Effectiveness of Memantine in Treating Cocaine-Dependent Individuals - 2: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effectiveness of Memantine in Treating Cocaine-Dependent Individuals - 2 (NCT00134901), a randomized phase 2 study. The primary outcome is Weekly Cocaine Use.

2. Clinical Question

The question concerns a continuous measure of weekly use, rather than simply whether a participant ever used cocaine or achieved abstinence. A lower group mean represents fewer reported cocaine-use days per week on average. The primary comparison estimates the difference between the group means; it does not by itself describe every participant’s pattern of use or establish why any observed difference arose.

The study summary also frames relapse prevention and several other clinical concerns as aims. The registry record lists one secondary outcome, cocaine abstinence based on daily self-reported cocaine use. It does not list separate posted outcomes for craving, psychological symptoms, functional impairment, or discontinuation of treatment. Their measurement and results are therefore not reported in the registry record. The clinical question that can be assessed from the posted results is narrower than the full set of aims described in the summary.

3. Trial Design

The registry identifies this as a phase 2, randomized, triple-masked, parallel-group trial with memantine and placebo arms. Its official title describes it as double-blind and placebo-controlled, while the record’s masking classification is triple. The record does not define which groups of people were masked, so the specific masking arrangements are not reported in the registry record. The trial is marked completed.

The registered enrollment is 81. The posted primary outcome results include 39 participants in the memantine arm and 42 in the placebo arm. Enrollment and the numbers attached to these outcome summaries are not interchangeable with a detailed account of recruitment, treatment exposure, or retention. The record does not provide a participant flow description that explains how people moved from enrollment to analysis, or whether every enrolled participant contributed to each outcome.

The record lists a start date of 2003-03 and primary completion date of 2007-05. It does not give further detail about recruitment duration, follow-up visits, treatment dose, treatment length, adherence, or how the length of each participant’s participation was determined. The primary endpoint timeframe is described as weekly use during the length of study participation, while the secondary abstinence timeframe refers to reported weekly cocaine use for 12 weeks or study participation.

The two arms are Memantine and Placebo. Random allocation and masking are design features intended to support fair comparisons, but the registry record does not state the randomization method, allocation concealment procedures, or masking procedures. It also does not report the trial’s original sample-size justification or its prespecified statistical analysis plan. These omissions limit how fully the design and analysis decisions can be evaluated from the registry record alone.

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 entered a 2-week, single-blind placebo lead-in period. To be eligible for randomization, patients were required to attend at least 2 of 4 scheduled therapy sessions and to submit at least 4 of 6 scheduled urine samples during the two weeks of the lead-in period.

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.

6. Primary Endpoint

The registry lists the primary outcome as Weekly Cocaine Use, measured over weekly use during length of study participation.

The primary endpoint is weekly cocaine use, measured as the mean number of cocaine-use days per week. It is a continuous outcome expressed in days. The registry describes the measure as based on self-reported use, verified by cocaine toxicology results, over the length of each participant’s study participation.

A weekly count captures frequency of use over time rather than only a single yes-or-no status. Comparing group means gives a concise summary of the average reported frequency in each arm. In the posted results, the memantine mean is 1.1 days per week and the placebo mean is 1.3 days per week. The reported mean difference, memantine minus placebo, is -0.2 days per week. The negative direction indicates a lower mean in the memantine arm, but the uncertainty around that estimate is important when interpreting its size and direction.

Using toxicology results to verify self-report can add an objective source of information to a measure that otherwise depends on participant reporting. The registry does not explain how reports and toxicology findings were reconciled, how often toxicology testing occurred, or how missing or discordant measurements were handled. It also does not report individual-level trajectories. Consequently, the group mean is useful for the registered comparison but cannot show the range of experiences or changes over time for individual participants.

7. Results Posted in the Registry

• Memantine: mean 1.10 versus 1.30 on control; difference -0.20 (95% CI -1.01 to 0.61), standardized effect d = -0.11; z-test p = 0.627.

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.

Cocaine Abstinence Based on Daily Self Reported Cocaine Use

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Memantine: 38.5% (15/39) versus 31.0% (13/42) on control; risk difference 7.5 percentage points (95% CI -13.2 to 28.2), risk ratio 1.24 (0.68–2.27); unadjusted two-proportion z-test p = 0.478.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Memantine and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists one secondary endpoint: cocaine abstinence based on daily self-reported cocaine use. The posted group counts and comparison have a direction numerically consistent with the primary point estimate, in that the memantine arm has the higher observed value on the respective outcome measures. The secondary comparison remains uncertain, however, and does not provide a clear independent confirmation of the primary result.

The two endpoints measure different aspects of use. Weekly cocaine-use days summarize frequency as a continuous mean, while abstinence classifies participants according to a binary outcome. It is possible for group patterns to align directionally without yielding precise estimates on either scale. In general, concordant directions can be descriptively informative, but they do not remove uncertainty or turn separate outcomes into one definitive result.

The secondary p-value is unadjusted, and the registry record lists no reported secondary analyses. The record does not state how many outcomes were prespecified for analysis beyond the outcome listings, whether there were other planned comparisons, or whether any multiplicity procedure was used. When several outcomes or comparisons are examined, the chance of at least one apparently unusual result can increase. General statistical reasoning: multiplicity adjustments and prespecified outcome hierarchies are ways to address that concern, but whether they were used in this trial is not reported in the registry record.

For that reason, an unadjusted secondary comparison generally carries less confirmatory weight than a prespecified primary comparison analyzed under a documented plan. This is not a claim that the secondary measure is unimportant; it reflects the limits of interpreting multiple results without a complete account of planning and analysis. The secondary outcome adds context about the pattern in the posted results, but it should not be used to override the primary estimate or to imply a conclusion not supported by the uncertainty around both comparisons.

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

• Cocaine Abstinence Based on Daily Self Reported Cocaine Use

9. Statistical Interpretation

10. Statistical Methods Explained

Statistical methods help translate group observations into estimates while making uncertainty visible. The primary outcome is continuous: each participant’s outcome is summarized in cocaine-use days per week, and the posted arm summaries give a mean and standard deviation. The comparison is a difference between means. In this record, the contrast is defined as memantine minus placebo, so a negative estimate points toward a lower mean in the memantine arm. The direction is a convention for describing the contrast, not a statement about clinical importance.

What does a mean difference describe?

A mean is the arithmetic average of the observations in a group. The difference between two means reports how far apart the group averages are in the endpoint’s measurement units. Here those units are days per week. A mean difference condenses the comparison into one value, but it does not show the distribution within either arm. Participants may vary substantially even when their group means are close. Standard deviations describe spread around each group mean; they are not confidence intervals and do not express uncertainty about the estimated difference in the same way.

The posted means are 1.1 days per week for memantine and 1.3 for placebo, with standard deviations of 1.9 and 1.8, respectively. Those arm-level summaries provide context for the estimate. The registry’s comparison gives a mean difference of -0.2. The reported uncertainty interval and p-value concern the between-group comparison, not the spread of individual outcomes. This distinction matters: the standard deviation describes observed variability among participants, whereas the confidence interval describes uncertainty in the estimated contrast under the analysis assumptions.

How should a confidence interval be read?

A confidence interval is a range generated by a statistical procedure that reflects sampling uncertainty in an estimate. In repeated use under its assumptions, the procedure has a specified coverage property. For a reader, the interval indicates which values are relatively compatible with the data and analysis, rather than identifying a set of values that are certainly true. The primary interval crosses zero, the value corresponding to no difference between the group means. It therefore includes the possibility of no mean difference and values on either side of zero.

The width of an interval is informative. A narrow interval indicates more precision than a broad interval, all else equal. The primary interval is broad relative to the point estimate, so the exact size and direction of the underlying difference remain uncertain. An interval should be assessed against the endpoint scale and against any prespecified threshold for meaningful differences. No such threshold is reported in the registry record, so statistical uncertainty can be described, but a formal judgment against a clinical threshold cannot be made from this record.

What does the p-value add?

The unadjusted p-value for the primary comparison is 0.627. A p-value measures how inconsistent the observed result, or a more extreme result, would be with a specified null model if that model and the test assumptions were correct. It is not an effect-size measure. A larger p-value does not demonstrate that there is no effect, and a smaller p-value would not by itself establish that a difference is important to participants. Here, the p-value and interval tell a compatible story: the data do not provide strong evidence against a no-difference model, while the interval also makes clear that the estimate is not highly precise.

The registry labels the p-value unadjusted. That means the value is reported without a stated correction for examining multiple outcomes or comparisons. Whether and how the trial’s original analysis handled multiplicity is not reported in the registry record. Therefore the p-value should be interpreted as the posted unadjusted comparison rather than as a fully documented confirmatory analysis.

Why do assumptions matter?

Any statistical comparison relies on assumptions about the outcome, sampling, and analysis. A mean-based comparison is often explained using a model in which the estimate has an approximately well-behaved sampling distribution. General statistical reasoning: when outcomes are highly variable, heavily skewed, or concentrated at a boundary, the model and method used to estimate uncertainty can matter. The primary endpoint is a count of days per week, which is bounded and may be variable. The registry reports means and standard deviations but does not report distributional checks, the precise test used, or any alternative model. Those analytic details are not reported in the registry record, so the calculations cannot be evaluated against unreported diagnostics.

General statistical reasoning: over-dispersion occurs when observed variability is greater than a simple model assumes. If a count outcome is over-dispersed, a model that assumes less variability may understate uncertainty, while a model designed to allow extra variability can produce more appropriate uncertainty estimates. This is a general consideration for count outcomes, not a claim that over-dispersion was present or that a particular model was used in this trial. The registry does not report a count-model analysis.

The reported standard deviations and comparison interval give useful descriptive and inferential summaries, but they do not reveal how repeated weekly observations were combined or whether within-person correlation was modeled. Because the endpoint describes weekly use during study participation, participants may have contributed observations across time. The record does not specify whether the mean was calculated first for each participant and then compared across arms, or whether another analysis used the repeated measurements directly. That distinction can affect the interpretation of the estimate and its standard error.

What is the secondary binary comparison?

The secondary endpoint records the number of participants classified as abstinent based on daily self-reported cocaine use. The posted counts are 15 of 39 in the memantine arm and 13 of 42 in the placebo arm. A risk difference expresses the difference in the proportions meeting that binary outcome, with the comparison direction defined as memantine minus placebo. Its confidence interval crosses zero, and the unadjusted p-value is 0.478. The interval is more informative than the point estimate alone because it shows the uncertainty around the proportion difference.

General statistical reasoning: binary outcomes can also be expressed using relative measures, such as a risk ratio or odds ratio. Such measures compare proportions on a relative scale, while a risk difference is expressed on an absolute proportion scale. The posted registry results include relative estimates, but the risk difference is often more direct for understanding the change in the proportion of participants meeting the outcome. Each scale answers a related but not identical question. None removes uncertainty, and a relative estimate should not be read as a direct change in the number of participants without considering the underlying group proportions.

The registry reports no secondary analyses, so it does not provide an adjusted model or a documented analysis plan for this endpoint. The posted comparison is unadjusted. As with the primary measure, it should be read as a summary of the posted data and not as a guarantee that the same estimate would hold under other analytic choices or populations.

What do planning calculations mean?

Sample-size calculations describe how many participants might be required under a specified set of assumptions to detect a target effect with a chosen power and significance level. They are planning illustrations, not a measure of how much power the completed trial actually had. The planning assumptions listed for these calculations are a two-sided significance level of 0.05, equal allocation between the arms, and a normal approximation. The assumed effect is set to a stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

General statistical reasoning: a normal approximation uses a mathematical distribution to approximate the variability of an estimator. The quality of that approximation depends on the outcome and assumptions. These planning calculations should therefore be understood as illustrations based on the specified method, not as evidence that the original trial used the same design target or calculations. The posted observed results, their uncertainty, and the planning illustrations answer different questions.

11. Confidence Intervals and Precision

Precision concerns how narrowly the data locate an estimated group difference. For the primary endpoint, the estimated mean difference is -0.2 days per week, and the confidence interval extends from -1.007 to 0.607. The interval crosses zero and includes differences in both directions. This means the point estimate alone gives a more definite impression than the interval supports: the data leave substantial uncertainty about the size and direction of the underlying difference.

The lower and upper limits are measured in weekly cocaine-use days, just like the mean difference. They are not limits for individual participants’ outcomes. Nor do they indicate that every value inside the interval is equally likely. A confidence interval is a procedure-based summary of uncertainty, conditional on the analysis and its assumptions. The registry does not state a threshold for a clinically important difference, so the interval cannot be classified against a registered importance margin.

The reported p-value of 0.627 is consistent with the interval including zero, but it does not replace the interval. The interval communicates a range of values compatible with the data under the method, while the p-value summarizes evidence relative to a null hypothesis. Neither says that the groups are equivalent. A study designed to demonstrate equivalence would need a defined margin and a corresponding analysis; neither is reported in the registry record.

The secondary abstinence comparison also has an interval crossing zero. Its uncertainty means the observed difference in the proportion classified as abstinent is not estimated precisely enough to rule out no difference or a difference in either direction within the interval. Precision is affected by the amount of information in the data and the variability of the outcome. The posted arm counts are modest, but the registry does not provide a precision target or a formal assessment of whether the trial could exclude effects of a particular size. It is more accurate to describe the estimates as uncertain than to treat a nonsignificant result as evidence of no 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 planning figures are participants per arm for 80% power and participants per arm for 90% power. They are not total events. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The calculations use the observed effect as a reference and consider planning for 100%, 75%, or 50% of that observed effect. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm that would be needed for an 80% chance of detecting the assumed effect at the stated significance level, if the planning assumptions held. The 90% column describes the corresponding planning target for a 90% chance. Higher target power generally calls for more participants when other assumptions remain fixed. These are design calculations; they do not report the probability that the completed trial found an effect, and they are not observed power.

For the primary endpoint, the figures increase as the assumed target effect becomes a smaller fraction of the observed effect: planning for 100% of the observed effect gives 1,342 participants per arm for 80% power and 1,796 for 90% power; planning for 75% gives 2,385 and 3,193; planning for 50% gives 5,366 and 7,183. The secondary endpoint’s corresponding figures are 630 and 843, then 1,106 and 1,480, then 2,455 and 3,286. Smaller effects are harder to distinguish from random variation, so they require more participants under these assumptions. These values are planning illustrations, not a reconstruction of the trial’s original design.

13. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Observed difference Δ = -0.200; pooled standard deviation σ = 1.849.
2. n = 2 × (1.96 + 0.842)² × 1.849² / -0.200² = 1,341.4, rounded up to 1,342 participants per arm for 80% power.

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

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -0.2                      # observed difference in means
sigma <- 1.8487765369412                     # 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(134901)
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 = 1,342 (80% power), 1,796 (90% power)
 75% of observed effect: n per arm = 2,385 (80% power), 3,193 (90% power)
 50% of observed effect: n per arm = 5,366 (80% power), 7,183 (90% power)
power.t.test: 1342.3 
Simulated power at n = 1,342 per arm: 0.798 (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. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several outcomes, time points, subgroups, or comparisons are tested. If each is considered separately at a chosen significance level, examining more tests can increase the chance of obtaining at least one result that appears statistically unusual by chance. This general concern is relevant to clinical trial interpretation, especially when a study summary describes multiple aims and the registry lists more than one outcome.

The record identifies weekly cocaine use as the primary outcome and lists cocaine abstinence as a secondary outcome. It does not report whether the primary endpoint was the only primary analysis, whether additional outcomes or time points were included in the original analysis plan, or whether any multiplicity adjustment or testing hierarchy was used. The posted p-values are unadjusted. The registry record therefore does not support a claim that multiplicity was controlled, nor does it support a claim that the investigators did not control it outside the posted information.

Missing-data handling is also not reported in the registry record. The outcome timeframe spans study participation, but the record does not describe how incomplete weekly reports, missed toxicology tests, withdrawal, or other unavailable measurements were treated. It does not state whether participants with missing data were excluded, whether an imputation method was applied, or whether sensitivity analyses assessed the effect of missingness. These choices can affect both the estimate and its uncertainty, particularly when missing outcomes are related to treatment or outcome status.

The analysis population is not reported in sufficient detail to identify whether the posted summaries use all randomized participants, only participants with observed outcomes, or another defined group. The arm-level sample counts are given for the outcomes, but the registry does not explain the relationship between those counts and enrollment or provide a participant flow account. The precise statistical model and the handling of repeated weekly observations are also not reported in the registry record.

These gaps do not invalidate the posted summaries, but they limit reproducibility and make it harder to assess possible sources of bias. A fuller interpretation would require the prespecified analysis plan, definitions of analysis populations, missing-data procedures, and an account of all planned and reported comparisons. Without those details, the posted unadjusted results are best treated as the record’s available estimates rather than a complete account of the trial’s inferential framework.

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: abdominal pain (2.4%), finger fracture (2.4%).

The registry lists serious adverse events for 0 of 39 participants in the memantine arm and 2 of 42 in the placebo arm. These are counts of participants reported as affected, not rates adjusted for time at risk. The listed serious-event terms are abdominal pain and finger fracture, each reported as 2.4%. The record does not provide additional detail about event timing, duration, resolution, relatedness to treatment, or how seriousness was defined for these cases.

The record also reports other adverse events affecting 24 of 39 participants receiving memantine and 28 of 42 receiving placebo. These are distinct from the serious adverse-event counts. The posted figures summarize what was recorded within each arm; they do not establish that a treatment caused an event or that an event was unrelated. The registry does not report a comparative statistical analysis of adverse events.

Deaths are not listed as a count in the registry record. Further details about safety monitoring and event ascertainment are not reported in the registry record. These figures should therefore be read as descriptive safety reporting, not as a complete assessment of relative risk or proof of a causal relationship.

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

The posted data show a small negative point estimate for the primary comparison: the mean number of weekly cocaine-use days is lower in the memantine arm than in the placebo arm. The interval around that difference crosses zero, and the unadjusted p-value is 0.627. This combination means the record does not establish a clear difference between the arms on the primary endpoint. It also does not establish that the treatments have exactly the same effect.

The secondary abstinence comparison is similarly uncertain. Its observed direction is consistent with the primary point estimate, but its confidence interval crosses zero and its unadjusted p-value is 0.478. Since the endpoint definitions differ, the secondary result is additional context rather than a direct replication of the primary measure. Neither result provides a precise estimate that rules out all meaningful differences in either direction.

These are group-level findings from a phase 2 randomized trial with 81 enrolled participants. Group averages and proportions cannot predict an individual person’s outcome. The registry record does not provide enough information to determine how missing outcome data, repeated weekly measurements, or the analysis population were handled. It also does not include the full statistical analysis plan or original sample-size justification. Those limitations affect how confidently the posted estimates can be generalized or reconstructed.

The results should not be read as proof of benefit, proof of no effect, or a basis for treatment decisions. The statistical interpretation is narrower: the posted comparisons estimate differences in the recorded outcomes, and the uncertainty intervals include no difference. The trial’s results do not answer every aim described in its study summary, because posted results for several aims are not reported in the registry record. Broader conclusions would require fuller reporting and evidence beyond these summaries.

17. Why This Trial Matters Statistically

This trial illustrates why a point estimate, its uncertainty interval, and the p-value need to be read together. The primary estimate points toward fewer weekly cocaine-use days in the memantine arm, but the interval crosses zero and the p-value is not small. Reporting only the direction of the mean difference could overstate what the data establish; reporting only that the comparison was not statistically significant could conceal the range of effects that remain compatible with the estimate.

It also highlights the difference between absence of strong evidence for a difference and evidence of equivalence. A confidence interval spanning zero does not prove that two arms are the same. Establishing equivalence or ruling out a difference of practical importance requires a meaningful threshold and an analysis designed around that threshold. The registry record does not report such a margin or analysis.

The posted secondary outcome offers another statistical lesson. A second measure can add context, but it does not automatically strengthen a conclusion, particularly when the analysis plan and multiplicity handling are not reported. Different outcome scales can point in similar directions while both remain imprecise. Transparent reporting of prespecified outcomes, analysis populations, and missing-data procedures helps readers distinguish confirmatory evidence from exploratory or descriptive patterns.

Finally, the planning illustrations demonstrate how assumptions shape sample-size requirements. The calculations use the observed effect as a reference, while smaller target effects require larger planned samples. Such illustrations are not evidence of the original trial’s rationale or of the completed trial’s observed power. Distinguishing design planning from result interpretation is essential when reading registry results, especially when the original sample-size assumptions are not reported.

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.

Several limitations are specific to the information available for this trial. The primary endpoint combines weekly self-report with toxicology verification, but the registry record does not describe the testing schedule or the rules used when self-report and toxicology results differed. It also does not explain how weekly observations were combined into the reported mean, or how participants with different lengths of participation contributed to the summary. The endpoint’s timeframe is therefore clear in broad terms but not in its operational details.

The registry does not report treatment dose, adherence, treatment duration, detailed eligibility criteria, or the participant flow from enrollment through outcome assessment. These omissions make it difficult to assess exposure, retention, and how representative the outcome summaries are of all randomized participants. Although arm-level outcome counts are posted, the record does not clarify whether they correspond to an intention-to-treat population or another analysis set.

The trial’s statistical analysis plan, original sample-size justification, and handling of missing data are not reported in the registry record. The exact model for the continuous endpoint, any method for repeated observations, distributional checks, and sensitivity analyses are likewise not reported. The posted comparisons are unadjusted; whether multiplicity was addressed in the original analysis is not reported. These gaps limit assessment of analytic choices and the robustness of the estimates.

The registry lists only one secondary outcome and does not post analyses for it, even though the study summary describes several additional aims. Results concerning craving, psychological symptoms, functional impairment, and treatment discontinuation are not reported in the registry record. Safety reporting is also limited: the record gives affected counts and serious-event terms, but not event timing, treatment relatedness, or a comparative safety analysis. These limitations call for a restrained interpretation of the registry results rather than assumptions about unreported procedures or outcomes.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Effectiveness of Memantine in Treating Cocaine-Dependent Individuals - 2 — NCT00134901.
• Registry record last updated 2019-04-24; status last verified 2019-04.
• Results first posted 2014-09-05.
• 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

NCT00134901 is a completed phase 2, randomized, triple-masked, parallel-group trial comparing memantine with placebo in people with cocaine dependence. The registry lists enrollment of 81 and dates of 2003-03 for study start and 2007-05 for primary completion.

The primary outcome was mean weekly cocaine-use days, based on self-report verified by toxicology results. The posted mean difference was -0.2 days per week, with a confidence interval crossing zero and an unadjusted p-value of 0.627. The one listed secondary outcome concerned cocaine abstinence based on daily self-report; its unadjusted comparison was also uncertain.

The registry record does not report the original sample-size justification, full statistical analysis plan, missing-data procedures, or analysis population. The available findings describe the posted comparisons and should not be taken as proof of either a treatment difference or equivalence.