Open-label Study of Safety and Tolerability of Memantine in Children With Autism: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Open-label Study of Safety and Tolerability of Memantine in Children With Autism (NCT01999894), a na phase 2 study. The primary outcome is Number of Patients Who Experienced a Treatment-emergent Adverse Event (TEAE).

2. Clinical Question

The study’s stated purpose is to evaluate the long-term safety and tolerability of memantine in pediatric patients with autism. The primary outcome concerns whether a patient experienced at least one treatment-emergent adverse event during the specified observation period. It is a patient count, not a measure of symptom improvement or treatment benefit.

The listed group labels indicate treatment histories, but they do not establish a randomized comparison. In particular, the registry record lists no allocation and identifies the model as single-group. A difference between the group summaries, if one were calculated, could not by itself be attributed to memantine or to a treatment contrast. The primary outcome’s count is not reported, and the record reports no analysis for it. The question that can be answered from the record is therefore limited: it describes what safety outcome was intended to be counted and provides selected adverse-event summaries, but it does not give the primary outcome result.

3. Trial Design

The registry identifies this as a phase 2 study of memantine in pediatric patients with autism. Its title describes an open-label safety and tolerability study, and the official title calls it an open-label extension study. The stated purpose is assessment of long-term safety and tolerability. The record lists the study as completed, with enrollment of 102.

The model is listed as single-group, allocation as NA, and masking as none. The arm descriptions nevertheless distinguish “Placebo to Memantine” from “Memantine to Memantine.” The descriptions state that memantine was given orally once daily for 48 weeks, including a 6-week double-blind dose-titration period and a 42-week maintenance period. These labels and details describe treatment history; they do not make the groups a randomized, like-for-like control comparison. The registry record does not explain the relationship between the open-label title and the double-blind period described in the arm information.

The primary outcome is the number of patients with one or more treatment-emergent adverse events from Visit 1 (Week 1) to 30 days after Visit 8 (Week 48). Its stated population is the safety population: 102 enrolled patients who received at least one dose of study drug. The record lists one outcome measure and no analyses. A numeric result for the primary outcome is not reported in the registry record.

The record gives a study start of 2009-11 and primary completion of 2013-02. It does not report random assignment, a concurrent control suitable for causal comparison, or a statistical analysis plan in the information available for this record. The study design is therefore best read as a descriptive safety follow-up rather than as an efficacy comparison.

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: Patients receiving placebo in the lead-in study underwent a double-blind up-titration targeting same weight-based dose assigned at lead-in study. Patients receiving a stable dose of active were not re-titrated but double-blind was maintained. Patients who were not dosed for more than 3 days between the studies underwent an open-label titration.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Safety Population consists of 102 enrolled patients who received at least one dose of study drug.

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 Number of Patients Who Experienced a Treatment-emergent Adverse Event (TEAE), measured over From Visit 1 (Week 1) to 30 days after Visit 8 (Week 48).

The primary endpoint counts patients who experienced at least one treatment-emergent adverse event during the defined observation period. The timeframe begins at Visit 1 (Week 1) and extends to 30 days after Visit 8 (Week 48). The registry describes the endpoint as a number of participants and defines an event as one or more treatment-emergent adverse events during the study.

Counting affected patients is a common way to summarize safety experience: each patient is counted as having an event or not having one, regardless of how many events that patient experienced. This makes the endpoint a patient-level count rather than a count of all event episodes. It does not convey the number, duration, severity, or clinical consequences of events for each patient. Those details would require additional outcome reporting.

The endpoint is used because the study’s stated aim is safety and tolerability over an extended treatment period. It can describe how many patients had an event within the observed period, but it cannot establish that the study drug caused an event. Nor does this endpoint measure whether autism-related symptoms changed. The registry record does not report the primary endpoint count or an analysis, so its result and precision cannot be assessed from the record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports a binary primary outcome expressed as a number of participants with an event. In general, a binary outcome classifies each participant according to whether the specified event occurred. For this study, the event is experiencing one or more treatment-emergent adverse events during the stated timeframe. The endpoint counts affected patients rather than counting every event episode.

What does a participant count tell the reader?

A count gives the number of patients meeting the endpoint definition. When a suitable denominator is also reported, a rate or proportion can describe how common the event was in the observed group. The primary outcome information in this record gives the population definition but not the observed count. It therefore does not support an estimate of the primary endpoint’s frequency. No rate or proportion should be inferred from the enrollment figure alone.

For the separate adverse-event summaries, the registry lists both counts and “at risk” values for each treatment-history group. These can be read as descriptive counts within the listed groups. They are not a substitute for a prespecified comparative analysis, especially when the record identifies the study as single-group, lists allocation as NA, and explicitly cautions that the arms are not a like-for-like control comparison.

Why does design matter for comparisons?

Random allocation, when used and properly implemented, helps make treatment groups comparable on average at the start of a study. That supports interpretation of a between-group difference as an estimate of a treatment contrast, subject to other design and analysis assumptions. Here, random allocation is not reported; allocation is listed as NA. The group labels describe transitions in treatment, and the record does not establish that the groups were formed to provide a concurrent randomized control. A numerical contrast between them would therefore be descriptive, not evidence of a causal effect.

Masking can also affect the reporting and assessment of outcomes, particularly outcomes involving clinical judgment or participant reporting. The registry lists masking as none, while the arm description mentions a double-blind dose-titration period. The record does not explain how these descriptions relate across the study. This leaves the masking history unclear from the registry information and should not be resolved by assumption.

How should uncertainty be assessed?

A confidence interval is a general statistical tool for expressing the precision of an estimate. A narrow interval indicates greater precision under the model used; a wide interval indicates more uncertainty. No confidence interval or primary outcome estimate is reported in the registry record. As a result, the precision of the primary safety outcome cannot be assessed. A p-value, if reported, would address compatibility with a statistical null model under stated assumptions; it would not by itself establish clinical importance or causation. No p-value or analysis result is reported.

The registry includes a description of a sample-size planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between groups, and a normal approximation. Such planning values are hypothetical illustrations of the sample size under stated assumptions; they are not observed power and do not document the design rationale for the completed trial. The record does not report the trial’s sample-size justification or the numerical planning results in the information summarized for this analysis. No planning illustration can fill in the missing primary outcome result.

What can be inferred from the safety summaries?

Counts of patients with events are useful for describing reported safety experience. They do not capture the complete clinical profile of each event. A count does not show whether events were mild or severe, how long they lasted, whether they recurred, or when they occurred. The primary endpoint definition combines patients with one or more events into a single patient-level outcome, so repeated events in the same patient would not be distinguished by that endpoint count.

Serious adverse events and other adverse events are listed in separate categories. These categories should not be conflated: the serious-event count is not the count for all adverse events, and the “other” category should not be assumed to provide detailed event severity or causality information. The registry lists lobar pneumonia with 2.0% among the serious-event terms, but does not report a group attribution for that term.

Finally, the registry identifies the safety population as patients who received at least one dose. This is an exposure-based analysis population, not necessarily the same as every person screened or enrolled under other possible definitions. No other analysis population or handling of missing outcome data is reported in the registry record. Without an outcome count, analysis description, and uncertainty estimate, the posted information supports a limited descriptive reading of selected safety counts rather than a full statistical assessment of the primary endpoint.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Memantine")
n      <- c(102)
events <- c(85)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Memantine  102  85  83.3  74.7  90.0  74.9  89.3  74.9  89.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and no analyses. It does not report a multiplicity adjustment, a hierarchy of outcomes, or a prespecified strategy for controlling false-positive findings across multiple statistical tests. Since no analysis results are listed, the record does not support an assessment of how multiplicity was handled in practice. A reader should not infer that a particular adjustment or testing sequence was used.

The safety population is described as 102 enrolled patients who received at least one dose of study drug. This defines the population for the primary outcome, but the registry record does not report a missing-data method, the amount of missing endpoint information, or how patients with incomplete follow-up were handled. The stated observation period extends from Visit 1 (Week 1) to 30 days after Visit 8 (Week 48), but that timeframe alone does not establish that all participants completed it or that all events were captured.

The record does not describe an intention-to-treat analysis population or another analysis set beyond the safety population definition. It also does not report sensitivity analyses addressing missing outcomes. Therefore, the effect of incomplete observation, if any, cannot be evaluated from the registry information. No imputation method or assumption about missing data should be presumed.

These omissions matter because incomplete follow-up can affect a safety count if event ascertainment differs across participants. The registry’s group-specific “at risk” counts and event counts provide descriptive information, but they do not document completeness of follow-up or resolve the primary endpoint’s missing result. The primary outcome value itself is not reported in the registry record.

12. 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: lobar pneumonia (2.0%).

The registry lists serious adverse-event counts separately for the treatment-history groups. The “Placebo to Memantine” group has 1 serious event among 50 at risk, and the “Memantine to Memantine” group has 0 among 52 at risk. These are descriptive summaries, not randomized comparisons. The record lists allocation as NA and identifies the study as single-group, so a difference between these counts should not be interpreted as a treatment effect.

The serious-event terms include lobar pneumonia with 2.0%. The record does not attribute this term to a particular group in the information reported. It is therefore inappropriate to assign it to either group. No death count is reported.

The other adverse-event counts are 37 among 50 at risk for “Placebo to Memantine” and 37 among 52 at risk for “Memantine to Memantine.” These are counts in the registry’s other adverse-event category. They should not be read as the serious-event count or as a complete account of event severity and clinical course. The record does not report the primary endpoint count, event timing, recurrence, or the analysis used for these summaries. The figures describe what is listed; they do not establish causality or comparative safety.

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

The record documents a phase 2 study intended to assess long-term safety and tolerability of memantine in pediatric patients with autism. It defines a primary safety endpoint based on whether participants experienced one or more treatment-emergent adverse events during the specified period. The record also reports selected serious and other adverse-event counts by treatment-history group. Those details help clarify the questions the study addressed and the type of safety information recorded.

However, the registry record does not report the numeric primary endpoint result, any analysis for that endpoint, or measures of statistical uncertainty. The primary outcome therefore cannot be used to estimate the frequency of treatment-emergent adverse events from the information reported. The listed safety counts are descriptive and do not establish that memantine caused, prevented, or changed the likelihood of an event.

The study’s group labels should not be treated as a randomized treatment comparison. Allocation is listed as NA, the model is single-group, and the record cautions that the groups are not a like-for-like control comparison. Differences in their counts, if considered, cannot be given a causal interpretation on the basis of this record. The record also does not report an efficacy outcome, so it does not answer whether treatment changed autism-related symptoms.

In general, a safety count can indicate how many patients met a defined event criterion, but it does not convey all relevant event details. It does not by itself describe severity, duration, recurrence, or causal attribution. Here, the serious-event term lobar pneumonia is listed with 2.0%, without group attribution. These limits mean the posted information supports a narrow descriptive account of selected adverse-event data, not a broad conclusion about safety or efficacy.

14. Why This Trial Matters Statistically

This study is statistically notable because it illustrates the difference between a safety question and a comparative efficacy question. Its stated purpose is to evaluate long-term safety and tolerability, and its primary endpoint counts patients with one or more treatment-emergent adverse events. That endpoint is designed to describe safety experience, not symptom improvement.

It also illustrates why group labels alone do not create a valid treatment comparison. The record calls the study single-group, lists allocation as NA, and cautions that the listed groups are not a like-for-like control comparison. Descriptive counts can be informative, but a causal interpretation requires a design and analysis that support comparison. Without that support, contrasts between treatment-history groups should not be treated as estimates of a treatment effect.

Another statistical lesson is the distinction between an endpoint definition and an endpoint result. The record defines the population and observation period for the primary measure, but it does not report the primary count or an analysis. A clearly specified outcome is not enough to evaluate its frequency or precision. The registry’s selected adverse-event summaries cannot replace that missing result.

Finally, a safety summary has limits even when event counts are available. Patient counts do not describe event episodes in detail, and a term-level listing without group attribution cannot identify which group experienced that event. These distinctions are important when reading registry records: they help separate what is described, what is estimated, and what remains unanswered.

15. 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 for interpretation is that the primary endpoint result is not reported in the registry record. Although the endpoint, timeframe, and safety population are defined, the record gives no numeric primary count and no analysis. It therefore does not permit assessment of the primary safety outcome’s frequency or statistical precision.

The design information also limits comparative interpretation. The record identifies a single-group model, lists allocation as NA, and states that the treatment-history groups are not a like-for-like control comparison. Group-specific adverse-event counts should not be interpreted as randomized evidence of a treatment difference. The record does not explain how participants entered each treatment-history group.

The record’s title describes an open-label extension, while the arm description mentions a double-blind dose-titration period. Masking is listed as none. The registry record does not reconcile these descriptions. The information also does not provide detailed eligibility criteria, a statistical analysis plan, sample-size justification, missing-data procedures, or an account of outcome ascertainment beyond the stated timeframe. These details are not reported in the registry record.

The adverse-event information is limited to selected counts and a serious-event term listing. The record does not provide event-level details such as onset, duration, recurrence, severity grading, or causal assessment. The lobar pneumonia listing is not assigned to a treatment-history group. No death count is reported. Finally, no efficacy outcome is reported, so the record cannot support conclusions about symptom change or treatment benefit.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Open-label Study of Safety and Tolerability of Memantine in Children With Autism — NCT01999894.
• Registry record last updated 2014-03-18; status last verified 2014-01.
• Results first posted 2014-03-18.
• 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.

19. Record Summary

NCT01999894 is a completed phase 2 study titled “Open-label Study of Safety and Tolerability of Memantine in Children With Autism.” The stated purpose was to evaluate long-term safety and tolerability in pediatric patients with autism. The record lists enrollment of 102 and describes once-daily oral memantine for 48 weeks, including a 6-week double-blind dose-titration period followed by a 42-week maintenance period.

The primary endpoint is the number of patients with one or more treatment-emergent adverse events from Visit 1 (Week 1) to 30 days after Visit 8 (Week 48). The safety population is described as 102 enrolled patients who received at least one dose. The primary endpoint result and any analysis are not reported in the registry record.

Selected serious and other adverse-event counts are listed for “Placebo to Memantine” and “Memantine to Memantine.” The record identifies a single-group design, lists allocation as NA, and cautions that these are not like-for-like control groups. The counts are therefore descriptive and do not establish a causal treatment comparison.