Kappa Opioid Receptor Antagonism for the Tx of AUD and Comorbid PTSD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Kappa Opioid Receptor Antagonism for the Tx of AUD and Comorbid PTSD (NCT03852628), a randomized phase 2 study. The primary outcome is Number of Participants With a Reduction in Alcohol Use Disorder (AUD), With TLFB Tool.

2. Clinical Question

The clinical question concerns a categorical change in drinking-risk level, rather than a direct comparison of average drinks consumed. Alcohol use was assessed with the Timeline Follow Back, a calendar-based method that records the frequency and amount of daily consumption. The registry describes the intervention as intended to evaluate efficacy and physiological effects in people with alcohol use disorder and comorbid post-traumatic stress disorder. The posted primary outcome addresses change in alcohol-use risk category; it does not, by itself, summarize every clinical or physiological outcome of interest.

3. Trial Design

The registry identifies NCT03852628 as a randomized, triple-masked, parallel-group Phase 2 trial. It lists an enrollment of 69 and a status of terminated. The recorded start date is 2019-05-20, and the primary completion date is 2022-05-02. The reason for termination is not reported in the registry record, so the status alone does not explain why the trial ended or whether termination affected the available outcome data.

The active regimen combined daily sublingual buprenorphine, listed in the intervention as 2 mg Buprenex, with 380 mg extended-release naltrexone, listed as Vivitrol, administered intramuscularly every 4 weeks. The control regimen used a daily sublingual placebo and intramuscular placebo injections on the corresponding schedule. The arm descriptions specify 12 weeks of daily pills and injections at baseline, week 4 and week 8. The registry’s brief summary also names Subutex and Vivitrol when describing the study objective; the arm listing uses Buprenex. The registry record does not explain this naming difference.

For the primary outcome, the registry reports 30 participants in the combination arm and 28 in the placebo arm. Those outcome counts are not the same as the overall enrollment figure. The primary outcome population required completion of both baseline and week 8 visits and sufficient consecutive days of alcohol-consumption data at week 8. The registry does not report why some enrolled participants are not represented in the posted primary outcome counts, and it does not report the original sample-size justification or the trial’s planned enrollment target separately from the enrollment recorded.

The design establishes a randomized comparison between the two assigned regimens, but the registry record does not give further details about the randomization procedure, the identities of the masked parties, or whether masking was maintained. Those details 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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Includes all randomized participants that completed the baseline and Week 8 visit, and reported enough consecutive days of alcohol consumption at the week 8 visit (defined as being within one standard deviation of the population mean for the 28-day look back period).

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With a Reduction in Alcohol Use Disorder (AUD), With TLFB Tool, measured over Baseline and 8 weeks.

The primary endpoint is the number of participants with a reduction in alcohol use disorder, assessed with the Timeline Follow Back tool. The registry defines a reduction as a decrease of at least one World Health Organization alcohol-use risk category between baseline and week 8. The categories are very high risk, high risk, medium risk and low risk, with abstinence included in the low-risk category. The endpoint therefore classifies each eligible participant according to whether their risk category moved down by at least one level.

The Timeline Follow Back is a calendar-based assessment of the frequency and amount of daily alcohol consumption. Its role in this outcome is to document reported drinking patterns and support the categorization of alcohol use into risk levels. Expressing the outcome as a count of participants makes the comparison about how many people met the defined reduction criterion, rather than the size of an individual’s change in grams of alcohol or drinking days.

A categorical endpoint can be clinically interpretable when the question is whether participants cross a defined risk threshold. However, it compresses information: two participants can both meet the reduction definition even if their underlying changes differ, while a participant with a meaningful change that does not cross a category boundary is counted as not meeting the endpoint. The registry’s outcome population also requires completed baseline and week 8 visits and a qualifying amount of week 8 reporting. Thus, the endpoint describes the qualifying participants with the required assessments, not necessarily every randomized participant.

7. Results Posted in the Registry

• 2mg Buprenex and 380mg Vivitrol: 56.7% (17/30) versus 82.1% (23/28) on control; risk difference -25.5 percentage points (95% CI -48.2 to -2.8), risk ratio 0.69 (0.48–0.99); unadjusted two-proportion z-test p = 0.036.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry lists a mixed models analysis for the sponsor’s primary comparison and also permits a simple unadjusted re-calculation from the posted arm counts. These approaches answer related but not identical statistical questions. Understanding their scales and assumptions helps explain why their estimates and p-values may differ.

What does the risk difference describe?

A risk difference compares the proportion meeting an outcome in one arm with the proportion meeting it in the other. Here, the event is a reduction of at least one WHO alcohol-use risk category. The unadjusted re-calculation uses the event counts and participant counts posted for the two arms, without model adjustment. A risk difference below zero in the treatment-minus-placebo direction means the event was less frequent in the treatment arm in these posted counts. This describes the observed comparison; it does not establish why the groups differed.

The risk difference is an absolute measure: it expresses the gap between the two arm-level event probabilities. It is directly tied to the chosen event definition and the analysis population. It can be useful for interpreting how frequently the endpoint occurred in each arm, but its precision depends on the amount of information in those groups and on whether the analysis population represents the intended target population.

What does the odds ratio describe?

The sponsor’s analysis reports an odds ratio. Odds compare the chance of an event with the chance of no event; an odds ratio compares those odds between groups. An odds ratio at its null value indicates no difference in the modeled odds. An odds ratio on one side of the null indicates lower modeled odds in one group, and a value on the other side indicates higher modeled odds. The odds ratio is not the same as a risk ratio or risk difference, and it should not be translated directly into a percentage-point change in risk.

The sponsor’s odds ratio comes from a mixed models analysis, whereas the posted-count re-calculation is unadjusted. In general, model-based results can depend on the model structure and the way repeated or clustered observations are handled. The registry record does not describe the mixed model’s covariates, random effects, correlation structure, or other specifications. It is therefore not possible to attribute the difference between the sponsor result and the re-calculation to a particular modeling choice.

How should the p-values be read?

A p-value describes how compatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The unadjusted re-calculation reports p = 0.036, while the sponsor analysis reports p = 0.08. These are results from different analyses, not two estimates that can be combined or averaged. Given the registry’s discrepancy flag, the sponsor’s reported analysis takes precedence.

Statistical interpretation also depends on the pre-specified analysis plan and on whether multiple outcomes or analyses were considered. The registry lists three outcome measures and three analyses, but no secondary outcomes are posted. A multiplicity adjustment or a hierarchy for interpreting outcomes is not reported in the registry record. Without those details, a single p-value should not be treated as a complete account of the trial’s statistical evidence.

What does a confidence interval add?

A confidence interval conveys the uncertainty around an estimate under the analysis model. A wider interval reflects less precision, while a narrower interval reflects greater precision, all else being equal. The sponsor’s reported interval extends from 0.02 to 1.22 for the odds ratio. Because it includes the null value for an odds ratio, the result is compatible with no modeled difference as well as with effects in either direction allowed by that interval. It does not prove that every value in the interval is equally plausible, and it is not a range containing a known true effect with certainty.

The confidence interval belongs to the sponsor’s model-based estimate. It should not be attached to the unadjusted risk difference or used as though it were an interval for that different measure. The registry’s summary data do not provide enough detail to reproduce the mixed models analysis independently.

Why does the endpoint definition matter?

The primary outcome is binary: each eligible participant either met the threshold for a reduction or did not. Dichotomizing a continuous pattern of drinking into categories makes the endpoint easier to interpret as a count, but it discards detail about the amount and timing of change. Results can also depend on the category boundaries and on the availability and quality of the daily drinking reports used to assign categories. The registry defines the required reporting threshold as being within one standard deviation of the population mean for the 28-day look-back period. How that rule was applied to participants with incomplete or unusual reporting is not reported in the registry record.

The primary analysis population excludes participants without the required completed visits or sufficient reporting, according to the posted definition. A comparison within this population may differ from one based on all randomized participants. The registry does not report a missing-data strategy, an analysis of all randomized participants for this endpoint, or sensitivity analyses addressing the population restriction. These omissions limit assessment of how robust the result is to missing outcomes or exclusions.

10. Confidence Intervals and Precision

The sponsor’s reported odds ratio is 0.17, with a 95% confidence interval from 0.02 to 1.22. The interval is wide and crosses the null value for an odds ratio. This means the posted estimate is imprecise enough that the interval remains compatible with no modeled difference as well as with effects on either side of the null. The point estimate alone should not be read as a settled description of the treatment contrast.

A confidence interval is a measure of uncertainty under the statistical method used to construct it. It is not a guarantee that the true effect falls within the interval, and its interpretation depends on the model and assumptions. Here the interval accompanies the sponsor’s mixed models odds ratio. It does not describe the unadjusted risk difference calculated from the event counts. The registry does not provide the model details needed to assess all assumptions behind the reported interval.

The posted arm counts are 17 of 30 and 23 of 28 for the primary outcome. These counts indicate that the event was observed in both groups, but do not remove uncertainty about the comparative effect. The trial enrolled 69 participants overall, while the endpoint population required completed baseline and week 8 visits and adequate reporting. The registry does not report a detailed account of why the primary outcome counts differ from overall enrollment or how missing observations were handled.

The unadjusted re-calculation has p = 0.036, while the sponsor analysis has p = 0.08. Their difference underscores that precision and inference depend on the analysis used. Because the registry flags a discrepancy, the sponsor’s result takes precedence. Taken together, the sponsor estimate and interval do not establish a precise treatment contrast, and the unadjusted result should not be used to imply greater certainty than the sponsor analysis supports.

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 a report of the trial’s original design calculation. The stated 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 to a stated fraction of the observed effect. The original design assumptions and sample-size justification are not reported in the registry record.

For planning for 100% of the observed effect, the illustration gives 51 participants per arm for 80% power and 67 participants per arm for 90% power. For planning for 75% of the observed effect, it gives 85 participants per arm for 80% power and 113 participants per arm for 90% power. For planning for 50% of the observed effect, it gives 177 participants per arm for 80% power and 236 participants per arm for 90% power.

In plain language, the 80% power column gives the illustrated number of participants in each arm under which the specified effect would be detected by the assumed test in 80% of repeated studies, if the assumptions held. The 90% column uses a higher chance of detection under those assumptions and therefore lists larger planning numbers. The effect fraction describes how large an effect the illustration aims to detect relative to the observed effect: the smaller the effect to be detected, the more participants are needed. These figures are not observed power for the completed trial; they are planning illustrations based on its observed effect.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.821; treatment proportion p₁ = 0.567; average p̄ = 0.694.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6517 = 1.2773.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6263 = 0.5271.
4. n = (1.2773 + 0.5271)² / (-0.2548)² = 50.2, rounded up to 51 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
p0 <- 0.821428571428571                          # control proportion
p1 <- 0.566666666666667                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3852628)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 51 (80% power), 67 (90% power)
 75% of observed effect: n per arm = 85 (80% power), 113 (90% power)
 50% of observed effect: n per arm = 177 (80% power), 236 (90% power)
power.prop.test: 50.2 
Simulated power at n = 51 per arm: 0.815 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 sponsor’s reported result is the mixed models odds ratio of 0.17, with a 95% confidence interval from 0.02 to 1.22 and p = 0.08. The unadjusted re-calculation from the posted arm counts instead yields a risk difference in the direction of fewer participants meeting the reduction criterion in the combination arm, with p = 0.036. The registry flags a discrepancy between these results.

They are not directly interchangeable estimates. The sponsor reports an odds ratio from a mixed models analysis; the re-calculation is a simple unadjusted comparison of the event counts. Different measures and analysis methods can produce different estimates and uncertainty. The registry record does not report enough detail about the model to identify the source of the discrepancy, so no particular explanation should be assumed.

Under the registry’s discrepancy designation, the sponsor’s own analysis takes precedence. The unadjusted calculation is a check on the posted summary counts, not a substitute for the sponsor analysis. The sponsor estimate’s confidence interval includes the null value for an odds ratio, and its p-value is 0.08. Interpretation should therefore preserve the uncertainty reflected in that analysis rather than selecting the more favorable-looking result from the re-calculation.

The registry does not report the full statistical analysis plan or detailed model specifications. It is not possible to determine from the posted analysis alone whether covariate adjustment, handling of repeated observations, analysis-population rules, or other features account for the difference.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists three outcome measures and three analyses, but reports zero posted secondary outcomes. The primary outcome is the reduction in alcohol-use risk category from baseline to week 8. The record does not describe a hierarchy among outcomes, a multiplicity correction, or a rule for controlling the overall type I error rate across analyses. Without that information, the displayed p-values cannot be interpreted as though they were necessarily the only statistical tests considered in the trial.

The reported sponsor analysis is a mixed models analysis, but the registry record does not describe its full model specification or analysis plan. It does not report whether the analysis was adjusted for baseline characteristics, how repeated measurements were represented, or what sensitivity analyses were planned. The original sample-size justification and design assumptions are also not reported in the registry record. The planning illustration based on the observed effect does not supply those missing original design details.

The primary outcome population is defined as randomized participants who completed both baseline and week 8 visits and reported enough consecutive days of alcohol consumption at the week 8 visit. The record defines sufficient reporting by a threshold based on being within one standard deviation of the population mean for the 28-day look-back period. This is a specific analysis-population rule and should be kept in view when interpreting the arm counts.

The registry does not report how many participants failed to meet each part of that population definition, why they were not included in the posted outcome counts, or how missing observations were handled. It does not describe imputation, a treatment-policy analysis, or sensitivity analyses for missing data. Therefore, the posted result cannot be assumed to represent an analysis of every randomized participant, and the potential effect of missing or excluded observations cannot be evaluated from the registry record.

The trial is listed as terminated, but the reason for termination and its relationship, if any, to the analysis or enrollment is not reported in the registry record. The record also does not report interim analyses or stopping rules. No inference about such procedures should be made from the termination status alone.

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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: congestive heart failure, exacerbation (2.9%), fractured vertebrae (2.9%), alcohol withdrawal syndrome (2.9%).

The registry reports serious adverse events for 3 of 35 participants at risk in the combination-treatment group and 0 of 34 at risk in the placebo group. It reports other adverse events for 22 of 35 participants at risk in the combination group and 16 of 34 in the placebo group. Deaths are reported as 0 in each group. These are descriptive safety counts in the populations listed for adverse-event reporting; they are not the same denominators as the primary outcome counts.

The serious-event terms listed with the highest reported percentages are congestive heart failure, exacerbation; fractured vertebrae; and alcohol withdrawal syndrome, each at 2.9%. The registry does not provide additional case descriptions in the registry record presented for this summary, so these terms and percentages should not be expanded into assumptions about severity, timing, relatedness, or outcome.

Counts alone do not establish that an intervention caused or prevented an adverse event. The reported figures do not adjust for time at risk or provide a causal comparison. The registry record also does not report detailed definitions, attribution judgments, or a full narrative of events. These figures should be read as the adverse-event totals recorded for the stated groups, alongside the limitations of the available reporting.

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

The registry’s primary outcome compares the number of eligible participants who moved down by at least one WHO alcohol-use risk category between baseline and week 8. The posted counts are 17 of 30 in the combination-treatment arm and 23 of 28 in the placebo arm. The unadjusted re-calculation points toward fewer participants meeting that criterion in the combination arm, but the registry flags a discrepancy with the sponsor’s mixed models analysis. The sponsor result is the analysis that takes precedence.

The sponsor reports an odds ratio of 0.17, with a 95% confidence interval from 0.02 to 1.22 and p = 0.08. Its interval includes the null value for an odds ratio, leaving considerable uncertainty about the direction and size of the modeled contrast. The unadjusted re-calculation reports p = 0.036, but it is based on a different measure and method and should not displace the sponsor result.

These data do not establish a definitive treatment effect. They describe a specific binary outcome in participants who completed baseline and week 8 assessments and met the reporting requirement. They do not directly quantify change in average drinking, describe all participants’ trajectories, or establish results for the other outcomes of interest. No secondary outcomes are posted in the registry record.

The sample-size figures are planning illustrations based on the observed effect and stated assumptions; they are not evidence of the trial’s original design rationale. The registry’s terminated status is recorded, but its reason is not reported. That status, the limited outcome population, the uncertainty in the sponsor interval, and the analysis discrepancy all matter to how cautiously the result should be interpreted. The figures are descriptive evidence from this trial record, not a standalone basis for conclusions about treatment effects in other populations or settings.

17. Why This Trial Matters Statistically

This trial raises several statistical issues that are common in clinical research. First, the headline outcome is binary even though drinking behavior is continuous and occurs over time. Defining success as a drop of at least one WHO risk category creates a clear event count, but it loses information about the magnitude and pattern of each participant’s drinking change. The event definition is therefore central to what the comparison means.

Second, the primary outcome population is restricted by completion of baseline and week 8 visits and by a reporting threshold. The resulting counts describe participants meeting those conditions, rather than necessarily all randomized participants. Because the registry does not report the reasons for exclusions or the missing-data strategy, readers cannot assess from this record how the population rule affected the result.

Third, the sponsor’s mixed models odds ratio and the unadjusted re-calculation from the counts do not agree in their inferential summaries. This illustrates why an analysis should be interpreted according to its measure, model and stated role, rather than choosing a result because its p-value is smaller. The registry marks the discrepancy, making the sponsor’s own analysis the primary result to interpret.

Finally, the confidence interval is broad and crosses the null value for the odds ratio. It communicates substantial uncertainty, even though the point estimate is on one side of the null. The planning illustration also shows how the required number of participants per arm changes when the target effect is smaller or the desired power is higher. These are useful statistical distinctions: the observed endpoint counts, model-based uncertainty, and future-study planning each answer different questions.

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 primary endpoint population is defined by completion of baseline and week 8 visits and sufficient consecutive days of reported alcohol consumption at week 8. The registry does not report how many participants were excluded by each requirement or how missing outcome data were handled. As a result, the posted arm counts cannot be assumed to represent all randomized participants, and the potential effect of missingness or selection into the analysis population is not assessable from the record.

The sponsor analysis is described as mixed models, but model details are not reported. The registry does not give the model specification, adjustment variables, or sensitivity analyses. The unadjusted re-calculation differs from the sponsor analysis, and the cause of that discrepancy is not reported. Interpretation should therefore prioritize the sponsor result while recognizing that the available registry description does not permit independent reconstruction of it.

The record lists three outcome measures and three analyses, but no secondary outcomes are posted. It does not report a multiplicity strategy, a full analysis plan, or the original sample-size justification. The planning figures based on the observed effect cannot fill those gaps. The registry also marks the trial as terminated without reporting the reason, so the termination cannot be linked to any particular result or design decision.

The sample is limited in size as reflected in the posted enrollment and arm counts, and the sponsor’s confidence interval is broad. The result is specific to the recorded population, intervention schedules, comparator and categorical outcome. The registry record does not establish whether the findings extend to other populations, treatment schedules, or measures of alcohol consumption.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Kappa Opioid Receptor Antagonism for the Tx of AUD and Comorbid PTSD — NCT03852628.
• Registry record last updated 2023-09-14; status last verified 2023-04.
• Results first posted 2023-06-07.
• 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

NCT03852628 is listed as a randomized, triple-masked, parallel-group Phase 2 trial with a recorded enrollment of 69 and a status of terminated. It compared daily sublingual buprenorphine plus extended-release injectable naltrexone with a daily sublingual placebo plus intramuscular placebo. The primary outcome was a reduction of at least one WHO alcohol-use risk category from baseline to week 8, assessed using the Timeline Follow Back.

The posted primary outcome counts are 17 of 30 in the combination arm and 23 of 28 in the placebo arm. The sponsor’s mixed models analysis reports an odds ratio of 0.17, a 95% confidence interval from 0.02 to 1.22, and p = 0.08. An unadjusted re-calculation differs; the registry flags that discrepancy, so the sponsor analysis takes precedence. The reason for termination, original sample-size justification, detailed model specification and missing-data methods are not reported in the registry record.