Long-term Efficacy, Safety and Tolerability of Pramipexole in Patients With Idiopathic Moderate to Severe Restless Legs Syndrome (RLS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Long-term Efficacy, Safety and Tolerability of Pramipexole in Patients With Idiopathic Moderate to Severe Restless Legs Syndrome (RLS) (NCT00472199), a randomized phase 4 study. The primary outcome is Change From Baseline in International Restless Legs Syndrome Study Group Rating Scale (IRLS) Total Score After 26 Weeks.

2. Clinical Question

The primary question focuses on the difference in symptom-score change between randomized groups, not simply whether either group changed from baseline. Both groups had lower mean scores at the endpoint than at baseline, and the posted comparison estimates the additional change associated with assignment to pramipexole relative to placebo. A negative difference reflects a larger reduction in score in the pramipexole group, given that higher IRLS scores indicate more severe symptoms.

The registry record also describes objectives concerning clinical and patient global impressions, RLS symptoms, augmentation and rebound, mood, limb pain, quality of life and safety. Those objectives broaden the clinical questions beyond the primary score, but they do not replace the prespecified primary endpoint. Results for these other questions should be interpreted separately and with attention to how many outcomes were assessed.

3. Trial Design

This was a Phase IV randomized, parallel-group trial comparing pramipexole with placebo. The registry record classifies masking as quadruple. It does not specify which particular roles were masked or describe the procedures used to maintain masking. The title describes the trial as double-blind, placebo-controlled and dose-titrated, while the trial record's masking classification is quadruple.

The primary comparison covered 26 weeks. Both groups followed a 4-week flexible titration period, after which the dose level was fixed for 22 weeks. In the pramipexole group, titration began at 0.125 mg once daily, with the potential to adjust in steps to 0.25 mg, 0.5 mg or 0.75 mg according to efficacy and tolerability. The placebo group followed a corresponding titration schedule. This design compares treatment strategies that include dose adjustment, rather than comparing one fixed pramipexole dose with placebo.

The official title describes a further 26-week open-label extension treatment period following the initial 26 weeks. The primary endpoint described in the registry record is change in IRLS total score after 26 weeks, so the reported primary comparison concerns the randomized, placebo-controlled treatment period. Extension-period results are not included in the primary result.

The registry record lists enrollment of 331 and a completed status. The primary endpoint analysis included 162 participants in the pramipexole group and 159 in the placebo group. Its analysis population is defined as randomized patients with a baseline and at least one non-missing post-baseline measure, with last observation carried forward (LOCF) used for post-baseline values. How the trial's original enrollment target was chosen, including the original sample-size justification, is not reported in the registry record.

Several design details that could help readers assess implementation are not reported in the registry record. These include the exact allocation procedure, the identities of the masked roles, and further information about conduct of the extension period. The registered design establishes a randomized placebo comparison, but the available record does not support assumptions about details beyond those stated.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Of 331 patients enrolled, 329 were treated with study drug. Two patients were randomized, but before their first intake of trial medication they decided to not participate in the study.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent to treat analysis. Number of randomized patients with a baseline and at least one non-missing post-baseline measure. Imputation by last observation carried forward (LOCF) for post-baseline values

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 International Restless Legs Syndrome Study Group Rating Scale (IRLS) Total Score After 26 Weeks, measured over Baseline and 26 weeks.

The primary endpoint is change from baseline in the International Restless Legs Syndrome Study Group Rating Scale (IRLS) total score after 26 weeks. The scale ranges from 0, for no RLS symptoms, to 40, for very severe RLS symptoms. It therefore measures symptom severity on a numeric scale, with lower scores representing fewer or less severe symptoms.

Change from baseline is used to describe how the score shifts during the treatment period. Comparing changes between randomized groups helps distinguish the relative pattern of change from improvement or worsening that occurs in both groups. The posted group means are least-squares means, with standard errors, and the reported primary comparison is a mean difference between groups.

In general, a continuous score can retain information across its range rather than reducing each participant's outcome to a yes-or-no category. That can make it useful for detecting differences in the degree of symptom change. Its interpretation still depends on the scale's meaning and on the analysis used. The registry record does not state a threshold for a clinically important IRLS change, so the numerical difference alone does not establish how important that difference would be to patients.

The primary analysis used an intent-to-treat population defined by availability of baseline and at least one post-baseline measure, with LOCF for post-baseline values. This description matters because the endpoint is not simply a comparison of scores from every randomized participant regardless of follow-up. The registry record does not provide further detail on missingness patterns or on alternative missing-data analyses.

7. Results Posted in the Registry

• Pramipexole: mean -13.70 versus -11.10 on control; difference -2.60 (95% CI -4.82 to -0.38), standardized effect d = -0.26; z-test p = 0.022.

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.

Diagnosis of Classified Augmentation According to Independent Expert Panel

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Pramipexole: 11.8% (18/152) versus 9.4% (14/149) on control; risk difference 2.4 percentage points (95% CI -4.5 to 9.4), risk ratio 1.26 (0.65–2.44); unadjusted two-proportion z-test p = 0.491.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Pramipexole and the control. The interval gives the range of effects that remain compatible with the data.

The registry record describes a broad set of secondary questions, including global impressions, responder rates, RLS symptom scales, mood, pain, quality of life, augmentation and safety. The numerical secondary result detailed in the available endpoint information concerns classified augmentation. Other registered objectives provide context for the primary symptom-score question, but their numerical estimates and uncertainty are not reported in the registry record.

For classified augmentation, the estimated risk difference is 0.025, with a 95% confidence interval from -0.045 to 0.094 and p = 0.491. This is not a precise result establishing either a difference or equivalence between groups. It does not contradict the primary IRLS result: the endpoints measure different aspects of the trial, use different outcome scales and can have different levels of uncertainty.

The primary endpoint assesses change in symptom severity on a continuous score, while augmentation is a classification of a particular treatment-associated phenomenon assessed after at least 4 weeks. A group comparison on one endpoint cannot be assumed to determine the result on another. The other secondary objectives likewise address distinct experiences, including global assessment and quality of life, rather than serving as interchangeable measures of the primary outcome.

In general, secondary endpoints can help describe whether a primary result is accompanied by related patterns, but their evidence depends on the quality and completeness of the reported estimates. The registry record lists 31 outcome measures and 24 analyses, and states that 30 secondary outcomes are posted. Without a full set of numerical results and a stated multiplicity strategy, it is not possible to assess all secondary findings together or determine how the trial handled the chance of false-positive results across many comparisons.

Unadjusted secondary results carry less weight because multiple comparisons can produce apparently unusual findings by chance. That is a general statistical concern, not a claim that any specific secondary result is false. The registry record does not report whether multiplicity was controlled. Secondary results are therefore best read as supportive or exploratory evidence rather than as a collection of equally definitive tests.

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

• Clinical Global Impression - Global Improvement (CGI-I) Responder Rate
• International Restless Legs Syndrome (IRLS) Study Group Rating Scale Responder Rate
• Patient Global Impression (PGI) Responder Rate
• Change From Baseline in Restless Legs Syndrome-6 (RLS-6) Score "Satisfaction With Sleep" After 26 Weeks
• Change From Baseline in RLS-6 Score "Severity Falling Asleep" After 26 Weeks
• Change From Baseline in RLS-6 Score "Severity During the Night" After 26 Weeks
• Change From Baseline in RLS-6 Score "Severity During the Day When at Rest" After 26 Weeks
• Change From Baseline RLS-6 Score "Severity During the Day Engaged in Activities" After 26 Weeks

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a continuous measure: change in IRLS total score from baseline to 26 weeks. The group results are reported as least-squares means and standard errors. A least-squares mean is a model-based group mean; it can differ from a simple arithmetic mean when the model adjusts for other variables. The registry record names ANCOVA as the sponsor's method for the reported comparison, but it does not list the covariates included in that model.

What does the mean difference represent?

The reported mean difference compares the two groups' changes in score. The estimate of -2.6 means the average change in the pramipexole group was 2.6 score points lower than the average change in the placebo group. Since a decline in the IRLS score represents less severe symptoms, the direction indicates a larger average reduction in the pramipexole group. It is a comparison of group averages; it does not mean every participant had a change of that size.

In general, ANCOVA models an outcome using treatment group and one or more covariates. For a baseline-adjusted endpoint analysis, baseline score is a common covariate, but whether it was used in this particular model is not reported in the registry record. Adjustment can account for modeled differences in baseline values and may improve precision. The exact model specification and its assumptions are not reported, so the posted ANCOVA result cannot be reconstructed in full from the summary statistics alone.

How does the unadjusted re-calculation differ?

The simple re-calculation uses the posted group means and standard errors to estimate an unadjusted difference and its uncertainty. Its estimated difference is -2.6, matching the reported mean difference at the displayed precision, while its unadjusted p-value is 0.022. The calculation uses an approximate pooled standard deviation and does not reproduce the sponsor's ANCOVA. The resulting inference is therefore a cross-check based on summary information, not a replacement for the trial's reported analysis.

In general, an unadjusted comparison does not account for covariates in the way an adjusted model may. It can be useful for understanding the size and direction implied by summary statistics, but its standard error and p-value may differ from those of a model that uses participant-level data and adjustment. The posted analysis from the sponsor is the direct report of the trial's stated method; the re-calculation provides a transparent, approximate comparison from the summaries.

What does the confidence interval add?

A confidence interval gives a range of estimates compatible with the observed data and the method's assumptions at a stated confidence level. The sponsor's 95% interval for the mean difference ranges from -4.6 to -0.7. Because all values in this interval are negative, it is consistent with a larger average decrease in IRLS score for the pramipexole group. The interval also shows that the estimate is not known exactly. A confidence interval is not a range containing a fixed percentage of individual participant outcomes.

In general, a standard error describes the estimated sampling uncertainty of a statistic. It is not the standard deviation of individual outcomes. The posted standard errors for the group means are 0.8 in each group. The re-calculation estimates a pooled standard deviation of 10.14 using an approximation from the posted summaries. That derived spread describes participant-level variation on the score scale more directly than the standard error does, but it is approximate and is not the adjusted ANCOVA residual variation.

How is the augmentation outcome analyzed?

Classified augmentation is a binary outcome: each participant is counted as having or not having the diagnosis according to an independent expert panel after at least 4 weeks of treatment. The posted comparison expresses the difference in risks between groups. The estimate is 0.025, and its 95% confidence interval runs from -0.045 to 0.094. A risk difference of zero would represent equal observed risks; the interval includes that value. The p-value is 0.491.

In general, binary endpoints can be summarized with risks, risk differences, risk ratios or odds ratios. These measures express different aspects of a group comparison, and their interpretation depends on the scale. Only the risk-difference estimate and its interval are needed to describe the posted result in plain terms. The registry record does not include a sponsor-reported statistical analysis for this secondary endpoint.

Multiplicity is relevant because the record describes numerous additional outcomes and analyses. In general, examining many outcomes increases the chance that at least one result appears statistically unusual by chance if each is assessed without adjustment. Whether any multiplicity-control procedure was used for the trial is not reported in the registry record. Accordingly, secondary p-values should not be treated as if they were the only tests conducted.

The primary analysis population required a baseline and at least one non-missing post-baseline measure, and post-baseline values were imputed by LOCF. In general, LOCF carries a participant's last observed value forward to fill a later missing value. This method makes an assumption about the unobserved outcome after the last measurement, and that assumption may not match the participant's actual course. The registry record does not report the amount or pattern of missing data, reasons for missingness, or sensitivity analyses using alternative assumptions.

11. Confidence Intervals and Precision

The sponsor's 95% confidence interval for the primary mean difference extends from -4.6 to -0.7. It is entirely below zero, consistent with a larger reduction in IRLS score in the pramipexole group under the reported ANCOVA analysis. Its width reflects uncertainty about the size of the between-group difference: the estimate is not precise enough to identify one exact value as the effect.

In general, a confidence interval is more informative than a point estimate alone because it displays a range of values compatible with the data and the method. It does not show the range of individual participant responses, nor does it assign a probability to the true value falling in this particular interval. Its interpretation relies on the model and analysis assumptions. The registry record does not provide the full ANCOVA specification, so the interval should be interpreted as the sponsor's reported model-based interval.

The unadjusted re-calculation also estimates a mean difference of -2.6, with p = 0.022. Its approximate interval is based on summary statistics rather than the sponsor's ANCOVA. Because the underlying calculation is approximate and its exact bounds are not needed to interpret the reported trial analysis, the sponsor's stated interval is the clearest precision estimate for the primary result.

For classified augmentation, the risk-difference estimate is 0.025 and the 95% confidence interval ranges from -0.045 to 0.094. This interval includes zero and extends in both directions, so the observed data are compatible with different possible group relationships. Its breadth cautions against reading the point estimate alone as a settled difference.

Precision depends on both the amount of information and the variability of outcomes. In general, more observations can narrow an interval when other features are similar, while greater outcome variability can widen it. The registry record does not report a clinical-importance boundary for the IRLS scale or for augmentation. Statistical precision therefore does not resolve whether a particular effect size is important to patients.

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 planning illustrations are based on the observed effects, not on the trial's original design assumptions. For the continuous primary endpoint and the binary augmentation outcome, the figures are participants per arm. Neither is a time-to-event outcome; for time-to-event planning, the stated convention is total events rather than participants per arm.

The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. The scenarios consider detecting 100%, 75% or 50% of the observed effect. The original trial's sample-size justification and design assumptions are not reported in the registry record. These illustrative assumptions should not be attributed to the trial's protocol.

The 80% power column gives the planned number of participants per arm that would, under those assumptions, yield an 80% chance of detecting the stated effect if that effect were the true one. The 90% power column gives the corresponding number for a 90% chance. Power is a design property under specified assumptions, not a measure of whether the trial's observed result is true.

The planning values increase as the effect to detect becomes smaller: detecting 75% or 50% of the observed effect requires more participants per arm than planning for the full observed effect. This is because a smaller difference is harder to distinguish from random variation. The values should be read as planning illustrations for hypothetical future studies rather than as a description of the number enrolled in this trial. These are 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 = -2.600; pooled standard deviation σ = 10.136.
2. n = 2 × (1.96 + 0.842)² × 10.136² / -2.600² = 238.6, rounded up to 239 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 <- -2.6                      # observed difference in means
sigma <- 10.1355329516097                     # 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(472199)
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 = 239 (80% power), 320 (90% power)
 75% of observed effect: n per arm = 425 (80% power), 568 (90% power)
 50% of observed effect: n per arm = 955 (80% power), 1,278 (90% power)
power.t.test: 239.5 
Simulated power at n = 239 per arm: 0.801 (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 reported primary analysis is an ANCOVA comparison. It gives a mean difference of -2.6, a 95% confidence interval from -4.6 to -0.7 and p = 0.008. These are the trial's reported model-based results for the primary endpoint.

The simple re-calculation from the posted summary statistics also estimates a difference of -2.6, but its unadjusted p-value is 0.022. The re-calculation is approximate, uses a pooled standard deviation estimated from summary information and does not reproduce the ANCOVA model. Similarity in the displayed point estimate does not imply that the analyses are identical: their uncertainty estimates and p-values can differ because the methods use different information and adjustments.

The registry record does not flag a discrepancy between the reported result and the re-calculation. The sponsor's result should be understood as the reported analysis, while the re-calculation is a check based on available summary statistics. It is not evidence that the sponsor's analysis used the same assumptions as the re-calculation, and it cannot reveal unreported details of the fitted model.

The registry record does not provide the ANCOVA covariates or a complete statistical analysis plan. That limits independent reproduction of the sponsor's estimate and interval from the posted summaries alone. The reported estimate and confidence interval remain the most direct account of the sponsor's analysis; the approximate unadjusted result is useful for illustrating how a simpler comparison relates to it.

15. Multiplicity, Missing Data and Analysis Populations

The trial record lists many outcomes beyond the primary IRLS score, including responder rates, symptom scales, quality-of-life measures, augmentation, rebound and safety. It also records 31 outcome measures and 24 analyses. Assessing multiple outcomes creates a multiplicity issue: in general, as the number of statistical tests increases, the chance of at least one apparently significant result arising by chance can increase when tests are considered separately.

Whether the trial used a hierarchy of endpoints, an adjustment for multiple comparisons, or another multiplicity strategy is not reported in the registry record. The primary result should therefore be distinguished from secondary findings, and secondary p-values should not be read as though they were the sole analysis. The record does not provide enough detail to determine whether any particular secondary analysis was formally confirmatory.

The primary population is described as an intent-to-treat analysis of randomized patients with a baseline and at least one non-missing post-baseline measure. This definition is specific: it does not say that every randomized participant contributed a measured outcome regardless of follow-up. The analysis included 162 participants receiving pramipexole and 159 receiving placebo.

For post-baseline values, the record states that last observation carried forward (LOCF) was used. In general, LOCF substitutes a participant's most recent observed measurement for a later missing measurement. It can keep participants with incomplete follow-up in an analysis, but it assumes the last observed value is a suitable stand-in for the unobserved value. That assumption may be questionable if symptoms change over time or if missingness is related to outcomes.

The amount and pattern of missing data, reasons for missingness, and sensitivity analyses using other approaches are not reported in the registry record. The exact handling of missing baseline data is also not described beyond the analysis-population definition. The record does not report the original sample-size justification, interim analyses, or a full statistical analysis plan. These omissions limit assessment of the planned analysis strategy and how robust the primary result is to alternative assumptions.

16. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: appendicitis (0.6%), viral infection (0.6%), orthostatic hypotension (0.6%), crohn's disease (0.6%), vomiting (0.6%).

Serious adverse events were reported for 8 participants in the pramipexole group, among 166 at risk, and 3 participants in the placebo group, among 163 at risk. These are counts of participants with serious adverse events, not counts of all events or a comparison adjusted for follow-up. The registry record also lists appendicitis, viral infection, orthostatic hypotension, Crohn's disease, vomiting and urinary tract infection among the serious-event terms, each at 0.6%.

Other adverse events were reported for 70 participants in the pramipexole group and 58 in the placebo group, with 166 and 163 at risk, respectively. These figures describe participants affected according to the registry record. The available information does not define the event threshold or provide detailed event-level descriptions, timing, duration or causal assessments, so the counts should not be interpreted as a complete account of the nature or severity of all events.

Deaths are not reported in the registry record. Safety counts should be read alongside the treatment period and population studied, and not as proof that an event was caused by or unrelated to a treatment. The record does not provide enough detail to establish such causal conclusions.

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

The primary result shows that the average IRLS score decreased in both groups over the 26-week comparison, with a larger average decrease in the pramipexole group. The sponsor's reported ANCOVA estimate of the between-group mean difference was -2.6 points, with a 95% confidence interval from -4.6 to -0.7. The interval indicates statistical uncertainty around the estimated difference, while remaining below zero under the reported analysis.

This result does not describe every participant's response. Group means can conceal a range of individual outcomes, and the registry record does not provide the distribution of participant-level changes or a responder analysis for the primary score. It also does not state a threshold for a clinically important difference on the IRLS scale. Consequently, the numerical result alone cannot establish how important the difference is from a patient perspective.

The comparison concerns the treatment schedule and population described in the registry record: patients with idiopathic moderate to severe RLS, flexible dose titration followed by a fixed dose, and a primary assessment after 26 weeks. It does not by itself answer questions about other populations, treatment schedules or longer-term outcomes. Although the official title describes a further 26-week open-label extension, extension results are not part of the primary comparison.

The augmentation estimate is uncertain and its confidence interval includes zero. That does not establish that the groups were equivalent, nor does it establish a difference. The safety counts likewise describe reported events in the recorded groups but cannot establish causation or provide a full clinical benefit-risk assessment.

In general, statistical evidence concerns compatibility between data and specified models; it is not a substitute for clinical judgment. A p-value is not the probability that a treatment effect is absent or present, and a statistically detectable difference does not automatically imply a clinically important one. The registry record provides a primary score comparison and selected safety and secondary information, but not every detail needed to answer all clinical or methodological questions.

18. Why This Trial Matters Statistically

This trial illustrates why a randomized placebo comparison is interpreted through the difference between groups rather than through within-group change alone. Both groups had reductions in mean IRLS score. The estimated mean difference of -2.6 describes the contrast between those changes, which is the central quantity for the primary comparative question.

It also illustrates the distinction between a sponsor-reported adjusted analysis and an unadjusted re-calculation. Both give the same displayed point estimate, while their p-values differ: 0.008 for the sponsor's ANCOVA and 0.022 for the approximate unadjusted re-calculation. The difference underscores that a point estimate does not fully characterize an analysis; model specification and uncertainty estimation matter.

The confidence interval adds information that a p-value alone cannot. The sponsor's interval from -4.6 to -0.7 shows the range of mean differences compatible with the reported analysis at its stated confidence level. At the same time, the absence of a reported clinically important threshold means statistical evidence cannot settle the practical importance of the estimated score difference.

The record also demonstrates why multiple outcomes need careful interpretation. The trial addressed symptom severity, global impressions, quality of life, augmentation and safety. These outcomes ask different questions, and secondary findings can be more vulnerable to chance patterns when many comparisons are made without a reported adjustment strategy.

Finally, the primary analysis used LOCF for post-baseline values. That method is easy to describe but depends on assumptions about later outcomes after a participant's last observed measurement. Information about missingness and sensitivity analyses would help readers assess how dependent the result is on that approach. These issues are central to interpreting a clinical trial result, even when the principal estimate is clearly stated.

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 registry record does not provide the covariates or full specification for the sponsor's ANCOVA. This limits independent reproduction of the reported model-based estimate and confidence interval. The unadjusted re-calculation is approximate and cannot replace an analysis of participant-level data.

The primary analysis used LOCF for post-baseline values. The record does not report how many values were missing, why they were missing, when they were missing, or whether conclusions were tested under other missing-data assumptions. Readers therefore cannot assess from the registry record how sensitive the primary estimate may be to the imputation method.

The original sample-size justification and trial-design assumptions are not reported in the registry record. The planning illustrations based on observed effects are not the trial's original design calculations. Interim analyses and a full statistical analysis plan are also not reported, so the planned handling of multiplicity and other analysis decisions cannot be independently reviewed.

The record describes numerous secondary objectives, but numerical estimates for most of those endpoints are not reported in the information available for interpretation. The augmentation comparison has a confidence interval that includes zero and is not a precise estimate of the group difference. Whether the secondary analyses were adjusted for multiple comparisons is not reported.

The primary result concerns change in an IRLS score after 26 weeks. A threshold for a clinically important change is not reported, and the score comparison does not provide the distribution of individual participant responses. The official title describes an open-label extension, but the primary comparison described in the results is limited to the 26-week randomized phase. These limitations narrow what can be concluded about clinical importance, individual outcomes, and longer-term results.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Long-term Efficacy, Safety and Tolerability of Pramipexole in Patients With Idiopathic Moderate to Severe Restless Legs Syndrome (RLS) — NCT00472199.
• Registry record last updated 2014-06-27; status last verified 2012-05.
• Results first posted 2009-11-17.
• 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

NCT00472199 is a completed Phase IV randomized, parallel-group, placebo-controlled trial of pramipexole in patients with idiopathic moderate to severe RLS. The primary endpoint was change in IRLS total score after 26 weeks. The registry record lists an enrollment of 331 and classifies masking as quadruple.

The sponsor-reported ANCOVA mean difference was -2.6 points, with a 95% confidence interval from -4.6 to -0.7 and p = 0.008. An approximate unadjusted re-calculation gave the same displayed estimate and p = 0.022. Classified augmentation was reported for 18 of 152 participants receiving pramipexole and 14 of 149 receiving placebo; its risk-difference interval included zero.

The record describes LOCF for post-baseline primary outcome values. The original sample-size rationale, multiplicity strategy, full model specification, and detailed missing-data information are not reported in the registry record.