A Study for Participants With Major Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study for Participants With Major Depression (NCT00420004), a randomized phase 2 study. The primary outcome is Change From Baseline to Week 8 in the 17-item Hamilton Depression Rating Scale (HAMD-17) Total Score.

2. Clinical Question

The central statistical contrast is the difference in change between LY2216684 and placebo. HAMD-17 scores measure depressive symptom severity, with higher scores indicating greater severity. A more negative change represents a larger decrease from baseline, but the between-group contrast and its uncertainty are needed to assess how clearly the groups differed in this outcome. The record also identifies secondary measures covering depressive symptoms, anxiety, global clinical impression, health status, suicidal ideation, and aggression. The results describe measured outcomes during the trial; they do not, by themselves, establish how relevant a particular difference would be to an individual participant.

3. Trial Design

The registry identifies this as a phase 2, randomized, quadruple-masked, parallel-group trial. Its stated enrollment was 469, and its status is completed. The listed study dates are 2006-12 for the start and 2008-03 for primary completion. Eli Lilly and Company is identified as the sponsor.

Three interventions were included: LY2216684, placebo, and escitalopram. LY2216684 was administered orally once daily for 8 weeks, with flexible doses of 3, 6, 9, or 12 mg. Participants began at 3 mg/day for the first week, with later dosing based on tolerability. Escitalopram was administered orally once daily for 8 weeks at flexible doses of 10 or 20 mg, starting at 10 mg/day for the first week. Placebo was given as tablet and capsule equivalents to the two active treatments.

The primary outcome was change in HAMD-17 total score from baseline to Week 8. The reported arm summaries are least-squares means, rather than simple raw averages, and the stated primary analysis used mixed model repeated measures. The outcome population was narrower than all enrolled participants: it required randomization, at least one dose of double-blind study drug, and both baseline and at least one post-baseline HAMD-17 value.

The registry record does not report the original sample-size justification, the intended allocation ratio, or the trial’s full statistical analysis plan. It also does not report detailed protocol information on interim analyses. Those omissions mean the planning illustrations based on observed effects should not be read as the design assumptions used when this trial was planned.

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: A total of 701 participants were included in a screening/washout period (from 3 to 30 days in duration) prior to group assignments; 232 participants discontinued during this period, and 469 participants were randomized. There was an 8-week Double-blind Phase followed by a 1-week Discontinuation Phase after abrupt discontinuation of treatment.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants who received at least one dose of double-blind study drug and who had a baseline and at least one post-baseline HAMD-17 value.

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.

Population: All randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline to Week 8 in the 17-item Hamilton Depression Rating Scale (HAMD-17) Total Score, measured over Baseline, Week 8.

The primary endpoint was change from baseline to Week 8 in the 17-item Hamilton Depression Rating Scale total score. The HAMD-17 assesses depression severity and change during treatment. Its 17 items use either a 5-point scoring scale from 0 to 4 or a 3-point scale from 0 to 2; the total is the sum of the item scores and ranges from 0 to 52. Higher scores indicate greater symptom severity.

Change from baseline is useful because it describes how the measured score shifted over the treatment period, while a comparison between groups helps distinguish change occurring in one treatment group from change occurring in the comparator group. Negative changes in the posted arm estimates indicate decreases in score from baseline. The primary registry analysis used least-squares means estimated with mixed model repeated measures, adjusting for treatment, investigator, visit, baseline score, treatment-by-visit, and baseline score-by-visit.

The primary analysis population included randomized participants who received at least one dose of double-blind study drug and had a baseline and at least one post-baseline HAMD-17 value. It therefore does not represent every person enrolled. HAMD-17 is a symptom scale; its score change is one measured outcome and does not alone capture every aspect of depression or participants’ experience.

7. Results Posted in the Registry

• LY2216684: mean -13.20 versus -12.60 on control; difference -0.60 (95% CI -2.38 to 1.18), standardized effect d = -0.07; z-test p = 0.510.
• Escitalopram: mean -14.70 versus -12.60 on control; difference -2.10 (95% CI -4.76 to 0.56), standardized effect d = -0.25; z-test p = 0.122.

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 Baseline to Week 8 in Maier-Philipp Subscale of the 17-item Hamilton Depression Rating Scale (HAMD-17)

• 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:
  • LY2216684: mean -6.70 versus -6.20 on control; difference -0.50 (95% CI -1.37 to 0.37), standardized effect d = -0.12; z-test p = 0.260.
  • Escitalopram: mean -7.60 versus -6.20 on control; difference -1.40 (95% CI -2.69 to -0.11), standardized effect d = -0.35; z-test p = 0.033.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LY2216684 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Escitalopram than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Clinical Global Impression of Improvement Score at Week 8

• 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:
  • LY2216684: mean 2.20 versus 2.30 on control; difference -0.10 (95% CI -0.39 to 0.19), standardized effect d = -0.07; z-test p = 0.505.
  • Escitalopram: mean 2.00 versus 2.30 on control; difference -0.30 (95% CI -0.74 to 0.14), standardized effect d = -0.22; z-test p = 0.182.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LY2216684 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Escitalopram and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Hamilton Anxiety Rating Scale (HAMA) Total Score up to Week 8 Endpoint

• 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:
  • LY2216684: mean -9.70 versus -9.60 on control; difference -0.10 (95% CI -2.11 to 1.91), standardized effect d = -0.01; z-test p = 0.923.
  • Escitalopram: mean -11.60 versus -9.60 on control; difference -2.00 (95% CI -4.83 to 0.83), standardized effect d = -0.24; z-test p = 0.166.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between LY2216684 and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Escitalopram and the control. The interval gives the range of effects that remain compatible with the data.

The registry record lists 18 secondary outcome measures. These span symptom subscales, response and remission, clinician-rated improvement, anxiety, health status, depressive symptoms, suicidal ideation, and aggression. The detailed numeric secondary results include continuous outcomes assessed at or through Week 8, alongside their arm estimates and re-calculated comparisons.

Across the secondary comparisons with detailed numeric results, most re-calculated intervals include zero. One unadjusted comparison has an interval that excludes zero. Taken together, the pattern is not a uniform set of differences across measures. It should be interpreted alongside the primary outcome, whose sponsor analysis reports p = 0.494 and whose re-calculated interval includes zero, rather than as a replacement for the primary result.

Secondary outcomes can add context by examining different aspects of symptoms and functioning, but multiple outcomes create multiple opportunities for apparently notable results. The registry record does not report a multiplicity adjustment or a hierarchy specifying how secondary comparisons were to be interpreted. An isolated unadjusted result therefore carries less weight than a finding supported by a prespecified analysis strategy that accounts for the number of comparisons.

The secondary results also use outcome-specific populations and estimates; their group sizes are not identical across measures. Differences in the number of observations may affect precision and mean that results across endpoints are not directly interchangeable. These results can describe the recorded measures, but the available information does not justify treating them as a single combined measure of treatment effect.

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

• Change From Baseline to Week 8 in Maier-Philipp Subscale of the 17-item Hamilton Depression Rating Scale (HAMD-17)
• Response and Remission Rates
• Clinical Global Impression of Improvement Score at Week 8
• Change From Baseline in Hamilton Anxiety Rating Scale (HAMA) Total Score up to Week 8 Endpoint
• Change From Baseline on the 36-item Short-Form (SF-36) Health Status Survey Mental and Physical Components up to Week 8 Endpoint
• Change From Baseline in Quick Inventory of Depressive Symptomatology Total Score up to Week 8 Endpoint
• Change From Baseline in Beck Scale for Suicide Ideation up to Week 8 Endpoint
• Change From Baseline in Modified Overt Aggression (OAS-M) Scale up to Week 8 Endpoint

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant’s HAMD-17 total score can take a range of values, and the analysis summarizes change on that scale. The posted arm estimates are least-squares means. In general, a least-squares mean is a model-based estimate that accounts for covariates or design factors specified in the analysis, rather than a simple unadjusted average. The registry reports adjustment for treatment, investigator, visit, baseline score, treatment-by-visit, and baseline score-by-visit.

What does mixed model repeated measures mean?

In general, a repeated-measures model analyzes measurements collected at multiple visits while accounting for the fact that measurements from the same participant are related. The registry names mixed model repeated measures (MMRM) for the primary endpoint and identifies baseline and Week 8 as the timeframe. The full model specification, including any assumptions about the correlation structure or missing observations, is not reported in the registry record. It is therefore not possible to infer those details from the reported estimates alone.

The sponsor’s posted primary analysis is a mixed models analysis comparing LY2216684 and placebo, with p = 0.494. A p-value describes how compatible the observed data are with a specified null hypothesis under the analysis assumptions; it is not the probability that the null hypothesis is true, nor a measure of the size or importance of an effect. The record does not report a sponsor estimate or interval for this comparison in the posted analysis summary.

What is the summary-statistic re-calculation?

The re-calculated comparison is an unadjusted comparison based on the posted arm-level estimates and standard errors. It gives a mean difference of -0.6 for LY2216684 versus placebo, with an interval from -2.384 to 1.184 and an unadjusted p-value of 0.51. The re-calculation uses an approximate pooled standard deviation of 8.29, and the calculation is explicitly flagged as approximate. It does not reproduce the sponsor’s adjusted MMRM analysis and should be interpreted as a separate, simpler analysis of the posted summaries.

In general, over-dispersion means that outcomes vary more than a particular statistical model expects. Accounting for such variability can affect standard errors and the apparent precision of estimates. The primary outcome is continuous, and the registry reports standard errors for the arm-level least-squares means; it does not report an over-dispersion parameter or an analysis specifically addressing over-dispersion. It would not be appropriate to claim that over-dispersion was or was not a feature of this analysis.

How should the standard errors and intervals be read?

A standard error describes the estimated uncertainty in an estimate, rather than the spread of individual participant scores. For the primary arm summaries, the posted standard errors are 0.53 for LY2216684 and 0.74 for placebo. The interval for the re-calculated difference describes a range of values compatible with that summary-based estimate under its calculation assumptions. Because the interval includes zero, it is compatible with no between-group difference as well as with differences in either direction. The registry record does not state the confidence level for the re-calculated interval, so no confidence percentage is attached.

The re-calculated p-value is unadjusted: it does not account for testing multiple outcomes or comparisons. It is distinct from the sponsor’s p-value and does not replace the sponsor’s analysis. When several endpoints are examined, the chance of seeing at least one apparently unusual result can increase; general statistical practice therefore gives more weight to prespecified primary analyses and to secondary findings interpreted with an explicit multiplicity strategy. The registry record does not report such a strategy.

Which participants contributed to the primary estimate?

The stated primary analysis population consisted of randomized participants who received at least one dose of double-blind study drug and had a baseline and at least one post-baseline HAMD-17 value. This is an outcome-specific population definition; it should not be confused with the stated overall enrollment. The record does not describe the number of participants excluded from this population, the reasons for missing measurements, or the handling of missing data in sufficient detail to reconstruct the analysis.

The clinical interpretation of a change score also depends on the scale and the comparison being considered. A decrease on HAMD-17 indicates a lower measured symptom-severity score, but a within-group decrease alone does not show that the treatment caused that change. The controlled comparison is more relevant to the question of differences between groups, and its uncertainty must be considered alongside the point estimate.

11. Confidence Intervals and Precision

The primary re-calculated LY2216684-versus-placebo mean difference is -0.6 HAMD-17 units. Its interval runs from -2.384 to 1.184. Since zero lies within that interval, the posted summary statistics do not pin down a single direction for the between-group difference. The plausible values under this calculation include a greater decrease in the LY2216684 group and a difference in the opposite direction.

The arm-level least-squares means have standard errors of 0.53 for LY2216684 and 0.74 for placebo. These standard errors describe uncertainty in the estimated means, not the range of individual participants’ scores. The re-calculated interval is based on a summary-statistic approach and approximate pooled variability; it is not the interval from the sponsor’s MMRM analysis. A confidence level for this re-calculated interval is not reported in the registry record.

Precision is also distinct from clinical importance. A narrow interval would indicate a more tightly estimated contrast, but it would not establish that a difference matters to participants. Conversely, an interval spanning zero does not prove that the groups are identical. The registry record does not report a threshold for clinical importance, so the statistical interval cannot be translated into a judgment about meaningful benefit.

The sponsor’s reported p-value is 0.494, while the unadjusted re-calculated p-value is 0.51. Both are consistent with substantial uncertainty in the primary comparison, though they come from different analyses. No sponsor estimate or interval is reported in the analysis summary, limiting a direct comparison of precision between the two methods.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total event counts. They are illustrations based on the observed effect and assume 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 original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the number of participants per arm that the planning calculation estimates would be needed for an 80% chance of detecting the assumed effect at the stated significance level, if the assumptions hold. The 90% power column gives the corresponding planning number for a 90% chance. These are design-planning quantities, not descriptions of what happened in the completed trial.

The rows that use 100%, 75%, and 50% of the observed effect illustrate the effect-size assumption. Smaller effects are harder to distinguish from random variation, so the planning calculations require more participants per arm when the assumed effect is smaller. Estimates based on the observed effect are sensitive to that observed estimate and should not be treated as a guarantee of future results.

These figures are not observed power. They do not measure the probability that the completed trial’s result is correct, and they do not change the interpretation of the posted primary comparison. The original design’s planning assumptions are not reported in the registry record; the stated assumptions apply to these planning illustrations only.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.600; pooled standard deviation σ = 8.290.
2. n = 2 × (1.96 + 0.842)² × 8.290² / -0.600² = 2,996.7, rounded up to 2,997 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.6                      # observed difference in means
sigma <- 8.29000209786516                     # 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(420004)
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 = 2,997 (80% power), 4,012 (90% power)
 75% of observed effect: n per arm = 5,328 (80% power), 7,132 (90% power)
 50% of observed effect: n per arm = 11,987 (80% power), 16,047 (90% power)
power.t.test: 2997.7 
Simulated power at n = 2,997 per arm: 0.800 (20000 replicates)

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

14. Analyses Reported by the Sponsor

The sponsor reports a mixed models analysis for the primary LY2216684-versus-placebo comparison, with p = 0.494. This is the registry’s reported primary analysis result. The re-calculation from posted summary statistics gives a mean difference of -0.6, an interval from -2.384 to 1.184, and an unadjusted p-value of 0.51.

The re-calculation is not a reproduction of the sponsor’s MMRM. It uses arm-level estimates and standard errors, with approximate pooled variability, and does not use the full participant-level data or the sponsor’s listed model adjustments. The two p-values are close, but their methods differ; neither should be substituted for the other. The sponsor analysis remains the appropriate reference for the sponsor’s stated model, while the re-calculation is a transparent, simpler comparison of the posted summaries.

The registry analysis summary does not provide the sponsor’s estimated treatment contrast or interval. That limits comparison of the magnitude and precision of the sponsor model result with the re-calculated contrast. The record does not report enough detail to identify a specific reason for the difference between the two p-values, so no more specific explanation can be established.

15. Multiplicity, Missing Data and Analysis Populations

The registry record names a primary HAMD-17 endpoint and lists numerous secondary outcomes, including multiple symptom, global-impression, anxiety, and health-status measures. Testing several outcomes can increase the chance of an apparently notable result occurring by chance. A prespecified strategy can address this issue through an outcome hierarchy or statistical adjustment, but the multiplicity strategy is not reported in the registry record.

The re-calculated p-values for comparisons are unadjusted. They do not account for the number of endpoints or comparisons. The sponsor’s primary mixed models p-value is reported separately; the record does not specify whether or how it relates to a broader multiplicity plan. Accordingly, the secondary comparison with an interval excluding zero should be treated as exploratory evidence in this record, rather than as a stand-alone confirmation.

The primary outcome population is defined as randomized participants who received at least one dose of double-blind study drug and had a baseline and at least one post-baseline HAMD-17 value. The record does not report a full flow of participants into and out of that analysis population, reasons for unavailable measurements, or a detailed missing-data strategy. Although the primary analysis is identified as MMRM, the record does not state the assumptions used for missing observations or sensitivity analyses evaluating those assumptions.

These details matter because missing outcomes can affect estimates if the participants with observed measurements differ systematically from those without them. Mixed models can use available repeated measurements, but the name of a model alone does not establish that missingness has no effect on the result. The record also does not report the original sample-size justification or interim-analysis details. These omissions limit assessment of the design and the robustness of the reported analyses; they do not, by themselves, show that the analyses are invalid.

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: gastroenteritis (1.6%), malaria (1.6%), suicide attempt (1.6%), near drowning (0.7%), anaemia (0.4%).

During the double-blind phase, the registry reports serious adverse events in 2 of 269 participants at risk in the LY2216684 group, 1 of 138 in the placebo group, and 2 of 62 in the escitalopram group. During the discontinuation phase, serious adverse events were reported in 0 of 269, 1 of 138, and 0 of 62, respectively. The record lists serious-event terms including gastroenteritis, malaria, and suicide attempt at 1.6% each, and near drowning at 0.7%; anaemia and depression are each listed at 0.4%.

These are counts recorded in the registry, not estimates of comparative risk adjusted for participant characteristics or time at risk. The numbers at risk differ by group, and the record does not provide enough detail to establish why. Serious adverse events are distinct from the other adverse-event counts also reported. The event counts alone do not establish that a study drug caused an event or that an event was unrelated to treatment. Death counts are not reported in the registry record.

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

The primary result describes change in HAMD-17 score through Week 8 among the specified analysis population. The adjusted arm estimates decreased in all three groups. The sponsor’s mixed models analysis reports p = 0.494 for LY2216684 versus placebo, while the summary-statistic re-calculation estimates a difference of -0.6 with an interval spanning zero. The data therefore do not precisely distinguish the primary group contrast in the re-calculation.

This does not establish that the treatments are identical. An interval containing zero is compatible with no difference, but it can also include differences in either direction. Nor does the direction of the point estimate establish that LY2216684 produced a meaningful change relative to placebo. The registry record does not state a clinical-importance threshold, and no such threshold can be inferred from the p-value or interval.

The result is specific to the recorded trial, its participants, treatment period, outcome definition, and analysis population. It does not establish results beyond the 8-week treatment period, for people outside the enrolled population, or for outcomes not measured. Secondary outcomes provide additional measurements, but their unadjusted comparisons must be considered in light of multiplicity and the missing statistical plan details.

Adverse-event counts describe events recorded in the trial groups. They do not, on their own, establish causation or provide a complete comparison of the treatments’ safety. Overall, the registry results support a cautious description of the measured data and their uncertainty, rather than a broad conclusion about whether a treatment works or does not work.

18. Why This Trial Matters Statistically

This trial illustrates why a treatment-group mean change is not the same as a treatment effect. The HAMD-17 scores decreased in the LY2216684 and placebo groups, but the statistical question is whether their changes differed. The relevant primary contrast is modest in the posted summary, and its re-calculated interval includes zero. Reading only the direction of each arm’s change would overlook the controlled comparison and its uncertainty.

It also demonstrates the distinction between a sponsor’s model-based analysis and a summary-statistic re-calculation. The sponsor reports an MMRM p-value, while the simple re-calculation uses arm-level estimates and approximate pooled variability. These analyses answer related but not identical statistical questions. Reporting both transparently helps readers understand which result is the sponsor’s stated analysis and which is a secondary reconstruction from summaries.

The trial’s secondary outcomes show why multiplicity matters. When many endpoints are examined, an isolated result can be harder to interpret without a reported adjustment plan or outcome hierarchy. The missing details about multiplicity and missing-data handling also demonstrate how registry reporting affects a reader’s ability to assess robustness, not just the point estimates themselves.

Finally, planning calculations based on an observed effect are illustrations rather than evidence about a completed trial’s power. Separating planning quantities from observed findings prevents a common misreading of sample-size analyses. These statistical distinctions are useful when interpreting clinical trial records beyond this particular study.

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 registry record does not report the original sample-size calculation or the assumptions used to plan enrollment. Its planning illustrations use the observed effect and cannot explain the design choices made for the trial. The record also does not report a complete statistical analysis plan, a multiplicity strategy, or details of interim analyses.

The sponsor’s primary analysis is identified as MMRM and its adjustment terms are listed, but the full model specification is not available in the registry record. Details on the handling of missing outcomes, assumptions about missingness, and sensitivity analyses are also not reported. The stated primary analysis population requires at least one post-baseline HAMD-17 value, so it is not simply all enrolled participants.

The re-calculated primary comparison is based on posted summary statistics and approximate pooled variability. It is unadjusted and is not a substitute for the sponsor’s participant-level model. The registry analysis summary does not give the sponsor’s primary effect estimate or interval, and the confidence level for the re-calculated interval is not stated.

The record includes many secondary outcomes, but no reported multiplicity approach to guide their interpretation. It also gives adverse-event counts without enough detail to establish causality or fully compare exposure and event timing. Finally, the primary endpoint covers baseline to Week 8; the posted primary result does not establish longer-term outcomes.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study for Participants With Major Depression — NCT00420004.
• Registry record last updated 2018-04-27; status last verified 2018-03.
• Results first posted 2018-04-27.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT00420004 is a completed phase 2 randomized, quadruple-masked, parallel-group study sponsored by Eli Lilly and Company. It enrolled 469 participants with major depressive disorder and compared LY2216684, placebo, and escitalopram. The listed start and primary completion dates are 2006-12 and 2008-03.

The primary outcome was change from baseline to Week 8 in HAMD-17 total score. The sponsor reports a mixed models comparison of LY2216684 with placebo with p = 0.494. A summary-statistic re-calculation gives a mean difference of -0.6, with an interval from -2.384 to 1.184 and an unadjusted p-value of 0.51. The record also lists secondary outcomes and serious adverse-event counts. Important design and analysis details, including the original sample-size justification, multiplicity strategy, and missing-data approach, are not reported in the registry record.