Safety and Efficacy of AVP-923 in the Treatment of Levodopa-induced Dyskinesia in Parkinson's Disease Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of AVP-923 in the Treatment of Levodopa-induced Dyskinesia in Parkinson's Disease Patients (NCT01767129), a randomized phase 2 study. The primary outcome is Least Squares Mean Dyskinesia Severity Area Under the Curve (AUC) Score As Assessed By Modified Movement Disorder Society-Unified Dyskinesia Rating Scale (MDS-UDysRS) Part 3.

2. Clinical Question

This was a Phase 2a study, randomized and triple-masked. The registry record lists dyskinesia and Parkinson’s disease as conditions and describes the goal as evaluating efficacy, safety, and tolerability. The main comparison is between treatment-period results, not simply between two independent groups of participants: the crossover structure is relevant to how results should be interpreted. However, details needed to understand the exact crossover analysis, including sequence and period effects, are not reported in the registry record.

3. Trial Design

The registry record describes a Phase 2a, randomized, triple-masked, placebo-controlled crossover study. It lists 14 participants as enrolled. The study status is completed, with a start date of 2013-10-16 and primary completion date of 2015-02-05. Avanir Pharmaceuticals is listed as the sponsor.

Participants received AVP-923-45 and placebo in treatment periods. Each listed treatment was taken twice daily for 14 days. The primary outcome was assessed on the last day of each period, Day 14 and Day 42, during a 2-hour levodopa infusion. Because the model was crossover, participants could contribute measurements under both treatment conditions; the registry record’s primary analysis population is described as all randomized participants who completed the study. The outcome results list 13 participants for each condition.

The record identifies triple masking, but does not specify which groups were masked. It also does not report the randomization sequence, whether treatment order was balanced, a washout interval, or how potential period and carryover effects were addressed. The protocol’s own sample-size justification, planned statistical model details, interim analyses, and missing-data procedures are not reported in the registry record. Those omissions matter because crossover analyses can depend on treatment order and on whether effects from one period persist into the next.

The primary outcome was a continuous measure reported as a least-squares mean with a standard error. The sponsor’s posted comparison is an ANOVA estimate of the mean difference in final values. The registry also lists 10 secondary outcomes, covering disability, motor movement, finger tapping, other MDS-UPDRS and MDS-UDysRS measures, motor-diary ratings, and quality-of-life measures. The record contains 11 outcome measures and 22 analyses overall.

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.

Treatment Period 1 (14 Days)

Reasons for not completing this period:

Treatment Period 2 (14 Days)

Analysis population for the primary outcome, as defined in the registry: Efficacy Evaluable (EE) Population: all randomized participants who completed the study

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Least Squares Mean Dyskinesia Severity Area Under the Curve (AUC) Score As Assessed By Modified Movement Disorder Society-Unified Dyskinesia Rating Scale (MDS-UDysRS) Part 3, measured over Over the 2-hour levodopa infusion period on the last day of each treatment period (Day 14 and Day 42).

The primary endpoint was the least-squares mean dyskinesia-severity area under the curve score, assessed using MDS-UDysRS Part 3. It covered the 2-hour levodopa infusion period on the final day of each treatment period, identified as Day 14 and Day 42. The measure was intended to capture the intensity of involuntary movements associated with treated Parkinson’s disease over that assessment period.

In the registry description, video recordings were evaluated by blinded central raters. The intensity scale considered seven body parts: face, neck, right arm or shoulder, left arm or shoulder, trunk, right leg or hip, and left leg or hip. For each body part, the highest disability score was summed. The scale included disability items related to communication, drinking, and walking, with scores from 0 (normal) to 4 (incapacitating dyskinesia) and a maximum total score of 28. Questions associated with a dressing task were assigned a score of 0 when the task was not performed because of the infusion line. Higher scores indicated more severe symptoms.

An area under the curve combines score information across the defined assessment interval, so the endpoint reflects severity over time rather than a single observation. The registry reports the outcome in score-hours. This endpoint was used to compare dyskinesia severity under the two treatment conditions. It is a clinical rating measure, not a direct measure of every aspect of Parkinson’s disease, daily function, or quality of life; those topics appear in several secondary outcomes.

7. Results Posted in the Registry

• AVP-923-45: mean 966.00 versus 1,049 on control; difference -83.40 (95% CI -200.43 to 33.63), standardized effect d = -0.55; z-test p = 0.162.

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.

Least Squares Mean Disability Area Under the Curve (AUC) Score As Assessed By Modified Movement Disorder Society-Unified Dyskinesia Rating Scale (MDS-UDysRS) Part 4

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • AVP-923-45: mean 409.40 versus 428.30 on control; difference -18.90 (95% CI -60.12 to 22.32), standardized effect d = -0.35; z-test p = 0.369.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AVP-923-45 and the control. The interval gives the range of effects that remain compatible with the data.

Least Squares Mean Motor Movement Area Under the Curve Score As Assessed by Modified Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS), Part III

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • AVP-923-45: mean 6,569 versus 6,397 on control; difference 171.90 (95% CI -890.42 to 1,234), standardized effect d = 0.12; z-test p = 0.751.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AVP-923-45 and the control. The interval gives the range of effects that remain compatible with the data.

Least Squares Mean Timed Finger Tapping Area Under the Curve (AUC) Score

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • AVP-923-45: mean 820.90 versus 772.10 on control; difference 48.80 (95% CI -64.29 to 161.89), standardized effect d = 0.33; z-test p = 0.398.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AVP-923-45 and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover several domains: disability, motor movement, timed finger tapping, other Parkinson’s disease and dyskinesia rating-scale components, motor-diary ratings, and quality of life. The three secondary outcomes with numerical results include estimates in different directions, and their reported confidence intervals include zero. Taken together, the posted pattern does not provide a consistent, precise signal across these distinct measures. Each outcome captures a different aspect of symptoms or function, so they are not interchangeable replications of the primary endpoint.

The primary result also has a point estimate in the direction of a lower dyskinesia-severity score under AVP-923-45, but its confidence interval includes zero. The secondary findings do not resolve that uncertainty. They should be read as supplementary context rather than as independent confirmation of a treatment effect. The registry lists 10 secondary outcomes and reports 22 analyses overall, making it important not to select only the most favorable result.

Secondary analyses are often less persuasive when many outcomes or comparisons are examined without a stated multiplicity adjustment. The chance of observing a seemingly notable result somewhere among multiple tests can increase as the number of tests grows. This is a general statistical issue, not evidence that any particular result is false. The registry record does not report a multiplicity strategy, and the posted secondary p-values should therefore be treated as unadjusted unless otherwise specified in the reported analysis.

The record does not post numerical results for every listed secondary outcome in the material described by these results. Their absence should not be interpreted as showing no effect; it means that the registry record does not report those numerical findings in this set of results. Overall, the secondary measures add breadth to the outcomes considered but do not overcome the small analyzed sample or the uncertainty in the primary comparison.

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

• Least Squares Mean Disability Area Under the Curve (AUC) Score As Assessed By Modified Movement Disorder Society-Unified Dyskinesia Rating Scale (MDS-UDysRS) Part 4
• Least Squares Mean Motor Movement Area Under the Curve Score As Assessed by Modified Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS), Part III
• Least Squares Mean Timed Finger Tapping Area Under the Curve (AUC) Score
• Change From Baseline in MDS-UPDRS Scores for Part I, II, and IV
• Change From Baseline in MDS-UDysRS Scores for Part 1 and 2
• Change From Baseline in PD Motor Diary Ratings Of Duration Of "On-time Without Bothersome Dyskinesia"
• Change From Baseline in Parkinson's Disease Questionnaire-39 (PDQ-39) Domain Scores at the End of Each Treatment Period
• Change From Baseline in PDQ-39 Single Index (PDQ-39-SI) Scores at the End of Each Treatment Period

9. Statistical Interpretation

10. Statistical Methods Explained

The primary measure is continuous: it records a numerical score over an assessment interval, and the registry reports its mean. A mean gives the arithmetic center of the analyzed observations, while a standard error describes the estimated uncertainty in that mean. The primary arm summaries are least-squares means with standard errors. Least-squares means are model-based averages; their exact interpretation depends on the statistical model used. The registry record does not report the full model specification.

What does the reported mean difference represent?

The sponsor’s reported analysis gives a mean difference of -83.4 for AVP-923-45 compared with placebo. The sign reflects the order of the comparison: a negative estimate means the reported value was lower under AVP-923-45. For this endpoint, higher scores indicate more severe symptoms, so a lower score is in the direction of lower measured severity. The difference is expressed in the endpoint’s score-hour units. It is an estimate, not a known fixed effect, and its uncertainty is described by the confidence interval.

The posted analysis method is ANOVA, and the reported parameter is the mean difference in final values. ANOVA is a method for comparing means while accounting for variation represented in a statistical model. The registry does not provide the model terms, assumptions checks, or specific handling of the crossover structure. It is therefore not possible to establish from the registry record exactly how repeated measurements, treatment order, or period were represented in the analysis.

Why does crossover design matter?

In a crossover study, participants receive more than one treatment condition across separate periods. Comparing conditions within the same participants can account for some stable differences between people. At the same time, results can depend on treatment sequence, changes over time, and whether an effect from one period carries into another. These are general considerations for crossover designs. The registry record does not report the sequence allocation, washout details, or the analysis used to assess these potential issues, so their handling cannot be evaluated from the posted information.

What does the confidence interval add?

A confidence interval gives a range of values reasonably compatible with the observed data under the assumptions of the analysis. It is not a guarantee that the true value lies in the interval, nor is it a range of individual participant responses. For the sponsor’s primary comparison, the 95% interval runs from -215.02 to 48.23. Since the interval includes zero, the posted result is compatible with no average difference as well as with differences in either direction. Its width also indicates that the estimate is not very precise.

The p-value reported for the sponsor analysis is 0.191. A p-value measures how unusual results at least as inconsistent with a specified null hypothesis would be if that null hypothesis and the analysis assumptions were true. It does not measure the size or importance of an effect, and it is not the probability that the null hypothesis is true. The p-value should be interpreted alongside the estimate, confidence interval, design, and analysis population rather than as a stand-alone verdict.

How should the recalculation be read?

The recalculated comparison is a simple, unadjusted calculation from the posted summary statistics. It gives the same point estimate, -83.4, but a 95% confidence interval from -200.4 to 33.63 and p = 0.162. Such a calculation cannot reproduce every feature of an analysis model when its details are not available. In particular, the posted summary statistics do not encode all information that may have been used in a crossover ANOVA. The sponsor’s own analysis is therefore the principal result to interpret.

The standard errors in the outcome summaries describe precision of the reported arm means, not the spread of individual observations. The registry does not post standard deviations for these results. A separate recalculation uses an approximate pooled standard deviation, as indicated in the statistical calculations, and should not be mistaken for a sponsor-reported model estimate. General statistical principle: greater variation and smaller analyzed samples tend to make estimates less precise, while model assumptions and design can also affect uncertainty.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is -83.4, and its 95% confidence interval extends from -215.02 to 48.23. This interval spans zero, so the posted analysis does not precisely determine whether the average endpoint value was lower or higher under AVP-923-45. It is also broad relative to the point estimate: a range of substantially different effects remains compatible with the data under the analysis. The p-value of 0.191 is consistent with this uncertainty, but the interval is more informative about the range and direction of effects compatible with the estimate.

The interval is not a prediction interval for individual participants. It describes uncertainty about the estimated group comparison under the statistical model, rather than the range of outcomes any one person might experience. Nor does it assign probabilities to particular values inside the interval. Its interpretation depends on the analysis assumptions, including assumptions of the ANOVA model, which are not detailed in the registry record.

The unadjusted recalculation yields the same point estimate, -83.4, and a 95% confidence interval from -200.4 to 33.63. This interval also includes zero but has different endpoints from the sponsor’s result. The difference underscores why the sponsor-reported analysis should be distinguished from a simple recalculation based on summary statistics. The registry does not provide enough model detail to explain the difference in interval construction.

With 13 participants reported in each treatment condition for the primary endpoint, the data offer limited precision for estimating the comparison. That does not mean the observed estimate is necessarily biased, but it means a wide range of effects remains plausible. The confidence interval should be read as evidence of uncertainty, not as proof that the treatment conditions are identical or that a particular effect size is established.

12. 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 sample-size display gives planning illustrations based on the observed effect, not the trial’s original design assumptions. The registry record does not report the original sample-size justification. For this continuous outcome, the figures are participants per arm, not total events. The calculations assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They consider detecting the stated fraction of the observed effect: 100%, 75%, or 50%.

The 80% power column describes the number of participants per arm that the planning calculation estimates would give an 80% chance of detecting the specified effect if that effect and the calculation assumptions were correct. The 90% power column gives the corresponding per-arm planning number for a 90% chance. These are conventional planning probabilities, not guarantees of a study result. Smaller effects require more participants to detect under the same assumptions, which is why the illustrations rise as the target effect decreases from 100% to 75% to 50% of the observed effect.

For example, the planning rows should be read as alternative assumptions about the effect size to target, not as a recommendation or a reconstruction of the trial’s enrollment plan. They are calculated from the observed effect in a small study, so that estimate itself is uncertain. The trial enrolled 14 participants, whereas the illustrations describe larger hypothetical per-arm sizes. Such calculations can help show how sample size relates to effect size and power, but they do not resolve limitations in the original design or analysis. This is not observed power.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -83.400; pooled standard deviation σ = 152.226.
2. n = 2 × (1.96 + 0.842)² × 152.226² / -83.400² = 52.3, rounded up to 53 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -83.4000000000001                      # observed difference in means
sigma <- 152.22637485009                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1767129)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 53 (80% power), 71 (90% power)
 75% of observed effect: n per arm = 93 (80% power), 125 (90% power)
 50% of observed effect: n per arm = 210 (80% power), 281 (90% power)
power.t.test: 53.3 
Simulated power at n = 53 per arm: 0.802 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s own primary analysis reports an ANOVA mean difference of -83.4, a 95% confidence interval from -215.02 to 48.23, and p = 0.191. The simple unadjusted recalculation from posted summary statistics also gives a point estimate of -83.4, but its 95% confidence interval is -200.4 to 33.63 and its p-value is 0.162. Thus, the estimated difference matches, but the uncertainty calculations do not.

The registry comparison is not marked as discrepant. Even so, a recalculation based on summary statistics is not a substitute for the trial’s own model-based analysis, particularly in a crossover study whose detailed model specification is not reported. The sponsor result should therefore take precedence as the posted analysis. The difference between the two intervals does not alter the broad reading that both include zero, while neither interval establishes that the treatment conditions are equivalent.

Similar distinctions apply to the secondary endpoints: sponsor analyses use ANOVA and report their own estimates, confidence intervals, and p-values, while recalculated quantities are simple unadjusted comparisons. Those calculations can be useful as a check on scale and direction, but they cannot reproduce unreported aspects of the analysis model. The registry record does not state whether the sponsor adjusted secondary analyses for multiple comparisons.

15. Multiplicity, Missing Data and Analysis Populations

The study lists a primary endpoint and multiple secondary outcomes. Testing more than one outcome creates a multiplicity concern: when many hypotheses are tested, the chance of at least one apparently notable result can increase, even if no genuine effect exists for any outcome. This is a general statistical consideration. Appropriate approaches may include prioritizing a prespecified outcome, adjusting the significance threshold, or using a structured testing sequence, but the registry record does not report a multiplicity procedure. The secondary p-values should consequently be interpreted cautiously and not treated as confirmatory evidence on their own.

The registry identifies the primary analysis population as the Efficacy Evaluable population, defined as all randomized participants who completed the study. It reports 13 participants per condition for the primary outcome. The record does not give a fuller account of exclusions, missing observations, or how missing data were handled. It also does not report whether all randomized participants were included in any intention-to-treat analysis. Because the endpoint data are tied to treatment periods in a crossover design, the impact of missing measurements could depend on which period or treatment condition was affected.

General statistical principle: excluding participants who do not complete a study can change the composition of the analyzed group, especially if completion is related to treatment response or tolerability. That possibility cannot be assessed from the registry information alone. The difference between enrollment and the reported efficacy-evaluable counts should not be used to infer a specific reason for noncompletion or a particular missing-data mechanism.

The registry record also does not report interim analyses, stopping rules, or detailed assumptions used in the primary model. These omissions do not establish that such procedures were or were not used; they mean the record does not document them. Overall, the available results support a limited interpretation of the posted comparison, with uncertainty about multiplicity control, missing-data handling, and some elements of the crossover analysis.

16. 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 non-serious adverse event terms

The registry reports serious adverse events for 0 of 14 participants at risk in the AVP-923-45 group and 0 of 13 participants at risk in the placebo group. It also reports 0 deaths in each group. These figures mean that no serious adverse events or deaths were recorded in the listed groups during the safety reporting period; they do not establish that serious risks are absent.

Other adverse events are reported for 11 of 14 participants at risk with AVP-923-45 and 2 of 13 at risk with placebo. These are counts of participants with other adverse events, not counts of individual event episodes. The registry record does not provide the event terms in the adverse-event information, and the list of top serious terms is empty. The small number of participants limits the ability to characterize uncommon harms or compare safety reliably between conditions.

Safety counts should be interpreted alongside the treatment duration and study design. Each treatment period was listed as 14 days, and the trial enrolled 14 participants. The registry record does not give enough detail to determine how adverse events were attributed in the crossover periods or to assess the full timing and severity of non-serious events. These descriptive figures are not a basis for individual medical decisions.

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

The posted primary result is an estimate of dyskinesia-severity AUC under AVP-923-45 compared with placebo in the analyzed population and assessment window. The estimated difference is -83.4, with a sponsor-reported 95% confidence interval from -215.02 to 48.23. This estimate points toward a lower score under AVP-923-45, but the interval includes zero and values in both directions. The study therefore leaves substantial uncertainty about the size and direction of the average difference.

The result does not prove that AVP-923-45 reduces dyskinesia, and it does not prove that the treatments have identical effects. A non-significant p-value is not evidence of equivalence. Establishing equivalence or non-inferiority would require an appropriate design and prespecified margins; such a framework is not reported in the registry record. Likewise, a confidence interval that includes zero does not show that every clinically important effect is ruled out.

The result is specific to the MDS-UDysRS Part 3 intensity-based AUC over the 2-hour levodopa infusion period on the final day of each treatment period. It cannot by itself answer questions about longer-term outcomes, every aspect of Parkinson’s disease, or quality of life. The trial’s crossover design and the registry’s limited description of the analysis also constrain interpretation.

Safety findings are similarly limited: no serious adverse events or deaths were recorded in the listed groups, while other adverse events were reported in both conditions. A small study cannot reliably characterize uncommon events. The registry results are statistical evidence from a small Phase 2a study, not a guarantee about outcomes for any individual or a substitute for clinical judgment.

18. Why This Trial Matters Statistically

This trial illustrates why a point estimate should not be read without its uncertainty. The primary estimate favors a lower dyskinesia-severity score under AVP-923-45 by direction, but the confidence interval is broad and includes zero. Looking only at the sign of the estimate would overstate what the data establish; looking only at the p-value would also lose information about the range of plausible effects.

The crossover structure makes design and analysis details especially relevant. Within-participant comparisons can be informative, but treatment sequence, period effects, and carryover are potential considerations. The registry record identifies the crossover model but does not report enough detail to assess how those considerations were handled. This is a reminder that a trial label alone does not fully describe its statistical analysis.

The study also shows the limits of small-sample evidence. The trial enrolled 14 participants, and the primary outcome results list 13 participants per condition. Such data can generate estimates and inform planning, but they may leave considerable uncertainty. Sample-size planning illustrations based on the observed effect help demonstrate that smaller target effects require larger studies; they are not a replacement for a prespecified design rationale or a guarantee of future results.

Finally, multiple secondary outcomes create a need for cautious interpretation. Results across different measures can add context, but unadjusted testing across many outcomes increases the scope for chance findings. Statistical interpretation depends on estimates, confidence intervals, design, analysis choices, and the hierarchy of outcomes together. No single number captures all of those considerations.

19. 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 analysis is reported for the Efficacy Evaluable population, defined as randomized participants who completed the study. The registry record does not describe a full intention-to-treat analysis, participant-level exclusions, or how missing measurements were handled. The primary outcome lists 13 participants per condition, but the reason for any difference between that count and enrollment is not reported in the registry record.

Although the study is identified as a crossover trial, the record does not report treatment sequence, washout details, or the model terms used to address period, sequence, or carryover effects. These details can affect the interpretation of crossover estimates. The registry reports an ANOVA comparison but does not provide its full specification or diagnostics.

The study enrolled 14 participants and the treatment periods lasted 14 days. The limited sample and treatment duration constrain the precision and scope of the evidence. The confidence interval for the primary comparison is broad, and the results cannot establish effects over longer periods. Safety counts are also too limited to characterize uncommon events reliably.

The registry lists multiple secondary outcomes but does not report a multiplicity adjustment. Numerical findings are not available in the posted results for every listed secondary outcome. The original sample-size justification, interim analyses, and missing-data procedures are not reported in the registry record. Finally, the simple recalculation uses posted summary statistics and is not a reproduction of the sponsor’s full crossover ANOVA; the sponsor-reported analysis is the principal analysis for interpretation.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy of AVP-923 in the Treatment of Levodopa-induced Dyskinesia in Parkinson's Disease Patients — NCT01767129.
• Registry record last updated 2022-04-29; status last verified 2022-04.
• Results first posted 2022-04-29.
• 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.

23. Record Summary

NCT01767129 is a completed Phase 2a randomized, triple-masked, placebo-controlled crossover study sponsored by Avanir Pharmaceuticals. It enrolled 14 participants and evaluated AVP-923-45, containing 45 mg dextromethorphan and 10 mg quinidine, against placebo. Each treatment was taken twice daily for 14 days.

The primary outcome was dyskinesia-severity AUC based on MDS-UDysRS Part 3 during a 2-hour levodopa infusion period on the last day of each treatment period. The sponsor reported a mean difference of -83.4, with a 95% confidence interval from -215.02 to 48.23 and p = 0.191. The estimate points toward a lower score under AVP-923-45, but the interval includes zero and leaves substantial uncertainty about the comparison.

The registry record does not report the original sample-size justification, detailed crossover model, multiplicity procedure, or missing-data approach. Serious adverse events and deaths were reported as 0 in both groups; other adverse events were reported for 11 of 14 participants with AVP-923-45 and 2 of 13 with placebo. The small study supports cautious interpretation of both efficacy and safety findings.