This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for A Within Subject Comparison of Opioid Withdrawal in Opioid Dependent Individuals (NCT01136356), a randomized phase 1, phase 2 study. The primary outcome is Mean Peak Opioid Withdrawal Assessed by the Clinical Opiate Withdrawal Scale (COWS).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT01136356 |
| Phase | Phase 1, Phase 2 |
| Condition | Opioid Abuse, Opioid Dependence |
| Allocation | Randomized |
| Design model | Crossover |
| Masking | Triple |
| Enrollment | 12 |
| Arms in the primary outcome | 2 |
| Lead sponsor | Johns Hopkins University |
| Results posted | Yes |
2. Clinical Question
Population
Opioid-dependent residential volunteers who were described as healthy and out of treatment. The registry lists an enrollment of 12. Each of the two sequence descriptions gives N=7, and the record does not explain how those sequence counts relate to total enrollment.
Intervention
Buprenorphine administered intramuscularly at 32 mg/day in four divided doses for 9 days, followed by an 18-day period of spontaneous withdrawal in the sequence description.
Comparator
Morphine administered intramuscularly at 120 mg/day in four divided doses for 9 days, followed by an 18-day period of spontaneous opioid withdrawal in the sequence description. Participants were randomized to the two treatment sequences.
Primary question
How did mean peak opioid withdrawal, assessed with the observer-rated Clinical Opiate Withdrawal Scale, compare during the study periods following buprenorphine and morphine administration? The registry reports arm-level means and full-range values, but no direct paired comparison estimate or significance test result.
The study also set out to characterize the timing and features of withdrawal using participant-rated and observer-rated measures. Its description identifies sleep disturbance and withdrawal-associated sensitivity to pain as particular areas of interest, and says that physiologic measures and psychomotor performance were collected during chronic opioid administration and placebo administration. Those aims provide context for the clinical question, but the posted primary result concerns COWS ratings.
The comparison is described as within-subject and crossover: participants were assigned to treatment sequences, rather than the record describing a simple parallel-group comparison. That structure matters when interpreting the arm summaries. The posted means alone do not reveal the paired differences within participants, how the sequence was handled analytically, or whether period and carryover effects were assessed. Those details are not reported in the registry record.
3. Trial Design
The registry describes a completed randomized, triple-masked crossover study in Phase 1 and Phase 2. Its stated purpose was to compare spontaneous withdrawal after chronic buprenorphine administration with withdrawal after chronic morphine administration in opioid-dependent individuals. The record lists an enrollment of 12. The arm descriptions each state N=7, one for the morphine-then-buprenorphine sequence and one for the buprenorphine-then-morphine sequence. How those counts correspond to the listed enrollment is not reported in the registry record, so the sequence counts should not be treated as a reconciled accounting of all enrolled participants.
In the morphine-first sequence description, residential volunteers received morphine at 120 mg/day intramuscularly, divided into four doses per day, for 9 days, then underwent an 18-day spontaneous opioid withdrawal period. The buprenorphine-first description gives buprenorphine at 32 mg/day intramuscularly, also divided into four daily doses, for 9 days, followed by an 18-day spontaneous withdrawal period. The registry labels the overall design crossover and says participants were randomized to these sequences.
The trial description says withdrawal was assessed with subject-rated and observer-rated measures. It also identifies physiologic measures and psychomotor performance during opioid administration and placebo administration, with special attention to sleep disturbance and withdrawal-associated hyperalgesia. These are descriptions of what the study set out to assess, not evidence that every such measure has a posted result.
The record lists a primary completion date of 2012-03 and status as completed. Further details about the randomization procedure, masking roles, sequence allocation, washout arrangements, eligibility criteria, adherence, protocol deviations, and conduct of the crossover periods are not reported in the registry record. In particular, the available description does not establish whether any interval between treatments was intended to limit carryover, or how the analysis addressed that possibility.
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.
First Intervention
| Stage | Morphine First, Then Buprenorphine | Buprenorphine First, Then Morphine |
|---|---|---|
| STARTED | 6 | 6 |
| COMPLETED | 6 | 6 |
| NOT COMPLETED | 0 | 0 |
First 18-day Withdrawal
| Stage | Morphine First, Then Buprenorphine | Buprenorphine First, Then Morphine |
|---|---|---|
| STARTED | 6 | 6 |
| COMPLETED | 2 | 6 |
| NOT COMPLETED | 4 | 0 |
Reasons for not completing this period:
| Reason | Morphine First, Then Buprenorphine | Buprenorphine First, Then Morphine |
|---|---|---|
| Withdrawal by Subject | 3 | 0 |
| Physician Decision | 1 | 0 |
Second Intervention
| Stage | Morphine First, Then Buprenorphine | Buprenorphine First, Then Morphine |
|---|---|---|
| STARTED | 2 | 6 |
| COMPLETED | 2 | 6 |
| NOT COMPLETED | 0 | 0 |
Second 18-day Withdrawal
| Stage | Morphine First, Then Buprenorphine | Buprenorphine First, Then Morphine |
|---|---|---|
| STARTED | 2 | 6 |
| COMPLETED | 2 | 5 |
| NOT COMPLETED | 0 | 1 |
Reasons for not completing this period:
| Reason | Morphine First, Then Buprenorphine | Buprenorphine First, Then Morphine |
|---|---|---|
| Withdrawal by Subject | 0 | 1 |
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.
| Characteristic | Unit | Within Subjects Design |
|---|---|---|
| Participants analysed | 12 | |
| Age, Categorical: <=18 years | Participants | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 12 |
| Age, Categorical: >=65 years | Participants | 0 |
| Gender: Female | Participants | 0 |
| Gender: Male | Participants | 12 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 1 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 11 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 |
| Race (NIH/OMB): Asian | Participants | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 |
| Race (NIH/OMB): Black or African American | Participants | 10 |
| Race (NIH/OMB): White | Participants | 2 |
| Race (NIH/OMB): More than one race | Participants | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 |
| Region of Enrollment: United States | participants | 12 |
6. Primary Endpoint
The registry lists the primary outcome as Mean Peak Opioid Withdrawal Assessed by the Clinical Opiate Withdrawal Scale (COWS), measured over Average mean peak opioid withdrawal thirty minutes before and after injection assessed up to 59 days.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | Full Range |
| Unit | units on a scale |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is mean peak opioid withdrawal assessed with the Clinical Opiate Withdrawal Scale (COWS). COWS is described as an observer-rated tool for quantifying opioid withdrawal, with possible scores from 0 to 48. The registry classifies scores from 5 to 12 as mild, 13 to 24 as moderate, 25 to 36 as moderately severe, and scores over 36 as severe withdrawal.
The recorded timeframe is average mean peak opioid withdrawal thirty minutes before and after injection, assessed up to 59 days. The outcome description says repeated measurements were analyzed as mean peak daily COWS ratings. Thus, the endpoint is intended to represent peak observer-rated withdrawal over the assessment period, rather than a single assessment at one moment. The record does not give a more detailed account of how the daily peaks were selected or combined to produce the reported arm means.
Peak ratings can describe the greatest observed withdrawal within an assessment period, which may capture episodes that an average across all observations could smooth out. This is a general feature of peak outcomes, not a claim about the particular participants. Peak measures can also be sensitive to timing and the completeness of repeated assessments; those details are not reported in the registry record.
The scale’s severity categories describe individual COWS scores. An arm-level mean is a group summary, not a classification of every participant’s withdrawal. The posted means therefore should not be read as showing that all participants in an arm experienced the category associated with that numerical value.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The posted arm-level mean for mean peak COWS was 12.6 for morphine and 1.3 for buprenorphine. The registry lists a range from 0.0 to 48.0 for each arm. These summaries describe the reported outcome values, but the record does not specify whether those ranges are observed participant values or scale limits. The scale itself ranges from 0 to 48, so the listed endpoints coincide with its stated limits.
The numerical summaries place the morphine arm mean above the buprenorphine arm mean. That is a descriptive contrast, not a reported estimate of the within-participant treatment effect. The registry does not provide a confidence interval, standard error, p-value, or direct paired contrast for this primary endpoint. It therefore does not establish the precision of the difference between treatment conditions or whether the reported means are compatible with a range of possible paired effects.
The study is described as a crossover, so the same-participant structure is central to interpretation. A suitable analysis would ordinarily need to preserve the relationship between repeated observations from the same person and account for the relevant period or sequence structure. The record states that repeated COWS measurements were analyzed by a two-factor ANOVA for mean peak daily ratings, but it does not name the factors, give the model terms, or report the resulting test statistics. No more specific conclusion about the analysis can be drawn from that description.
The arm descriptions report N=7 for each randomized sequence, while overall enrollment is 12. The registry does not reconcile these counts or state the number included in the primary analysis. Without clear analysis-population counts and participant-level completion information, it is difficult to assess how much information contributed to the posted means or whether missing measurements affected the summaries.
The COWS scale categories offer context for individual scores, but the group means do not reveal the distribution of scores across participants or days. A mean can obscure variability, differing time courses, and participants with high or low peak values. The full-range entries do not resolve that uncertainty, particularly because the registry does not clarify whether they are observed ranges or simply the scale’s limits. The posted result should therefore be read as a pair of descriptive arm summaries, not as a complete account of the withdrawal profiles.
No secondary outcome values are presented in the record, although two secondary outcomes are named: mean daily peak pain ratings using the Visual Analog Scale and mean peak sleep assessed by the Pittsburgh Sleep Quality Index. The primary result does not answer those secondary questions. Overall, the posted data describe different arm-level means on the primary measure, while the registry record does not report the inferential or distributional information needed to quantify the treatment contrast precisely.
9. Statistical Methods Explained
The registry says repeated COWS ratings were analyzed with a two-factor ANOVA for mean peak daily ratings. ANOVA is a general statistical approach for comparing mean outcomes across modeled groups or conditions. A two-factor model includes two factors, but the record does not identify which factors were used, how they were coded, or whether interaction terms were included. It also does not report the estimated effects, test statistics, degrees of freedom, or p-values. The description therefore identifies a method family without enough detail to reconstruct the analysis.
Why the crossover structure matters
In a crossover study, participants receive or are assigned to more than one treatment condition in a planned sequence. Measurements from the same participant are correlated: a person’s ratings may tend to be more similar to their own ratings across periods than to ratings from another person. General statistical practice accounts for this within-person dependence rather than treating every observation as independent. The registry describes this study as within-subject and crossover, but it does not state how within-participant correlation was represented in the two-factor ANOVA.
Sequence and period can also matter in crossover studies. A period effect is a systematic difference associated with when an observation is made, while a carryover effect is influence from an earlier treatment that persists into a later period. These are general design considerations, not findings about this trial. The registry does not report whether period, sequence, or carryover effects were tested or included in the analysis. It also does not describe a washout period. The posted arm means alone cannot show whether those design features affected the comparison.
Repeated measurements and peak scores
The primary outcome is described as repeated, with daily peak COWS ratings and assessment up to 59 days. Repeated measurements can reveal change over time, but they also create dependence among observations from the same participant. Reducing repeated assessments to a peak can make the outcome easier to summarize, while discarding information about the timing and pattern of ratings. The registry does not explain the exact rule for identifying a daily peak, handling multiple assessments around an injection, or calculating the final arm-level mean.
Different approaches to repeated data can answer different questions. A comparison of overall means does not necessarily describe the shape or timing of withdrawal over time. A model with time represented explicitly could address a time course, whereas a summary of peak values emphasizes maximum observed ratings. The record states that mean peak daily ratings were analyzed, but it does not provide enough model detail to tell how time entered the analysis or what estimand the statistical test targeted.
Precision and variability
A confidence interval is a statistical range expressing uncertainty around an estimate under a specified analysis. It helps readers assess which effect sizes are consistent with the data and model, rather than relying only on a point estimate. No confidence interval is reported for the arm means or their comparison. Dispersion values are also not given in the posted arm summaries. The listed full-range values are not a substitute for a standard deviation, standard error, or confidence interval, and the record does not clarify whether those endpoints reflect observed values or the possible scale limits.
Small studies can yield imprecise estimates, but no additional numerical precision assessment can be made from these summaries. The registry lists enrollment of 12 and sequence descriptions of N=7 each; it does not state the primary analysis sample size or report the number of observations contributing to each mean. The relationship among enrollment, sequence counts, and analyzed participants is not explained.
Over-dispersion and count outcomes
Over-dispersion is a general issue for count data when variability exceeds what a chosen model assumes. It can make standard errors too small if the model does not allow for the extra variability. The primary endpoint is a score, and the registry identifies a two-factor ANOVA rather than a count model. No over-dispersion assessment is reported, and there is not enough information to say whether it was relevant to any other analyses.
The record gives sample-size planning assumptions for an illustration based on the observed effect: a two-sided significance level of 0.05, equal allocation, a normal approximation, and effects expressed as a fraction of the observed effect. Such illustrations are planning scenarios, not evidence of observed power and not the trial’s original sample-size justification. The registry does not report the trial’s own sample-size justification. Specific illustration values are not stated in the record, so none can be interpreted as actual enrollment targets or results.
What the analysis description leaves open
The registry does not report diagnostic checks for the ANOVA, the handling of incomplete repeated records, transformations, covariate adjustment, the contrast used for the main treatment comparison, or whether analyses followed an intention-to-treat principle. Nor does it report an interim analysis plan. These omissions do not establish that a procedure was or was not performed; they mean that the procedure is not reported in the registry record. The most defensible reading is to distinguish the stated analysis family from the unreported details required to evaluate its implementation.
10. Multiplicity, Missing Data and Analysis Populations
The registry names one primary outcome and two secondary outcomes. It does not report a multiplicity strategy for interpreting tests across outcomes, time points, or any other comparisons. The number of analyses is listed as 0, while three outcome measures are listed. Those record entries do not supply a set of inferential results or explain the relationship between planned and reported analyses. No adjustment method, hierarchy, or prespecified testing sequence is reported in the registry record.
Multiplicity is a general concern when many statistical questions are tested: the chance of at least one apparently notable result can rise as the number of tests increases. A prespecified strategy can clarify which comparisons are confirmatory and how results should be interpreted. In this record, the primary outcome is identified, but a detailed statistical analysis plan and multiplicity handling are not reported. No interim analyses are reported either.
Missing-data handling is not reported in the registry record. The record does not state whether every randomized participant completed each period, how many COWS observations were missing, why any measurements were absent, or whether incomplete records were excluded or otherwise handled. It also does not report sensitivity analyses for missing data.
The analysis population is not described. Overall enrollment is 12, while each sequence description gives N=7; the record does not reconcile those numbers or specify how many participants contributed to the primary outcome. An intention-to-treat, per-protocol, or other analysis population is not reported. These uncertainties limit assessment of attrition, selection into the analysis, and the extent to which the summaries represent all randomized participants.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| All Participants | 0 | 12 | 0.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| All Participants | 4 of 12 |
Most frequent non-serious adverse event terms
| Event term | Organ system | All Participants |
|---|---|---|
| Nausea | Gastrointestinal disorders | 4 of 12 |
| Insomnia | Psychiatric disorders | 4 of 12 |
| Headache | General disorders | 4 of 12 |
| Indigestion | Gastrointestinal disorders | 3 of 12 |
| Lacrimation Increased | Eye disorders | 2 of 12 |
| Chills | General disorders | 2 of 12 |
The registry lists 12 participants at risk for serious adverse events and 0 affected. It lists 12 at risk for other adverse events and 4 affected. These figures count participants reported as affected, not necessarily the total number of event episodes. The registry does not give serious adverse-event terms; no serious terms are listed.
Deaths are not reported in the registry record. The record also does not provide the types, timing, severity, or treatment attribution of the four participants affected by other adverse events. It is therefore not possible to characterize those events further from the posted figures.
These counts are descriptive safety reporting, not a comparative safety analysis. The registry does not provide arm-specific adverse-event counts or a statistical comparison between treatment conditions. Because the study is described as crossover, the overall counts also do not show which treatment period or exposure was associated with an event. The figures should be read within the limits of the recorded at-risk counts and the missing detail, rather than as a complete account of safety experience.
12. What the Result Does, and Does Not, Mean
The primary posted summaries report mean peak COWS values of 12.6 for morphine and 1.3 for buprenorphine. They show that the arm-level summaries differ numerically on the observer-rated withdrawal scale. They do not, on their own, provide a precise estimate of the within-participant contrast, because the registry does not report a paired effect estimate or confidence interval.
The endpoint captures peak daily withdrawal ratings, not every dimension of withdrawal or the full time course. The study description also discusses participant ratings, pain, sleep, physiologic measures, and psychomotor performance, but the record does not post results for those measures in the material described. The named secondary outcomes are peak pain ratings and peak sleep scores; their numerical results are not reported in the registry record.
The scale ranges from 0 to 48 and has categories for individual scores. Group means cannot be used to assign a severity category to each participant. They also cannot establish how many participants had low, moderate, or high ratings, whether ratings changed consistently over time, or how the distribution differed between sequences. The range entries are listed as 0.0 to 48.0 for both arms, but their interpretation as observed values versus scale limits is unclear.
The study’s randomized crossover design is relevant to the intended comparison, but information needed to assess the analysis is absent: the exact model specification, analysis population, missing-data handling, and paired comparison results are not reported. The sequence counts are also not reconciled with overall enrollment. These gaps prevent the descriptive means from serving as a complete statistical conclusion about the treatment contrast.
Accordingly, the result is best understood as a record of the reported arm-level COWS summaries, with limited information about precision and analysis. It does not establish an individual participant’s withdrawal experience, explain differences in sleep or pain, or provide a complete safety comparison. It is not a basis for medical advice.
13. Why This Trial Matters Statistically
This trial is statistically interesting because its stated question is within-subject, while the readily available primary results are presented as arm-level summaries. A crossover design can use each participant as their own comparison, which may account for stable differences between people. To interpret that advantage, however, readers need a treatment contrast that preserves the paired structure and enough information about sequence, period, and repeated observations. The registry does not report those details for the primary comparison.
The outcome also emphasizes peaks over repeated assessments. Peak scores can be useful when the research question concerns the most pronounced observed withdrawal, but they are not interchangeable with an average over time or a trajectory. The record’s description of mean peak daily ratings and assessments up to 59 days indicates a longitudinal measurement context, while the posted arm means compress that context into summary values. Without time-specific estimates or participant-level distributions, the timing and variability behind the summaries remain unclear.
The difference between a descriptive result and an inferential result is central. Means of 12.6 and 1.3 report the observed arm summaries; without uncertainty estimates and a direct paired comparison, they do not quantify how precisely the contrast is estimated. A reported analysis method, such as the named two-factor ANOVA, is useful context, but the model factors and test results are needed to assess what the method concluded.
Finally, the enrollment and sequence counts illustrate why analysis-population reporting matters. The registry lists enrollment of 12 and N=7 in each sequence description, without explaining the relationship. Clear participant flow, missing-data accounting, and endpoint-specific denominators help readers judge the basis of each estimate. These statistical reporting issues do not negate the posted summaries; they define what those summaries can and cannot support.
14. 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 trial-specific details needed to evaluate the primary comparison are not reported in the registry record. The record does not give a paired treatment-effect estimate, confidence interval, standard error, or p-value. It names a two-factor ANOVA but does not identify the factors, model terms, or test statistics. Thus, the statistical evidence behind the two arm means cannot be independently assessed from the posted result.
The relationship between enrollment of 12 and the N=7 sequence descriptions is unexplained. No participant flow, endpoint-specific analysis counts, or account of withdrawals and incomplete periods is reported. The record does not state how missing COWS observations were handled or identify the primary analysis population. These omissions make it difficult to determine how broadly the posted summaries represent randomized participants.
The registry does not report how randomization was implemented, which roles were masked, or how sequence and period effects were handled. It does not describe a washout period or report assessment of carryover. These are particularly relevant considerations for a crossover comparison, but their absence from the record is not evidence that they were absent from trial conduct.
The posted ranges of 0.0 to 48.0 for both arms match the scale limits, and the record does not clarify whether they represent observed ranges or possible values. No dispersion statistic is posted. The named secondary outcomes lack numerical results, and the adverse-event figures lack event descriptions and treatment-specific counts. Together, these limits constrain interpretation of outcome variability, time course, secondary questions, and safety.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: A Within Subject Comparison of Opioid Withdrawal in Opioid Dependent Individuals — NCT01136356.
- Registry record last updated 2017-03-03; status last verified 2017-01.
- Results first posted 2017-03-03.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry identifies a completed randomized, triple-masked crossover study in Phase 1 and Phase 2, with enrollment of 12. It compares sequences involving intramuscular morphine or buprenorphine administration in opioid-dependent residential volunteers described as healthy and out of treatment. The primary outcome is mean peak opioid withdrawal measured with COWS.
The posted primary means are 12.6 for morphine and 1.3 for buprenorphine; the listed range for each is 0.0 to 48.0. The record states that repeated mean peak daily COWS ratings were analyzed using a two-factor ANOVA, but it does not report a direct comparison estimate, uncertainty interval, or test result. Two secondary outcomes are named, with no numerical results reported in the registry record.
Overall enrollment and sequence counts are not reconciled, and details of the analysis population, missing-data handling, and crossover-period modeling are not reported. Safety entries list 0 of 12 participants affected by serious adverse events and 4 of 12 affected by other adverse events; further event details are not reported in the registry record.