A Study of Combination Product (Sumatriptan Succinate and Naproxen Sodium) in Migraine Subjects Who Report Poor Response or Intolerance to Short Acting Triptans (Study 1 of 2): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Combination Product (Sumatriptan Succinate and Naproxen Sodium) in Migraine Subjects Who Report Poor Response or Intolerance to Short Acting Triptans (Study 1 of 2) (NCT00383162), a randomized phase 3 study. The primary outcome is Sustained Freedom From Migraine Pain Between 2-24 Hours Post-dose.

2. Clinical Question

The study addressed a group selected for a history of poor response or intolerance to short-acting triptans. Its question was not whether any treatment would prevent migraine, but whether the recorded primary outcome differed between the combination product and placebo when each was used for separate attacks. The registry describes treatment during the mild phase of each attack, an early-intervention approach.

The primary endpoint required more than being pain-free at a single assessment. It required no pain at 2 hours, no rescue medication, and no recurrence of pain or use of rescue medication from 2 to 24 hours after dosing. That definition makes the outcome a measure of sustained response over a specified interval. It also means that rescue medication use was part of how primary success was defined, rather than a separate issue that could be ignored when interpreting the endpoint.

The record reports 10 participants with the primary outcome in the placebo group and 36 in the sumatriptan–naproxen sodium group. Those are recorded arm-level counts. Because the trial used a crossover design, the results arise from participants treating separate attacks under both treatment conditions, but the registry record does not report the participant-level paired outcome pattern.

3. Trial Design

This was a Phase 3, randomized, double-blind, placebo-controlled, outpatient crossover study. The registry describes a two-attack evaluation: participants treated two separate migraine attacks during the mild phase, using the combination product for one attack and placebo for the other. The sequence could be combination product followed by placebo, or placebo followed by combination product. The registry record does not report the randomization method or further details of how the crossover sequence was implemented.

The study enrolled 173 participants. The registry lists the start as 2006-11, primary completion as 2007-10, and status as completed. The primary analysis was defined for an Intent-to-Treat population consisting of people in the Safety Population who provided an evaluation of the investigational product for at least one treated attack. The record does not give a detailed account of exclusions from that population or explain how participants who evaluated only one attack contributed to each endpoint.

For the primary outcome, the registry records 133 participants in the placebo group and 136 in the sumatriptan–naproxen sodium group. The outcome is expressed as participant counts by treatment, even though treatments were assigned in a crossover arrangement across two attacks. These group counts should not be mistaken for evidence that the trial was a conventional parallel-group study: the sequence descriptions establish that each participant was assigned to a treatment order and treated separate attacks under both conditions.

The registry reports that the primary analysis used Generalized Estimating Equations and gives an odds ratio with a confidence interval and p-value. It does not report the model’s covariates, working correlation structure, handling of period or sequence effects, or other implementation details. The protocol’s original sample-size assumptions and any interim analyses 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.

Migraine Attack 1 (Period 1)

Reasons for not completing this period:

Migraine Attack 2 (Period 2)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) population included subjects in the Safety Population who provided an evaluation of their Investigational Product (IP) for at least one treated attack.

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 Sustained Freedom From Migraine Pain Between 2-24 Hours Post-dose, measured over 2 - 24 hours post-dose.

The primary endpoint was sustained freedom from migraine pain between 2 and 24 hours after dosing. The registry defines success as having no pain at 2 hours without using rescue medication, followed by no recurrence of pain and no use of rescue medication from 2 to 24 hours after dosing. The measure is therefore a binary outcome: a participant either met the full set of conditions or did not.

This endpoint is used to capture both an early pain-free state and its maintenance over time. A single pain assessment at 2 hours would not, by itself, establish that relief continued through the remainder of the observation period. Requiring no rescue medication also makes the endpoint sensitive to the need for additional symptom management during that interval. These features give the outcome a sustained-response interpretation, but they also mean that failure can arise from any of its components.

The registry reports participant counts for this outcome and identifies the analysis population as the Intent-to-Treat population: members of the Safety Population who evaluated the investigational product for at least one treated attack. The record does not report component-specific counts, such as how many participants failed because of pain at 2 hours versus later recurrence or rescue medication use. Thus, the overall endpoint can be interpreted as the recorded frequency of meeting the complete definition, but not as a breakdown of which condition most often determined the result.

7. Results Posted in the Registry

• Sumatriptan-Naproxen Sodium: 26.5% (36/136) versus 7.5% (10/133) on control; risk difference 19.0 percentage points (95% CI 10.3 to 27.6), risk ratio 3.52 (1.82–6.80); unadjusted two-proportion z-test p = <0.001.

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.

Pain-Free Assessment at 2 Hours Post-dose

• 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:
  • Sumatriptan-Naproxen Sodium: 39.7% (54/136) versus 17.3% (23/133) on control; risk difference 22.4 percentage points (95% CI 12.0 to 32.8), risk ratio 2.30 (1.50–3.51); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Sumatriptan-Naproxen Sodium than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Rescue Medication Used up to 24 Hours Post-dose

• 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:
  • Sumatriptan-Naproxen Sodium: 29.4% (40/136) versus 63.2% (84/133) on control; risk difference -33.7 percentage points (95% CI -45.0 to -22.5), risk ratio 0.47 (0.35–0.62); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Sumatriptan-Naproxen Sodium than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Sustained Freedom From Migraine

• 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:
  • Sumatriptan-Naproxen Sodium: 23.5% (32/136) versus 8.3% (11/133) on control; risk difference 15.3 percentage points (95% CI 6.7 to 23.8), risk ratio 2.84 (1.50–5.41); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Sumatriptan-Naproxen Sodium than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary results cover several related dimensions of migraine response, including pain-free assessments at specified times, rescue medication use, and sustained freedom outcomes. Across the results described in the registry, the direction of the unadjusted comparisons generally aligns with the primary finding: recorded response outcomes tend to be more frequent under sumatriptan–naproxen sodium, while rescue medication use is less frequent. This broad pattern offers context for the primary endpoint, which itself includes both sustained pain freedom and absence of rescue medication.

These measures are related but not interchangeable. A pain-free assessment at a particular time reflects status at that assessment, whereas a sustained outcome requires a response to persist across an interval. Rescue medication use is a distinct event and can also affect whether a participant meets a sustained endpoint definition. The registry’s secondary outcomes further include assessments of migraine and migraine-associated symptoms. Taken together, they describe multiple facets of the recorded response rather than several copies of the same endpoint.

The registry lists 18 posted secondary outcomes and reports one primary analysis. Multiple outcomes create multiple opportunities for a result to appear noteworthy by chance. The record does not report a multiplicity adjustment or a hierarchy for interpreting secondary outcomes. Accordingly, unadjusted secondary p-values carry less weight than the prespecified primary analysis unless the analysis plan establishes how those tests were controlled. The secondary pattern can support contextual understanding, but it should not be treated as a set of independent confirmations or as proof that every measured outcome has the same level of evidence.

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

• Pain-Free Assessment at 2 Hours Post-dose
• Rescue Medication Used up to 24 Hours Post-dose
• Pain-Free Assessment at 1/2, 1, 4, 8 Hours Post-dose
• Sustained Freedom From Migraine
• Migraine-Free Assessment at 2, 4, and 8 Hours Post-dose
• Sustained Freedom From Migraine-Associated Sinus Pain
• Migraine-Associated Sinus Pain Assessed at Baseline, 2, 4, and 8 Hours Post-dose
• Sustained Freedom From Migraine-Associated Neck Pain

9. Statistical Interpretation

10. Statistical Methods Explained

What does the risk difference describe?

A risk difference compares the proportion of participants meeting an outcome in one group with the proportion in another. A value on the probability scale of zero would indicate equal observed proportions; positive values mean a higher proportion in the treatment group for outcomes coded as success, while negative values mean a lower proportion. For the primary outcome, the unadjusted recalculation is a positive risk difference. This gives an intuitive comparison on the probability scale, but it does not account for the trial’s crossover structure or any model adjustments.

The arm counts help ground that comparison: the registry records 10 participants meeting the primary endpoint with placebo and 36 with sumatriptan–naproxen sodium. A count alone is not a proportion, because the denominators differ. The accompanying denominators are 133 and 136. The posted percentages and recalculated effect measures should be read in relation to these arm-specific denominators, not as raw counts alone.

What does the odds ratio describe?

The sponsor’s reported analysis uses an odds ratio. Odds are the chance of meeting the outcome divided by the chance of not meeting it; an odds ratio compares those odds across treatment conditions. The reported odds ratio is 4.50. This indicates higher estimated odds of meeting the sustained-freedom endpoint under the combination product in the fitted analysis. Odds and probabilities are different scales, especially when an outcome is not rare, so the odds ratio should not be described as a probability ratio.

The registry identifies Generalized Estimating Equations as the analysis method. In general, GEE methods estimate population-average relationships and can accommodate correlated observations, such as repeated measurements from the same participant. That feature is relevant to a crossover study in which participants treat separate attacks under both conditions. The exact correlation structure, covariates, and handling of period or sequence are not reported in the registry record, so the model cannot be reconstructed from the method label alone.

How should the p-value be read?

The sponsor reports p < 0.001 for the primary analysis. In general, a p-value describes how incompatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. Here it accompanies an odds ratio and confidence interval; it should be interpreted alongside the outcome definition, group counts, and uncertainty interval.

The unadjusted recalculation also reports p < 0.001. It is a simple comparison based on posted summary statistics, not a reproduction of the crossover model. The sponsor’s analysis is the analysis reported by the registry, while the recalculation is useful as a separate, transparent summary on the risk-difference scale. Agreement in direction does not make the methods interchangeable.

Why does the crossover design matter statistically?

In a parallel-group study, each participant generally contributes observations under one assigned treatment. In a crossover study, a participant receives more than one treatment in different periods or, in this trial, for separate attacks. Responses from the same person may be correlated. General methods that treat every treatment observation as independent can understate or otherwise misrepresent uncertainty if that correlation is ignored. GEE is one general approach for correlated observations, but the registry record does not give the details needed to determine precisely how this trial’s repeated treatment observations were modeled.

The registry gives treatment-specific participant counts, but it does not provide a paired cross-tabulation showing how each participant’s outcomes compared across the two attacks. Without that paired information, the simple recalculation cannot use within-person response patterns. It is therefore appropriately described as an unadjusted recalculation from posted summary statistics, rather than as a reanalysis of participant-level crossover data.

What does a confidence interval add?

A confidence interval conveys the precision of an estimate under the statistical model. A relatively narrow interval indicates greater precision than a much wider interval, while the interval’s location indicates the range of effect values compatible with the data and assumptions. It does not guarantee that the true effect lies inside the interval, and it does not account for sources of bias or uncertainty omitted from the model.

For the sponsor’s primary odds ratio, the 95% confidence interval is 2.17 to 9.36. This interval excludes the no-association value of 1, consistent with the reported p < 0.001. The unadjusted risk-difference interval is also reported in the statistical results, but its endpoints are not repeated. The distinction between effect scale and analysis method remains important: an interval for an odds ratio is not directly interchangeable with an interval for a risk difference.

What can the secondary analyses contribute?

Secondary outcomes examine other aspects of response, including pain-free status at different times, rescue medication, and sustained outcomes. They can add context about the pattern of recorded results, but each outcome has its own definition and comparison. Unless a suitable multiplicity strategy is specified, testing several outcomes increases the chance of at least one apparently notable result arising by chance. The registry record does not report whether or how such adjustment was planned. Consequently, secondary p-values should be viewed as exploratory or supportive rather than as independent confirmation of the primary finding.

11. Confidence Intervals and Precision

The sponsor’s primary analysis reports an odds ratio of 4.50 with a 95% confidence interval from 2.17 to 9.36. The interval expresses uncertainty about the estimated odds ratio under the fitted model. Its lower and upper limits are both above the no-association value of 1, and the reported p-value is p < 0.001. The interval nevertheless spans a range of effect sizes, so the point estimate should not be treated as exact.

In general, confidence intervals describe the sampling uncertainty represented by a statistical model; they do not include every source of uncertainty. They do not correct for possible bias, incomplete outcome information, or misspecification of a model. They also do not establish that the analyzed population and trial setting represent other populations or treatment circumstances.

The unadjusted recalculation uses the posted treatment-group counts to describe a risk difference, while the sponsor’s estimate is an odds ratio from Generalized Estimating Equations. The two estimates are on different scales and use different analysis approaches. Their precision intervals therefore answer related but not identical questions. In particular, the unadjusted calculation does not use the participant-level paired outcome information that would be relevant to the crossover structure.

The registry reports denominators of 133 for placebo and 136 for sumatriptan–naproxen sodium for the primary endpoint. The record defines the analysis population, but detailed information about missing outcomes and their handling is not reported in the registry record. That limits the ability to assess how incomplete observations may affect precision or introduce bias.

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 figures are planning illustrations based on the observed effect, not the original design calculations for this trial. The trial’s own sample-size justification and original design assumptions are not reported in the registry record. Because these outcomes are participant counts, the figures are participants per arm, not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect.

The 80% power column describes the number of participants per arm that a hypothetical study would require to have an 80% chance of detecting the specified effect under those assumptions, if that effect were the true effect and the planning model were appropriate. The 90% power column uses a higher target chance of detection and therefore generally calls for more participants per arm. These are design calculations, not descriptions of what happened in the completed trial.

The rows for 100%, 75%, and 50% of the observed effect illustrate how the required sample size changes when the target effect is made smaller. Smaller effects are harder to distinguish from random variation and need more participants. For example, for the primary outcome, planning for 100% of the observed effect gives 61 participants per arm for 80% power and 81 participants per arm for 90% power; planning for 50% of the observed effect gives 187 and 250 participants per arm, respectively.

These figures are not observed power. They should not be used to claim that the completed trial had a particular probability of detecting its observed result. They are conditional illustrations for planning future studies, and their relevance depends on whether the assumed effect and statistical conditions are reasonable.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.075; treatment proportion p₁ = 0.265; average p̄ = 0.170.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5312 = 1.0411.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5140 = 0.4326.
4. n = (1.0411 + 0.4326)² / (0.1895)² = 60.5, rounded up to 61 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
p0 <- 0.075187969924812                          # control proportion
p1 <- 0.264705882352941                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(383162)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 61 (80% power), 81 (90% power)
 75% of observed effect: n per arm = 96 (80% power), 128 (90% power)
 50% of observed effect: n per arm = 187 (80% power), 250 (90% power)
power.prop.test: 60.5 
Simulated power at n = 61 per arm: 0.816 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 registry’s primary analysis uses Generalized Estimating Equations and reports an odds ratio of 4.50, a 95% confidence interval from 2.17 to 9.36, and p < 0.001. The unadjusted recalculation uses the posted group summaries to compare the outcome on the risk-difference scale. Its direction is consistent with the sponsor’s reported analysis: the recorded primary outcome is more frequent in the sumatriptan–naproxen sodium group than in the placebo group.

The estimates should not be expected to match numerically because an odds ratio and a risk difference measure effects on different scales, and because the sponsor’s model is not the same as an unadjusted comparison of summary counts. The crossover structure also means the model may account for information that the simple recalculation cannot recover from the arm-level counts. The registry does not report enough model detail to reproduce the sponsor’s estimate.

The registry record identifies the sponsor’s analysis as the reported primary analysis, and no discrepancy is flagged between it and the recalculation. The recalculation is best read as a complementary, unadjusted summary, not as a substitute for the reported analysis. The reported confidence interval and p-value belong to the sponsor’s fitted model.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary endpoint and reports 18 secondary outcomes. It does not report whether a formal multiplicity procedure was used, how the primary and secondary tests were ordered, or whether any secondary outcomes were designated as confirmatory. Without those details, the effect of testing many outcomes on the interpretation of unadjusted p-values cannot be assessed. The secondary comparisons are therefore best treated as supportive or exploratory context rather than as separate, definitive findings.

The registry defines the primary analysis population as the Intent-to-Treat population, consisting of people in the Safety Population who evaluated the investigational product for at least one treated attack. This definition gives some information about who could enter the analysis, but the record does not describe the amount or pattern of missing outcome data. It also does not report whether missing values were imputed, excluded, treated as failures, or handled through another method. Those details are not reported in the registry record.

The crossover setting makes information about incomplete participation particularly relevant: the study involved treatment of two separate attacks, and the defined primary population includes participants who evaluated the product for at least one treated attack. The record does not explain how observations from participants with only one evaluated attack contributed to each treatment comparison, nor does it report a sensitivity analysis for incomplete crossover data. No conclusion should be inferred about the direction or size of any resulting bias.

The registry names Generalized Estimating Equations for the primary analysis, but the model’s correlation structure, covariate adjustment, and handling of treatment sequence or period are not reported. The original statistical analysis plan is also not reported in the registry record. These omissions limit independent assessment of the modeling choices and of the robustness of the reported estimate.

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.

For serious adverse events, the registry records 0 affected participants among 135 at risk in the placebo group and 0 among 140 at risk in the sumatriptan–naproxen sodium group. The registry also records 0 participants with other adverse events in each of those groups. These are counts for the safety information posted in the record; they should not be interpreted as proof that no adverse events occurred beyond the recorded categories or observation period.

The registry does not provide death counts for these groups, so deaths are not reported in the registry record. It also lists no top serious adverse-event terms. The record does not give additional detail about event definitions, duration of safety follow-up, or event ascertainment. Zero recorded serious events in these groups is a description of the posted safety counts, not a guarantee about risk or a comparison capable of ruling out uncommon events.

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

The posted results show that, among the analyzed participants, the recorded primary outcome of sustained freedom from migraine pain was more frequent in the sumatriptan–naproxen sodium group than in the placebo group. The sponsor’s model reports an odds ratio of 4.50 with a 95% confidence interval from 2.17 to 9.36 and p < 0.001. The arm counts are 36 of 136 for the combination product and 10 of 133 for placebo. These statistics describe this trial’s observed results and model-based estimate.

The result does not establish what any individual participant would experience, and it does not identify which component of the endpoint drove a response. The primary definition combines pain status at 2 hours, later recurrence, and rescue medication use through the stated observation period. The registry does not report counts for each component separately. Nor does the result establish that every secondary symptom or outcome changed in the same way.

The finding is specific to the enrolled population, the mild-phase treatment approach, and the two-attack crossover design. The registry describes participants with migraine who had stopped short-acting triptans because of non-response or intolerance. Results from this population and trial arrangement should not automatically be generalized to people with different characteristics or to other treatment settings.

Statistical evidence is not a substitute for clinical judgment. The confidence interval represents model-based uncertainty, and missing details about the model, incomplete observations, and multiplicity limit what can be concluded beyond the reported comparison. These results do not provide medical advice, guarantee an outcome, or resolve questions that the registry record did not address.

18. Why This Trial Matters Statistically

This trial is statistically notable for its crossover structure and binary sustained-response endpoint. Participants treated separate attacks under both treatment conditions, so observations may be correlated within a person. The registry’s use of Generalized Estimating Equations is relevant to that repeated-observation setting, while the unadjusted recalculation from arm-level summaries illustrates the limits of analyses that cannot use participant-level pairing.

The endpoint is also more informative than a single-time-point measure in one specific way: it requires pain freedom at 2 hours and maintenance of that state without recurrence or rescue medication through 24 hours. That compound definition makes the estimand clinically interpretable as sustained response, but it also means the count cannot reveal which element was most responsible for meeting or failing the endpoint.

The contrast between the sponsor’s odds ratio and the recalculated risk difference illustrates why effect measures should be identified by name and not treated as interchangeable. The reported model-based odds ratio and the summary-statistic risk difference tell related but distinct parts of the statistical story. Their interpretation is strengthened by keeping the design, endpoint definition, and uncertainty interval in view.

Finally, the multiple secondary outcomes illustrate why a coherent pattern of results is useful context but does not remove the need to consider multiplicity. The registry record gives substantial detail on endpoints and posted counts, but leaves important aspects of model specification, missing data, and multiplicity unreported. Recognizing both the observed signal and those limits is central to reading clinical trial statistics carefully.

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 reports treatment-specific counts for a crossover study, but it does not provide participant-level paired outcomes. That prevents an independent recalculation that fully reflects within-person comparisons across the two treated attacks. The sponsor’s analysis method is identified as Generalized Estimating Equations, yet the precise model specification and its treatment of correlation, sequence, or period are not reported.

The analysis population includes participants who evaluated the investigational product for at least one treated attack. The record does not give the number with complete evaluations for both attacks, describe missing outcomes by treatment, or explain how incomplete observations were handled. This makes it difficult to judge the potential effects of incomplete crossover data on the reported comparisons.

The registry lists many secondary outcomes, but does not report a multiplicity strategy or the role of those outcomes in the original analysis plan. Unadjusted secondary comparisons can be difficult to interpret when numerous outcomes are examined. The record also does not provide the protocol’s sample-size justification, original design assumptions, interim-analysis details, or a full statistical analysis plan.

The safety listing reports no serious adverse events in the posted group counts, but the record does not provide detailed safety follow-up or additional event-ascertainment information in these results. The trial’s enrollment and specific eligibility group also limit how far the findings can be generalized. These limitations do not negate the reported results; they identify what cannot be independently evaluated from the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Combination Product (Sumatriptan Succinate and Naproxen Sodium) in Migraine Subjects Who Report Poor Response or Intolerance to Short Acting Triptans (Study 1 of 2) — NCT00383162.
• Registry record last updated 2017-02-02; status last verified 2016-12.
• Results first posted 2010-02-17.
• 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

The completed Phase 3 study enrolled 173 participants and used a randomized, double-blind, placebo-controlled crossover design. Participants treated two separate migraine attacks, one with sumatriptan–naproxen sodium and one with placebo, during the mild phase of each attack.

The primary endpoint was sustained freedom from migraine pain from 2 to 24 hours after dosing. The registry reports 36 participants meeting this endpoint in the combination-product group and 10 in the placebo group. The sponsor’s Generalized Estimating Equations analysis reports an odds ratio of 4.50, a 95% confidence interval from 2.17 to 9.36, and p < 0.001.

Important details about the crossover model, missing-data handling, multiplicity, and original sample-size assumptions are not reported in the registry record. The posted results should be interpreted in light of those limits and the specific population and trial design.