Baclofen for Treating Anxiety and Alcoholism: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Baclofen for Treating Anxiety and Alcoholism (NCT01751386), a randomized phase 2 study. The primary outcome is Total Amount of Alcohol Consumed During the Alcohol Self Administration (ASA) Session.

2. Clinical Question

3. Trial Design

The registry record describes a phase 2, randomized, quadruple-masked, parallel-group, placebo-controlled human laboratory pilot study. Baclofen was compared with placebo in people described as having alcoholism and high anxiety. The registry lists enrollment as 39 and the status as completed. It names the National Institute on Alcohol Abuse and Alcoholism (NIAAA) as sponsor.

The brief summary describes screening with a physical examination, medical history, blood and urine samples, and assessments of alcohol dependence and anxiety. Participants were assigned to baclofen or placebo. The study summary says that, after a study visit, they began taking the assigned treatment three times daily. After one week on study medication, the summary describes an overnight visit at the National Institutes of Health, with blood tests, mood and behavior questions, and alcohol-related laboratory tests. The brief summary ends during its description of those tests, so further procedural details are not reported in the registry record.

The arm descriptions specify baclofen 10 mg three times daily and placebo three times daily. The record gives the study start as 2012-12-03 and primary completion as 2016-08-17. It lists alcoholism, alcohol dependence, alcohol drinking-related problems, alcohol drinking, and anxiety disorder among the conditions. These design details frame the results as a comparison of assigned groups in a controlled laboratory setting. The registry record does not report the trial's original sample-size justification or the assumptions used to choose its enrollment target.

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.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The analysis included all subjects who took the medication (either baclofen or placebo) and finished the alcohol laboratory session

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 Total Amount of Alcohol Consumed During the Alcohol Self Administration (ASA) Session, measured over 2 hours.

The primary endpoint was the total amount of alcohol consumed during the alcohol self-administration session, summarized as a mean. The measure counted the mini-drinks each participant chose to drink during a two-hour session. The endpoint description states that the available choice ranged from 0 to 8 mini-drinks. The alcohol content of each mini-drink was calculated using the participant's total body water and was designed to raise blood alcohol concentration by 0.015 g/dL.

A laboratory self-administration measure can capture drinking choices under the conditions of the study session. It provides a defined setting in which the groups' mean consumption can be compared. It is not the same measure as drinking over everyday life, nor does the registry record establish how well the laboratory measure predicts outcomes outside the session.

The primary analysis population was described as all participants who took the medication, whether baclofen or placebo, and finished the alcohol laboratory session. The posted result is therefore tied to completion of that session and medication-taking, rather than simply to enrollment or assignment. Further details about participants who did not meet that description, and how any missing endpoint observations were handled, are not reported in the registry record.

7. Results Posted in the Registry

• Baclofen: mean 4.11 versus 4.50 on control; difference -0.39 (95% CI -2.30 to 1.52), standardized effect d = -0.14; z-test p = 0.690.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the mean represent?

The primary outcome is continuous: it records the amount of alcohol consumed as a count of mini-drinks during the session, and the result is summarized by a mean. A mean describes the group average, but does not show how individual observations were distributed. Participants may have had the same or different session amounts around that average; the registry record does not post participant-level data or a distributional summary for the outcome.

The posted arm means are 4.11 mini-drinks for baclofen and 4.5 mini-drinks for placebo. The comparison is defined as baclofen minus placebo, giving a mean difference of -0.39 mini-drinks. With this direction of subtraction, a negative estimate means the observed baclofen-group average was lower than the placebo-group average. The estimate describes the observed group contrast; it does not, on its own, determine whether the groups differ beyond sampling variation.

What do the standard errors describe?

The arm summaries report standard errors of 0.66 for baclofen and 0.72 for placebo. In general, a standard error describes the uncertainty in an estimated quantity, such as a group mean, due to sampling variability. It is not a description of the spread of individual participants' outcomes. The registry reports standard error as the dispersion measure, not standard deviation, so it should not be read as if it were the within-group spread of drinking amounts.

A standard error is useful for understanding the precision of a mean estimate, but it does not convey the complete outcome distribution. The record does not report a histogram, median, range of observed values, or other participant-level distribution details beyond the endpoint's possible 0-to-8 mini-drink measure. The listed mini-drink range describes the measure, not necessarily the range observed in each arm.

How should the mean difference be read?

The mean difference expresses the contrast in the original endpoint units, mini-drinks. Its estimate of -0.39 means the baclofen sample mean was lower by that amount than the placebo sample mean in the posted comparison. This is a difference in group averages, not a difference for every participant. The registry also posts a standardized effect estimate of -0.137 and a pooled standard deviation of 2.838. A standardized estimate expresses a group contrast relative to a pooled measure of spread; it does not replace the original-unit difference, and the registry record does not provide further detail about its calculation.

Statistical significance and practical importance are separate questions. A p-value evaluates how compatible the observed data are with a specified null model under the analysis assumptions; it does not establish clinical importance. The registry does not report a threshold for a meaningful difference in mini-drinks, so the result cannot be classified against such a threshold from the record alone.

What is a confidence interval?

In general, a confidence interval describes the uncertainty around an estimated effect under a statistical procedure. It gives a range of effect values that are compatible with the data and that procedure; it is not a range containing a stated proportion of individual outcomes. The interval posted for this mean difference extends from -2.304 to 1.524 mini-drinks. Because it includes zero, it is compatible with no mean difference as well as with differences in either direction. Its breadth indicates that the point estimate is not a highly precise estimate of the group contrast.

What can be inferred from the p-value?

The posted unadjusted p-value is 0.69. In general, a p-value describes how unusual data at least as discordant with a null hypothesis would be if that hypothesis and the test's assumptions held. It is not the probability that the null is correct, nor the probability that the result occurred by chance in an everyday sense. The record does not describe the test procedure in enough detail to identify every modeling or distributional assumption behind this p-value. No adjusted p-value is reported.

These posted summaries support a cautious description of the observed mean contrast and its uncertainty. They do not show whether assumptions were checked, whether alternative models were explored, or how the original analysis plan specified the test. Those details are not reported in the registry record.

10. Confidence Intervals and Precision

The estimated mean difference was -0.39 mini-drinks, with a reported interval from -2.304 to 1.524. The interval spans zero, so the posted result is compatible with no difference in the group means as well as with a negative or positive difference. The point estimate alone should not be treated as a precise estimate of the effect.

The interval is broad relative to the point estimate and extends in both directions. That means the posted data do not pin down the size or direction of a group difference with much precision. The result is not evidence that the true group means are identical: an interval that includes zero does not demonstrate equivalence. An equivalence conclusion would require a prespecified margin and an analysis designed to assess equivalence, neither of which is reported in the registry record.

Standard errors of 0.66 and 0.72 accompany the baclofen and placebo group means, respectively. These describe uncertainty in the estimated means, not the variability among individual participants. The trial's outcome analysis lists 18 participants in the baclofen group and 16 in the placebo group. The registry does not report additional interval details, such as a named confidence level or a fuller account of the method used to construct the interval.

The p-value of 0.69 is unadjusted. Its interpretation depends on the test and assumptions used, details that are not fully reported in the registry record. Taken together, the mean difference, interval, and p-value describe an uncertain laboratory-session comparison, not a conclusive estimate of a broader treatment effect.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not the trial's original enrollment calculation. The registry record does not report the trial's original sample-size assumptions or justification.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as a stated fraction of the observed effect. For planning for 100% of the observed effect, the estimates are 832 participants per arm for 80% power and 1,113 participants per arm for 90% power. For planning for 75% of the observed effect, the estimates are 1,478 participants per arm for 80% power and 1,979 participants per arm for 90% power. For planning for 50% of the observed effect, the estimates are 3,325 participants per arm for 80% power and 4,451 participants per arm for 90% power.

The 80% power column represents the estimated participants per arm for a design with an 80% chance of detecting the specified effect under the stated planning assumptions, if that effect and assumptions hold. The 90% power column represents the corresponding estimate for a design with a 90% chance. These are planning illustrations, not observed power.

Smaller effects require more participants per arm to detect under the same assumptions, which is why the estimates increase when the target is reduced from 100% to 75% or 50% of the observed effect. The figures should not be read as recommendations or as evidence that a future study would reproduce the observed effect. They depend on the observed effect estimate, which is uncertain, and on the planning assumptions.

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 = -0.390; pooled standard deviation σ = 2.838.
2. n = 2 × (1.96 + 0.842)² × 2.838² / -0.390² = 831.2, rounded up to 832 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 <- -0.39                      # observed difference in means
sigma <- 2.83785570457696                     # 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(1751386)
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 = 832 (80% power), 1,113 (90% power)
 75% of observed effect: n per arm = 1,478 (80% power), 1,979 (90% power)
 50% of observed effect: n per arm = 3,325 (80% power), 4,451 (90% power)
power.t.test: 832.1 
Simulated power at n = 832 per arm: 0.803 (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. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary outcome: total alcohol consumed during the alcohol self-administration session. It lists no secondary outcome measures. The record also gives zero posted analysis records, although a primary outcome result and comparison are reported. No further information about the number of planned analyses or outcomes is reported in the registry record.

The comparison's p-value is unadjusted. The registry record does not report a multiplicity strategy, such as whether any adjustment was planned for multiple outcomes, time points, subgroup comparisons, or repeated analyses. Without that information, the p-value should be read as the unadjusted result for the posted comparison, not as an estimate already corrected for a broader set of tests.

The stated primary analysis population includes participants who took the assigned medication and completed the alcohol laboratory session. This differs from an analysis defined simply by randomized assignment. The registry record does not explain why this analysis population was chosen, how many assigned participants did not complete the session, or whether any participants were excluded for other reasons.

Missing-data methods are not reported in the registry record. It does not say whether missing endpoint observations occurred, whether data were imputed, or whether sensitivity analyses assessed the effect of missingness. The record also does not describe the detailed statistical model, covariate adjustment, or handling of protocol deviations. These omissions limit the ability to assess the potential influence of analysis choices on the posted estimate.

The absent detail does not establish that a particular procedure was or was not used; it means that the registry record does not report it. Interpretation should therefore stay with the analysis population and unadjusted comparison that are posted, without assuming a specific multiplicity or missing-data plan.

14. 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 5% reporting threshold.

Most frequent non-serious adverse event terms

The adverse-event record lists separate at-risk groups for the safety summaries. In the baclofen group, 20 participants were at risk for serious adverse events and 0 were reported as affected; 20 were at risk for other adverse events and 9 were reported as affected. In the placebo group, 19 participants were at risk for serious adverse events and 0 were reported as affected; 19 were at risk for other adverse events and 7 were reported as affected. Deaths were reported as 0 in each group.

These are counts, not calculated rates. The at-risk counts are the denominators associated with the registry's safety summaries and should not be replaced with the outcome-analysis group sizes. The record does not list serious adverse event terms, and it does not provide more detail about the nature, severity, timing, or relationship of the other adverse events to study treatment.

Zero reported serious adverse events or deaths in these groups does not establish that such events are impossible or that the study could rule out uncommon harms. The record's safety figures describe what was posted for this trial; they do not establish causation or predict safety in other populations or settings.

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

The result describes a comparison of mean mini-drink consumption during a controlled, two-hour alcohol self-administration session among participants who took study medication and completed that session. The baclofen-group mean was 4.11 mini-drinks, and the placebo-group mean was 4.5 mini-drinks. The estimated difference was -0.39 mini-drinks, with a reported interval from -2.304 to 1.524 and an unadjusted p-value of 0.69.

The observed direction of the point estimate is toward a lower session mean in the baclofen group. But the interval includes zero and extends to values in both directions. The posted result therefore does not establish a clear group difference, and it also does not establish that the groups are equivalent. The data are compatible with a range of possible mean contrasts, including no difference.

The endpoint is a laboratory measure. It captures choices made during the study session under the specific conditions described in the registry record. It is not a direct measure of drinking across daily life, sustained alcohol use, anxiety symptoms, or longer-term outcomes. The record does not post secondary outcome measures that would address those broader questions.

The findings also need to be considered in light of the pilot-study description, the outcome-analysis population, and the limited information about missing data and analysis procedures. The registry record does not report the original sample-size justification, detailed statistical plan, or multiplicity strategy. Consequently, the posted result should not be expanded into a claim about all people with alcohol dependence or anxiety, or about treatment effects beyond this laboratory endpoint.

The most defensible interpretation is descriptive: the posted sample means differed by the stated amount, but the uncertainty around that comparison is substantial. The result alone does not establish that baclofen reduces drinking or anxiety, and it does not show that baclofen has no effect. It reports a specific group comparison in a specific study context.

16. Why This Trial Matters Statistically

This trial is statistically instructive because its primary outcome is a clearly defined laboratory measure, while the estimate of the group contrast remains uncertain. The endpoint records the number of mini-drinks chosen in a two-hour session, and the analysis compares group means. That makes the result interpretable in its original units, while the interval shows why a point estimate alone is not enough to characterize the evidence.

The estimated difference is -0.39 mini-drinks, but the reported interval extends from -2.304 to 1.524 and includes zero. This illustrates the distinction between the direction of a point estimate and the precision of the evidence. A negative estimate does not by itself establish a group difference; an interval spanning both directions leaves substantial uncertainty about the size and direction of the contrast.

The posted planning illustrations also show how sample-size requirements depend on the effect chosen for planning. Under the listed assumptions, targeting a smaller fraction of the observed effect requires more participants per arm. These figures are not the trial's original design justification and are not observed power. Their relevance is methodological: planning depends on an assumed effect, and an estimate from a pilot study may be uncertain.

The record further makes the analysis population and safety denominators important to distinguish. The primary outcome includes participants who took medication and completed the laboratory session; the safety summaries use their own at-risk counts. Reading each result with its stated population avoids treating different analyses as if they used the same denominator.

Finally, the absence of detailed reporting on missing data, multiplicity, and the original analysis plan limits how far the statistical result can be audited from the registry record alone. Those limits do not erase the posted comparison, but they define the degree of confidence and scope appropriate to its interpretation.

17. 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 primary endpoint is specific to an alcohol self-administration laboratory session lasting two hours. Although the endpoint description defines how mini-drinks were constructed and measured, the registry record does not establish that this session measure represents alcohol consumption in everyday settings or over a longer period. The record also does not report secondary outcome measures that could broaden the assessment of alcohol use or anxiety.

The primary outcome analysis population is restricted to participants who took the assigned medication and completed the alcohol laboratory session. The record does not report the number of randomized participants who failed to meet that description, the reasons for non-completion, or how missing endpoint data were handled. That limits assessment of whether the analysis population could differ from all randomized participants.

The posted mean difference has a broad interval spanning zero and values in both directions. This limits precision about the size and direction of the group contrast. The registry record does not provide a threshold for a clinically meaningful difference in mini-drinks, so the estimate cannot be judged against a stated importance criterion.

The record describes the trial as a pilot and reports enrollment as 39, while the outcome-analysis groups list 18 for baclofen and 16 for placebo. It does not report the original sample-size justification, detailed statistical plan, or multiplicity strategy. Nor does it provide detailed analysis methods or participant-level outcome distributions. These are reporting limitations; the registry record does not establish whether particular procedures were absent or simply unreported.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Baclofen for Treating Anxiety and Alcoholism — NCT01751386.
• Registry record last updated 2017-09-18; status last verified 2017-07-17.
• Results first posted 2017-08-16.
• 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.

21. Record Summary

The registry record describes NCT01751386, Baclofen for Treating Anxiety and Alcoholism, as a completed phase 2 randomized, quadruple-masked, parallel-group pilot study. Baclofen 10 mg three times daily was compared with placebo. Enrollment was listed as 39, and the sponsor was the National Institute on Alcohol Abuse and Alcoholism.

The primary outcome was total alcohol consumed during a two-hour alcohol self-administration session, measured as mean mini-drinks. The posted means were 4.11 for baclofen and 4.5 for placebo; the mean difference was -0.39, with a reported interval from -2.304 to 1.524 and an unadjusted p-value of 0.69. The registry record lists no secondary outcome measures. It does not report the original sample-size justification, multiplicity plan, or missing-data methods.