N-acetylcysteine Treatment of Alcohol Use Disorder In Veterans With TBI: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for N-acetylcysteine Treatment of Alcohol Use Disorder In Veterans With TBI (NCT02791945), a randomized phase 2 study. The primary outcome is Change in Percent of Heavy Drinking Days Per Week as Assessed by the Timeline Followback (TFLB).

2. Clinical Question

This question concerns a short-term behavioral outcome measured over the period from baseline to Week 8. The endpoint is based on participants’ reports about alcohol use during the prior week, gathered using a calendar. The posted comparison is between randomized arms, not a comparison of each arm with an external benchmark. The outcome is one measure of alcohol-use frequency; it does not, on its own, describe every aspect of alcohol use, traumatic brain injury, or participants’ broader health. The record does not report a separate clinical threshold for interpreting the size of a change.

3. Trial Design

This was a phase 2, randomized, double-masked, parallel-group trial sponsored by the University of California, San Francisco. The registry lists enrollment of 30. The study status is completed, with a start date of 2016-08 and primary completion in 2018-05. The intervention list includes N-acetylcysteine, placebo, and Medical Management Counseling. The arm descriptions specify daily capsules for the two study arms, up to 3200 mg for each description. The registry record does not explain how the counseling intervention was incorporated into the randomized comparison.

The posted primary outcome includes 12 participants in the N-acetylcysteine arm and 14 in the placebo arm. The record does not explain the relationship between these outcome counts and the enrollment figure, or state why the counts differ. The primary endpoint is identified as change in percent of heavy drinking days per week from baseline to Week 8. It is assessed using a Timeline Followback interview, in which a calendar is used to ask about the prior week’s frequency of alcohol use on each day.

A parallel design assigns participants to separate groups whose outcomes are compared. Randomization is intended, in general, to reduce systematic differences between groups at the start of a study; it does not guarantee that the groups will be identical, particularly in a small trial. Double masking indicates that the record classifies the study as double-masked, but the specific roles masked are not reported. Allocation details, the randomization method, concealment procedures, protocol deviations, and the original sample-size justification are not reported in the registry record. The posted results include one primary outcome comparison and no posted secondary outcomes or analyses.

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

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 Change in Percent of Heavy Drinking Days Per Week as Assessed by the Timeline Followback (TFLB), measured over Baseline to Week 8.

The primary endpoint is the change in percent of heavy drinking days per week from baseline to Week 8, measured with the Timeline Followback (TLFB) interview. In this interview, participants use a calendar to report the prior week’s frequency of alcohol use on each day. The outcome therefore represents a weekly pattern of reported drinking rather than a direct biological measurement.

Tracking change over time can describe how an outcome differs from its baseline level during the study period. Comparing the mean change between randomized arms can help quantify the difference in average change associated with assignment to the study arms, subject to the trial’s design and limitations. The registry record reports this endpoint as a continuous measure and presents arm values as means with standard deviations. It labels the comparison as a mean difference.

As a general measurement principle, a repeated report can give a structured way to assess a behavior over a defined interval. A calendar-based interview may help organize recall by day, but the record does not report information about interview reliability, missing responses, or how incomplete recall was handled. The endpoint is limited to the stated weekly heavy-drinking measure and the baseline-to-Week 8 timeframe. The registry record does not define a threshold for a clinically important change or explain how the reported values should be interpreted in relation to other alcohol-use outcomes.

7. Results Posted in the Registry

• N-acetylcysteine: mean 26.20 versus 26.50 on control; difference -0.30 (95% CI -0.57 to -0.03), standardized effect d = -0.84; z-test p = 0.029.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes the primary endpoint as a continuous measure and reports means and standard deviations for the two arms. The comparison is presented as a mean difference: the estimated difference between the N-acetylcysteine and placebo arm values on the stated endpoint. The posted estimate is -0.3. Its sign depends on the order of the comparison; the record identifies N-acetylcysteine as the arm being compared with placebo. The sign should not be separated from the outcome’s definition and scale when interpreting direction.

What does a mean summarize?

In general, a mean is an average that combines the observed values in a group into one central summary. It does not show the full distribution of individual observations. Two groups can have similar means while differing in how their observations are spread, and a mean can be influenced by unusually high or low values. The registry reports means of 26.2 and 26.5, with standard deviations of 0.31 and 0.39. The record does not give a distribution plot or individual observations in the registry record described for this explanation.

What does the standard deviation describe?

In general, a standard deviation describes how dispersed observations are around their group mean. It is not the same as the standard error or the uncertainty interval for a group comparison. The posted standard deviations describe within-arm variation; they do not directly indicate how precisely the mean difference is estimated. The comparison’s confidence interval is the relevant posted summary for uncertainty in the difference.

How should the mean difference be read?

A mean difference expresses the separation between two group means in the endpoint’s units. In this registry result, the reported difference is -0.3. The result is labeled as a change in percent heavy drinking days per week, assessed from baseline to Week 8. The registry record does not explain in further detail whether the endpoint is represented in percentage points or another transformation, or how the individual change values were constructed. For that reason, the estimate should be described using the registry’s own outcome label rather than converted into a more specific unit or clinical interpretation.

What does the p-value contribute?

In general, a p-value measures how unusual data at least as incompatible with a null hypothesis would be if the null model and its assumptions were correct. The posted unadjusted p-value is 0.029. It is not a measure of effect size, the probability that a treatment effect exists, or the likelihood that the finding will replicate. Its interpretation depends on the analysis and assumptions used. The registry record does not report the exact statistical test or model behind this value, so it is not possible to verify those assumptions from the registry entries alone.

What does the confidence interval add?

In general, a confidence interval gives a range of estimates that are compatible with the observed data under a specified statistical procedure. Its width conveys precision: a narrower interval indicates less statistical uncertainty under that procedure, while a wider interval indicates more. The posted interval for the mean difference lies entirely below zero. The registry record does not state the interval method, confidence level, or whether it was adjusted for other analyses. Its interpretation should therefore be kept to the information the record provides: it is the reported interval for the comparison, and its bounds are on the same side of zero.

Which analysis details are unavailable?

The registry record does not report whether the comparison used a simple two-arm test, a model adjusted for baseline values, or another approach. It does not state whether assumptions were checked, whether observations were weighted, or whether the p-value and interval came from the same model. It also does not report how missing endpoint values were handled, how participants were assigned to the analysis population, or how deviations from assigned treatment were addressed. These are important details because analysis choices can affect the estimate, uncertainty, and interpretation. They should not be inferred from the reported mean difference alone.

As a general statistical concern, small samples can make estimates sensitive to the values contributed by individual participants and can leave substantial uncertainty about the underlying group difference. The outcome counts are 12 and 14, and the record reports no analyses beyond the posted comparison. The statistical summaries describe the observed trial data; they do not establish that the analysis assumptions held or that the same estimate would appear in another study. The registry record does not report an interim analysis plan, multiplicity strategy, or detailed statistical analysis plan.

10. Confidence Intervals and Precision

The mean difference is reported as -0.3, and its confidence interval has both bounds below zero. This interval conveys uncertainty around the estimated difference, not a range containing a stated proportion of individual participants’ outcomes. In general, a confidence interval is constructed by a statistical procedure that would cover the underlying parameter at a specified rate over repeated studies if the procedure’s assumptions hold. The registry record does not report the interval’s confidence level or construction method, so those details cannot be assessed from the posted record.

The interval’s position relative to zero is consistent with the unadjusted p-value of 0.029 for the reported comparison. That relationship does not show that the estimate is precise in every practically relevant sense. Precision depends on the interval’s width and on the scale of the outcome, while practical importance depends on how the outcome and a meaningful difference are defined. The record does not report a threshold for a clinically important difference.

The outcome counts are 12 for N-acetylcysteine and 14 for placebo. The registry lists arm standard deviations of 0.31 and 0.39. Together, these summaries give some information about observed variation and the basis of the comparison, but they do not substitute for the confidence interval or reveal the individual data. The record does not provide subgroup-specific estimates or additional primary-outcome analyses that would show how stable the estimate is across alternative specifications.

Precision should also be considered in light of the unreported analysis details. The registry does not state how missing observations were handled or which analysis population was used. Without those details, the interval can be described as the posted uncertainty summary, but its sensitivity to analysis choices cannot be evaluated. The result therefore supports a bounded statistical reading of the reported comparison, not a definitive account of the range of effects that could occur in other populations or settings.

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 expressed as participants per arm, not as total events. They are illustrations based on the observed effect, rather than the trial’s original design calculation. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The registry record does not report the trial’s original sample-size assumptions or justification.

In general, power is the probability that a specified analysis will detect an effect of a stated size if that effect is present and the model assumptions hold. The 80% power column gives the number of participants per arm used in the illustration to target 80% power for the corresponding assumed effect; the 90% power column does the same for a 90% target. These are planning targets, not guarantees about a future study’s result. They also do not tell how likely the posted trial result is to be true.

The displayed planning scenarios use 100%, 75%, and 50% of the observed effect. At 100% of the observed effect, the illustrations list 23 participants per arm for 80% power and 30 for 90% power. At 75%, they list 40 and 53; at 50%, they list 89 and 119. The pattern reflects a general property of study planning: smaller effects are harder to detect, so more participants are needed to reach the same target power, all else being equal.

These figures are not observed power. Observed power is not a useful substitute for interpreting the estimated effect and its confidence interval after a study has been completed. The planning values depend on assumptions about the effect and analysis, and the original study design assumptions are not reported in the registry record. They should be read as conditional illustrations rather than as evidence about whether the completed trial was adequately sized.

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 = -0.300; pooled standard deviation σ = 0.356.
2. n = 2 × (1.96 + 0.842)² × 0.356² / -0.300² = 22.1, rounded up to 23 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 <- -0.300000000000001                      # observed difference in means
sigma <- 0.355574652264941                     # 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(2791945)
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 = 23 (80% power), 30 (90% power)
 75% of observed effect: n per arm = 40 (80% power), 53 (90% power)
 50% of observed effect: n per arm = 89 (80% power), 119 (90% power)
power.t.test: 23.1 
Simulated power at n = 23 per arm: 0.801 (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 identifies one primary outcome: change in percent of heavy drinking days per week from baseline to Week 8. It lists three outcome measures overall, but no secondary outcomes are posted in the record. The record also lists zero analyses. This means the available registry information does not provide a set of additional posted outcome analyses against which to assess the breadth of statistical testing.

Multiplicity arises, in general, when multiple outcomes, time points, subgroups, or statistical comparisons are examined. If many hypotheses are tested, the chance of at least one small p-value can increase unless the analysis plan accounts for the number and structure of tests. The registry record does not report a multiplicity adjustment, a prespecified hierarchy of outcomes, or a plan for interpreting multiple tests. The unadjusted p-value of 0.029 should therefore be understood as the posted unadjusted result, not as evidence that a multiplicity strategy was or was not used elsewhere.

The record does not report the analysis population, such as whether the comparison included all randomized participants, only participants with available outcome data, or another group. It also does not report how participants with missing baseline or Week 8 values were treated, whether imputation was used, or whether sensitivity analyses were conducted. These details matter because the participants included and the handling of missing observations can affect the result and its uncertainty.

The registry record does not report a detailed statistical analysis plan, interim analysis procedures, covariate-adjustment plan, or rules for handling protocol deviations. Their absence from the record should not be taken as proof that a procedure was or was not used. It means that the posted information does not support evaluating those features. Consequently, the primary comparison can be described as registered, but the full context needed to judge the analysis strategy and the risk of selective reporting is not available in the registry record.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: inpatient hospitalization for self-reported "broken back in 3 places". (7.1%), inpatient hospitalization/5150 (<72 hours) for suicidal threats while intoxicated (7.1%).

The registry lists serious adverse event fields for 14 participants at risk in the N-acetylcysteine group and 16 in the placebo group. It records 2 participants affected in the N-acetylcysteine group and 0 in the placebo group. It reports 14 participants affected by other adverse events in the N-acetylcysteine group and 16 in the placebo group. Deaths are listed as 0 in both groups.

Two serious-event terms are listed, each with a reported frequency of 7.1%: “Inpatient hospitalization for self-reported ‘broken back in 3 places’.” and “Inpatient hospitalization/5150 (<72 hours) for suicidal threats while intoxicated”. The record does not explicitly assign these terms to an arm in the listed serious-event terms, so they should not be attributed to a particular group based on the available information.

In general, adverse-event counts describe reported events or affected participants under the registry’s definitions; they do not by themselves establish that an event was caused by the study intervention. The record does not provide event timing, severity details beyond the serious-event designation, exposure-adjusted rates, or an assessment of relatedness. The at-risk counts and affected counts should be kept distinct when reading the safety results.

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

The posted result describes a mean difference of -0.3 in change in percent of heavy drinking days per week between the N-acetylcysteine and placebo arms from baseline to Week 8. The confidence interval lies entirely below zero, and the unadjusted p-value is 0.029. These summaries indicate that the estimated difference is not centered on zero in the posted analysis. They do not determine whether the size of the difference is important in practice, because the registry record does not state a threshold for meaningful change or provide further detail about the endpoint scale.

The result applies to the endpoint, comparison, and timeframe recorded for this phase 2 trial. It does not describe outcomes beyond Week 8, other measures of alcohol use, or longer-term follow-up. Nor does it resolve questions about the full study population: the posted outcome counts are 12 and 14, while enrollment is listed as 30, and the record does not explain the difference. The record also does not report the analysis population or missing-data approach.

A statistical difference is not the same as proof of a causal mechanism or a complete assessment of benefit and harm. Although randomization can support a comparison of assigned groups in general, the details of allocation, analysis, and participant follow-up are not reported in the registry record. The findings should be read as the reported trial comparison, not as a conclusion about all veterans with traumatic brain injury or all people with unhealthy alcohol use.

The safety fields report serious events in the N-acetylcysteine group and none in the placebo group, while other adverse events are listed for participants in both groups and deaths are listed as 0 in both. Those counts are descriptive; the record does not establish event relatedness or provide enough detail for a broader safety evaluation. Taken together, the posted data offer a statistical description of one small trial’s primary outcome and adverse-event counts, with important design and analysis details unreported.

16. Why This Trial Matters Statistically

This trial illustrates why the estimate, its uncertainty, and the analysis context should be read together. The registry reports a mean difference of -0.3 and a confidence interval whose bounds are both below zero, alongside an unadjusted p-value of 0.029. The point estimate communicates the direction and scale as recorded; the interval communicates uncertainty under an unreported procedure; and the p-value addresses compatibility with a null model under assumptions that are not fully described. None of these summaries alone conveys clinical importance.

The endpoint is a participant-reported weekly measure collected through a calendar-based interview. That makes the outcome meaning clear at a broad level, but its interpretation still depends on the exact construction and scale of the change measure. The registry record does not report a clinically important threshold, detailed analysis method, missing-data handling, or analysis population. This distinction between a clearly named outcome and incomplete methodological detail is important when readers assess what can and cannot be inferred.

The record also gives readers a way to consider planning without confusing it with evidence from the completed trial. The participant-per-arm illustrations show how planning targets change under different assumed fractions of the observed effect. They are conditional calculations, not a retrospective measure of the study’s success or adequacy. The original sample-size justification is not reported.

Finally, the safety counts show why adverse-event summaries need context. The registry reports affected counts and serious-event descriptions, but not event attribution, timing, or relatedness. Statistical interpretation is strongest when outcome estimates, uncertainty, design information, and limitations are considered together. In this record, the posted comparison is informative as a defined Week 8 result, while several details required for a fuller assessment are not reported.

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 registry record lists enrollment of 30, while the primary-outcome arm counts are 12 and 14. It does not explain the difference or state how many participants were randomized, completed follow-up, or contributed to each analysis. The primary outcome is a participant report, and the record does not describe missing responses, incomplete calendars, or methods used to address missing data.

The endpoint is described as change in percent heavy drinking days per week, but the registry does not give further calculation details or a threshold for a meaningful change. It reports the arm means, standard deviations, mean difference, confidence interval, and unadjusted p-value, but does not identify the statistical test or model, interval construction, confidence level, or whether baseline adjustment was used. The analysis population and any sensitivity analyses are also not reported.

The record lists Medical Management Counseling as an intervention but does not describe how it was incorporated into the arms or whether its delivery was the same across groups. It records double masking without specifying which roles were masked. Details of allocation concealment and the randomization procedure are not reported. The original sample-size justification, interim analysis procedures, and full statistical analysis plan are not reported in the registry record.

Safety information is limited to the posted counts and serious-event terms. The registry does not report event timing, relatedness, or further clinical detail. The trial is identified as phase 2, and the primary endpoint timeframe is baseline to Week 8; longer-term outcomes are not described in the posted primary result. These limitations restrict interpretation of the result to the reported comparison and available registry entries.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: N-acetylcysteine Treatment of Alcohol Use Disorder In Veterans With TBI — NCT02791945.
• Registry record last updated 2021-05-11; status last verified 2021-04.
• Results first posted 2020-12-16.
• 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

NCT02791945 is a completed, randomized, double-masked, parallel phase 2 trial sponsored by the University of California, San Francisco. Enrollment is listed as 30; the start date is 2016-08 and primary completion is 2018-05. The study concerns veterans with traumatic brain injury and hazardous or harmful alcohol use. The listed interventions include N-acetylcysteine, placebo, and Medical Management Counseling.

The primary outcome is change in percent of heavy drinking days per week from baseline to Week 8, assessed by Timeline Followback. The posted arm values are 26.2 for N-acetylcysteine and 26.5 for placebo, with a mean difference of -0.3 and unadjusted p-value of 0.029. The registry lists 12 and 14 participants in the outcome arms. Serious adverse events affected 2 participants in the N-acetylcysteine group and 0 in the placebo group; deaths are listed as 0 in both groups.

The registry record does not report several analysis and design details, including the statistical model, missing-data handling, analysis population, and original sample-size justification.