Riluzole to Treat Child and Adolescent Obsessive-Compulsive Disorder With or Without Autism Spectrum Disorders: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Riluzole to Treat Child and Adolescent Obsessive-Compulsive Disorder With or Without Autism Spectrum Disorders (NCT00251303), a randomized phase 2 study. The primary outcome is Much/Very Much Improved on Clinical Global Impressions - Improvement Score (CGI-I).

2. Clinical Question

3. Trial Design

The registry describes a randomized, quadruple-masked, parallel-group phase 2 study. It lists riluzole and placebo as the interventions, with participants assigned to one of the two groups. The study title specifies child and adolescent obsessive-compulsive disorder, with or without autism spectrum disorders. The conditions listed in the record also include autism, Asperger disorder, and developmental disorder. Detailed eligibility criteria, the age range, and the diagnostic procedures are not reported in the registry record.

The recorded enrollment is 78. The primary outcome is reported for 22 participants in the riluzole group and 29 in the placebo group. The registry record does not explain why these outcome-group counts do not account for the full enrollment, nor does it report participant flow or the reasons for missing outcome data. Safety information uses groups of 30 participants per arm, so the efficacy and safety figures have different reported denominators.

The primary outcome timeframe is 12 weeks. The study status is completed; the recorded start is 2005-08 and primary completion is 2012-02. The sponsor is the National Institute of Mental Health (NIMH). The registry record does not report the original sample-size justification, the intended analysis population, or detailed procedures for allocation, masking, or follow-up. The randomized and masking labels describe design features, but they do not by themselves establish how consistently those procedures were implemented.

There are two posted outcome measures and one posted analysis. No secondary outcome results are posted. The record does not report whether additional analyses were planned or conducted, or how participants who discontinued or had incomplete assessments were handled.

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: Subjects had to be symptomatic for at least 6 weeks, with a score of at least 20 CY-BOCS; had to have had at least one standard-of-care treatment for childhood OCD for adequate periods of time; had to have been stable for at least 6 weeks on any medicine; had to meet entrance criteria at a screening and 2 weeks later.

Overall Study

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: These data apply only to double-blind phase.

6. Primary Endpoint

The registry lists the primary outcome as Much/Very Much Improved on Clinical Global Impressions - Improvement Score (CGI-I), measured over 12 weeks.

The primary endpoint counts participants rated much or very much improved on the Clinical Global Impressions - Improvement score, abbreviated CGI-I. The registry gives a 12-week timeframe and reports the outcome as a number of participants. This is a binary outcome: each participant either meets the stated improvement categories or does not. The record does not give further details about who made the ratings, how often they were assessed, or how ratings were converted into the two categories.

A categorical endpoint can be useful when a study aims to compare the proportion of participants reaching a defined clinical status rather than comparing average scores. It gives a direct count of participants meeting the specified threshold. However, collapsing ratings into two categories discards distinctions among participants who fall on the same side of the threshold. It also does not describe changes in symptoms outside that classification.

The posted counts are 3 participants in the riluzole group and 4 in the placebo group. Those counts need to be read with their group denominators: 22 for riluzole and 29 for placebo. These are the numbers reported for the primary outcome, not the overall enrollment. The registry record does not report additional endpoint details that would clarify assessment timing beyond the 12-week timeframe or explain the outcome-data availability.

7. Results Posted in the Registry

• Riluzole: 13.6% (3/22) versus 13.8% (4/29) on control; risk difference -0.2 percentage points (95% CI -19.2 to 18.9), risk ratio 0.99 (0.25–3.97); unadjusted two-proportion z-test p = 0.987.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary, meaning it records whether a participant met the specified improvement category. For this type of outcome, group results can be expressed as event counts, event proportions, and contrasts between those proportions. Here, the event is being rated much or very much improved on CGI-I by 12 weeks. The registry reports 3 events among 22 riluzole participants and 4 among 29 placebo participants.

What does a risk difference describe?

A risk difference subtracts the event proportion in one group from the event proportion in the other. The recalculated estimate is -0.002 for riluzole relative to placebo. A value near zero indicates that the observed proportions are close. A negative value indicates a lower observed event proportion in the first-named group under this contrast convention. Unlike a ratio, a risk difference is expressed on the event-proportion scale, making it a direct measure of the absolute separation between groups. It does not adjust for other participant characteristics.

How should the confidence interval be read?

A confidence interval describes uncertainty around an estimated effect under the statistical method used. It gives a range of values compatible with the data and method, rather than a guarantee that the true effect lies within the particular interval. The posted 95% interval for the risk difference runs from -0.192 to 0.189. Because the interval includes zero, it does not exclude no difference. Its span also includes differences in either direction, which signals limited precision. The result should not be reduced to a yes-or-no conclusion based only on whether the interval includes zero.

What does the p-value test?

The registry reports a chi-squared analysis with p = 0.959. A chi-squared test evaluates how unusual a difference in categorical outcomes would be under a null hypothesis of no association between group and outcome, according to the test's assumptions. The p-value is not the probability that the null hypothesis is true, and it is not a measure of the size or clinical importance of an effect. A large p-value does not demonstrate equivalence. Equivalence or non-inferiority conclusions require a prespecified framework and margin; those details are not reported in the registry record.

The unadjusted re-calculation gives p = 0.987. This is a separate calculation from posted summary statistics, not a substitute for the registry's own analysis. Differences between p-values can arise from analytical choices, but the detailed procedure underlying the registry result is not reported beyond the chi-squared label. The registry record does not flag a discrepancy between the reported result and the recalculation.

Why do counts and denominators matter?

Counts alone can be misleading when group sizes differ. The number of participants meeting the endpoint must be interpreted alongside the number assessed in each arm. The primary outcome counts use 22 participants in the riluzole group and 29 in the placebo group. These denominators differ from the recorded enrollment of 78. The record does not describe the participant flow or missing-data rules, so the reason for the difference cannot be determined from the registry information.

General statistical context: sparse event counts can make estimates unstable and intervals wide. When outcomes are uncommon, moving only a small number of participants between categories can materially change the observed proportions. That general property is relevant to interpreting a result based on the posted counts, but it does not identify what happened to participants without a posted primary outcome.

What does masking contribute?

The registry labels the study quadruple-masked. Masking is intended in general to reduce the chance that knowledge of assigned treatment influences care, assessment, or other trial procedures. The record does not identify which roles were masked or describe how masking was maintained. The label therefore conveys the registered design classification, but the available information does not permit an assessment of masking success.

Likewise, randomization is a design feature intended in general to support comparable groups at baseline, subject to implementation and chance variation. The record identifies the allocation as randomized but does not report the sequence-generation method, concealment process, or baseline characteristics. Those details are needed to assess the conduct and balance of the groups rather than to infer them from the design label alone.

10. Confidence Intervals and Precision

The recalculated risk difference is -0.002, with a 95% confidence interval from -0.192 to 0.189. The point estimate is close to zero, but the interval is broad and includes zero. It also extends in both directions, so the data are compatible with a lower or higher event proportion in the riluzole group relative to placebo. The interval communicates uncertainty about the effect size; it should not be read as evidence that the groups are equivalent.

In general, a confidence interval is more informative than a point estimate alone because it shows how much uncertainty remains under the specified method. Here, the limited outcome counts—3 of 22 and 4 of 29—contribute to the need for caution about precision. A p-value of 0.959 in the registry's chi-squared analysis does not resolve this imprecision. A large p-value and a wide interval can coexist: one describes the test's compatibility with a null comparison, while the other displays the range of effect sizes supported by the data and method.

The interval is an unadjusted re-calculation from the posted summary statistics. It does not account for possible differences in participant characteristics, missing outcomes, or other analyses. The registry record does not report adjusted estimates or enough information to determine how missing data were handled. The confidence interval therefore characterizes the stated unadjusted contrast, not every possible analysis of the trial.

11. 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 events. They illustrate the sample size that would be needed under the stated planning assumptions to detect a chosen fraction of the observed effect. The assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The calculation uses the observed effect as its reference, then considers 100%, 75%, and 50% of that effect. The original trial's own sample-size assumptions and justification are not reported in the registry record.

For 100% of the observed effect, the illustration gives 756,141 participants per arm for 80% power and 1,012,257 participants per arm for 90% power. For 75% of the observed effect, the corresponding figures are 1,345,865 and 1,801,731 participants per arm. For 50% of the observed effect, they are 3,031,829 and 4,058,759 participants per arm. These are planning illustrations based on the observed effect, not the sample sizes used in the trial.

In plain terms, 80% power means that, if the assumed effect and other planning assumptions were correct, a study of that size would have an 80% chance of meeting its specified statistical detection criterion. The 90% column uses a higher chance of detection under the same assumed framework and therefore calls for a larger sample size. Smaller target effects require more participants because they are harder to distinguish from random variation. These calculations are not observed power and should not be interpreted as a description of the trial's achieved power.

12. 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 = 0.138; treatment proportion p₁ = 0.136; average p̄ = 0.137.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4865 = 0.9535.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4865 = 0.4094.
4. n = (0.9535 + 0.4094)² / (-0.0016)² = 756,140.3, rounded up to 756,141 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 <- 0.137931034482759                          # control proportion
p1 <- 0.136363636363636                          # 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(251303)
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 = 756,141 (80% power), 1,012,257 (90% power)
 75% of observed effect: n per arm = 1,345,865 (80% power), 1,801,731 (90% power)
 50% of observed effect: n per arm = 3,031,829 (80% power), 4,058,759 (90% power)
power.prop.test: 756140.3 
Simulated power at n = 756,141 per arm: 0.799 (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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor's chi-squared analysis with p = 0.959. A separate unadjusted re-calculation from the posted counts gives a risk difference of -0.002, a 95% confidence interval from -0.192 to 0.189, and p = 0.987. The registry record does not flag a discrepancy between the reported analysis and the re-calculation.

The two p-values come from differently described results: the registry identifies the reported method as chi-squared, while the re-calculation is an unadjusted analysis based on the posted summary statistics. The reported analysis remains the trial's own posted analysis; the recalculated estimate and p-value are supplementary descriptions of the available counts. Neither p-value alone establishes equivalence or determines the clinical importance of a difference.

The registry record does not report enough detail about the sponsor's statistical procedure to compare every analytical choice, such as adjustments or handling of incomplete outcomes. The re-calculation should therefore be read as a simple unadjusted comparison, not as a correction or replacement for the sponsor's analysis.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists two outcome measures and one analysis, with no secondary outcome results posted. It does not report a multiplicity strategy, such as whether a primary analysis was protected from additional testing or how multiple outcomes were addressed. Without that information, the relationship between the posted primary result and any broader testing plan is unclear. The record does not establish whether other outcomes or analyses were planned but not posted.

The primary outcome is reported for 22 riluzole participants and 29 placebo participants, while total enrollment is 78. The registry does not report participant flow, reasons for missing endpoint data, or the approach used to handle missing observations. It is therefore not possible to determine whether the outcome counts reflect all randomized participants, a defined subset, or another analysis population. No assumptions about exclusion, imputation, or treatment discontinuation should be made from the available record.

The analysis population is not reported in the registry record. In general, an intention-to-treat analysis aims to retain participants in their randomized groups, while a per-protocol analysis applies specified adherence or eligibility criteria. The record does not indicate which framework was used, so neither label can be assigned to the posted result. The statistical plan, including any prespecified analysis populations, is also not reported.

These gaps affect interpretation because missing outcomes and the choice of analysis population can influence a categorical comparison. They do not establish that the posted analysis is biased; they limit the ability to assess its handling of incomplete data and its place within the full analysis plan. Details about interim analyses, stopping rules, and multiplicity adjustments are not reported in the registry record.

15. 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: pancreatitis (3.3%).

The serious adverse event figures list 1 affected participant among 30 at risk in the riluzole group and 0 among 30 at risk in the placebo group. The registry also lists pancreatitis as a serious adverse event term at 3.3%. The figures describe events recorded in the registry; they do not establish whether an event was caused by study treatment. The criteria used to classify seriousness and details about event timing are not reported in the registry record.

Other adverse events are listed for 27 of 30 participants in the riluzole group and 29 of 30 in the placebo group. These are separate figures from the serious adverse event counts. The registry record does not provide further detail about event types, severity, duration, or how events were collected. Death counts are not reported in the registry record.

Safety denominators are 30 per arm, unlike the primary outcome denominators of 22 and 29. The record does not explain the difference, so safety and efficacy counts should not be treated as if they describe an identical analysis population. These registry figures are descriptive and should be interpreted alongside the limits of the posted safety information.

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

The posted primary result is a close comparison in the observed proportion of participants meeting the much-or-very-much-improved CGI-I threshold at 12 weeks. The riluzole group had 3 such participants among 22, and the placebo group had 4 among 29. The recalculated risk difference is close to zero, but its 95% confidence interval extends from -0.192 to 0.189. The data therefore do not determine a narrow range for the possible difference.

The registry's chi-squared p-value is 0.959. This result does not establish that riluzole and placebo are equivalent, and it does not prove that no effect exists. A nonsignificant comparison can reflect a small difference, substantial uncertainty, or both. Here, the broad interval and limited endpoint counts are important parts of the interpretation. The observed result should be described as an imprecise comparison rather than as a definitive finding of sameness or difference.

The endpoint captures whether participants crossed a particular improvement threshold. It does not show the full distribution of CGI-I ratings, the amount of symptom change among participants, or outcomes beyond the 12-week timeframe. The registry record does not report additional secondary outcome results that would broaden the interpretation.

The posted enrollment is 78, but the primary endpoint counts cover fewer participants. Because the participant flow and missing-data approach are not reported, the effect of incomplete outcome information cannot be assessed. The result also cannot be generalized beyond the studied population solely from the registry entry. Eligibility details and participant characteristics are not reported in the registry record.

Overall, these data describe the primary comparison recorded for this completed phase 2 trial. They do not by themselves answer questions about other treatment durations, other patient groups, long-term outcomes, or the balance of benefits and harms. The registry record contains limited safety detail and does not provide a complete analysis plan.

17. Why This Trial Matters Statistically

This trial illustrates why event counts, denominators, effect estimates, and uncertainty need to be read together. The primary outcome counts are 3 of 22 and 4 of 29. Looking only at the event counts would hide the group sizes; looking only at a p-value would hide the estimated magnitude and its uncertainty. The risk difference and confidence interval make the near-zero point estimate and broad range of compatible values more visible.

It also demonstrates the distinction between a test result and a conclusion about equivalence. The registry's chi-squared analysis has p = 0.959, but that value does not prove that the groups are the same. A confidence interval that spans differences in both directions gives a more direct sense of what remains uncertain. When event counts are limited, precision can be poor even when a point estimate is close to zero.

The planning illustration further shows how sample-size requirements depend on the effect targeted for detection. Smaller assumed effects require larger studies under the stated assumptions. The figures are not a reconstruction of the trial's original design rationale, which is not reported in the registry record, and are not observed power.

Finally, differences between recorded enrollment, endpoint denominators, and safety denominators highlight the importance of participant flow and analysis-population reporting. Without those details, readers cannot fully evaluate how the analyzed groups relate to everyone enrolled. These are general lessons in interpreting registered trial results, grounded in the counts and reporting gaps of this study.

18. 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 main limitation of the posted primary result is the small number of participants meeting the endpoint: 3 in the riluzole group and 4 in the placebo group. The resulting confidence interval is broad, spanning negative and positive risk differences. This limits precision and makes it difficult to distinguish a small difference from a larger one in either direction.

The primary outcome denominators are 22 and 29, while the registry records enrollment of 78. The reasons for this difference, participant disposition, and handling of missing outcomes are not reported in the registry record. Without these details, readers cannot assess whether endpoint availability differed by group or how the analysis population was defined. Safety results use 30 participants per arm, a separate denominator whose relationship to the primary outcome groups is not explained.

Detailed eligibility criteria, participant characteristics, allocation procedures, masking procedures, and the identities of those covered by quadruple masking are not reported. The registry also does not describe the statistical plan, the original sample-size rationale, interim analyses, multiplicity handling, or the analysis population. No secondary outcome results are posted, despite the record listing two outcome measures.

The endpoint is a thresholded improvement classification at 12 weeks. The registry record does not provide detailed rating procedures or the distribution of scores beyond that classification. Safety reporting is also limited: adverse event detail and death counts are not reported in the registry record. These limitations restrict interpretation of the posted efficacy and safety results; they do not, on their own, establish that the trial was improperly conducted.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Riluzole to Treat Child and Adolescent Obsessive-Compulsive Disorder With or Without Autism Spectrum Disorders — NCT00251303.
• Registry record last updated 2014-07-15; status last verified 2014-07.
• Results first posted 2014-07-15.
• 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.

22. Record Summary

NCT00251303 is a completed randomized, quadruple-masked, parallel-group phase 2 trial sponsored by the National Institute of Mental Health (NIMH). It compared riluzole with placebo in a study of child and adolescent obsessive-compulsive disorder, including participants with autism spectrum disorders. Recorded enrollment was 78.

The primary outcome was the number of participants rated much or very much improved on CGI-I at 12 weeks. The posted counts were 3 of 22 for riluzole and 4 of 29 for placebo. The registry reports a chi-squared p-value of 0.959; the unadjusted re-calculation gives a risk difference of -0.002 with a 95% confidence interval from -0.192 to 0.189.

Participant flow, missing-data handling, analysis populations, the original sample-size rationale, and detailed safety procedures are not reported in the registry record. No secondary outcome results are posted.