Cognitive Remediation for Cocaine Dependence: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cognitive Remediation for Cocaine Dependence (NCT01393457), a randomized phase 2 study. The primary outcome is Cocaine Use Based on Urine Drug Screening.

2. Clinical Question

3. Trial Design

This was a completed Phase 2 randomized, parallel-group trial of medication strategies for cocaine dependence. The registry record lists enrollment of 119 and four groups: levodopa/carbidopa 800/200 mg/d with ropinirole 2 mg/d, levodopa/carbidopa 800/200 mg/d with ropinirole 4 mg/d, levodopa/carbidopa 800/200 mg/d alone, and placebo. The stated rationale was to test treatments intended to improve cocaine treatment success by combining medications that target cocaine-related cognitive impairments.

The trial is recorded as quadruple-masked. The record does not specify which trial roles were masked, how masking was maintained, or whether it changed during the study. It also does not report details of sequence generation, allocation concealment, treatment adherence, or the criteria for discontinuing treatment. These details are not reported in the registry record.

The registry lists a start of 2011-06 and primary completion of 2017-02. Those dates describe the record’s trial timeline; they do not, by themselves, explain why the study took that length of time. A protocol schedule, recruitment timeline, or reasons for the timing are not reported in the registry record.

The primary outcome was assessed over 10 weeks in an intention-to-treat population. The planned endpoint description refers to predicted probabilities derived from generalized linear mixed models and averaged over the 10 weeks. A secondary outcome recorded the number of participants who completed the 10 Week Trial. The registry lists two outcome measures, one secondary outcome, and no posted comparative analyses. Its record does not state the original sample-size justification or other design assumptions.

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: 119 were enrolled (that is, signed the consent); however, only 110 were randomized. Of the 9 that were not enrolled but not randomized, 1 was no longer interested after signing the consent, and 8 were lost to follow up.

Overall Study

Analysis population for the primary outcome, as defined in the registry: intention to treat

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 Cocaine Use Based on Urine Drug Screening, measured over 10 weeks.

The primary endpoint is titled Cocaine Use Based on Urine Drug Screening. It is expressed as the number of negative drug screens, with a mean and standard deviation reported for each group. The endpoint description says that the reported quantity is the mean of predicted probabilities of a negative screen across all 10 weeks, derived from generalized linear mixed models, as proposed in the protocol. The analysis population is listed as intention to treat.

Urine screening provides an objective measure of cocaine-use status at the times samples are collected. A negative result is the endpoint’s favorable screening state, but the registry’s summary does not provide the week-by-week observations, the number of samples available per participant, or a detailed explanation of how missing screens entered the model. The model-based average is intended to represent the repeated screening outcome across the assessment period, rather than a single test at one visit.

The registry’s reported unit and its description of averaging predicted probabilities should be read together; the record does not further clarify the scale transformation or aggregation. The model can account for the repeated structure of observations in principle, but the specific fixed effects, random effects, covariance structure, and missing-data assumptions are not reported in the registry record. Consequently, the endpoint summaries alone do not reveal those modeling choices.

7. Results Posted in the Registry

• Ldopa + Ropinirole Low Dose: mean 20.36 versus 27.80 on control; difference -7.44 (95% CI -23.56 to 8.68), standardized effect d = -0.25; z-test p = 0.366.
• Ldopa + Ropinirole High Dose: mean 15.57 versus 27.80 on control; difference -12.23 (95% CI -26.45 to 1.99), standardized effect d = -0.46; z-test p = 0.092.
• Ldopa: mean 22.30 versus 27.80 on control; difference -5.50 (95% CI -20.61 to 9.61), standardized effect d = -0.19; z-test p = 0.475.

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.

Number of Participants Who Completed the 10 Week Trial

• 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:
  • Ldopa + Ropinirole Low Dose: 53.8% (14/26) versus 55.6% (15/27) on control; risk difference -1.7 percentage points (95% CI -28.5 to 25.1), risk ratio 0.97 (0.59–1.58); unadjusted two-proportion z-test p = 0.901.
  • Ldopa + Ropinirole High Dose: 53.8% (14/26) versus 55.6% (15/27) on control; risk difference -1.7 percentage points (95% CI -28.5 to 25.1), risk ratio 0.97 (0.59–1.58); unadjusted two-proportion z-test p = 0.901.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Ldopa + Ropinirole Low Dose and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Ldopa + Ropinirole High Dose and the control. The interval gives the range of effects that remain compatible with the data.

The registry record lists one secondary outcome: the number of participants who completed the 10 Week Trial. Completion is a retention measure, whereas the primary endpoint reflects the urine-screen outcome across the assessment period. These outcomes address different aspects of the trial. Completion information can provide context about whether participants remained through the study period, but it does not itself explain the pattern of urine-screen results or show how missing screens affected the primary model.

Across the recorded groups, completion counts appear broadly similar in relation to the group sizes. This overall pattern offers a limited descriptive view of retention; it does not establish that missing outcome data were balanced or that the primary analysis was unaffected by attrition. Completion at the end of the trial does not necessarily mean that every scheduled screen was available, and the registry record does not provide participant-level completion histories or reasons for non-completion.

The secondary comparisons are unadjusted and no reported analyses are listed in the registry record. General statistical practice gives less weight to unadjusted secondary findings when many outcomes or contrasts may be examined, because repeated testing can increase the chance of apparently notable results arising by chance. The record does not report a multiplicity plan, a hierarchy of outcomes, or adjusted secondary tests. The secondary outcome should therefore be read as descriptive context rather than as independent confirmation of the primary result.

Completion is relevant to interpreting a 10-week repeated-screen endpoint, but it is not a substitute for a stated missing-data method. The primary population is labeled intention to treat, yet the registry record does not explain whether all randomized participants contributed an outcome, how incomplete screening histories were modeled, or whether sensitivity analyses were performed. No conclusions about the robustness of the primary result to missing data can be drawn from the completion measure alone.

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

• Number of Participants Who Completed the 10 Week Trial

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous as posted: each group has a mean and a standard deviation. A mean locates the center of a group’s observed endpoint values, while a standard deviation describes their spread. The standard deviation is not the uncertainty of the mean. Uncertainty about a group mean or a between-group contrast depends on the sample size and the variability, among other factors.

Repeated screening and the stated model

The endpoint description says the mean was formed from predicted probabilities of negative drug screens across all 10 weeks, using generalized linear mixed models. In general, a generalized linear mixed model can relate an outcome to explanatory variables while representing dependence among repeated observations from the same participant. Its generalized component accommodates outcomes that do not follow the ordinary continuous normal model, and its mixed component can account for participant-level clustering. These are general properties of the method, not a complete description of the particular analysis.

The registry record does not report the model formula, link function, covariates, random-effects structure, or covariance assumptions. It also does not say how the predicted probabilities were averaged, whether treatment-by-time terms were included, or how missing weekly screens were handled. Therefore, the description establishes the intended broad method, but not enough detail to reproduce or fully evaluate its implementation.

How to read the recalculated contrasts

The posted pairwise comparisons are labeled mean differences and use placebo as the reference group. Each recalculated estimate is the active group’s mean minus the placebo mean. A negative difference therefore indicates a lower reported mean for that active group. This direction alone does not establish the meaning of the model-based treatment effect, because the contrast is recalculated from posted summary statistics rather than taken from a reported mixed-model analysis.

The confidence intervals and p-values accompanying these comparisons are unadjusted recalculations based on the posted group summaries. In general, a confidence interval represents a range of values reasonably compatible with the data and the calculation’s assumptions. An interval that includes zero indicates that the data do not rule out a zero mean difference at the corresponding confidence level. A p-value summarizes how unusual the observed data or a more extreme result would be under a specified null hypothesis and statistical model; it is not the probability that the null hypothesis is true.

These calculations do not adjust for potential covariates or for repeated within-person measurements. Nor do they incorporate any model terms that may have been used in the protocol analysis. Although the endpoint is described as model-derived, the registry record does not post model-based comparison estimates or test results. The simple recalculations should not be presented as the primary analysis that the protocol proposed.

Why over-dispersion matters

In general, over-dispersion occurs when an outcome varies more than a chosen statistical model expects. If it is not accommodated, standard errors can be too small and confidence intervals can be too narrow, making evidence appear more precise than it is. A mixed model or another suitable approach may address aspects of variation and dependence, but that depends on the model specification and fit. The registry record does not report a dispersion assessment or diagnostics, so it is not possible to say whether over-dispersion was present or how it was handled in the planned analysis.

The posted standard deviations describe variation in the group summaries, but they do not diagnose over-dispersion in the repeated screening data. That assessment would require details or results from the fitted model. General statistical considerations about dispersion therefore explain why model reporting matters; they do not establish a problem with this trial’s analysis.

Tests and analysis reporting

The registry record lists no reported analyses and no analysis count. The primary endpoint description refers to an analysis proposed in the protocol, but the actual model estimates, standard errors, test statistics, and adjusted contrasts are not reported in the registry record. The available pairwise p-values are explicitly unadjusted recalculations. They can describe the posted summaries but do not fill the gap left by absent protocol-analysis results.

Because three active groups are compared with a single control, an analysis plan would ordinarily need to clarify whether the contrasts were co-primary, how their error rates were controlled, and whether any overall test was planned. These are general statistical considerations. The registry record does not report such a plan, and no inference about the intended multiplicity procedure is warranted.

11. Confidence Intervals and Precision

Precision concerns how tightly the data estimate a quantity, not whether an estimate points in a particular direction. The recalculated mean differences have wide 95% confidence intervals: -7.44 with an interval from -23.56 to 8.676 for the low-dose combination; -12.23 with an interval from -26.45 to 1.993 for the high-dose combination; and -5.5 with an interval from -20.61 to 9.606 for levodopa/carbidopa alone. Every interval includes zero. Each also spans values on both sides of zero, so the recalculations leave substantial uncertainty about the direction and size of the contrast.

In general, a confidence interval is a range generated by a statistical procedure that, under its assumptions, has the stated coverage over repeated use. It is not a range containing a fixed percentage of individual participants’ outcomes, and it does not assign a probability to the true effect after the data are observed. Here the intervals are unadjusted recalculations from group summaries, not intervals reported for the protocol’s generalized linear mixed model.

The arm-level standard deviations are 29.51, 21.96, 28.0, and 30.35. The comparison groups contain 26.0, 26.0, 31.0, and 27.0 participants, respectively. In general, greater variability and limited sample size make estimates less precise. The registry record does not post arm-specific confidence intervals for the means or uncertainty measures from the planned repeated-measures model.

Accordingly, the posted comparisons do not precisely establish a difference or equivalence between any active group and placebo. A failure to exclude zero is not proof that the groups are identical; an interval that includes zero can also remain compatible with potentially meaningful effects in either direction. The registry record does not state a threshold for a clinically important difference, so the clinical importance of values within these intervals cannot be determined from the posted information alone.

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 based on the observed effects, not on the trial’s original design rationale. For the primary continuous endpoint, the figures mean participants per arm. They use the posted sample summaries, a two-sided significance level of 0.05, equal allocation between the two comparison arms, and a normal approximation. The effect being planned for is the stated fraction of the observed effect: 100%, 75%, or 50%. The registry record does not report the trial’s original sample-size assumptions or justification.

The 80% power column describes the number of participants per arm that the planning calculation estimates would be needed for an 80% chance of detecting the specified effect under those assumptions, if that effect were real. The 90% power column gives the corresponding planning target for a 90% chance. These are design calculations conditional on the assumptions, not guarantees about a future study’s result. They should not be read as an assessment of the probability that the observed trial result is correct.

For each comparison, planning for 75% or 50% of the observed effect requires more participants than planning for 100% of the observed effect. In general, smaller effects are harder to distinguish from random variation, so they require more participants to achieve the same planned power. The large variability in the posted primary outcome summaries also matters to these calculations. These numbers are not observed power.

The figures are not total events: the primary endpoint is continuous, and the planning output is participants per arm. The calculation is not a reconstruction of what the investigators originally expected when they designed the trial. A curve associated with these planning values should be read as illustrating how the required sample size changes as the target effect changes under the stated assumptions, rather than as a record of actual enrollment or results.

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 = -7.440; pooled standard deviation σ = 29.941.
2. n = 2 × (1.96 + 0.842)² × 29.941² / -7.440² = 254.2, rounded up to 255 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 <- -7.44                      # observed difference in means
sigma <- 29.9411800820522                     # 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(1393457)
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 = 255 (80% power), 341 (90% power)
 75% of observed effect: n per arm = 452 (80% power), 606 (90% power)
 50% of observed effect: n per arm = 1,017 (80% power), 1,362 (90% power)
power.t.test: 255.2 
Simulated power at n = 255 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

The registry record does not report how multiplicity was addressed. There are three active-versus-placebo comparisons for the primary endpoint, and a secondary completion outcome is also posted. When multiple hypotheses are tested, the chance of at least one apparently unusual result can increase. General approaches include defining a testing hierarchy or adjusting the testing threshold, but the record does not state that any particular approach was used. The recalculated p-values are unadjusted and should not be interpreted as multiplicity-controlled results.

The record describes the primary endpoint analysis as based on generalized linear mixed models, but lists no reported analyses. It does not provide model estimates, tests, covariate adjustments, or an analysis plan for the multiple treatment groups. The endpoint description says the analysis was proposed in the protocol; details of the proposal beyond the broad method and intention-to-treat population are not reported in the registry record. Consequently, the recalculations cannot establish what the protocol’s primary inference would have been.

The intention-to-treat label identifies the stated primary analysis population, but does not specify the analysis set in operational detail. The record does not say whether participants were analyzed according to random assignment, how participants with no post-baseline screens were handled, or how incomplete weekly observations contributed. It also does not report imputation, likelihood-based missing-data assumptions, or sensitivity analyses. In general, missing-data methods rely on assumptions about why observations are unavailable; those assumptions can affect estimates and uncertainty.

The completion outcome offers a group-level count of participants who finished the 10-week trial, but that information does not describe missingness at each screening visit. The registry record does not give reasons for discontinuation or the timing and amount of missing screening data. It is therefore not possible to judge whether attrition was related to treatment group or outcome status, or whether the planned model adequately addressed it.

The original sample-size justification, interim-analysis plan, and detailed statistical analysis plan are not reported in the registry record. No assumptions about them should be inferred from the planning illustrations, which are separate calculations based on observed effects. These gaps limit assessment of prespecification, multiplicity, and the potential sensitivity of the primary result to analysis choices.

15. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: suicide ideation (3.8%), chest pain (3.7%), weight gain (3.7%), gout flare up (3.7%).

The adverse-event summary reports serious adverse events affecting 0 participants in the low-dose combination group, 1 participant in the high-dose combination group, 0 participants receiving levodopa/carbidopa alone, and 3 participants receiving placebo. The numbers at risk were 26, 26, 31, and 27, respectively. The record reports 0 deaths in each group. It also reports 0 participants affected by other adverse events in every group.

These are counts of participants affected, not a full account of all events, their timing, duration, or severity. The registry lists suicide ideation at 3.8%, chest pain at 3.7%, weight gain at 3.7%, and gout flare up at 3.7% among the top serious terms. The record does not assign those listed terms to treatment groups in the available summary, so they should not be attributed to a particular arm.

In general, safety counts in a trial of this size provide limited precision for uncommon events. Differences in counts alone do not establish a treatment-related cause, and the registry record does not provide a comparative safety analysis. The adverse-event summary should be interpreted as descriptive reporting for the recorded trial groups.

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

The registry results describe a randomized, quadruple-masked, parallel-group Phase 2 trial comparing three medication strategies with placebo in people with cocaine dependence. The primary outcome concerns negative urine drug screens averaged over 10 weeks using predicted probabilities from a generalized linear mixed model, as described in the record. The posted arm summaries show variation in group means, and the unadjusted recalculated contrasts have confidence intervals that include zero.

This means the posted summary-based comparisons do not precisely distinguish the active groups from placebo on the reported primary measure. It does not establish that the groups are the same, nor does it establish that one regimen is superior. Wide intervals can remain compatible with effects in either direction. The registry does not define a clinically important difference, so statistical uncertainty cannot be translated into a judgment about clinical importance from these results alone.

The recalculated estimates are not the model-based comparisons described by the endpoint methodology. The record lists no reported comparative analyses, and therefore does not show the planned model results or their uncertainty. Interpretation should keep that distinction clear: descriptive arm summaries and unadjusted recalculations are available, while the actual comparative findings from the stated repeated-measures model are not reported in the registry record.

The trial’s outcome period was 10 weeks, and completion was recorded as a secondary measure. The results do not establish what would happen beyond that period. Nor do they answer questions about all people with cocaine dependence, because the registry record does not provide further details about the enrolled population and eligibility criteria. The reported safety counts are also descriptive and do not establish causation or rule out uncommon risks.

Overall, the record supports a cautious account of observed data and their imprecision, not a treatment recommendation. It leaves important questions about the model-based analysis, missing screening observations, multiplicity, and the original design assumptions unanswered. Those limitations constrain the strength and scope of any interpretation.

17. Why This Trial Matters Statistically

Several statistical features make this trial useful to examine. First, the primary outcome is based on repeated urine screens over 10 weeks, while the registry describes a model that produces predicted probabilities and averages them across time. This illustrates why a repeated-measures endpoint needs a clear account of both the outcome scale and the model used to combine observations within participants.

Second, the record contains four randomized groups but reports no comparative analysis from the stated model. Summary means and standard deviations allow simple unadjusted contrasts to be calculated, yet those contrasts cannot capture unreported features of a repeated-measures analysis. The distinction between a posted arm summary and a protocol-based model estimate is important when evaluating how directly the available numbers answer the primary question.

Third, the uncertainty is central to interpretation. The recalculated confidence intervals are broad and include zero, while the observed group variability is substantial. An estimate’s direction alone does not convey how precisely the data locate the underlying contrast. Confidence intervals make that limitation visible and discourage overstating small-trial findings.

Finally, the registry lists a secondary completion outcome but does not report detailed missing-data methods, multiplicity procedures, or the original sample-size rationale. These omissions show why transparent statistical reporting is part of evidence assessment: readers need to know how repeated observations, multiple comparisons, incomplete data, and design assumptions were handled. The sample-size illustrations can help show the planning implications of different target effects, but they do not reconstruct the trial’s original design or replace the missing model results.

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 trial-specific limitations follow from the registry record. The randomized groups were modest in size, with 26.0 participants in each combination group, 31.0 in the levodopa/carbidopa group, and 27.0 in placebo. The primary outcome summaries have substantial reported standard deviations, and the recalculated confidence intervals are broad. The available summary therefore gives limited precision for each active-versus-placebo contrast.

The primary endpoint is described as an average of model-predicted probabilities across repeated weekly assessments, but the registry does not report the model specification or its comparative estimates. The posted unadjusted recalculations use group summaries and cannot reproduce the stated generalized linear mixed-model analysis. No diagnostics, model checks, or sensitivity analyses are reported.

Further limitations concern incomplete reporting. The registry record does not describe missing screening-data handling, reasons for trial non-completion, the original sample-size justification, multiplicity control, or interim analyses. Although intention to treat is listed as the primary population, the record does not clarify how that population was implemented in the analysis. The secondary completion outcome cannot resolve these uncertainties about missing repeated screens.

The record also gives limited detail about eligibility, treatment adherence, and the masking process. Safety summaries provide counts but not a comparative analysis or arm-specific attribution for the listed serious-event terms. The primary outcome covers 10 weeks; the record does not provide evidence about outcomes beyond that period. These limitations do not invalidate the posted summaries, but they restrict what can be concluded from them.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Cognitive Remediation for Cocaine Dependence — NCT01393457.
• Registry record last updated 2018-03-29; status last verified 2018-03.
• Results first posted 2018-03-29.
• 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

NCT01393457, titled Cognitive Remediation for Cocaine Dependence, is recorded as a completed Phase 2 randomized, quadruple-masked, parallel-group trial sponsored by The University of Texas Health Science Center, Houston. The registry lists enrollment of 119, a start of 2011-06, and primary completion of 2017-02.

Four groups received levodopa/carbidopa with ropinirole 2 mg/d, levodopa/carbidopa with ropinirole 4 mg/d, levodopa/carbidopa alone, or placebo. The primary outcome was cocaine use based on urine drug screening over 10 weeks, described as an intention-to-treat average of predicted probabilities from generalized linear mixed models.

Arm summaries and recalculated unadjusted comparisons are available, but the registry record lists no reported comparative analyses. The primary comparison intervals are wide and include zero. One secondary outcome records trial completion. The record does not report detailed model specifications, missing-data methods, multiplicity procedures, or the original sample-size rationale.