Mecamylamine for the Treatment of Patients With Depression and Alcohol Dependence: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Mecamylamine for the Treatment of Patients With Depression and Alcohol Dependence (NCT00563797), a randomized phase 3 study. The primary outcome is Number of Drinking Days.

2. Clinical Question

The registry describes a superiority hypothesis: mecamylamine was expected to reduce depressive symptoms and alcohol consumption relative to placebo. That is the study’s stated hypothesis, not a conclusion about what the posted results establish. The primary endpoint is a continuous measure, reported as a group mean with standard deviation, and the comparison is a difference in means.

The distinction between the planned question and the reported evidence matters. The registry identifies the primary endpoint’s measurement period and method, but the information does not give a full endpoint definition beyond drinking days measured with time line follow back measures. It also does not report a primary outcome estimate from the sponsor’s model. The numerical group means and an unadjusted re-calculation offer one view of the posted endpoint, while the registry lists a separate sponsor analysis with a p-value and no corresponding estimate or interval. Those results should not be treated as interchangeable.

3. Trial Design

This was a Phase 3, randomized, double-blind, parallel-group trial sponsored by Yale University. The registry lists the study as completed, with enrollment of 21. The recorded start was 2007-08 and primary completion was 2014-07. The trial compared drug interventions: mecamylamine and placebo.

Random allocation means assignment was intended to occur by chance, while double masking means the registry identifies the study as double-blind. The parallel design indicates that participants were assigned to one intervention group rather than crossing between both interventions. The placebo capsules were described as identical in size and color to the medication capsules. The registry does not describe the allocation sequence, concealment procedures, who specifically was masked, or whether masking was assessed.

The mecamylamine arm description reports an initial dose of 2.5 mg once daily, with a second dose of placebo, followed by an increase to 5.0 mg twice daily over 3 weeks. The brief summary describes the comparison as mecamylamine at 10 mg/day versus placebo. The registry does not explain the dosing description beyond those statements. The population is described as patients with depression and alcohol dependence; the stated hypothesis refers to patients on a stable SSRI dose.

The primary endpoint was the number of drinking days over 25 weeks, measured with time line follow back measures. Group results are reported for 10 participants in the placebo group and 11 in the mecamylamine group. The registry also names secondary outcomes concerning percentage of drinking days and heavy drinking days by smoking status. The information does not give their numerical results in this record summary. Other details of the trial conduct, including follow-up completeness and the prespecified statistical plan, are not reported in the registry record.

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: 284 subjects were assessed for eligibility, but 263 were excluded because they did not meet inclusion criteria or did not want to participate.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Participants were included in the study if they met criteria for major depression and alcohol dependence, had been on a stable antidepressant dose for 2 weeks, and were either smokers or nonsmokers.

6. Primary Endpoint

The registry lists the primary outcome as Number of Drinking Days, measured over 25 weeks.

The primary endpoint is the number of drinking days, reported as a mean in days over a timeframe of 25 weeks. The registry says it was measured with time line follow back measures. These measures use a retrospective timeline approach to record drinking behavior across a specified period. The endpoint therefore describes frequency of drinking days, rather than amount consumed on a particular day or a diagnosis alone.

A continuous endpoint such as this can be summarized by a group mean and a measure of spread. The posted results use the standard deviation to describe how much individual observations vary around each group’s mean. The primary comparison is expressed as the difference between the group means. A negative difference in the posted comparison corresponds to a lower mean number of drinking days in the mecamylamine group than in the placebo group, based on the ordering of those groups in the reported means.

Drinking days is relevant to the study’s alcohol-related objective because it captures how often drinking occurred during the measurement period. It is not, on its own, a complete description of alcohol-related outcomes, depressive symptoms, or overall clinical status. The registry names other outcomes, including heavy drinking days and depressive symptoms in the study objective, but the primary endpoint is drinking days. Details such as how a drinking day was defined, how missing diary information was handled, and whether the measure was aggregated across all 25 weeks or assessed at particular visits are not reported in the registry record.

7. Results Posted in the Registry

• Mecamylamine: mean 12.60 versus 17.60 on control; difference -5.00 (95% CI -10.63 to 0.63), standardized effect d = -0.76; z-test p = 0.082.

8. Statistical Interpretation

9. Statistical Methods Explained

A randomized comparison aims to make treatment groups comparable, on average, at the start of a trial. Randomization supports a causal interpretation when assignment is implemented as intended and follow-up and analysis preserve the randomized comparison. It does not guarantee balance on every characteristic, particularly in a small study. The registry identifies this study as randomized, but the sequence generation and allocation concealment methods are not reported in the registry record.

What does a mean difference represent?

The primary endpoint is continuous, so a difference in means is a natural summary of the contrast between groups. The posted placebo mean is 17.6 drinking days and the mecamylamine mean is 12.6 drinking days. The unadjusted mean difference is -5.0 days, using the direction mecamylamine minus placebo. A negative value indicates a lower observed group mean in the mecamylamine arm. The estimate is a descriptive contrast based on the posted summaries; it does not by itself account for a mixed-model structure or other adjustments.

The standard deviation describes spread among participant measurements within a group. It is different from the standard error, which describes uncertainty in an estimated group mean. The registry lists a standard deviation of 6.57 in each arm, as well as standard errors for the two means in the recalculation. General statistical principle: standard errors tend to be smaller than the underlying participant-level variability when there is more information, but they also depend on the number of observations and the analysis assumptions.

How should the confidence interval and p-value be read?

A confidence interval gives a range of values compatible with the observed data under the method used to calculate it. In repeated use of the same method under its assumptions, intervals at a stated confidence level have a long-run coverage interpretation; it is not a probability statement about this one fixed interval. For this unadjusted re-calculation, the 95% interval includes zero. Zero represents no difference in group means. Its inclusion means the interval does not rule out no difference at the conventional two-sided 0.05 threshold, although the interval also permits nonzero differences.

The unadjusted p-value of 0.082 measures how unusual a result at least as inconsistent with a no-difference null hypothesis would be under that hypothesis and the calculation’s assumptions. It is not the probability that the null hypothesis is true, nor does it measure the probability that the observed difference arose by chance in an ordinary-language sense. It is also not a measure of effect size or clinical importance. Here the p-value and interval are consistent with each other in indicating uncertainty around the unadjusted contrast.

What does the standardized effect add?

The standardized effect of -0.761 expresses the mean difference in units of the pooled standard deviation. It allows comparisons across outcomes measured on different scales in a general statistical sense, but it does not preserve the direct unit of days and should not replace the mean difference. Its interpretation depends on the observed variability, and it does not provide an independent source of evidence. The registry does not report a clinical threshold for an important difference in drinking days, so clinical importance cannot be inferred from the standardized estimate alone.

What is a mixed model?

General statistical explanation: mixed models can account for correlated observations, such as repeated measurements from the same participant, by representing within-participant dependence and other sources of variation. They can also incorporate fixed effects for group, time, or other modeled factors. The exact model matters: its outcome definition, included effects, covariance structure, handling of missing observations, and target contrast determine what the result estimates.

The registry lists a Mixed Models Analysis with p = 0.014, but does not report its estimate, confidence interval, specification, time structure, covariates, or analysis population. It is therefore not possible to say whether this was a repeated-measures group comparison, a particular time-point contrast, or another model-based test. The association of the reported analysis with the placebo group identifier only adds to the ambiguity. The registry’s listed p-value cannot be treated as a model-based estimate of the same contrast as the unadjusted re-calculation without further detail.

Why can dispersion matter?

General statistical explanation: a mean-based analysis is most straightforward when the distribution and residual variation are adequately represented by its assumptions. Over-dispersion refers to variability greater than a model expects. If the model understates variability, uncertainty may be understated; if it allows for appropriate extra variability, precision estimates may change. The registry reports standard deviations but does not report diagnostics or a model assessment, so the adequacy of a distributional assumption cannot be judged from this record.

Drinking-day measures are bounded by the observation period and may have distributions unlike a normal distribution. That general feature can matter when selecting or checking a model, but it does not establish that the analysis used was inappropriate. The endpoint’s detailed construction, the raw distribution, and diagnostic results are not reported in the registry record.

What is the role of the recalculation?

The recalculation is a simple unadjusted comparison from posted summary statistics. It is useful for seeing the direction, scale, and uncertainty of the group-mean contrast under a straightforward method. It is not a substitute for a prespecified sponsor analysis, and it does not reconstruct an unavailable mixed model. Because the registry record flags a discrepancy between this recalculation and the sponsor analysis, the sponsor’s own analysis takes precedence as the trial’s reported inferential analysis. The difference between the results remains unexplained in the registry record.

10. Confidence Intervals and Precision

The unadjusted mean difference is -5.0 days, with a 95% confidence interval from -10.63 to a positive upper endpoint. The interval crosses zero, so the data and calculation are compatible with no mean difference as well as with differences on either side of zero. This is a wide range relative to the point estimate and indicates limited precision for the unadjusted contrast.

Confidence intervals are more informative than a point estimate alone because they show the uncertainty associated with the estimate under the method’s assumptions. The lower endpoint represents a larger reduction in mean drinking days for mecamylamine relative to placebo; the positive upper endpoint permits a difference in the opposite direction. The interval should not be interpreted as a set of equally likely values, and it does not account for uncertainty arising from unreported choices in the sponsor’s model or any multiplicity across outcomes.

The reported group standard deviations are 6.57 in both arms. The group summaries list 10 participants for placebo and 11 for mecamylamine. In general, precision depends on the amount of information and the variability of the measurements. The trial’s small group summaries and the observed variability help explain why a point estimate can coexist with a broad range of plausible differences, although the registry does not report the completeness of outcome data or model-based precision details.

The registry’s sponsor analysis gives p = 0.014 but no estimate or confidence interval. Its precision and effect magnitude therefore cannot be assessed from the reported value alone. A p-value does not substitute for an interval: it does not show the range of effect sizes compatible with the model or the direction and magnitude of the associated contrast. The record’s discrepancy means the unadjusted interval should not be presented as if it were the sponsor model’s interval.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total events. The calculation uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers effects equal to 100%, 75%, and 50% of the observed effect. The trial’s own original design assumptions, including its intended sample-size target and design effect, are not reported in the registry record.

For an assumed effect equal to 100% of the observed effect, the planning illustration gives 28 participants per arm for 80% power and 37 participants per arm for 90% power. For 75% of the observed effect, it gives 49 participants per arm for 80% power and 65 participants per arm for 90% power. For 50% of the observed effect, it gives 109 participants per arm for 80% power and 146 participants per arm for 90% power.

In plain terms, 80% power describes the chance a study would reach the specified statistical threshold if the assumed effect were the true effect and the calculation assumptions held. The 90% column describes a higher chance under the same conditions, which calls for more participants per arm. Smaller effects are harder to distinguish from random variation, so the planning figures rise as the assumed fraction of the observed effect gets smaller. These are planning illustrations based on the observed effect, not observed power, and they do not establish what sample size the original trial planned or justified.

The accompanying curve is read using the same planning assumptions: lower assumed effects correspond to larger participant requirements. The observed effect in this small study may itself be uncertain, so planning directly from it should not be treated as a definitive design recommendation. The registry does not report whether the trial’s original design used these assumptions or another approach.

12. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Observed difference Δ = -5.000; pooled standard deviation σ = 6.570.
2. n = 2 × (1.96 + 0.842)² × 6.570² / -5.000² = 27.1, rounded up to 28 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 <- -5                      # observed difference in means
sigma <- 6.57                     # 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(563797)
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 = 28 (80% power), 37 (90% power)
 75% of observed effect: n per arm = 49 (80% power), 65 (90% power)
 50% of observed effect: n per arm = 109 (80% power), 146 (90% power)
power.t.test: 28.1 
Simulated power at n = 28 per arm: 0.799 (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.

13. Analyses Reported by the Sponsor

The registry lists the sponsor’s own result as a Mixed Models Analysis with p = 0.014. It does not give the model’s estimated treatment contrast, its confidence interval, the model specification, or a clear description of the contrast tested. The listed analysis is associated only with the placebo group identifier, so the meaning of that association is not evident from the record.

The unadjusted re-calculation uses the posted group means and standard deviations. It estimates a mean difference of -5.0 days, with a 95% confidence interval that crosses zero and an unadjusted p-value of 0.082. This is a distinct calculation, not a reproduction of the sponsor’s mixed model. Differences may arise when analyses use different models, populations, time points, or assumptions, but the registry does not provide enough information to determine the reason in this case.

The discrepancy is flagged in the registry record. Accordingly, the sponsor’s own analysis takes precedence when identifying the trial’s reported inferential result. That precedence does not make the result fully interpretable: without the estimate, interval, and analysis details, the size and direction of the sponsor’s modeled contrast are not reported in the registry record. The unadjusted calculation remains useful as a transparent description of the posted group summaries, but it should not be substituted for or presented as the sponsor’s result.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists four outcome measures and four analyses, and names secondary outcomes concerning the mean percentage of drinking days by smoking status and the mean percentage of heavy drinking days by smoking. The primary endpoint is number of drinking days. The registry record does not report a multiplicity adjustment, a hierarchy of hypotheses, or which outcome analyses were designated confirmatory. When several outcomes or contrasts are tested, the chance of at least one apparently unusual result can increase; general statistical practice may address this through prespecified prioritization or adjustment. No particular approach can be attributed to this trial from the record.

Missing-data handling is also not reported. The registry does not say how incomplete time line follow back information, missed visits, or withdrawal from the study was handled, or whether the sponsor’s analysis used all randomized participants or a restricted analysis population. These choices can affect both the estimated contrast and its precision. A mixed model can use incomplete repeated measurements under specified assumptions, but the model and assumptions are not reported, so that possibility should not be taken as evidence of how missing information was handled.

The registry gives group sizes for the posted primary endpoint summary but does not identify the population used for the sponsor analysis. It does not state whether this was an intention-to-treat population, a per-protocol population, or another set of participants. Nor does it describe exclusions after randomization. These omissions make it difficult to connect the sponsor p-value with the group means or assess whether the analysis preserved the randomized comparison.

The reported sponsor p-value is not accompanied by the estimate or interval needed to judge its magnitude and precision. The absence of those details, alongside the discrepancy with the unadjusted recalculation, limits interpretation. The registry record also does not report interim analyses, stopping rules, or a detailed statistical analysis plan. Their presence or absence cannot be inferred from the completed status.

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 non-serious adverse event terms

The registry reports zero serious adverse events among 11 participants at risk in the mecamylamine group and zero among 10 participants at risk in the placebo group. It lists no serious adverse event terms. These figures describe the serious-event counts recorded in the registry, not a comprehensive assessment of comparative safety.

For other adverse events, the registry reports 3 affected participants among 11 at risk in the mecamylamine group and 4 affected participants among 10 at risk in the placebo group. These are counts of participants affected, not counts of individual event episodes. The registry does not give event types, severity, timing, or the method used to determine attribution in the information.

Death counts are not reported in the registry record. The zero serious-event figures should not be read as proof that serious events could not occur, particularly in a trial with a small number of participants. These observations describe what was recorded for the study groups; they do not establish a general safety profile or support treatment advice.

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

The posted primary summaries show a lower mean number of drinking days in the mecamylamine group than in the placebo group: 12.6 versus 17.6. The unadjusted re-calculation estimates a mean difference of -5.0 days. Its confidence interval crosses zero and its p-value is 0.082, indicating substantial uncertainty around that unadjusted estimate. These figures do not alone establish whether the groups truly differed in the population from which the trial participants came.

The sponsor’s registry-listed mixed-model analysis reports p = 0.014, but the estimated contrast and confidence interval are not reported, and the analysis description is insufficient to determine exactly what was tested. The discrepancy with the re-calculation is explicitly flagged. The sponsor’s result takes precedence as the sponsor analysis, but the available record does not allow its direction or magnitude to be interpreted with confidence from the p-value alone.

The endpoint concerns drinking days over 25 weeks. It does not by itself answer the full study objective concerning depressive symptoms and alcohol symptoms, and the record information does not provide the numerical results for those additional outcomes. The registry names secondary outcomes related to drinking days and heavy drinking days by smoking status, but those results are not included in the registry record.

The trial was randomized and double-blind, which are design features intended to reduce bias, but the registry information does not provide enough detail to assess all aspects of trial conduct and analysis. The enrollment was 21, and the posted group summaries are based on small groups. Missing-data methods, analysis populations, endpoint completeness, and multiplicity handling are not reported. These uncertainties matter when judging the strength and generalizability of the evidence.

The result should therefore be understood as a small randomized comparison with a reported sponsor p-value that cannot be fully evaluated, alongside a separate unadjusted estimate that is imprecise. Neither result justifies a claim that mecamylamine works or does not work. They describe the evidence recorded for this trial and the limits of what can be concluded from its registry information.

17. Why This Trial Matters Statistically

This trial illustrates why an effect estimate, its uncertainty, and the analysis that produced it should be read together. The unadjusted re-calculation gives a direct contrast in days, making the direction and scale understandable. Its confidence interval, however, is broad and crosses zero. The sponsor’s reported p-value is different, but without the model estimate and interval it cannot communicate how large the modeled contrast was or how precisely it was estimated.

The contrast between the two results is also a practical lesson in reproducibility. Summary statistics permit a transparent unadjusted re-calculation, but they cannot recreate an analysis whose model details and target contrast are unavailable. A mixed model may account for repeated observations or other structures, but the record does not say how this one was specified. Reporting a p-value without its estimate, uncertainty, population, and model context leaves important statistical questions unresolved.

Small trials can produce unstable estimates, and their uncertainty is important both for interpretation and for planning later studies. The planning illustrations show how required participants per arm increase when the assumed detectable effect is reduced, but those calculations begin with the observed effect and are not the trial’s documented design justification. Planning should not be mistaken for evidence that a result is reliable or for a retrospective measure of power.

The study also highlights the value of reporting missing-data handling, analysis populations, multiplicity decisions, and adverse event details. Without these, readers cannot fully assess whether the analysis corresponds to the randomized question or how broadly to interpret the result. The statistical importance lies not in a single threshold, but in seeing how design, estimates, uncertainty, and reporting detail jointly shape what the evidence can support.

18. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The most immediate limitation is the unresolved difference between the unadjusted re-calculation and the sponsor’s mixed-model p-value. The registry gives no model estimate or confidence interval and does not describe the model sufficiently to identify the contrast. Although the sponsor result takes precedence as the trial’s own analysis, the reported information is not enough to judge its effect size or reconcile it with the group summaries.

The endpoint details are limited. The registry identifies number of drinking days, a 25-week timeframe, and time line follow back measurement, but does not define a drinking day, specify the measurement schedule, or describe how the period-level value was constructed. It does not report the completeness of these measurements, missing-data handling, or the number of participants contributing to each analysis.

The registry does not report the analysis population, model specification, covariate adjustment, interim analyses, or the prespecified statistical plan. It also does not state how multiplicity was handled across the listed outcomes and analyses. These omissions prevent a complete assessment of the sponsor’s inferential result and whether it addresses the stated primary question.

Enrollment was 21, and the primary group summaries include 10 placebo participants and 11 mecamylamine participants. With a small study, estimates can be imprecise, as reflected by the broad unadjusted confidence interval. The record does not provide enough detail to assess how representative the enrolled participants were or how well the findings generalize beyond the described population.

Safety information is brief: serious adverse event counts are reported, but event descriptions are absent, other adverse events are not characterized by type, and death counts are not reported. The registry also names outcomes relating to depressive symptoms and smoking-status subgroups without including their numerical results in this record summary. Those gaps restrict interpretation to the posted primary endpoint summaries and the limited safety counts.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Mecamylamine for the Treatment of Patients With Depression and Alcohol Dependence — NCT00563797.
• Registry record last updated 2016-04-18; status last verified 2016-03.
• Results first posted 2016-04-18.
• 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.

22. Record Summary

The registry record describes a completed, randomized, double-blind, parallel-group Phase 3 trial sponsored by Yale University. It enrolled 21 participants and compared mecamylamine with placebo in patients with depression and alcohol dependence. The primary endpoint was mean number of drinking days over 25 weeks, measured with time line follow back measures.

The posted means were 17.6 days for placebo and 12.6 days for mecamylamine. An unadjusted re-calculation estimated a mean difference of -5.0 days, with a 95% confidence interval crossing zero and p = 0.082. The registry also lists a sponsor mixed-model p-value of 0.014, without an estimate or sufficient model details; the discrepancy is unresolved, and the sponsor’s analysis takes precedence as the reported analysis.

The registry does not report key details including the analysis population, missing-data approach, multiplicity handling, and the sponsor model specification. Serious adverse event counts were zero in both groups; other adverse events affected 3 participants in the mecamylamine group and 4 in the placebo group. Death counts are not reported in the registry record.