Study of Neural Responses Induced by Antidepressant Effects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Neural Responses Induced by Antidepressant Effects (NCT02674529), a randomized phase 2, phase 3 study. The primary outcome is Change in Montgomery-Åsberg Depression Rating Scale (MADRS) Scores.

2. Clinical Question

3. Trial Design

The study, NCT02674529, is titled “Study of Neural Responses Induced by Antidepressant Effects”; its official title refers to simulated antidepressants. It is recorded as a randomized, triple-masked, parallel trial, with phase listed as PHASE2 and PHASE3. The condition is major depressive disorder. Enrollment was 60, and the registry status is completed. The record gives a start date of 2016-09 and a primary completion date of 2021-05.

The randomized comparison was between an antidepressant-treatment arm and a placebo arm. The treatment description specifies escitalopram over an 8-week period, beginning with 10mg for the first week. The placebo description specifies a placebo pill over an 8-week period. The listed interventions also include a real-time neurofeedback fMRI task before and after the randomized trial. The record’s brief summary describes an aim to examine neural changes associated with fast-acting antidepressant treatments and develop imaging-based biomarkers of treatment response. That stated aim does not establish that the study tested a fast-acting antidepressant as its randomized intervention.

The primary outcome was change in MADRS scores at baseline and week 8. The registry reports arm-level means and standard deviations, along with a sponsor-reported regression analysis. It says 46 participants were analyzed for the primary outcome and 8 were lost to follow up. The arm-level outcome records show 22 in the antidepressant-treatment arm and 24 in the placebo arm. The registry does not describe the reasons for loss to follow up, the handling of missing values, the details of the randomization process, or the original sample-size justification.

QIDS score change and neural responses during the sham neurofeedback fMRI task are listed as secondary outcomes. The posted secondary results include QIDS and another outcome record, but the record does not provide a numerical result for neural responses during the sham task. Details of the analysis population, protocol deviations, and any interim analyses are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: Six participants were excluded after enrollment for the following reasons: leaving the study site without informing study team, asking for extra-time to consider the study, feeling uncomfortable in the fMRI and asking to exit the scanner, testing positive for pregnancy tests, feeling indecisive, and not responding to contact attempts.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Scores can range from 0-60, with higher scores meaning greater the depression severity. The greater the change in MADRS scores from baseline, the better the outcome. 46 participants were analyzed, with 8 participants lost to follow up.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Change in Montgomery-Åsberg Depression Rating Scale (MADRS) Scores, measured over baseline and week 8.

The primary endpoint is change in Montgomery-Åsberg Depression Rating Scale scores from baseline to week 8. MADRS is a ten-item questionnaire used by psychiatrists to measure the severity of depressive episodes in people with mood disorders. The registry description says that scores range from 0-60 and that higher scores mean greater depression severity. It also describes a greater change from baseline as a better outcome. The outcome is recorded as a mean on a score scale, with standard deviation as the measure of spread.

Change scores are intended to describe how the score at follow-up differs from the baseline score. Comparing the average change between randomized arms can summarize the contrast associated with assignment to the treatment or placebo group. In general, using a recognized symptom scale gives researchers a structured way to quantify a clinical outcome, while the spread of scores indicates how much participants’ observed changes vary. These are general statistical explanations; the registry does not report the full scoring or analysis protocol.

The record notes that MADRS was designed as an adjunct to the Hamilton Rating Scale for Depression and was intended to be more sensitive to changes associated with antidepressants and other forms of treatment. It also states that scores on the two measures are highly correlated. That background explains the scale’s stated purpose, but it does not determine how precisely the trial estimated the treatment contrast. Precision depends on the observed variability and the uncertainty around the comparison, considered separately from the endpoint’s clinical rationale.

7. Results Posted in the Registry

• Changes in MADRS Scores From Baseline to Week 8 Within the Antidepressant Treatment Arm: mean 11.00 versus 12.00 on control; difference -1.00 (95% CI -6.21 to 4.21), standardized effect d = -0.11; z-test p = 0.707.

8. Secondary Efficacy Endpoints

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

Change in Quick Inventory of Depressive Symptomatology (QIDS) Scores

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Changes in QIDS Scores From Baseline to Week 8 Within the Antidepressant Treatment Arm: mean 6.82 versus 5.54 on control; difference 1.28 (95% CI -2.05 to 4.61), standardized effect d = 0.22; z-test p = 0.451.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Changes in QIDS Scores From Baseline to Week 8 Within the Antidepressant Treatment Arm and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists change in QIDS scores and neural responses during the sham neurofeedback fMRI task as secondary outcomes. The posted secondary results include a QIDS outcome and one other analysis record, but the numerical result for neural responses during the sham task is not reported in the registry record. This limits how completely the secondary outcome pattern can be assessed.

At a broad level, the posted secondary results do not provide a clear, independent confirmation of the primary MADRS result. The QIDS comparison is another symptom-scale outcome, so it may offer related information, but it is not a substitute for the prespecified primary endpoint. The registry’s limited reporting also means that the secondary outcomes cannot be treated as a comprehensive account of the study’s neural-response aims. The brief summary describes development of imaging-based biomarkers as an aim; an aim is not a posted biomarker finding.

Secondary analyses generally carry less weight than a prespecified primary analysis because there may be multiple outcomes or analyses, and the chance of a nominally notable result can increase when many comparisons are examined. This is a general statistical concern, not a claim that any particular adjustment was or was not performed in this trial. The registry does not report a multiplicity strategy, a hierarchy among secondary outcomes, or a complete statistical analysis plan. Unadjusted secondary results should therefore be interpreted as exploratory or supportive evidence rather than as stand-alone proof.

The secondary information should be considered alongside the primary result, not used to overturn it selectively. Different endpoints can capture different aspects of an outcome, and their estimates can vary. Without a fully reported analysis plan and complete numerical results for the listed secondary outcomes, the record does not allow a strong conclusion about consistency across all secondary measures.

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

• Change in Quick Inventory of Depressive Symptomatology (QIDS) Scores
• Neural Responses During the Sham Neurofeedback fMRI Task.

9. Statistical Interpretation

10. Statistical Methods Explained

For a continuous endpoint such as a rating-scale change, the mean describes the average observed value in an arm. The standard deviation describes variation among observed values around that mean. They answer different questions: an arm mean describes its center, while a standard deviation describes how dispersed the individual scores are. Neither measure alone gives the uncertainty in the difference between arms.

What does the mean difference represent?

The recalculated comparison subtracts the placebo-arm mean from the antidepressant-treatment-arm mean, following the direction used in the posted comparison. The resulting mean difference is -1.0. A negative value therefore indicates that the posted change-score mean is lower in the antidepressant-treatment arm than in the placebo arm under this coding. This is an unadjusted comparison calculated from the posted summary statistics, not the sponsor’s regression estimate. The registry record flags a discrepancy, so the sponsor’s own analysis takes precedence.

A mean difference is expressed in the endpoint’s units, which are score points in this case. It preserves the scale of the measurement and can be read alongside the arm means. The estimate is a single summary of the observed contrast; it does not say that every participant had a change of that size, nor does it describe the range of individual responses.

What does a confidence interval add?

In general, a confidence interval gives a range of values compatible with the data and the statistical procedure used. It conveys uncertainty around an estimate rather than only reporting the estimate itself. The recalculated interval is -6.207 to 4.207. Because it spans zero, values corresponding to no average difference remain compatible with this unadjusted calculation, alongside values on either side of zero. The interval’s width also indicates that the summary data do not pin down the contrast narrowly. The sponsor’s regression analysis has no confidence limits reported in the registry record.

A confidence interval is not a guarantee that the true effect lies within its limits, and it does not assign equal probability to all values in the interval. In general, its interpretation depends on the assumptions of the analysis and the way the data were collected and analyzed. The registry record does not provide enough detail to assess all such assumptions for the sponsor’s regression model.

How should a p-value be read?

In general, a p-value describes how unusual data at least as incompatible with a specified null model would be, if that model and the analysis assumptions were correct. It is not the probability that the null hypothesis is true, nor is it a measure of the size or importance of an effect. The recalculated unadjusted p-value is 0.707. The sponsor’s regression analysis reports p = 0.58. These values do not provide evidence of a statistically clear primary contrast under their respective analyses, but they do not prove equivalence or rule out meaningful differences.

The record also lists a correlation result of r = 0.02 and p = 0.05. It does not provide enough context to identify the role of that correlation in the primary treatment comparison. It should not be substituted for the reported regression result or treated as a second primary test without supporting information.

Why distinguish the recalculation from the sponsor analysis?

The recalculation is an unadjusted comparison derived from posted group summaries. It does not reproduce a model that may account for covariates or other features of the trial analysis. The registry’s sponsor-reported method is linear regression, with a regression coefficient of -1.299 and p = 0.58. Since the record flags a discrepancy, the sponsor’s analysis takes precedence; the recalculation is useful as a transparent descriptive check, not as a replacement for the sponsor’s result.

The record does not report the regression model’s covariates, coding details, handling of missing observations, or diagnostic checks. Those omissions prevent a fuller evaluation of how the reported coefficient was obtained. Randomization is recorded, but the registry record does not describe allocation procedures or explain the masking roles. Such details matter when assessing whether the analysis and study conduct support a causal interpretation.

What do standard deviations and sample sizes contribute?

The posted arm records give a standard deviation of 9.0 for each arm. In general, greater outcome variability makes an average contrast harder to estimate precisely when other features are held constant. The number analyzed in each arm also affects precision. These quantities are part of why an estimate must be read together with its confidence interval rather than in isolation. The registry says 46 participants were analyzed for the primary outcome and 8 were lost to follow up, but does not report the missing-data method or the reasons for loss.

The sample-size illustrations use the observed effect as a planning reference and apply stated assumptions. They are not evidence about the trial’s original design calculation. The record does not report the original sample-size justification, so it is not possible to determine from this registry entry whether the trial was planned around a particular target effect or precision.

All of these explanations of means, uncertainty, and testing are general statistical guidance. They do not add a new analysis of the trial. The quantitative interpretation remains anchored to the posted arm summaries and sponsor analysis, with the discrepancy and missing analysis details kept in view.

11. Confidence Intervals and Precision

The unadjusted mean difference for the primary outcome is -1.0, with a confidence interval from -6.207 to 4.207. This interval is wide enough to include zero and values on either side of it. As a result, the recalculated estimate does not locate the average contrast precisely. It is consistent with a range of possible contrasts, not just the single point estimate.

In general, a confidence interval communicates sampling uncertainty under the method used; it does not include every source of uncertainty about trial design, outcome measurement, or missing data. The arm standard deviations are 9.0 in both groups, and the primary outcome was analyzed for 22 in the antidepressant-treatment arm and 24 in the placebo arm. The registry also reports 8 lost to follow up. Without the missing-data method and further details of the analysis population, the interval cannot resolve concerns about how unavailable outcomes may have affected the estimate.

The sponsor’s regression coefficient is -1.299, with p = 0.58, but the sponsor confidence interval is not reported in the registry record. Therefore, the sponsor result cannot be assessed for precision using a reported interval. The record flags a discrepancy, and the sponsor’s own analysis takes precedence over the simple unadjusted recalculation. The recalculated interval remains informative as a descriptive, unadjusted assessment, but it should not be presented as the sponsor’s model-based uncertainty estimate.

Precision is distinct from whether a result crosses a conventional significance threshold. A non-significant test does not by itself show that the effect is negligible, and an estimate near zero does not establish equivalence. The registry record does not report an equivalence margin or a precision target. The data therefore support a cautious description of an uncertain primary contrast rather than a definitive conclusion about the size of any underlying difference.

12. Planning a Similar Trial: Sample Size

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

The sample-size illustrations are planning calculations based on fractions of the observed effect, not the trial’s original design assumptions. They 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 record’s original sample-size justification and original design assumptions are not reported in the registry record.

Each planning figure is expressed as participants per arm, not as total participants and not as total events. The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the assumed effect under the stated assumptions, if that effect and the assumptions were correct. The 90% power column gives the corresponding planning number for a 90% chance. These are design illustrations, not guarantees about a future study’s results.

The rows describe planning for 100%, 75%, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so the planning calculations require more participants per arm as the target effect gets smaller. This is a general relationship under the stated assumptions; it does not establish what effect size a future study should target.

These figures are not observed power. They are not a retrospective measure of the probability that the completed trial detected an effect. The observed trial result and its uncertainty should instead be read from the estimated contrast, confidence interval, and sponsor-reported analysis. The planning calculations also do not substitute for a justification based on a prespecified clinically meaningful effect or a complete design protocol, neither of which is 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 = -1.000; pooled standard deviation σ = 9.000.
2. n = 2 × (1.96 + 0.842)² × 9.000² / -1.000² = 1,271.5, rounded up to 1,272 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 <- -1                      # observed difference in means
sigma <- 9                     # 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(2674529)
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 = 1,272 (80% power), 1,703 (90% power)
 75% of observed effect: n per arm = 2,261 (80% power), 3,027 (90% power)
 50% of observed effect: n per arm = 5,087 (80% power), 6,809 (90% power)
power.t.test: 1272.5 
Simulated power at n = 1,272 per arm: 0.799 (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 posted primary analysis is a linear regression with a reported coefficient of -1.299 and p = 0.58. The registry record does not report confidence limits for that coefficient. It also lists a correlation result, r = 0.02 and p = 0.05, but does not explain enough about the correlation’s purpose or coding to connect it confidently to the primary treatment comparison.

The unadjusted recalculation from the posted group summaries gives a mean difference of -1.0, a confidence interval from -6.207 to 4.207, and p = 0.707. These are not the same analysis as the sponsor’s regression. They are a simple comparison of posted summaries and do not reproduce any model adjustment. The registry record explicitly flags a discrepancy between the recalculation and sponsor analysis. Accordingly, the sponsor’s own analysis takes precedence, while the recalculation is best treated as a descriptive check.

The two estimates are numerically different, and the record does not explain why. Model specification, coding, data handling, or other analysis details could matter in general, but the registry does not provide a basis to identify the reason in this case. No explanation should be inferred. The absence of a reported sponsor confidence interval also limits assessment of the sponsor estimate’s precision. The sponsor p-value does not establish that the treatment arms are identical; it indicates that the reported regression test did not find a statistically clear difference under that analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary endpoint and lists secondary outcomes, but it does not report a complete multiplicity strategy. Multiplicity arises when several outcomes, comparisons, or analyses are considered: even if each test is interpreted separately, the chance of at least one apparently notable finding can increase as the number of tests grows. This is a general statistical issue. The record does not state whether tests were adjusted, ordered in a hierarchy, or designated as exploratory.

The primary outcome record says that 46 participants were analyzed and 8 were lost to follow up. The arm records list 22 analyzed in the antidepressant-treatment arm and 24 in the placebo arm. The registry does not give reasons for missing follow-up, explain whether those participants were excluded from the analysis, or describe imputation or other missing-data procedures. Consequently, the potential influence of missing outcomes cannot be evaluated from the record alone. The reported results should not be assumed to use an intention-to-treat population or any other specific analysis population unless the record says so.

The registry also does not report the original sample-size justification, planned interim analyses, or a full prespecified statistical analysis plan. These details are important for assessing whether analyses and decisions were set before results were known. Their absence does not establish that such procedures were not used; it means they are not reported in the registry record.

The sponsor’s regression result and the recalculated unadjusted contrast differ, and the record flags that discrepancy without explaining its source. The sponsor’s analysis takes precedence, but the missing model details limit independent evaluation. The listed correlation result also lacks sufficient context for interpretation. Taken together, incomplete reporting of missing-data handling, analysis populations, multiplicity, and model specification constrains how far the registry results can support claims beyond the posted estimates.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports serious adverse events separately from other adverse events. In the antidepressant-treatment group, 25 were at risk for serious adverse events and 0 were affected; in the placebo group, 29 were at risk and 0 were affected. Deaths are recorded as 0 in each group. The serious-event terms list is empty, so no specific serious adverse event term is reported there.

For other adverse events, the record says 22 of 25 at risk were affected in the antidepressant-treatment group and 18 of 29 in the placebo group. These are registry counts, not an adjusted comparison. The record does not provide event descriptions, severity details, or a statistical comparison for these figures. The serious-event and other-event categories should not be conflated: the absence of reported serious events does not mean there were no other adverse events.

These figures describe the safety information entered in the registry for the recorded groups. They do not establish that the groups have identical safety profiles, and the record does not provide enough detail to assess event timing, causality judgments, or completeness of reporting. Safety interpretation should remain limited to the posted counts.

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

The primary outcome is an average change in a symptom rating scale over the study’s baseline-to-week-8 timeframe. The posted arm means and unadjusted recalculation describe a small numerical separation between the averages, while the recalculated confidence interval includes zero and extends in both directions. The sponsor’s regression analysis reports p = 0.58 and takes precedence because the registry flags a discrepancy. Together, the available primary results do not establish a statistically clear difference in the sponsor’s analysis, but they also do not prove that the treatments have identical effects.

A result that is not statistically clear can reflect substantial uncertainty, a modest underlying contrast, or a combination of factors. The estimate should therefore not be converted into a categorical claim that the intervention works or does not work. The registry record does not report a prespecified threshold for a clinically important difference, so the practical importance of the estimated contrast cannot be determined from statistical significance alone.

The findings apply to the recorded trial, its randomized comparison, and its measured outcomes. The record describes escitalopram and placebo over an 8-week period, with a neurofeedback fMRI task also listed among the interventions. The primary numerical outcome is MADRS change. It does not directly establish a result for the neural-response or biomarker aims described in the study summary. A listed research aim should not be treated as a demonstrated finding.

Interpretation is further limited by missing follow-up information and the absence of a reported missing-data method, full analysis plan, original sample-size justification, and sponsor confidence interval. The record also does not explain the discrepancy between the sponsor analysis and the unadjusted recalculation. These limits affect how precisely and completely the result can be understood. The evidence supports a careful account of the posted estimates and their uncertainty, not a broader conclusion beyond the outcomes and analyses reported.

18. Why This Trial Matters Statistically

This trial illustrates why a point estimate, uncertainty interval, and analysis method need to be read together. The primary mean difference alone gives a compact summary of the observed arm contrast, but its interval shows that the unadjusted estimate remains compatible with a range of values. Reporting the interval alongside the estimate makes the uncertainty visible rather than encouraging an overly definite reading of a single number.

The record also highlights the distinction between a summary-statistic recalculation and a sponsor-reported model. The unadjusted comparison is transparent and can be checked against the posted arm means, but it is not interchangeable with linear regression. Because the registry flags a discrepancy, the sponsor’s own analysis takes precedence. At the same time, missing regression confidence limits and model details make it harder to evaluate that analysis independently.

Finally, this record shows why registry completeness matters for statistical interpretation. The primary outcome analysis population is described in part, but missing-data handling, multiplicity planning, the original sample-size rationale, and some secondary numerical results are not reported. Those omissions do not prove that the analyses were inappropriate; they limit what can be concluded from the registry entry. Clear reporting of estimates, precision, analysis populations, and planned methods helps readers distinguish a result that is uncertain from one that has been established with precision.

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 trial constrain interpretation. The primary analysis reports 46 participants analyzed and 8 lost to follow up, but the registry does not explain why follow-up was lost or how missing outcomes were handled. The record also does not say whether the reported analysis follows an intention-to-treat approach or another defined population. These gaps make it difficult to assess possible effects of missing data on the estimate.

The registry flags a discrepancy between the simple unadjusted recalculation and the sponsor-reported regression analysis, but gives no explanation. The sponsor analysis takes precedence, yet its confidence limits, covariates, coding, and model diagnostics are not reported. The accompanying correlation result has no adequate context for interpretation as part of the primary comparison. As a result, independent evaluation of the sponsor model is limited.

The original sample-size justification, original design assumptions, multiplicity plan, interim-analysis information, and full statistical analysis plan are not reported in the registry record. The planning illustrations based on the observed effect cannot fill that gap. The record lists neural responses during the sham neurofeedback fMRI task as a secondary outcome but does not report a numerical result for that outcome. The study summary describes an imaging-biomarker aim, but the available record does not establish a biomarker finding.

These limitations concern reporting and interpretability; they do not establish that the trial procedures were deficient. They mean the posted record supports a narrower reading than a complete protocol and analysis report would allow.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study of Neural Responses Induced by Antidepressant Effects — NCT02674529.
• Registry record last updated 2022-11-03; status last verified 2022-10.
• Results first posted 2022-11-03.
• 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

NCT02674529 is a completed, randomized, triple-masked, parallel study of escitalopram versus placebo in major depressive disorder, with phase listed as PHASE2 and PHASE3. The primary endpoint was change in MADRS scores from baseline to week 8. The registry reports arm means of 11.0 and 12.0, and an unadjusted recalculated mean difference of -1.0 with a confidence interval from -6.207 to 4.207. The sponsor’s linear regression result is a coefficient of -1.299 with p = 0.58; because the record flags a discrepancy, the sponsor analysis takes precedence.

The primary result does not establish a statistically clear difference in the sponsor analysis, and the unadjusted interval indicates substantial uncertainty around the comparison. The record reports 46 participants analyzed and 8 lost to follow up, but does not describe missing-data handling, the original sample-size justification, or a complete analysis plan. Secondary outcomes include QIDS change and neural responses during a sham neurofeedback fMRI task, with incomplete numerical reporting for the latter.