A Study to Determine the Maintenance of Effect After Long-term Treatment of Sativex® in Subjects With Neuropathic Pain: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Determine the Maintenance of Effect After Long-term Treatment of Sativex® in Subjects With Neuropathic Pain (NCT00713817), a randomized phase 3 study. The primary outcome is Change From Baseline in Mean Daily Pain Severity on a 0-10 Numerical Rating Scale Score at the End of Treatment (Average of Last 7 Days Treatment).

2. Clinical Question

The central statistical question is about the estimated difference in change between the two groups, not simply whether either group’s average score moved. A within-group change can describe the direction of change among the observed subjects, but it does not by itself separate treatment effects from changes that might also occur under placebo or over time. The between-group analysis is intended to address that comparison.

The record describes a study focused on maintenance of effect after long-term treatment, while the primary outcome timeframe is Day 0-35. It does not report additional details about the prior treatment period, the transition into the follow-on study, or the exact timing of random assignment relative to treatment withdrawal. These details would help readers understand the clinical setting of the comparison, but they cannot be inferred from the stated outcomes alone.

3. Trial Design

The registry identifies this as a Phase 3, randomized, parallel study with quadruple masking. It records an enrollment of 19 and lists Sativex® and placebo as drug interventions. The sponsor is Jazz Pharmaceuticals, and the study status is completed. The record gives a start date of 2007-03 and a primary completion date of 2007-07.

The official title describes a multicentre, open-label follow-on study to assess maintenance of effect, tolerance and safety, followed by a randomized-withdrawal phase for a subset of patients. This wording does not match all the design descriptors recorded elsewhere: the registry also labels allocation as randomized, masking as quadruple, and the model as parallel. These descriptions may refer to different parts or aspects of the study, but the record does not explain their relationship. It does not identify which people were masked, how allocation was carried out, or how the randomized-withdrawal phase was organized.

The posted primary outcome concerns change from baseline in mean daily pain severity at the end of treatment, averaged over the last 7 days of treatment. It is reported for 10 subjects receiving Sativex and 9 receiving placebo. The efficacy population is described as randomized subjects who received at least one dose and yielded on-treatment efficacy data. The record does not state whether all enrolled subjects met those criteria or describe the disposition of any subject who did not.

Six secondary outcomes are named, covering pain, treatment failure, loss of response, sleep disruption, global impression of change and adverse events. The registry lists 7 outcome measures overall and 6 analyses, but does not provide results for the secondary outcomes in the registry record described by this record. Detailed eligibility criteria, treatment duration beyond the primary timeframe, dose, adherence, allocation concealment and protocol deviations 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

Analysis population for the primary outcome, as defined in the registry: The efficacy analyses were conducted on data from all subjects who were randomised, received at least one dose of study medication and yielded on-treatment efficacy data.

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 From Baseline in Mean Daily Pain Severity on a 0-10 Numerical Rating Scale Score at the End of Treatment (Average of Last 7 Days Treatment), measured over Day 0-35.

The primary endpoint is change from baseline in mean daily pain severity on a 0–10 numerical rating scale at the end of treatment. The reported endpoint averages scores from the last 7 days of treatment, within the stated Day 0-35 timeframe. Subjects recorded pain severity at the same time each day, at bedtime, in response to a question about pain severity during the preceding 24 hours. The scale anchors were 0 for no pain and 10 for pain as bad as the subject could imagine.

The registry’s endpoint description says that a negative value indicates improvement from baseline. This makes the sign of a change important: a negative mean represents a reduction in the recorded pain score, while a positive mean represents an increase. The posted group mean change is 0.57 for Sativex, with a standard deviation of 1.06, and -0.49 for placebo, with a standard deviation of 2.32. These are group summaries, not individual responses; the standard deviations describe variation in observed changes within each group.

A repeated daily rating averaged over several treatment days can describe pain near the end of the treatment period rather than relying on a single day’s score. Comparing changes between groups is intended to account for baseline levels while assessing the relative pattern of change. The record does not explain the choice of this endpoint, specify a clinically important difference, or report validation details for the scale beyond its wording and anchors. The endpoint therefore supports interpretation of recorded pain-score changes, not broader conclusions beyond the measured outcome and study population.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the mean change represent?

A continuous outcome records measurements on a numeric scale, and a mean summarizes the average observed value or change for a group. Here, the outcome is change from baseline in pain severity, expressed in units on a scale. A negative change is defined in the registry as improvement. The group means therefore describe average changes, not the proportion of subjects who improved, and they do not reveal the full pattern of individual responses.

The standard deviation describes how spread out the observed values are around the group mean. The posted standard deviations are 1.06 for Sativex and 2.32 for placebo. A larger standard deviation indicates greater spread in the observed changes within that group. It is not the same as the standard error of the mean, and it is not a confidence interval for the mean. The record does not post confidence intervals for the separate group means.

Why compare changes between groups?

Changes within each group can be informative descriptively, but a treatment comparison focuses on how the groups differ in their changes. A comparison is needed because changes can occur over time for reasons that are not specific to the study drug. The registry reports an ANCOVA estimate of the mean treatment difference, rather than relying only on the two arm means. ANCOVA is a general regression approach that compares outcomes while accounting for one or more covariates. The exact covariates, model specification and contrast definition are not reported in the registry record, so those features should not be presumed.

The reported estimate is 0.89 units, with a 95% confidence interval from -0.78 to 2.56. A confidence interval conveys uncertainty around an estimated quantity under the model and its assumptions. Its width reflects that the estimate is not known with exact precision. Because this interval contains zero, the reported results are compatible with a range of possible between-group differences that includes no difference, as well as values on either side. The confidence interval does not assign probabilities to individual values within it, and it does not establish that every value in the interval is equally plausible.

What does the p-value say?

The reported p-value is 0.273. In general, a p-value describes how unusual results at least as incompatible with a specified null hypothesis would be under that hypothesis and the statistical model. It is not the probability that the null hypothesis is true, nor the probability that the result occurred by chance in a broad everyday sense. The registry gives the p-value but does not state the full hypothesis wording, test details beyond ANCOVA, or whether this test was adjusted for multiple comparisons.

A p-value should be read alongside the estimate and confidence interval. The estimate conveys the fitted difference, while the interval conveys uncertainty about its size and direction. The p-value alone does not describe magnitude, precision or clinical importance. No threshold for clinical importance is reported for this endpoint, and no decision rule for interpreting the result is stated in the registry record.

Why can dispersion matter?

In general, variability affects how precisely a group mean or group difference can be estimated. Greater variation relative to the amount of information in a study tends to make estimates less precise. For count outcomes, over-dispersion means that observed variability exceeds what a simple count model expects; this can make standard errors too small if the excess variation is ignored. That general issue is not the model used for the primary endpoint: the primary outcome is continuous, and the reported analysis is ANCOVA. The record does not report a count-model analysis for the primary outcome.

For this continuous endpoint, the standard deviations give a direct description of within-group spread. Their difference is a reason to avoid treating the two arm summaries as if they had identical variability, but the registry does not provide enough information to assess model diagnostics or the handling of unequal variances. The reported ANCOVA result should be understood as the result of the stated analysis, not as a model-free comparison.

What does the analysis population mean?

The registry says the efficacy analyses used subjects who were randomized, received at least one dose of study medication and yielded on-treatment efficacy data. This definition narrows the analysis to subjects meeting all three conditions. It is not described as an analysis of every randomized subject regardless of treatment received or outcome availability. The record does not report how many randomized subjects failed to meet the criteria, why data might have been unavailable, or how missing outcome measurements were handled.

Missingness can affect interpretation when subjects without observed outcomes differ from those with observed outcomes. Statistical methods for missing data rely on assumptions, and the choice of method can affect an estimate. No imputation approach, missing-data sensitivity analysis or detailed observation schedule is reported in the registry record. Consequently, the posted efficacy result should be read as pertaining to the stated analysis population, with the handling and impact of unavailable measurements unspecified.

How should the sample-size illustrations be read?

General planning calculations can show how a hypothetical study might be sized to detect a chosen fraction of an observed effect under stated assumptions. Such calculations are planning illustrations, not measurements of how much power the completed study actually had. The record’s stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between arms and a normal approximation; the assumed effect is a stated fraction of the observed effect. No numerical planning results are included in the registry record described for this record, and the trial’s own sample-size justification is not reported in the registry record.

In general, a planned sample size depends on assumptions about the effect to detect, outcome variability, significance level, allocation and analysis. If those assumptions differ from what is observed, the resulting design calculation may not reflect the study’s actual operating characteristics. A post hoc calculation should not be described as observed power or used to replace interpretation of the estimate and its confidence interval. For this study, the directly reported primary analysis remains the ANCOVA estimate, its interval and its p-value.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Sativex", "Placebo")
n   <- c(10, 9)
m   <- c(0.57, -0.49)
sp  <- c(1.06, 2.32)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Sativex  10  0.570  1.060  -0.188  1.328
CARM  Placebo  9  -0.490  2.320  -2.273  1.293

11. Analyses Reported by the Sponsor

The registry reports the sponsor’s analysis as an ANCOVA estimate of the mean treatment difference: 0.89 units, with a 95% confidence interval from -0.78 to 2.56 and p = 0.273. The group summaries are a mean change of 0.57 with standard deviation 1.06 for Sativex and -0.49 with standard deviation 2.32 for placebo. The registry record does not provide a separate unadjusted recalculation of the treatment contrast, so there is no posted recalculated estimate to compare numerically with the sponsor’s result.

A recalculation from summary statistics, when available, is generally unadjusted and may not reproduce an ANCOVA that accounts for covariates. It can also differ because it may use a different variance estimate or analysis population. The registry does not report the covariates or other model details needed to reproduce the sponsor’s analysis. The sponsor’s reported ANCOVA is therefore the available inferential result, while the arm means and standard deviations provide descriptive context.

The record does not identify a discrepancy between the posted summaries and analysis. No claim of agreement or disagreement with an independent recalculation is warranted without a recalculated result and compatible analysis details.

12. Multiplicity, Missing Data and Analysis Populations

The registry names six secondary outcomes in addition to the primary pain-severity endpoint. They concern change in neuropathic pain score, treatment failure, loss of response, sleep disruption, subject global impression of change and adverse events. The record lists 7 outcome measures overall and 6 analyses, but the numerical results for the secondary outcomes are not reported in the registry record.

When several outcomes or comparisons are tested, the chance of obtaining at least one apparently unusual result can increase. In general, multiplicity procedures may adjust the interpretation or testing thresholds to account for multiple analyses. The registry record does not report whether a multiplicity adjustment was planned or used, how the primary endpoint was distinguished from other analyses for decision-making, or whether secondary outcomes were considered confirmatory or exploratory. No adjustment should be assumed.

The stated efficacy analysis population includes randomized subjects who received at least one dose and yielded on-treatment efficacy data. This differs from a definition that automatically retains every randomized subject regardless of treatment or outcome availability. The record does not report the number of randomized subjects excluded from the efficacy population, the reasons for exclusion, or subject flow by analysis population.

Missing outcome data can introduce uncertainty if availability is related to outcomes or treatment. The registry record does not specify how missing daily ratings were handled, whether a minimum number of daily records was required for the end-of-treatment average, or whether sensitivity analyses assessed alternative assumptions. It also does not report the frequency or pattern of missingness. These omissions limit assessment of how robust the reported estimate is to incomplete outcome data.

The sample-size justification, interim analyses and detailed statistical analysis plan are not reported in the registry record. Their absence means that the prespecified hierarchy of analyses, if any, cannot be reconstructed from the posted information. Interpretation should therefore stay with the stated primary result and avoid treating unreported secondary findings or analysis choices as known.

13. 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 1% reporting threshold.

Most frequent non-serious adverse event terms

The posted safety figures list 10 subjects at risk for serious adverse events in the Sativex group and 9 in the placebo group. The number recorded as affected by a serious adverse event is 0 in each group. These figures describe the counts recorded for the listed groups; they do not establish that serious events are impossible or that no event occurred outside the scope of what was captured and reported.

For other adverse events, the record lists 3 affected subjects among 10 at risk in the Sativex group and 2 among 9 at risk in the placebo group. These are counts, not an adjusted comparison, and the record does not provide event-specific detail for these other events in the registry record described. The top serious adverse-event terms are not listed. Death counts are not reported in the registry record.

Safety counts from a study with these group sizes provide limited information about uncommon events. The record does not report confidence intervals for safety outcomes, event severity or relatedness, exposure-adjusted rates, or a fuller breakdown by event type. The figures should be read as the recorded subject counts, not as a comprehensive estimate of risk or a basis for medical decisions.

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

The result establishes what the registry reports for the primary pain-score analysis: the Sativex group had a mean change of 0.57 units and the placebo group had a mean change of -0.49 units, with the endpoint defining negative change as improvement. The reported ANCOVA estimate of the mean treatment difference is 0.89, with a 95% confidence interval from -0.78 to 2.56 and p = 0.273. The interval spans zero, so the analysis does not precisely determine the direction of the adjusted contrast.

This result does not show that the treatments are equivalent. A confidence interval that includes zero can still include nonzero differences, and equivalence requires a defined margin and an analysis designed to assess it. No equivalence margin is reported. Nor does the result show that one group’s observed mean is a treatment effect by itself: a treatment effect is a comparative interpretation, not merely a within-group change.

The estimate does not establish the importance of a difference to subjects, because the registry record gives no threshold for a clinically important change. The confidence interval describes statistical uncertainty under the analysis and its assumptions; it does not explain how subjects experienced their pain, nor does it indicate the probability that an individual will improve.

The inference is limited to the stated endpoint and efficacy analysis population in this study. It cannot be extended from this record to other pain conditions, populations, treatment durations or dosing practices. The registry’s purpose refers to maintenance of effect after long-term treatment, but the primary endpoint is measured over Day 0-35, and details of any prior treatment period are not reported in the registry record.

Secondary outcomes could address other questions, including sleep disruption, treatment failure, loss of response, global impression and adverse events, but their numerical results are not reported in the registry record. The primary result cannot substitute for those outcomes. Overall, the posted evidence describes an uncertain between-group estimate from a small comparison, rather than a definitive conclusion about treatment benefit, lack of benefit or safety.

15. Why This Trial Matters Statistically

This trial is statistically instructive because it distinguishes descriptive group changes from an adjusted comparison. The arm means describe average change within each group, while the ANCOVA estimate addresses a between-group contrast. Those quantities answer related but different questions; neither should be mistaken for the other.

The endpoint also makes direction important. A negative change is explicitly defined as improvement, so the signs of the posted means carry meaning. At the same time, the treatment-difference estimate’s subtraction convention is not stated. This illustrates why numerical contrasts need a clear definition of group order and sign before their direction is interpreted.

Precision is central to reading the result. The estimate is accompanied by a 95% confidence interval from -0.78 to 2.56, which includes zero. Rather than treating the p-value as a verdict, readers can consider the estimate, interval and observed variability together. This is especially relevant with 10 and 9 subjects in the two groups, where the posted results offer limited information about the range of plausible effects.

The record also highlights the importance of understanding analysis populations and missing-data procedures. Efficacy analyses included randomized subjects who received at least one dose and yielded on-treatment efficacy data, but subject flow and missing-data handling are not described. Without those details, readers cannot fully assess the impact of unavailable observations or exclusions.

Finally, the record’s design descriptions raise a reporting question: it labels the study randomized, quadruple-masked and parallel, while the official title describes an open-label follow-on study followed by a randomized-withdrawal phase. Recognizing that the relationship is unexplained is more appropriate than filling in missing design details by assumption. The trial therefore offers a useful example of careful interpretation under limited registry reporting.

16. 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 comparison includes 10 subjects in the Sativex group and 9 in the placebo group. This small group size limits the amount of information available for estimating the between-group difference and makes the reported uncertainty important to interpretation. The record does not report a sample-size justification, so the adequacy of the planned enrollment cannot be assessed from the registry record.

The design descriptions are not fully reconciled. The registry records randomization, quadruple masking and a parallel model, whereas the official title describes an open-label follow-on study followed by a randomized-withdrawal phase for a subset. The record does not explain which design features applied to which phase, who was masked, or how subjects entered the withdrawal phase. This leaves uncertainty about the exact structure of the primary comparison.

The ANCOVA model is identified, but its covariates, assumptions, missing-data handling and contrast definition are not reported. The absence of those details limits independent reproduction and makes it difficult to assess whether model assumptions were suitable. The reported treatment-difference estimate should be read as the registry’s stated analysis rather than as a fully reproducible model specification.

The efficacy population requires randomization, at least one dose and on-treatment efficacy data. Subject disposition, exclusions and missing daily pain ratings are not detailed. The record also does not state how the average of the last 7 days was formed when some daily entries were absent. These gaps prevent assessment of the possible influence of missingness or departures from the intended analysis population.

Numerical results for the named secondary outcomes are not reported in the registry record, and multiplicity procedures are not described. The record does not provide a clinically important difference for the pain scale, detailed safety-event terms, deaths, full dosing information or the sample-size rationale. These limitations restrict conclusions to the posted primary analysis and the safety counts reported.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: A Study to Determine the Maintenance of Effect After Long-term Treatment of Sativex® in Subjects With Neuropathic Pain — NCT00713817.
• Registry record last updated 2023-05-03; status last verified 2023-04.
• Results first posted 2012-08-16.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

20. Record Summary

This completed Phase 3 study record concerns Sativex® and placebo in subjects with neuropathic pain. It lists randomized allocation, quadruple masking and a parallel model, while the official title describes an open-label follow-on study followed by a randomized-withdrawal phase for a subset. Enrollment is 19.

The primary outcome is change from baseline in mean daily pain severity on a 0–10 numerical rating scale, averaged over the last 7 days of treatment within Day 0-35. The posted group means are 0.57 for Sativex and -0.49 for placebo. The ANCOVA estimate is 0.89, with a 95% confidence interval from -0.78 to 2.56 and p = 0.273.

The efficacy population is described as randomized subjects who received at least one dose and yielded on-treatment efficacy data. Serious adverse events affected 0 subjects in each group. Secondary outcome results, detailed model specifications, missing-data methods and the sample-size justification are not reported in the registry record.