posted group means and summary statistics. They do not reproduce the sponsor’s covariate-adjusted analysis. In addition, the primary arm summaries concern baseline-to-Week 8 change, while sponsor comparisons are labeled final values. Those distinctions mean that the estimates need not match exactly. Because the registry identifies a discrepancy, the sponsor’s own analysis takes precedence for its reported comparisons.

What does the confidence interval tell the reader?

A confidence interval describes uncertainty around an estimated contrast under the statistical model and its assumptions. A narrower interval indicates greater precision than a wider interval. If an interval includes zero for a difference, the data are compatible with no difference as well as with differences in either direction at the stated confidence level. Excluding zero is evidence of a nonzero contrast under that model, but does not by itself establish clinical importance or remove concerns about analysis choices. The interval is not a range containing every individual outcome.

The unadjusted recalculations give interval estimates for the active-versus-placebo mean changes. The sponsor’s reported ANCOVA intervals instead correspond to its final-value comparisons. These should be read as different analyses, rather than interchangeable versions of a single calculation. The p-values also reflect the particular analysis from which they come. A small p-value is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect.

Why does missing-data handling matter?

The primary population is described as the Full-analysis Set and the method for missing values as Last Observation Carried Forward (LOCF). LOCF substitutes a participant’s last available observation for a later missing measurement. As a general method, this can make analysis easier to implement, but it relies on an assumption that the last observed value is a reasonable stand-in for the unobserved endpoint. If symptoms would have changed after the last observation, that assumption may not hold. The registry record does not report the amount or pattern of missingness, reasons for missing values, or sensitivity analyses using other missing-data methods. The treatment of missing outcomes in the sponsor’s reported comparisons is not reported in the registry record.

What do the recalculations and planning illustrations represent?

Recalculated contrasts use the posted summary statistics and are simple, unadjusted comparisons, not replacements for the trial’s fitted ANCOVA. The recalculated standard deviations and standardized effects are derived for statistical comparison, but do not add adjustment for baseline or other covariates. The registry indicates a discrepancy, so sponsor analyses take precedence when their corresponding comparison is reported.

Sample-size illustrations use the observed effect as a planning reference and apply the stated assumptions. They are not a retrospective assessment of the probability that the completed study detected an effect. General planning calculations are sensitive to the assumed effect, variability, allocation, significance level, and model. The trial’s original sample-size justification and design assumptions are not reported in the registry record. The posted illustrations therefore answer a hypothetical planning question, not what the trial was originally designed to guarantee.

What statistical plan details are absent?

The registry record does not report whether interim analyses were conducted, the full hierarchy of hypotheses, a multiplicity-control method, or a complete statistical analysis plan. These omissions matter because the study includes multiple arms and multiple outcome measures. When many comparisons are examined, some may appear noteworthy by chance. Without a reported plan for prioritizing or adjusting those tests, individual secondary p-values warrant cautious interpretation. No further assumptions about the protocol, analysis choices, or sample-size rationale can be inferred from the posted results.

11. Confidence Intervals and Precision

The sponsor’s primary-outcome comparisons provide confidence intervals that make the uncertainty around the estimated contrasts visible. For placebo versus escitalopram 20 mg with placebo, the reported adjusted mean difference in final MADRS values was -5.55, with a 95% confidence interval from -7.98 to -3.11. For placebo versus escitalopram 20 mg with gaboxadol 5 mg, it was -5.09, with a 95% confidence interval from -7.53 to -2.66. Both intervals lie below zero, and the corresponding sponsor p-values are < 0.001.

The sponsor’s direct combination-versus-escitalopram comparisons are less conclusive about a difference between those regimens. For escitalopram plus placebo versus the gaboxadol 10 mg combination, the estimate was -0.45 and the 95% confidence interval ran from -2.34 to 1.44. For escitalopram plus placebo versus the gaboxadol 5 mg combination, the estimate was 0.47 and the interval ran from -1.41 to 2.35. Both intervals include zero and allow estimates in either direction. They do not establish equivalence: an equivalence conclusion would require an appropriate prespecified margin and analysis, which are not reported in the registry record.

A confidence interval represents uncertainty in an estimated group contrast under the analysis model; it is not a guarantee about the true value and does not describe the spread of individual participant outcomes. Wider intervals allow a broader range of plausible contrasts and indicate less precision. Precision also does not resolve potential bias from missing data, multiplicity, or analysis choices.

The recalculated intervals are unadjusted summaries-based estimates, whereas sponsor intervals come from ANCOVA. Given the recorded discrepancy, sponsor analyses take precedence where available. The registry record does not report a clinically meaningful difference threshold, so the precision of the estimates cannot by itself determine whether a difference would matter clinically.

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 figures are planning illustrations based on the observed effect, not the original enrollment plan. For these continuous-outcome comparisons, the entries are participants per arm, not total events. The stated assumptions 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. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column gives the participants per arm estimated for an 80% chance of detecting the specified effect under those assumptions, if that effect were the true contrast and the model assumptions held. The 90% power column applies the higher target probability of detection. These are planning probabilities for a hypothetical future design; this is not observed power.

The table includes scenarios based on 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. Smaller target effects require more participants per arm to reach the same power target, because a smaller signal is harder to distinguish from random variation. The accompanying curve should be read as the same relationship across assumed sample sizes and effects; it does not alter the underlying planning assumptions or show the trial’s actual probability of success.

Observed effects are uncertain and can be larger or smaller than the effect a future study should target. Using an observed estimate as a planning anchor is therefore illustrative rather than a design recommendation. The registry record does not report whether the trial’s original enrollment target was based on these assumptions or another rationale.

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 = -5.600; pooled standard deviation σ = 10.179.
2. n = 2 × (1.96 + 0.842)² × 10.179² / -5.600² = 51.9, rounded up to 52 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 <- -5.6                      # observed difference in means
sigma <- 10.1789564222544                     # 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(807248)
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 = 52 (80% power), 70 (90% power)
 75% of observed effect: n per arm = 93 (80% power), 124 (90% power)
 50% of observed effect: n per arm = 208 (80% power), 278 (90% power)
power.t.test: 52.8 
Simulated power at n = 52 per arm: 0.797 (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 analyses are adjusted ANCOVA comparisons of final MADRS values. The primary outcome description also identifies an ANCOVA analysis, but the posted arm summaries and recalculated contrasts concern mean change from baseline. The recalculations are unadjusted and based on summary statistics, so they are not expected to reproduce every feature of the sponsor’s model.

For the placebo comparisons that are reported by the sponsor, the sponsor estimates are -5.55 for escitalopram with placebo versus placebo and -5.09 for the gaboxadol 5 mg combination versus placebo. Their 95% confidence intervals are -7.98 to -3.11 and -7.53 to -2.66, respectively, with p-values < 0.001. These are the sponsor’s own adjusted results and take precedence over recalculated contrasts because the registry records a discrepancy.

The registry also gives sponsor analyses comparing each gaboxadol combination with escitalopram plus placebo. Those estimates, confidence intervals, and p-values address the added-combination question directly. The intervals include zero, and the p-values are 0.640 for the 10 mg comparison and 0.623 for the 5 mg comparison. These results do not establish a difference between the combination groups and escitalopram alone. They also do not establish that the groups are equivalent.

Comparisons should be interpreted only in the direction and parameter specified in the sponsor analysis. The record does not provide the full ANCOVA model, covariate set, or a complete account of analysis populations for these comparisons, so those details are not inferred.

15. Multiplicity, Missing Data and Analysis Populations

The trial has multiple randomized groups and several outcome measures, and the registry reports numerous analyses. When many comparisons are made, the chance of at least one result appearing statistically notable by chance increases. Statistical approaches can address this by adjusting significance thresholds, specifying a hierarchy of tests, or identifying a limited set of confirmatory hypotheses. The registry record does not report which, if any, approach was used for multiplicity.

This omission is especially relevant when reading secondary outcomes and the recalculated p-values. Those recalculations are unadjusted. Their p-values describe individual contrasts under their respective calculations and do not account for the broader set of tests. The sponsor’s reported p-values should be distinguished from the recalculated ones, but the registry record does not supply a full multiplicity strategy for the sponsor analyses either. Consequently, the posted findings should not be interpreted as a family of independently confirmatory conclusions.

The primary analysis description identifies the Full-analysis Set, LOCF, and ANCOVA. The registry does not report a detailed definition of the Full-analysis Set, the extent and pattern of missing data, reasons for missingness, or sensitivity analyses. LOCF is the named approach, but the record does not report the assumptions used to justify it or whether alternative approaches were evaluated. Missing-data assumptions can influence estimated group contrasts, so the absence of such details limits assessment of robustness.

Further analysis-plan details are also not reported in the registry record: whether interim analyses were planned or performed, which covariates were included in each ANCOVA, whether any subgroup analyses were prespecified, and how deviations from the analysis plan were handled. The original sample-size justification is not reported either. These gaps do not invalidate the posted results, but they prevent a full audit of how hypotheses and analytic choices were set before outcomes were examined.

The statistical reading should therefore separate the prespecified primary endpoint as described in the record, the sponsor’s reported analyses, unadjusted recalculations, and secondary results. Each is informative in a different way, and none resolves the unreported multiplicity and missing-data details.

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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: appendicitis (1.4%).

The registry posts serious adverse-event counts and counts for other reported adverse events by group. Serious events are listed for 1 participant in the placebo group and 0 in each of the three escitalopram-based groups. The corresponding numbers at risk are 71, 140, 139, and 140. The top serious-event term is Appendicitis, listed as 1.4%. Death counts are not reported in the registry record.

Other reported adverse-event counts are 12 for placebo, 22 for escitalopram with placebo, 31 for the gaboxadol 5 mg combination, and 19 for the gaboxadol 10 mg combination. These counts should be read with their reported at-risk group sizes and according to the registry’s adverse-event categories. The record details do not define the category details or provide event-level descriptions beyond the listed serious-event term.

Counts alone do not establish whether an event was caused by treatment, and a zero posted count does not show that no risk exists. Comparisons of safety also depend on event definitions, follow-up, completeness of reporting, and the number of participants observed. The registry record does not provide the additional information needed for a broader comparative safety assessment.

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

The posted results describe average scores and estimated group contrasts over an 8-week treatment period in this randomized study of major depressive disorder. The primary endpoint is change in MADRS, and the arm summaries show negative mean changes in all groups. Sponsor ANCOVA results report differences for placebo comparisons and direct comparisons between the gaboxadol combinations and escitalopram plus placebo. The direct combination comparisons have confidence intervals that include zero, so the sponsor analyses do not establish a difference between those groups.

That statistical conclusion is narrower than saying the treatments are the same. Failure to establish a difference is not proof of equivalence or noninferiority; those conclusions require an appropriate design, a prespecified margin, and analysis. Such a margin is not reported in the registry record. Conversely, a statistically distinguishable contrast does not by itself show that the size of the difference is important to patients. The registry record does not report a clinically meaningful MADRS threshold.

These results also do not establish what happens beyond Week 8, in other populations, or under different treatment conditions. The record provides no long-term follow-up results in the registry record. Findings from a particular analysis population and missing-data method may not represent outcomes under other assumptions, and sensitivity analyses are not reported in the registry record.

The trial’s design labels and posted data support a statistical description, not personalized medical guidance. The results should not be used alone to make a treatment choice. They also cannot determine the mechanism of any observed score changes or establish the cause of adverse events. Statistical significance, effect size, precision, clinical importance, and safety are distinct considerations; the registry record does not contain all information needed to resolve each of them.

18. Why This Trial Matters Statistically

This trial illustrates why the question attached to a comparison matters. A comparison of an active regimen with placebo addresses a different question from a direct comparison of a combination regimen with escitalopram alone. The sponsor’s analyses include both types of comparison, and the direct combination-versus-escitalopram estimates do not establish a difference. Reading only the active-versus-placebo results would not answer whether gaboxadol adds a detectable change relative to escitalopram.

It also demonstrates why the estimand and analysis need to be kept clear. The registry’s primary outcome summary is mean change from baseline, while sponsor comparisons are reported as adjusted mean differences in final values. Recalculations using posted summaries can be useful for understanding scale and direction, but they cannot substitute for the sponsor’s adjusted model. In this record, a discrepancy is explicitly identified, making it important to give precedence to the sponsor’s own analyses.

The example further shows how precision shapes interpretation. The intervals for the sponsor’s direct combination comparisons include zero and allow effects in either direction; they do not prove the regimens identical. A nonsignificant result and an equivalence result are not the same claim. Similarly, statistical evidence in a placebo comparison does not by itself determine clinical importance, especially when no meaningful-difference threshold is reported.

Finally, multiple outcomes, unreported multiplicity handling, and limited missing-data detail influence how strongly the results can be interpreted. Planning illustrations based on observed effects are not observed power and are not the trial’s documented design rationale. Together, these considerations make the study a useful example of separating the statistical question, analysis method, uncertainty, and clinical interpretation.

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 limitations specific to this record constrain interpretation. The registry reports a discrepancy between recalculated summaries and sponsor analyses. The recalculations are unadjusted and based on posted group summaries, whereas sponsor comparisons use ANCOVA and final values. Sponsor results take precedence where the comparison is reported, but the complete model specification and covariates are not reported in the registry record.

The primary population is described as the Full-analysis Set and the missing-data approach as LOCF, but the record does not report the population definition, extent or pattern of missingness, reasons for missingness, or sensitivity analyses. These omissions make it difficult to judge how dependent the estimates are on the handling of unobserved Week 8 scores. Further details of allocation concealment, recruitment setting, eligibility criteria, adherence, and outcome assessment are also not reported in the registry record.

Multiplicity is another limitation: the record lists multiple outcomes and analyses but does not report a correction strategy or testing hierarchy. Secondary p-values should therefore be interpreted cautiously. The numerical secondary results described include only some of the listed outcomes, and the remaining outcome findings are not reported in the registry record.

The direct combination comparisons have intervals that include zero, but the record does not report an equivalence or noninferiority margin. Thus, those results cannot establish similarity. Nor does the record give a clinically meaningful MADRS difference threshold, so clinical importance cannot be judged from statistical estimates alone.

Finally, the primary comparison concerns an 8-week timeframe. Longer-term outcomes and any broader follow-up are not reported in the registry record. The sample-size illustrations use observed effects and stated planning assumptions; they are not the original trial justification, which is not reported.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Short-term Study of Combination Treatment of Escitalopram and Gaboxadol in Major Depressive Disorder — NCT00807248.
• Registry record last updated 2012-12-17; status last verified 2012-12.
• Results first posted 2011-08-11.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT00807248 is a completed, randomized, double-blind, parallel-group, placebo-controlled Phase 2 study in major depressive disorder, sponsored by H. Lundbeck A/S. The registry lists enrollment of 490 and an 8-week comparison of escitalopram 20 mg with gaboxadol 5 mg or 10 mg against escitalopram with placebo, alongside a placebo-only group.

The primary endpoint was MADRS mean change from baseline to Week 8. The registry describes a Full-analysis Set, LOCF, and ANCOVA. Sponsor analyses report placebo comparisons and direct comparisons of each combination with escitalopram alone. The direct comparisons do not establish a difference; they do not establish equivalence. Because a discrepancy is recorded, sponsor analyses take precedence over unadjusted recalculations where results overlap.

Interpretation is limited by unreported multiplicity handling, missing-data detail, full model specifications, and the original sample-size justification. The registry record does not report a clinically meaningful MADRS threshold or longer-term outcome results.