Riluzole Augmentation in Treatment-refractory Obsessive-compulsive Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Riluzole Augmentation in Treatment-refractory Obsessive-compulsive Disorder (NCT00523718), a randomized phase 2 study. The primary outcome is Partial Responders by Yale-Brown Obsessive-Compulsive Scale (Y-BOCS).

2. Clinical Question

The trial addressed whether adding riluzole was associated with a different number of participants reaching this symptom-improvement threshold than adding placebo. It was a comparison of augmentation strategies: participants continued their existing medication regimen while receiving the assigned study treatment. The primary endpoint was a count of participants meeting a threshold, rather than an average change in Y-BOCS score. The record reports five partial responders among 19 participants in the riluzole group and two among 18 in the placebo group. Those observed counts describe the participants with posted outcome data; they do not by themselves establish how precisely the groups differ or whether any difference is attributable to assignment.

3. Trial Design

NCT00523718, titled “Riluzole Augmentation in Treatment-refractory Obsessive-compulsive Disorder,” is recorded as a completed phase 2 trial sponsored by Yale University. The registry lists randomized allocation, quadruple masking, and a parallel-group model. The intervention groups were riluzole and placebo, each added to participants’ medication regimens. The registry record does not specify the identities of all parties covered by the quadruple-masking label, so the precise masking arrangements are not reported in the registry record.

The recorded enrollment was 40. The primary endpoint’s posted counts include 19 participants in the riluzole group and 18 in the placebo group. These outcome counts need not be treated as the same quantity as overall enrollment. The reasons for any difference between the enrollment figure and the posted endpoint denominators, and the handling of participants without outcome data, are not reported in the registry record.

The recorded study start was 2006-09, and primary completion was 2015-08. The primary outcome timeframe was 14 weeks. The registry describes the study population as having obsessive-compulsive disorder; its official title also identifies serotonin reuptake inhibitor-refractory obsessive-compulsive disorder and depression. Further eligibility criteria, the planned analysis population, and the original statistical design assumptions are not reported in the registry record. The recorded phase and design labels describe the trial’s classification and structure, but do not establish details about the analysis that was conducted.

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: Forty subjects with treatment-refractory OCD were consented; 1 dropped out after consent, and one was excluded after it was revealed they were not taking an SSRI. These individuals were not randomized.

Overall Study

Reasons for not completing this period:

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: One individual was excluded from analysis (they were in the riluzole group) before unblinding due to a protocol violation.

6. Primary Endpoint

The registry lists the primary outcome as Partial Responders by Yale-Brown Obsessive-Compulsive Scale (Y-BOCS), measured over 14 weeks.

The primary endpoint counted participants classified as partial responders by the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) at 14 weeks. The Y-BOCS is described in the registry as a clinician-rated, 10-item scale of obsessive-compulsive disorder symptom severity. Each item ranges from 0, indicating no symptoms, to 4, indicating extreme symptoms; the total possible score ranges from 0 to 40. The record gives interpretive bands from sub-clinical through extreme severity.

For this outcome, improvement was defined a priori as a 25% improvement from baseline. Thus the endpoint reduces the scale information to whether a participant met the specified improvement threshold. The posted result is a count of participants meeting that definition in each group, not the average change in total Y-BOCS score. The registry record does not report the underlying baseline and follow-up Y-BOCS scores or a separate continuous-score comparison.

General statistical context: a threshold-based endpoint can be easier to interpret as a responder count, but it does not show how far above or below the threshold individual participants were. Participants with different degrees of score change can be placed in the same category, while small differences around the cutoff can lead to different classifications. The endpoint therefore addresses the likelihood of meeting the registry’s stated response definition, rather than every aspect of symptom change.

7. Results Posted in the Registry

• Riluzole: 26.3% (5/19) versus 11.1% (2/18) on control; risk difference 15.2 percentage points (95% CI -9.3 to 39.8), risk ratio 2.37 (0.52–10.70); unadjusted two-proportion z-test p = 0.238.

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.

Average Hamilton Depression Inventory (HAM-D)

• 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:
  • Riluzole: mean 11.94 versus 12.53 on control; difference -0.59 (95% CI -4.65 to 3.47), standardized effect d = -0.09; z-test p = 0.776.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Riluzole and the control. The interval gives the range of effects that remain compatible with the data.

Average Hamilton Anxiety Inventory (HAM-A)

• 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:
  • Riluzole: mean 14.18 versus 14.26 on control; difference -0.09 (95% CI -4.27 to 4.10), standardized effect d = -0.01; z-test p = 0.968.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Riluzole and the control. The interval gives the range of effects that remain compatible with the data.

Clinical Global Impression (CGI) - Severity of Illness Item

• 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:
  • Riluzole: mean 3.94 versus 3.95 on control; difference -0.01 (95% CI -0.69 to 0.67), standardized effect d = -0.01; z-test p = 0.977.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Riluzole and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover average HAM-D, average HAM-A, and the CGI severity-of-illness item at 14 weeks. Across these measures, the unadjusted comparisons show estimates close to zero, and each reported recalculated confidence interval spans zero. The overall pattern does not supply a precise contrast that changes the cautious reading of the primary responder result. The registry lists three secondary outcomes and does not report sponsor analyses for them.

General statistical context: secondary outcomes can provide additional views of a trial, but they do not automatically confirm or overturn the primary result. Their interpretation depends on whether they were prespecified, how they were analyzed, and whether the analysis considered the number of outcomes examined. The registry record does not report a multiplicity adjustment for these secondary comparisons. Without such an adjustment or a stated hierarchy of hypotheses, unadjusted secondary p-values, where calculated, carry less weight as confirmatory evidence because examining multiple outcomes can increase the chance of apparently notable findings by chance.

The secondary endpoints also measure different constructs from the primary Y-BOCS responder definition. Depression, anxiety, and global severity scores are not interchangeable with a threshold-based measure of obsessive-compulsive symptoms. Their results can be considered as complementary descriptive information, but the registry record does not establish that they were intended to test a unified claim or that they were powered as independent confirmatory outcomes. Their estimates and uncertainty should therefore be interpreted separately and in context, rather than treated as a set of decisive tests.

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

• Average Hamilton Depression Inventory (HAM-D)
• Average Hamilton Anxiety Inventory (HAM-A)
• Clinical Global Impression (CGI) - Severity of Illness Item

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary at the participant level: a participant either met the stated partial-response definition or did not. The posted group results are counts, with the number assessed in each group as the denominator. The comparison is therefore based on response proportions, although the registry’s reported analysis is identified as chi-squared. The registry record does not give additional detail about the exact implementation of that test or any adjustment to it.

What does a risk difference describe?

General statistical context: a risk difference subtracts the response proportion in the comparison group from the response proportion in the intervention group. It is expressed on the probability scale, so zero represents no difference in proportions. The recalculated risk difference is 0.152. Its confidence interval, from -0.093 to 0.398, crosses zero. This means the posted data are compatible with a range that includes no difference as well as differences in either direction. The estimate is unadjusted, meaning it compares the posted group summaries without adjustment for other variables.

How do risk ratios and odds ratios differ?

General statistical context: a risk ratio compares the response proportions by division, while an odds ratio compares the odds of response. Both are relative measures, and a value of one is their no-difference reference. The recalculated risk ratio is 2.368, with a confidence interval from 0.524 to 10.7. The recalculated odds ratio is 2.857, with a confidence interval from 0.477 to 17.11. Their intervals extend across the no-difference reference, and their breadth signals imprecision. Relative measures can look large when the event counts are small; they should be considered alongside the observed counts and the risk-difference estimate.

What does the p-value add?

General statistical context: a p-value measures how unusual data at least as incompatible with a specified null model would be if that model and the assumptions of the test were correct. It is not the probability that the null model is true, and it is not a measure of clinical importance. The registry reports p = .24 for its chi-squared analysis; the unadjusted recalculation is p = 0.238. The p-value does not resolve the uncertainty represented by the confidence interval or make the point estimate precise.

Why can the estimate be unstable?

General statistical context: with few observed events, the estimate can be sensitive to the outcome status of individual participants. Confidence intervals communicate uncertainty in the estimated effect; they are not guarantees that the true value lies inside a particular interval. For binary outcomes, the choice of effect measure also matters because a difference in proportions, a ratio of proportions, and a ratio of odds answer related but distinct questions. In this trial, the posted intervals are wide across these measures. The registry record does not report a more detailed model, covariate adjustment, or sensitivity analysis for the primary endpoint.

What does the reported test identify?

The registry labels the sponsor’s primary comparison as chi-squared. General statistical context: a chi-squared test assesses whether the observed distribution across groups is compatible with a null hypothesis of no association, subject to its assumptions. The record does not report whether an alternative test was considered or how assumptions were checked. The recalculated effect estimates provide interpretable magnitudes and intervals from posted summary counts, but they do not replace the trial’s reported analysis. As the discrepancy flag is false, the registry record does not indicate a conflict requiring the sponsor’s analysis to be set aside. Interpretation should still emphasize the limited outcome counts and imprecision.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.152, with a confidence interval from -0.093 to 0.398. The interval spans zero, so the posted data do not rule out no difference under this interval calculation. It also extends to values on either side of zero, leaving uncertainty about the direction and magnitude of the group contrast. The range is wide relative to the point estimate, so the estimate should not be treated as precise.

The relative-effect intervals reinforce that uncertainty. The risk ratio is 2.368, with a confidence interval from 0.524 to 10.7, and the odds ratio is 2.857, with a confidence interval from 0.477 to 17.11. Both intervals include their no-difference reference value. These estimates are based on five events in the riluzole group and two in the placebo group, which makes the apparent contrast sensitive to limited counts.

General statistical context: a confidence interval conveys the range of effect values reasonably compatible with the data and the calculation’s assumptions. A wide interval indicates that the data have not pinned down the effect closely; an interval spanning the null value indicates that the data do not distinguish the estimated contrast from no difference at the chosen confidence level. Neither fact proves that the groups are equal. Precision is separate from the point estimate: an estimate can suggest a direction while remaining too uncertain to support a narrow conclusion.

The registry record does not provide a sponsor-reported confidence interval for the primary contrast. The intervals are unadjusted recalculations from the posted summary counts. The registry’s reported chi-squared p-value is .24. Taken together, the counts, intervals, and test result call for a cautious description of the observed pattern rather than a definitive claim about the underlying effect.

12. Planning a Similar Trial: Sample Size

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

The sample-size figures are planning illustrations expressed as participants per arm. They are not total event counts. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is a stated fraction of the observed effect. The original design assumptions used to plan this trial are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for a design with an 80% chance of detecting the specified effect under the planning assumptions, if that effect is the true one and the model assumptions hold. The 90% power column gives the corresponding per-arm planning number for a 90% chance. These are design concepts, not guarantees about a future result. For example, planning for 100% of the observed primary effect gives 103 participants per arm for 80% power and 137 participants per arm for 90% power. Planning for 75% of the observed effect gives 168 participants per arm for 80% power and 224 participants per arm for 90% power; planning for 50% gives 344 participants per arm for 80% power and 460 participants per arm for 90% power.

Smaller target effects require more participants because a smaller difference is harder to distinguish from random variation. The calculations use the observed effect as their starting point, and the observed effect in a small trial can itself be uncertain. These figures are not observed power: they are planning illustrations under the stated assumptions, not a measure of what the completed trial achieved.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.111; treatment proportion p₁ = 0.263; average p̄ = 0.187.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5516 = 1.0811.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5410 = 0.4553.
4. n = (1.0811 + 0.4553)² / (0.1520)² = 102.1, rounded up to 103 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
p0 <- 0.111111111111111                          # control proportion
p1 <- 0.263157894736842                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(523718)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 103 (80% power), 137 (90% power)
 75% of observed effect: n per arm = 168 (80% power), 224 (90% power)
 50% of observed effect: n per arm = 344 (80% power), 460 (90% power)
power.prop.test: 102.1 
Simulated power at n = 103 per arm: 0.809 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 a chi-squared comparison for the primary outcome with p = .24. The unadjusted recalculation from the posted counts gives p = 0.238 and a risk difference of 0.152, with a confidence interval from -0.093 to 0.398. The reported and recalculated p-values are numerically close. The registry record does not explain the difference in their displayed precision or provide further details of the sponsor’s calculation.

The sponsor’s reported result should remain the primary account of the trial’s analysis. The recalculated estimates are simple, unadjusted calculations from the posted summary counts; they help describe the size and uncertainty of the observed contrast but do not supersede the sponsor’s analysis. The discrepancy flag is false, so the registry record does not identify a substantive conflict between the reported analysis and the recalculation.

The registry does not report a sponsor-derived confidence interval for the primary comparison. The confidence interval discussed for the recalculated risk difference therefore describes the uncertainty under that unadjusted calculation, not a sponsor-reported interval. Both the reported test and the recalculation should be read alongside the small outcome counts and the interval’s breadth.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures in total: the primary Y-BOCS partial-response outcome and three secondary outcomes. It reports one primary analysis, identified as chi-squared. The registry record does not report a multiplicity strategy, such as an adjustment for testing several outcomes, a hierarchy among outcomes, or a distinction between confirmatory and exploratory analyses. It is therefore not possible to determine from the record how the risk of false-positive findings across outcomes was addressed.

Missing-data handling is also not reported in the registry record. The primary outcome counts use 19 participants in the riluzole group and 18 in the placebo group, while recorded enrollment is 40. The record does not explain the difference between enrollment and the endpoint denominators, state how participants without a posted endpoint were handled, or describe imputation or sensitivity analyses. The denominators alone do not identify why an outcome is absent or what assumptions were used in the analysis.

The analysis population is not reported in the registry record. In particular, the record does not state whether the primary comparison followed an intention-to-treat approach, a per-protocol approach, or another population definition. It also does not specify which participants were included in each secondary analysis beyond the posted arm counts. These omissions matter because the population analyzed and the treatment of missing observations can affect both the estimated contrast and its uncertainty.

General statistical context: multiplicity concerns arise when several hypotheses are tested, because the chance of at least one apparently unusual result can increase across a collection of tests. Missing-data methods matter because results can be biased if outcome availability is related to participant characteristics or outcomes. Neither concern can be resolved from group summaries alone. The registry record does not report enough detail to assess these aspects of the analysis.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: overdose of narcotics (5.6%), passively suicidal (5.3%).

The posted serious-adverse-event figures report one affected participant among 19 at risk in the riluzole group and one among 18 at risk in the placebo group. These are group-level counts in a small trial and should be read as descriptive registry results. They do not, on their own, establish whether an event was caused by the assigned study treatment.

For other adverse events, the registry reports seven affected participants among 16 at risk in the riluzole group and 11 among 17 at risk in the placebo group. The at-risk denominators differ from the serious-event denominators. The record does not explain these denominator differences or provide further detail in the reported figures to support a causal interpretation.

The listed serious-event terms include “Overdose of Narcotics” at 5.6% and “Passively Suicidal” at 5.3%. The registry does not identify those percentages by treatment group in the posted terms, so they should not be assigned to a particular arm. Death counts are not reported in the registry record. Overall, these figures describe recorded events; they do not establish comparative safety with precision.

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

The registry results describe a small randomized comparison of riluzole augmentation and placebo augmentation in participants with treatment-refractory obsessive-compulsive disorder. At 14 weeks, the riluzole group had five participants meeting the Y-BOCS partial-response definition among 19, compared with two among 18 in the placebo group. The unadjusted point estimate favors a higher observed response proportion with riluzole, but the confidence intervals are wide and include values consistent with no difference and with a difference in the opposite direction.

These findings do not establish that riluzole caused the observed difference, nor do they establish that the groups had the same underlying response rate. The limited event counts and broad uncertainty leave the size and direction of the underlying contrast unresolved. The registry’s reported chi-squared p-value, .24, is consistent with an imprecise comparison rather than a precise estimate.

The result is specific to the trial population, augmentation approach, endpoint definition, and 14-week timeframe recorded. The endpoint represents meeting a 25% improvement threshold, not the full pattern of individual symptom changes. The record also contains secondary measures and adverse-event counts, but the available registry results do not resolve the primary uncertainty or provide a basis for a broader claim.

General statistical context: absence of a result that clearly separates groups is not proof of equivalence, and a point estimate in one direction is not proof of a treatment effect. The confidence interval is important because it shows that the data allow a range of plausible contrasts. Interpretation should preserve that uncertainty and distinguish the observed sample pattern from a definitive conclusion about the intervention.

18. Why This Trial Matters Statistically

Statistically, this trial illustrates why the observed direction of a group contrast is not enough to characterize a result. The riluzole group had more recorded partial responders, but the recalculated risk-difference interval extends across zero and is broad. The result therefore combines a point estimate with substantial uncertainty. Reporting the estimate, interval, and event counts together gives a more complete account than relying on the p-value alone.

The binary endpoint also shows how a threshold definition shapes interpretation. The primary question concerns who met the stated 25% improvement criterion, rather than the average change in symptom score. That is a direct and interpretable question, but it discards information about the magnitude of changes within each response category. The registry does not post the underlying Y-BOCS score summaries needed to assess that dimension.

The trial’s recorded enrollment of 40 and its smaller posted primary-outcome denominators make analysis-population and missing-data details relevant, yet those details are not reported in the registry record. The listed secondary outcomes add broader clinical measures, but multiplicity handling is also not reported. These omissions limit assessment of the full analytic process, beyond the uncertainty inherent in the primary counts.

The planning illustrations further show the relationship between target effect size and required enrollment under their stated assumptions: planning for smaller effects requires more participants. They are not evidence about the trial’s original design or a retrospective measure of its success. Overall, the record is useful for examining how sparse events, endpoint definitions, confidence intervals, and unreported analytic details shape interpretation of a randomized comparison.

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 trial-specific limitations affect interpretation. The primary endpoint is based on a thresholded Y-BOCS response definition, and the registry does not report the group-level baseline and follow-up Y-BOCS scores or average change. Consequently, the posted result does not show how symptom changes were distributed within the responder and nonresponder categories.

The primary comparison has few observed responders, and its recalculated confidence interval is broad. The registry record also does not explain why the primary outcome denominators are 19 and 18 when recorded enrollment is 40. Missing-data handling, the analysis population, and sensitivity analyses are not reported in the registry record. These omissions prevent assessment of whether the posted counts represent all randomized participants or how incomplete outcomes were addressed.

The record identifies a chi-squared analysis but gives limited detail on its implementation. It does not report a sponsor-derived primary confidence interval, covariate adjustment, or additional analysis methods. It also does not report a multiplicity plan for the secondary outcomes, and no sponsor analyses are listed for those outcomes. Thus the secondary results should be treated as unadjusted descriptive comparisons rather than as a confirmed set of additional findings.

The recorded study start and primary completion span a substantial calendar interval, but the registry information does not explain that interval or its implications. Other limitations include the absence of further eligibility details in the posted results and the limited detail linking the serious-event terms to treatment groups. These are limits of what the registry record reports; they do not establish that particular procedures were absent from the conduct of the trial.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Riluzole Augmentation in Treatment-refractory Obsessive-compulsive Disorder — NCT00523718.
• Registry record last updated 2020-03-06; status last verified 2020-03.
• Results first posted 2016-03-23.
• 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

NCT00523718 is a completed, randomized, quadruple-masked, parallel-group phase 2 trial sponsored by Yale University. It compared riluzole augmentation with placebo augmentation in serotonin reuptake inhibitor-refractory obsessive-compulsive disorder. Recorded enrollment was 40, and the primary endpoint was partial response on the Y-BOCS at 14 weeks, defined as a 25% improvement from baseline.

The posted primary counts were five responders among 19 participants assigned to riluzole and two among 18 assigned to placebo. The registry reports a chi-squared p-value of .24. The unadjusted recalculated risk difference is 0.152, with a confidence interval from -0.093 to 0.398. The interval is broad and includes zero, so the observed direction is uncertain in magnitude.

Three secondary outcomes and adverse-event counts are also recorded. The registry does not report multiplicity handling, missing-data methods, or the analysis population. Overall, the posted results describe an observed responder-count contrast but leave substantial uncertainty about its size and the details of the analysis.