Exploratory Study of SPD489 in Adults With Major Depressive Disorder (MDD) as Augmentation Therapy to an Antidepressant: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exploratory Study of SPD489 in Adults With Major Depressive Disorder (MDD) as Augmentation Therapy to an Antidepressant (NCT00905424), a randomized phase 2 study. The primary outcome is Change From Augmentation Baseline for Non-Remitters in Montgomery-Ǻsberg Depression Rating Scale (MADRS) Total Score at Week 6 - Last Observation Carried Forward (LOCF).

2. Clinical Question

The registry record lists the condition as major depressive disorder and describes the study as exploratory. Its stated objective was to evaluate SPD489 augmentation using mean change in MADRS total score. The primary efficacy population was narrower than all adults with major depressive disorder: it focused on non-remitters at the start of augmentation who met the specified treatment and measurement requirements. This distinction matters when interpreting the results. They describe the analysis population and endpoint defined in the record, not every person with depression or everyone enrolled in the study.

3. Trial Design

The registry record describes a Phase 2, multicenter, randomized, triple-masked, parallel-group, placebo-controlled exploratory efficacy and safety study. Participants were assigned to antidepressant plus SPD489 or antidepressant plus placebo. The record lists enrollment as 246 and the study status as completed. The study start date is 2009-07-30, and the primary completion date is 2010-08-04. The record identifies Shire as the sponsor.

The primary efficacy analysis concerned non-remitters at augmentation baseline, rather than the full enrolled group. Its definition required a MADRS total score > 10 at augmentation baseline, at least one dose of randomized augmentation treatment, and at least one primary efficacy measurement after randomization. The reported primary arm summaries include 65 participants in the SPD489 group and 64 in the placebo group. These are analysis counts for the primary outcome, not a restatement of total enrollment.

The registry record describes LOCF for the primary Week 6 outcome and identifies an ANCOVA analysis among the sponsor's reported analyses. It does not report the original sample-size justification, the planned allocation ratio, the identities of the roles masked, or further details of the ANCOVA specification. The registry record also does not report whether interim analyses were planned or conducted. The available record therefore establishes the broad design and the named analysis method, but not every operational or statistical detail that would be needed to reconstruct the full protocol.

Secondary outcomes included symptom, functioning, fatigue, executive-function, and global-impression measures among non-remitters. Some were assessed at Week 6, and several outcome titles specify LOCF. The registry record lists 18 outcome measures and 18 analyses, with 17 secondary outcomes posted. These additional endpoints broaden the range of questions considered, but they do not replace the primary outcome or establish that every secondary comparison had the same analysis population or missing-data handling.

4. Participant Flow and Accounting

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

Before assignment: 246 subjects entered the antidepressant lead-in phase receiving escitalopram oxalate (antidepressant) 20 mg/day for 8 weeks. After 8 weeks, subjects with residual depressive symptoms (n = 177) were randomly assigned (stratified by remission, either remitters or non-remitters) to receive augmentation therapy (either SPD489 or placebo).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Primary Efficacy Analysis Set defined as all non-remitters (MADRS total score greater than 10 at augmentation baseline) who take at least 1 dose of randomized augmentation treatment and have at least 1 primary efficacy measurement after randomization.

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 From Augmentation Baseline for Non-Remitters in Montgomery-Ǻsberg Depression Rating Scale (MADRS) Total Score at Week 6 - Last Observation Carried Forward (LOCF), measured over Augmentation Baseline, 6 weeks.

The primary endpoint was change from augmentation baseline to Week 6 in total score on the Montgomery-Åsberg Depression Rating Scale (MADRS), among non-remitters. The registry describes MADRS as a validated, 10-item rating scale. Each item is scored from 0-6, and the total score ranges from 0-60. Lower scores indicate decreased severity of depression. The endpoint is continuous: it records a score and evaluates its change, rather than classifying each participant only as improved or not improved.

In this trial, MADRS was used to assess the stated efficacy question about adding SPD489 to an antidepressant. Comparing mean change between the augmentation groups makes the average change in this measure the focus of the primary analysis. A negative change represents a lower score relative to augmentation baseline, which corresponds to reduced symptom severity according to the scale description in the registry record. The group mean does not show the distribution of individual responses, and it cannot by itself establish how many participants experienced a particular degree of change.

The endpoint is specifically a Week 6 change measure for non-remitters, with LOCF used for the primary outcome. Thus, its interpretation is tied to that time point, population, and missing-data approach. It is not a result for all enrolled participants or a direct measure of outcomes beyond the recorded six-week timeframe. The registry record does not report a threshold for deciding whether a change is clinically important, so the reported score differences should be read as scale-based statistical results rather than as proof of a particular level of benefit for an individual.

7. Results Posted in the Registry

• Antidepressant + SPD489 (Non-remitters): mean -7.10 versus -4.90 on control; difference -2.20 (95% CI -4.79 to 0.39), standardized effect d = -0.29; z-test p = 0.096.

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 From Augmentation Baseline for Non-Remitters in the Hamilton Depression Scale (HAM-D) Total Score at Week 6 - LOCF

• 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:
  • Antidepressant + SPD489 (Non-remitters): mean -4.90 versus -4.00 on control; difference -0.90 (95% CI -2.69 to 0.89), standardized effect d = -0.17; z-test p = 0.324.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Antidepressant + SPD489 (Non-remitters) and the control. The interval gives the range of effects that remain compatible with the data.

Change From Augmentation Baseline for Non-Remitters in the Sheehan Disability Scale (SDS) Total Score at Week 6

• 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:
  • Antidepressant + SPD489 (Non-remitters): mean -3.70 versus -1.70 on control; difference -2.00 (95% CI -4.04 to 0.04), standardized effect d = -0.34; z-test p = 0.054.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Antidepressant + SPD489 (Non-remitters) and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Non-Remitters With Improvement on Clinical Global Impression-Improvement (CGI-I) at Week 6 - LOCF

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Antidepressant + SPD489 (Non-remitters): 60.0% (39/65) versus 45.3% (29/64) on control; risk difference 14.7 percentage points (95% CI -2.4 to 31.7), risk ratio 1.32 (0.95–1.85); unadjusted two-proportion z-test p = 0.095.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Antidepressant + SPD489 (Non-remitters) and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes extend beyond the primary MADRS change to other symptom scales, disability, fatigue, executive function, and global impressions. Their overall pattern is mixed in precision: some posted comparisons have estimates in the same general direction as the primary comparison, while others leave greater uncertainty. This range is compatible with the fact that different measures capture different aspects of symptoms and functioning. It does not constitute a single, uniform secondary finding.

These endpoints provide additional context, but they do not alter what the primary result measures. The primary question concerns Week 6 MADRS change among non-remitters; secondary outcomes address other scales or summaries and may have different analysis details. A result on a secondary measure is not a substitute for the primary endpoint, and concordance across measures would not by itself establish a treatment effect.

The registry lists 18 outcome measures and 18 analyses, with 17 secondary outcomes posted. When many outcomes are examined, some may appear notable because of random variation even when no multiplicity adjustment is used. The registry record does not report a multiplicity-adjustment plan. Consequently, unadjusted secondary p-values and intervals carry less confirmatory weight than a prespecified primary comparison interpreted under its stated analysis. The record does not establish which secondary comparisons, if any, were intended as confirmatory.

The secondary results are most useful as exploratory descriptions that can help identify patterns for further study. They should be considered with their endpoint definitions, analysis populations, missing-data approaches, and uncertainty. The results should not be selectively treated as proof based on whether an individual comparison appears more or less compelling than the primary result.

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

• Change From Augmentation Baseline for Non-Remitters in the Hamilton Depression Scale (HAM-D) Total Score at Week 6 - LOCF
• Change From Augmentation Baseline for Non-Remitters in the Sheehan Disability Scale (SDS) Total Score at Week 6
• Percentage of Non-Remitters With Improvement on Clinical Global Impression-Improvement (CGI-I) at Week 6 - LOCF
• Assessment in Non-Remitters of Clinical Global Impression-Severity of Illness (CGI-S) at Augmentation Baseline
• Assessment in Non-Remitters of Clinical Global Impression-Severity of Illness (CGI-S) at Week 6
• Change From Augmentation Baseline for Non-Remitters in the Behavior Rating Inventory of Executive Function - Adult Version (BRIEF-A) Scale Total Score at Week 6
• Change From Augmentation Baseline for Non-Remitters in the Multidimensional Assessment of Fatigue (MAF) Scale Total Score at Week 6
• Change From Augmentation Baseline for Non-Remitters in the Quick Inventory of Depressive Symptomatology - Self-Report (QIDS-SR) Scale Total Score at Week 6

9. Statistical Interpretation

10. Statistical Methods Explained

What does the primary measure compare?

General: A continuous outcome records a quantity on a scale. For a change endpoint, each participant's follow-up score is considered relative to a baseline score; the group summaries then describe average change. A negative mean change is not inherently favorable or unfavorable across all measures, so interpretation depends on the scale's direction. Here, the registry says lower MADRS scores indicate decreased depression severity. The primary summaries are least-squares means, and the record identifies standard error as their dispersion measure.

The primary comparison is between the two augmentation groups among non-remitters. The reported arm means are -7.1 and -4.9, and the recalculated difference is -2.2. In a difference framed as SPD489 minus placebo, a negative estimate means the numerical change was more negative in the SPD489 group. This describes the direction of the group contrast; it does not tell how consistently individual participants changed or whether the difference is clinically important.

What is an ANCOVA result?

General: Analysis of covariance (ANCOVA) compares groups while accounting for one or more covariates included in the model. Its estimate can differ from a simple comparison of posted group summaries because it is model-based and may use information beyond the two displayed means. The registry identifies the sponsor's reported primary analysis as ANCOVA and gives a mean difference of -2.3. The record does not specify the covariates or other details of the model, so those should not be inferred.

The recalculated comparison is explicitly an unadjusted re-calculation from posted summary statistics. It is useful as a transparent check of the basic contrast, but it is not a reconstruction of the sponsor's ANCOVA. The close but nonidentical estimates do not imply a data error: the analyses are not the same calculation. The registry record does not flag a discrepancy between them.

How should a p-value be read?

General: A p-value describes 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, the probability that the result occurred by chance, or a measure of clinical importance. The recalculated unadjusted primary p-value is 0.096. It should be read alongside the estimate, interval, analysis method, and exploratory context rather than as a stand-alone verdict.

The registry also reports a p-value for the ANCOVA analysis, but the main distinction for interpretation is that this is the sponsor's model-based result while the recalculation is unadjusted. A p-value does not communicate the range of effects compatible with the data; an interval provides information about uncertainty and precision. Neither alone answers whether an effect matters to patients.

Why can over-dispersion matter?

General: For continuous outcomes, variation among participants affects how precisely a group mean or difference can be estimated. Larger variability generally makes it harder to distinguish a group contrast from sampling uncertainty when other features are held fixed. For count or event outcomes, over-dispersion means variation exceeds what a simple model assumes; that can make standard errors too small if not addressed. The primary outcome is continuous, and the record reports standard errors for the group means. It does not report an over-dispersion analysis or a need for one, so this general concept should not be taken as a claim about the trial's data.

What does LOCF mean in this analysis?

General: Last observation carried forward uses a participant's most recent recorded measurement as a substitute for a later missing measurement, according to the analysis definition. The primary outcome title specifies LOCF at Week 6. This identifies the stated approach to missing follow-up information; it does not show how many observations were missing or whether the assumptions behind carrying a prior value forward were appropriate. The registry record does not report those missingness counts or a sensitivity analysis for the primary outcome.

The primary efficacy population also has explicit entry requirements: non-remission at augmentation baseline, at least one randomized augmentation dose, and at least one post-randomization primary efficacy measurement. Such a defined analysis population can differ from the set of everyone enrolled. The results therefore need to be understood in the context of these eligibility conditions rather than generalized automatically to all enrolled participants.

11. Confidence Intervals and Precision

A confidence interval expresses uncertainty around an estimated group contrast under the analysis that produced it. General: A narrower interval indicates greater precision, while a wider interval leaves a broader range of values compatible with the data and method. An interval crossing zero includes both a contrast in one direction and no difference; it does not by itself prove that the groups are identical.

For the primary endpoint, the registry reports the sponsor's ANCOVA mean difference as -2.3, with a 90% confidence interval from -4.5 to -0.1. This is the interval associated with the sponsor's reported analysis. The recalculated unadjusted interval spans zero. Because the recalculation and ANCOVA are different analyses, their interval results should be interpreted as method-specific rather than combined into a single uncertainty statement.

The arm-level primary summaries report standard errors of 0.93 and 0.94. A standard error describes the precision of an estimated mean, not the spread of individual participant scores. It should not be read as a range containing most participants. The registry record does not provide arm-level confidence limits for those summaries.

Precision is also distinct from clinical importance. Even a relatively precise estimate would need a relevant threshold to determine whether its magnitude matters clinically; such a threshold is not reported in the registry record. Conversely, an interval that spans zero does not show that the true difference is exactly zero. The appropriate reading is that the posted data and analysis leave uncertainty about the size and direction of the group contrast, with the sponsor's ANCOVA and the unadjusted recalculation giving different interval 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 express the required number of participants per arm, not a total across both arms and not a total number of events. The listed planning 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 as the stated fraction of the observed effect. These are planning illustrations based on the observed result; the trial's original design assumptions and sample-size justification are not reported in the registry record.

The 80% power column describes a planning target under which a study would have an 80% chance of detecting the specified effect if that effect and the other assumptions held. The 90% power column uses a higher chance-of-detection target for the same specified effect, and therefore generally calls for more participants. Each row changes the effect size used for planning: a smaller fraction of the observed effect represents a smaller target contrast. Smaller effects are harder to distinguish from random variation, so they require more participants under the stated assumptions.

The figures are not observed power. They are not a measure of how likely the completed trial was to have detected its result, and they do not revise the actual study enrollment or analysis population. Because the observed effect is used as the starting point, the illustrations should be read as conditional scenarios rather than as a guarantee that a future study of those sizes would reproduce the same effect. The displayed curve is another way to consider the relationship between planning assumptions and sample size, not evidence about what the original investigators planned.

Power calculations also depend on assumptions about the outcome, variability, allocation, analysis, and target effect. The record specifies the assumptions used for these illustrations but does not provide the original design calculation. It is therefore important to keep the retrospective planning exercise separate from the Phase 2 trial design: the former is an illustration tied to the posted effect, while the latter's full sample-size rationale is not reported in the registry record.

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 = -2.200; pooled standard deviation σ = 7.509.
2. n = 2 × (1.96 + 0.842)² × 7.509² / -2.200² = 182.9, rounded up to 183 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 <- -2.2                      # observed difference in means
sigma <- 7.50887097419356                     # 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(905424)
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 = 183 (80% power), 245 (90% power)
 75% of observed effect: n per arm = 326 (80% power), 436 (90% power)
 50% of observed effect: n per arm = 732 (80% power), 980 (90% power)
power.t.test: 183.8 
Simulated power at n = 183 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 registry reports the sponsor's primary analysis as ANCOVA, with a mean difference of -2.3 and a 90% confidence interval from -4.5 to -0.1. The simple recalculation from the posted arm summaries gives a mean difference of -2.2 and an interval that spans zero. The sponsor's reported analysis and the recalculation are not the same procedure: ANCOVA is the named model-based analysis, while the recalculation is an unadjusted comparison of summary statistics.

The registry record does not flag a discrepancy between these results. The recalculation can help readers see how the posted group means relate to a basic contrast, but it should not be treated as a replacement for the sponsor's analysis. The sponsor's estimate, interval, and method describe the reported primary analysis. Differences between the intervals may reflect the methods and interval specifications; details of the ANCOVA covariates are not reported in the registry record.

The sponsor's interval excludes zero, whereas the unadjusted recalculated interval spans zero. That distinction is important when describing statistical uncertainty. It does not establish clinical importance, and the reported 90% interval should not be directly compared with an interval of an unstated confidence level as if both were identical in construction. The registry does not report enough model detail to reproduce the ANCOVA independently.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes or comparisons are tested. General: As the number of analyses increases, the chance of at least one apparently notable result can increase even if there is no true effect across the set. Adjustment procedures can control some forms of this problem, but they may change the interpretation of individual p-values and confidence intervals. The registry record lists 18 outcome measures and 18 analyses, including 17 posted secondary outcomes. It does not report whether a multiplicity strategy was specified or used.

The absence of a reported adjustment plan means that secondary p-values should be treated as unadjusted unless the record states otherwise. The registry reports individual analyses, but does not establish an overall family of hypotheses, a testing hierarchy, or a procedure for controlling error across endpoints. These details are not reported in the registry record. Thus, the secondary findings are informative descriptively but should not be read as a set of independent confirmatory tests.

Missing-data handling is partly specified. The primary endpoint and some secondary outcome titles identify LOCF. The record does not report the number or pattern of missing observations, the reasons for missingness, or sensitivity analyses testing alternative assumptions for the primary endpoint. It also does not specify missing-data methods for every secondary outcome in the details available. LOCF is a stated analysis method, not evidence that missing outcomes were absent.

The primary efficacy population is defined by non-remission at augmentation baseline, receipt of at least one dose of randomized augmentation treatment, and at least one post-randomization primary efficacy measurement. This is the population used for the stated primary analysis, and it is not automatically equivalent to all randomized or enrolled participants. The registry record does not report further analysis-population details beyond this definition, such as additional exclusions or a separate intention-to-treat analysis. The implications of those details cannot be assessed from the record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: rhabdomyolysis (1.2%).

The registry reports serious adverse events among the safety groups: 0 affected participants among 88 at risk in the antidepressant plus SPD489 group, and 1 among 85 at risk in the antidepressant plus placebo group. These are the posted figures for participants with serious adverse events, not a measure of all adverse experiences or their severity distribution. The record lists rhabdomyolysis at 1.2% among the top serious terms; the available summary does not identify the arm for that term.

Other adverse events were reported for 35 of 88 participants in the SPD489 group and 15 of 85 in the placebo group. These are listed separately from serious adverse events and should not be combined with them. The registry record does not report deaths in the available safety summary. It also does not provide enough detail to assess event timing, duration, recurrence, or a causal relationship to treatment.

Safety counts are descriptive and depend on the trial's observation and reporting procedures. They do not establish that an event was caused by an assigned treatment, and the posted figures alone do not support a comprehensive comparison of safety. Interpretation should remain specific to the groups, outcomes, and record details stated.

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

The posted results describe a randomized comparison of adding SPD489 or placebo to an antidepressant in a Phase 2 exploratory study. For the primary outcome, the analysis focused on non-remitters at augmentation baseline and assessed mean MADRS change at Week 6. The arm summaries show decreases in both groups, with a larger numerical decrease in the SPD489 group. The sponsor's ANCOVA estimate and the simple unadjusted recalculation are similar in direction but have different interval results.

This does not establish that SPD489 works or does not work. Statistical estimates describe group averages under particular methods and populations; they do not determine what happened to every participant. Nor does a difference in average scale change establish how important that difference is to patients. The registry record does not report a clinical-importance threshold for MADRS change.

The findings are limited to the recorded study conditions and time frame. The primary endpoint was assessed at six weeks, and the primary population excluded people who had already remitted at augmentation baseline as well as those who did not meet the stated treatment and measurement requirements. The results do not directly describe longer-term outcomes or other populations.

Secondary endpoints add descriptive evidence but involve multiple measures, and a multiplicity-adjustment plan is not reported. Missing-data methods are not fully described across outcomes. Safety figures likewise summarize recorded event counts and do not by themselves establish causation. Taken together, the registry record supports a careful account of the trial's estimates and their uncertainty, but it does not answer every question about clinical relevance, durability, generalizability, or comparative safety.

18. Why This Trial Matters Statistically

This trial illustrates why a treatment comparison cannot be reduced to whether one group mean is numerically lower. The primary endpoint is a change score on a defined scale, measured in a specified non-remitter population and analyzed at a specified time. Its interpretation depends on the scale's direction, the analysis method, and the uncertainty around the contrast. Those elements should be read together.

It also shows why reported model-based results and simple recalculations need to be distinguished. The sponsor reports an ANCOVA comparison, while the recalculation uses posted summaries without adjustment. Their estimates are close, but their intervals differ in whether they span zero. Reporting the method alongside the estimate helps readers understand that an apparent discrepancy can arise from different analyses rather than from a direct contradiction.

The record's secondary outcomes and planning illustrations raise two further statistical issues. Multiple outcomes create a need to consider multiplicity, especially when adjustment plans are not reported. Sample-size scenarios based on an observed effect can help illustrate how smaller targets demand more participants, but they are not evidence of the original design rationale or of the completed trial's observed power.

Finally, defining the analysis population and handling missing measurements matter. The primary results apply to non-remitters meeting the stated dose and measurement criteria, and LOCF is named for the primary endpoint. These details shape what the estimate represents. Keeping the question, population, method, and uncertainty explicit makes the findings more interpretable without extending them beyond what the registry record reports.

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 record describes the study as exploratory and Phase 2, so the posted results should be read as evidence from that study context rather than as a definitive answer to broader questions. The primary analysis concerns non-remitters who met the stated treatment and measurement conditions; it does not represent every enrolled participant or every person with major depressive disorder. The record does not report the original sample-size justification, which limits assessment of how the planned study size related to its intended primary question.

The primary outcome uses LOCF, but the registry record does not report missingness counts, reasons for missing measurements, or sensitivity analyses for alternative missing-data assumptions. The ANCOVA result is reported without details of the covariates or full model specification. This limits independent reconstruction of the sponsor's analysis. The simple recalculation is useful as an unadjusted comparison, but it cannot supply those missing model details.

Many secondary outcomes are listed, while a multiplicity-adjustment plan is not reported. Their individual results therefore need cautious interpretation, and the record does not establish which were intended as confirmatory. The six-week primary timeframe also limits what can be inferred about persistence of changes beyond the measured period. A clinically important threshold for MADRS change is not reported.

The safety summary provides counts for serious and other adverse events, but the available details do not describe timing, duration, attribution, or all aspects of event reporting. The listed serious term is not assigned to an arm in the summary. Together, these limitations restrict conclusions about clinical relevance, longer-term outcomes, missing-data sensitivity, and comprehensive safety.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Exploratory Study of SPD489 in Adults With Major Depressive Disorder (MDD) as Augmentation Therapy to an Antidepressant — NCT00905424.
• Registry record last updated 2021-06-14; status last verified 2021-06.
• Results first posted 2011-07-14.
• 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 record describes NCT00905424 as a completed, randomized, triple-masked, parallel-group, placebo-controlled Phase 2 exploratory study of SPD489 augmentation in adults with major depressive disorder. It lists enrollment as 246 and identifies Shire as the sponsor. The primary outcome was Week 6 change in MADRS total score among non-remitters at augmentation baseline, using LOCF.

The posted primary arm means were -7.1 for antidepressant plus SPD489 and -4.9 for antidepressant plus placebo. The recalculated unadjusted difference was -2.2, while the sponsor's ANCOVA analysis reported -2.3 with a 90% confidence interval from -4.5 to -0.1. The registry lists multiple secondary outcomes and safety summaries. The record does not report the original sample-size justification or a multiplicity-adjustment plan.