Oral vs. Injectable Naltrexone for Hospitalized Veterans With Alcohol Dependence: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Oral vs. Injectable Naltrexone for Hospitalized Veterans With Alcohol Dependence (NCT01856712), a randomized phase 3 study. The primary outcome is Retention Rate: Percentage of Participants Attended an Initial Behavioral Treatment Visit Within 2 Weeks of Hospital Discharge..

2. Clinical Question

3. Trial Design

The registry describes NCT01856712 as a Phase 3, randomized, parallel-group trial with no masking. The sponsor was the University of Wisconsin, Madison. The recorded enrollment was 54. The study started in 2013-05, reached primary completion in 2016-01, and is listed as completed. The brief summary characterizes the work as a pilot feasibility study, so the record contains both that description and a Phase 3 designation.

Participants were assigned to oral or injectable naltrexone. The oral regimen began with a 50 mg dose before discharge and continued with a 30-day prescription. The injectable regimen began with a 380 mg intramuscular dose before discharge and included a second injection one month later. With no masking, participants and study personnel were not blinded to the assigned regimen, according to the registry’s masking designation. The registry record does not report further details about how the randomization sequence was generated or concealed.

The primary outcome was attendance at an initial behavioral treatment visit within 2 weeks of discharge. Its registry timeframe is listed as 12 months, while the outcome definition specifies the visit window after discharge. The record also lists three secondary outcomes: attendance at recommended outpatient substance abuse treatment, medication adherence, and ongoing alcohol consumption. The relationship between the recorded timeframe and the precise assessment schedule for each outcome is not reported in the registry record.

The primary result lists 23 participants in the oral group and 22 in the injectable group. These outcome counts do not account for the full recorded enrollment of 54, and participant flow, exclusions, and the analysis population are not reported in the registry record. The study status is completed, but completion alone does not establish that all enrolled participants contributed to each outcome.

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: 9 participants who didn't receive medication were discharged from the hospital before they could get medication. Therefore, only 45 , out of 54 participants got the medication.

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.

Population: Out of 54 enrolled participants, 9 participants were discharged before they could get medication. 23 participants received oral naltrexone, 22 participants received injectable naltrexone (1 received oral in error and was analyzed as randomized). Total 45 participants received intervention.

6. Primary Endpoint

The registry lists the primary outcome as Retention Rate: Percentage of Participants Attended an Initial Behavioral Treatment Visit Within 2 Weeks of Hospital Discharge., measured over 12 months.

The primary endpoint records how many participants attended an initial behavioral treatment visit within 2 weeks after leaving the hospital. The registry calls this a retention rate, but the posted measure is expressed as a count of participants rather than a percentage. The primary results list 23 participants for oral naltrexone and 22 for injectable naltrexone, with corresponding outcome values of 23 and 22.

This endpoint reflects engagement with behavioral treatment after discharge, not alcohol consumption itself. That distinction matters because the study rationale connected medication adherence to subsequent treatment engagement, while the stated central hypothesis focused on reducing alcohol use in the days immediately after hospitalization. Attendance at an initial visit can describe whether participants made contact with recommended treatment, but it does not by itself establish the amount of treatment received or whether alcohol use changed.

A post-discharge attendance measure is relevant to the study’s stated interest in treatment engagement. It is also a binary event at the participant level: a participant either attended an initial visit within the specified window or did not. The registry record does not report further details about how attendance was verified, how the recorded 12-month timeframe relates to the 2-week visit window, or how missing attendance information was handled.

7. Results Posted in the Registry

• Injectable Naltrexone: 100.0% (22/22) versus 100.0% (23/23) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0), risk ratio 1.00 (1.00–1.00); unadjusted two-proportion z-test p = n/a.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Percentage of Patients Attended Recommended Outpatient Substance Abuse Treatment

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Injectable Naltrexone: 81.8% (18/22) versus 82.6% (19/23) on control; risk difference -0.8 percentage points (95% CI -23.1 to 21.6), risk ratio 0.99 (0.75–1.30); unadjusted two-proportion z-test p = 0.945.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Injectable Naltrexone and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Adhered to Medication

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Injectable Naltrexone: 63.6% (14/22) versus 60.9% (14/23) on control; risk difference 2.8 percentage points (95% CI -25.6 to 31.1), risk ratio 1.05 (0.66–1.65); unadjusted two-proportion z-test p = 0.848.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Injectable Naltrexone and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes extend beyond initial behavioral treatment attendance to recommended outpatient substance abuse treatment, medication adherence, and ongoing alcohol consumption. The posted attendance and adherence comparisons show little separation in their unadjusted estimates, and their confidence intervals are broad. The outcome summaries therefore do not establish a clear, precise difference between the assigned regimens on those measures. Ongoing alcohol consumption is listed as a secondary outcome, but a result for it is not included among the posted secondary results described in the registry record.

This pattern is broadly consistent with caution about drawing strong distinctions between groups from a small set of observed participants, but it does not confirm the primary result. The primary outcome is specifically attendance at an initial behavioral treatment visit within 2 weeks of discharge; the secondary outpatient-treatment measure is a separate outcome. Likewise, medication adherence and ongoing alcohol consumption address different aspects of the study rationale. They should not be treated as interchangeable evidence or combined into a single conclusion.

The secondary p-values are unadjusted. In general, when multiple outcomes are examined, unadjusted statistical tests do not account for the increased opportunity for apparently notable findings across that set of tests. The registry record does not report a multiplicity plan, confirmatory hierarchy, or adjustment. Secondary comparisons are consequently best treated as exploratory descriptions rather than independent confirmation of the primary result. Their interpretation is also constrained by the absence of reported analysis methods, missing-data handling, and analysis populations. The overall pattern is limited evidence about engagement and adherence in the posted counts, not a definitive test of the broader hypothesis about post-discharge alcohol use.

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

• Percentage of Patients Attended Recommended Outpatient Substance Abuse Treatment
• Percentage of Participants Adhered to Medication
• Ongoing Alcohol Consumption

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary outcome is a participant count, and the two groups are presented as a randomized parallel comparison. For a binary outcome, each participant contributes an event or a non-event. Here, the event is attendance at an initial behavioral treatment visit within 2 weeks of discharge. The registry lists 23 events among 23 participants in the oral group and 22 among 22 participants in the injectable group.

What does a risk difference describe?

A risk difference compares the event proportions between groups by subtracting one group’s proportion from the other’s. It is expressed on the proportion scale, with zero indicating equal observed proportions. The recalculated primary risk difference is 0.0, with limits of 0.0 and 0.0. In these posted counts, that result follows from every listed participant in each arm having the event. It should not be taken as evidence of exact equality in the underlying populations: when the observed data contain no non-events, the simple calculation has no observed variation to convey uncertainty.

That limitation is especially important for a binary outcome with all participants counted as events. A conventional estimate of an odds ratio from the raw counts encounters a zero cell because there are no non-events in either group. The registry nevertheless reports an odds ratio of 1.71 with a 95% confidence interval from 0.35 to 9.98. Without a reported method, it is not possible to determine how that odds ratio was obtained or why it differs from the direct unadjusted picture. The registry record does not report whether the analysis used an adjustment, continuity correction, or another approach; none should be inferred.

How does an odds ratio differ?

An odds ratio compares the odds of an event between groups, rather than comparing event proportions directly. A value of 1 represents equal odds on that scale. Values on either side of 1 indicate different estimated odds, but the meaning and precision depend on the data and analysis method. The posted interval is broad and includes 1. It therefore leaves substantial uncertainty about the association represented by the reported estimate. The absence of a reported p-value also means that a p-value-based conclusion for the registry analysis cannot be described from the record.

A confidence interval describes a range of estimates compatible with the data and the method used, subject to the method’s assumptions. It is not a range that guarantees the true value lies within it, and it is not a measure of the probability that a treatment works. For a sparse binary outcome or an outcome with no observed non-events, the choice of method can materially affect the interval. Because the primary analysis method is not reported, readers cannot assess that method’s assumptions from the registry record.

What do the secondary calculations add?

The secondary comparisons include unadjusted risk differences, confidence intervals, and unadjusted p-values. These are simple comparisons of the posted summary counts and do not establish that the original trial used those methods. General statistical practice treats multiple outcome comparisons cautiously: when several outcomes are examined, the chance of at least one apparently unusual result can increase. The record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting the secondary outcomes.

The registry also contains planning illustrations for the secondary attendance and adherence outcomes. They are based on the observed effects, use a two-sided significance level of 0.05, assume equal allocation, and use a normal approximation. The rows describe planning for the full observed effect and for fractions of that observed effect. These are hypothetical sample-size illustrations, not the trial’s design justification and not observed power. The registry record does not report a sample-size justification for the actual trial, and the illustrations do not show the probability that this completed trial detected an effect.

More generally, over-dispersion refers to variability greater than a statistical model expects. It is relevant to some outcome types and models, but the registry’s primary outcome is posted as a participant count for a binary event, and no dispersion estimate or model details are reported. A reader should not assume that a particular model, variance correction, or treatment of repeated measurements was used. The registry record also does not report the randomization procedure, analysis population, missing-data strategy, or whether the reported analysis followed the randomized assignment.

11. Confidence Intervals and Precision

The registry reports an odds ratio of 1.71 for the primary comparison, with a 95% confidence interval from 0.35 to 9.98. This interval spans values below and above 1 and is wide, indicating that the estimate is imprecise. It is compatible with markedly different odds associations, so the point estimate alone should not be read as a settled description of the comparison. The analysis method and p-value are not reported, limiting further interpretation of its precision.

The unadjusted recalculation from the posted counts gives a risk difference of 0.0 and confidence limits of 0.0 and 0.0. This apparent precision is a feature of the observed counts: all listed participants in both groups had the event. It does not mean that the underlying attendance probabilities are known without uncertainty. The raw data contain no non-events with which to estimate variation in this simple contrast, and the recalculated interval should not be treated as a robust measure of population-level precision.

In general, confidence intervals communicate uncertainty around an estimate under the assumptions of the analysis method. They do not resolve uncertainty about missing outcomes, participant selection, or unreported analytic choices. Here, the difference between the posted counts and the reported odds ratio, together with the absence of method details, makes it particularly important to distinguish the reported result from the simple recalculation. The record does not support a precise claim of equivalence between regimens.

12. Analyses Reported by the Sponsor

The registry’s reported primary analysis is an odds ratio of 1.71, with a 95% confidence interval from 0.35 to 9.98. No p-value or analysis method is reported for that result. The direct unadjusted recalculation from the posted counts gives a risk difference of 0.0, with confidence limits of 0.0 and 0.0, because all 23 participants counted in the oral group and all 22 counted in the injectable group attended.

These estimates describe different effect measures, and the raw counts alone do not explain the reported odds ratio. In particular, the registry does not report the analytic method or other details that would make the result reproducible from the posted summary. The recalculation is a simple check of the counts, not a replacement for the registry’s own analysis. Since the record does not identify a reason for the difference, neither an adjusted model nor any other explanation should be assumed. The reported odds ratio should be read as the registry’s analysis, while the recalculation should be read only as the unadjusted implication of the displayed counts. The wide reported interval emphasizes that the primary analysis does not offer a precise estimate.

13. Multiplicity, Missing Data and Analysis Populations

The record lists a primary outcome and three secondary outcomes, but it does not report how the analysis accounted for examining multiple outcomes. A multiplicity plan might specify which outcomes are confirmatory, how tests are ordered, or whether significance thresholds are adjusted. Without such information, the secondary p-values should be regarded as unadjusted, and isolated findings would require particular caution. The registry does not report whether any outcomes or analyses were prespecified beyond the descriptions available in the record.

Details of missing data are also not reported in the registry record. It does not describe whether participants with unavailable attendance, adherence, or alcohol-use information were excluded, treated as non-events, or handled by another method. The recorded enrollment is 54, while the primary outcome arms list 23 and 22 participants. The record does not explain participant flow or the relationship between these counts, so it is not possible to determine the reasons for the difference or its implications for the analysis.

The analysis population is not reported. For a randomized trial, it matters whether results include all participants according to their original assignment, only those with available outcomes, or another defined group. The record does not state which approach was used. It also does not report how protocol deviations, treatment discontinuation, or crossovers were handled. No conclusion about an intention-to-treat or per-protocol analysis can be made from the available record.

The registry lists a timeframe of 12 months for the primary outcome, while its definition focuses on attendance within 2 weeks of discharge. It does not explain the relationship between those descriptions or report timing details for the secondary assessments. These omissions limit assessment of outcome ascertainment and analytic consistency. None of these missing details should be filled in by assuming standard trial practice.

14. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The posted serious adverse event figures list zero affected participants in each group: 0 of 23 at risk in the oral naltrexone group and 0 of 22 at risk in the injectable naltrexone group. The registry lists no serious adverse event terms. These figures describe the recorded serious adverse events for the participants represented in those groups; they do not establish that no serious events occurred outside the recorded data or that the regimens have equal safety.

For other adverse events, the record lists 1 affected participant among 23 at risk in the oral group and 0 among 22 at risk in the injectable group. The registry does not provide further descriptions of the other event, and the death counts are not reported in the registry record. These small event counts offer limited information for comparing safety. They should be interpreted descriptively, especially because the record does not report follow-up completeness, event-collection methods, or a statistical comparison of adverse events.

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

The results describe the participants and outcomes recorded in this trial. In the primary outcome counts, all listed participants in both groups attended an initial behavioral treatment visit within 2 weeks after discharge. The registry also reports an odds ratio of 1.71 with a 95% confidence interval from 0.35 to 9.98. That interval is broad, and the analysis method is not reported. The direct unadjusted recalculation gives a risk difference of 0.0, but this reflects the absence of any recorded non-events in either listed group and should not be mistaken for proof that the regimens have identical effects.

The results do not show that assignment to either regimen caused attendance, nor do they establish equivalence or superiority. Randomization is recorded, but details of the randomization procedure and the analysis population are not reported. The relationship between the enrollment count and the outcome counts is also unexplained. Those omissions prevent a full assessment of who contributed to the results and how the comparison was conducted.

Attendance is not itself a measure of alcohol consumption, long-term treatment engagement, or medication adherence. The primary outcome therefore does not directly test the stated hypothesis about reducing alcohol use immediately after hospitalization. Secondary outcomes are relevant to other parts of the study rationale, but their unadjusted comparisons and limited reporting do not establish a broader treatment effect. Nor do the safety counts establish that the regimens have equivalent safety.

The most defensible conclusion is descriptive: the posted primary counts record attendance for every participant represented in both groups, while the reported odds-ratio estimate is imprecise and cannot be reconciled with the counts from the information in the registry. The record leaves uncertainty about missing outcomes, participant flow, and analytic methods. These results should be understood within those limits, rather than as a definitive answer about comparative effectiveness.

16. Why This Trial Matters Statistically

This trial is statistically interesting because its primary outcome has no observed non-events in either group. That creates a sharp contrast between the simple count-based description, where every listed participant attended, and the reported odds ratio with a wide confidence interval. The example illustrates why effect measures and analytic methods matter: a result can look complete in the observed sample while still offering limited information about the population-level difference.

The registry also juxtaposes a recorded enrollment of 54 with primary outcome groups of 23 and 22. Without participant flow and analysis-population details, readers cannot tell how those counts arose. This is not a minor bookkeeping issue: the population included in an analysis determines what the estimate describes and how broadly it can be interpreted.

Finally, the study’s stated rationale spans medication adherence, behavioral treatment engagement, and alcohol use, while the primary endpoint measures initial treatment attendance. These concepts are related in the research question but are not interchangeable outcomes. Statistical interpretation depends on keeping the measured endpoint distinct from the broader clinical hypothesis. The record’s missing analysis details and multiplicity information further show why transparent reporting is essential for evaluating small, multi-outcome trials.

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.

Several limitations specific to this trial constrain interpretation. The study is described in its summary as a pilot feasibility study, but the registry also designates it Phase 3. The record does not explain the relationship between those descriptions or provide a sample-size justification for the trial. The planning illustrations available for secondary outcomes are not a record of the original design rationale.

The primary outcome counts include 23 participants in the oral group and 22 in the injectable group, while recorded enrollment is 54. Participant disposition and the analysis population are not reported, so the scope of the primary results is uncertain. The registry’s primary analysis method is also unspecified. This is especially consequential because all participants in the posted primary counts had the event, yet the registry reports an odds ratio of 1.71 with a broad confidence interval. The record does not provide enough information to reconcile those results.

The trial was unmasked, according to the registry, and no additional information about outcome verification or attendance ascertainment is reported. The primary outcome timeframe is listed as 12 months, although the measure is defined by an attendance window within 2 weeks of discharge. The timing and handling of secondary outcomes are not fully described. Missing-data methods, multiplicity procedures, randomization details, and analysis populations are not reported.

Finally, the posted primary endpoint captures initial behavioral treatment attendance rather than the alcohol-use outcome named in the central hypothesis. The secondary outcomes address other relevant questions, but their unadjusted comparisons do not resolve the limitations of the primary result or establish a comprehensive effect on post-discharge alcohol use.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Oral vs. Injectable Naltrexone for Hospitalized Veterans With Alcohol Dependence — NCT01856712.
• Registry record last updated 2019-03-05; status last verified 2019-02.
• Results first posted 2019-03-05.
• 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

NCT01856712 is a completed, randomized, open-label, parallel-group Phase 3 trial sponsored by the University of Wisconsin, Madison. It compared oral and injectable naltrexone among hospitalized veterans with alcohol dependence; recorded enrollment was 54. The primary outcome was attendance at an initial behavioral treatment visit within 2 weeks of discharge. The posted primary outcome counts are 23 of 23 for oral naltrexone and 22 of 22 for injectable naltrexone. The registry reports an odds ratio of 1.71 with a 95% confidence interval from 0.35 to 9.98, but does not report the analysis method or a p-value. The simple unadjusted recalculation gives a risk difference of 0.0. Participant flow, missing-data handling, and analysis-population details are not reported in the registry record, limiting conclusions about the comparison.