Drug and Non-Drug Treatment Of Severe Migraine: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Drug and Non-Drug Treatment Of Severe Migraine (NCT00910689), a randomized phase 4 study. The primary outcome is Change in Number of Migraine Episodes Per 30 Days at Month 10..

2. Clinical Question

3. Trial Design

This was a Phase 4, randomized, quadruple-masked, parallel trial sponsored by Ohio University. The registry lists enrollment of 232 and status as completed. The recorded start was 2001-07 and the primary completion was 2005-11. The masking category is reported as quadruple, but the identities of the masked parties and the practical masking procedures are not reported in the registry record.

The four treatment groups shared Optimal Acute Therapy (OAT). The reference group received OAT plus beta-blocker placebo. Another group received OAT plus propranolol or nadolol. A third received OAT plus Behavioral Migraine Management (BMM) and placebo. The remaining group received OAT plus BMM and propranolol or nadolol. The registry describes the study purpose as evaluating whether preventive medication, behavioral migraine management, or their combination improves the outcome of optimal acute therapy for frequent migraines.

Randomization and parallel assignment are recorded, but the sequence-generation method, allocation concealment procedures, and details of the intended masking are not reported in the registry record. The registry lists the primary analysis population as intent-to-treat and says that it included all randomized participants (N = 232). The endpoint summaries are reported by treatment group, with group sizes of 55, 53, 55, and 69. These are the group sizes for the primary outcome summaries, not a substitute for further information about participant flow or missing outcome data.

The primary completion date concerns the primary outcome timeline; it does not, by itself, describe when every secondary outcome or safety follow-up ended. The registry also lists secondary outcomes for migraine days, quality of life, and migraine episodes at later time points. Detailed scheduling and measurement procedures for those outcomes beyond their listed names and time points 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.

Before assignment: 5 week Optimal Acute Therapy (OAT) Run in (M1) precedes random assignment (see detailed design description).

Month 1: Optimal Acute Therapy Run in

Reasons for not completing this period:

Month 2- 4: Treatment Period

Reasons for not completing this period:

Months 5 - 10: Evaluation Period I

Reasons for not completing this period:

Months 11- 16: Evaluation Period II

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy analyses were intent-to-treat analyses that included all randomized (N = 232) participants.

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 in Number of Migraine Episodes Per 30 Days at Month 10., measured over Change from Month 1 to Month 10.

The primary endpoint was change in the number of migraine episodes per 30 days at Month 10, measured as change from Month 1 to Month 10. An episode required a 24-hour pain-free period between episodes. The measure was obtained from a daily electronic diary, and the registry describes the reported parameter as a mean with a 95% confidence interval. The unit is number of migraine episodes.

Change scores describe how the recorded episode frequency at the later assessment differs from the Month 1 reference. The negative group estimates indicate a decrease in the measured number of episodes over that interval. The size of a group’s mean change is a summary of participants in that group; it does not show each participant’s individual course or the distribution of their changes.

Using a defined interval and diary-based episode records gives the endpoint a direct connection to the trial’s stated question about frequent migraine. Requiring a 24-hour pain-free period helps define when one episode ends and another can count as a distinct episode. In general, repeated diary recording can capture events that may be difficult to reconstruct later from memory, although the registry record does not report diary completion rates or procedures for incomplete entries.

The primary endpoint is distinct from migraine days, migraine episodes at Month 16, and quality of life, which appear among the secondary outcomes. A change in episodes does not automatically imply a corresponding change in migraine days or quality of life. Those outcomes require their own estimates and analyses.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a continuous measure summarized as a mean change. The arm estimates are reported with 95% confidence intervals. In general, a confidence interval describes uncertainty around an estimate under the model and sampling assumptions used to construct it. It is not a range containing every participant’s outcome, and it does not guarantee that the true value falls within the interval. The interval limits for the four group means are reported; limits for the pairwise mean differences are not reported in the registry record.

What does the mean change summarize?

In general, a mean is an average that combines observations into a single summary. A mean change can show the average direction and magnitude of change in a group, while concealing variation between participants. Here, the reported negative means indicate that the average episode count decreased from Month 1 to Month 10 in each group. The estimates do not describe the range of individual changes, the number of participants who improved, or the number whose episode count increased. Those details are not reported in the registry record.

The measure is based on episodes per 30 days, and the registry specifies a 24-hour pain-free period between episodes. The diary method is reported, but the registry does not report how diary observations were transformed into monthly counts, how incomplete diaries were handled, or whether the mean changes were adjusted for baseline values or other covariates. Those details matter when interpreting the exact estimand and the assumptions behind the estimates.

How should the reported tests be read?

The registry reports an overall Mixed Models Analysis across the four groups, with p < .01. In general, a mixed model can account for structured data such as repeated observations, but the model’s specification, covariance structure, included time points, and adjustment variables are not reported in the registry record. The overall p-value indicates evidence against the model’s overall null hypothesis as implemented by the sponsor; the record does not describe that hypothesis in more detail.

The pairwise comparisons are reported as two-sided t-tests, with a mean difference (net), a 95% confidence level, and a p-value. In general, a two-sided test evaluates departures in either direction from a specified null value. A p-value quantifies how incompatible the observed data are with that null under the test assumptions. It does not measure the size or practical importance of an effect. The comparison-specific p-values should be considered alongside the reported mean differences and the group estimates.

The net comparison estimates are 1.2 for combined BMM and beta blocker versus OAT plus placebo, 0.1 for BMM and placebo versus OAT plus placebo, 0.0 for beta blocker versus OAT plus placebo, 1.1 for combined treatment versus BMM and placebo, 1.3 for combined treatment versus beta blocker, and 0.1 for beta blocker versus BMM and placebo. Their associated p-values are, respectively, p < .001, p > .25, p > .98, p < .001, p < .001, and p > .29. The record identifies the tests as two-sided but does not give the pairwise confidence-interval limits.

Why does over-dispersion matter?

In general, over-dispersion occurs when count data vary more than a model’s assumed distribution allows. If present and not handled appropriately, it can affect standard errors and uncertainty estimates. The primary outcome is a count-based episode measure, but the registry describes the reported parameter as a mean and the reported pairwise method as a t-test. Whether over-dispersion was assessed, whether a count model was considered, or how the mixed model accommodated the outcome’s distribution is not reported in the registry record. This is a general consideration for count outcomes, not evidence that over-dispersion occurred in this trial.

What do the planning illustrations mean?

The sample-size material is described as a planning illustration computed from the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It considers effects defined as stated fractions of the observed effect. Such figures are hypothetical planning quantities, not observed trial outcomes and not evidence that the trial had a particular probability of detecting an effect. The registry record does not report the trial’s own sample-size justification, and the planning illustrations should not be interpreted as observed power.

For a continuous endpoint, the planning material is framed as participants per arm. In general, a planning calculation depends on its assumed effect, variability, significance level, allocation, and analysis method. The listed assumptions do not establish that these were the design assumptions used by the sponsor. The registry record does not report the sponsor’s original sample-size calculation or explain whether any interim analyses were planned or conducted.

10. Analyses Reported by the Sponsor

The registry reports the group mean changes and the sponsor’s analysis results separately. The group summaries show mean changes of -2.1 for OAT plus placebo, -2.1 for OAT plus beta blocker, -2.2 for OAT plus BMM and placebo, and -3.3 for OAT plus BMM and beta blocker. The sponsor-reported mixed models analysis across the groups has p < .01, and the sponsor-reported pairwise tests include mean differences (net) and p-values.

For the combined BMM and beta-blocker group, the sponsor reports differences of 1.2 versus OAT plus placebo, 1.1 versus OAT plus BMM and placebo, and 1.3 versus OAT plus beta blocker, each with p < .001. The other reported pairwise comparisons have differences of 0.1, 0.0, and 0.1, with p > .25, p > .98, and p > .29, respectively. The record labels the pairwise procedure as a two-sided t-test and identifies a 95% confidence level, but does not include interval limits for these contrasts.

The registry record does not flag a discrepancy between the sponsor’s results and the posted values. A separate set of recalculated contrast estimates is not described in the registry record, so there is no additional recalculation result to compare with the sponsor’s analyses. The sponsor’s reported tests should therefore be read as the trial’s reported inferential analyses, with their methods and uncertainty details limited to what the registry records.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists an overall mixed models analysis and six pairwise t-test comparisons for the primary endpoint. It does not report whether the pairwise p-values were adjusted for multiple comparisons, whether a testing hierarchy was prespecified, or whether one comparison was designated as the primary contrast. Without that information, the p-values should be understood as reported, but the record does not establish how the family of comparisons was handled. In general, multiple testing can increase the chance of at least one apparently strong result when many hypotheses are assessed; the extent depends on the testing plan and any adjustment used.

The registry identifies the efficacy analyses as intent-to-treat and says they included all randomized participants (N = 232). That description establishes the stated analysis population, but it does not explain how the outcome data were available for every randomized participant. The registry record does not report missing-data counts for the primary outcome, reasons for missingness, imputation procedures, sensitivity analyses, or assumptions about missing observations. It is therefore not possible to determine from the record whether all randomized participants contributed observed Month 10 diary data or how incomplete records affected the analysis.

Details of the mixed model are not reported in the registry record, including how repeated measurements were represented, which observations entered the model, and whether baseline or other covariates were included. The pairwise comparison reports identify t-tests but do not supply further details about assumptions, variance handling, or adjustments. These omissions limit independent assessment of the model assumptions and the robustness of the reported p-values.

The registry lists five secondary outcomes, including changes in migraine days, quality of life, and migraine episodes at Month 16. Results for those outcomes are not described in the registry record for this analysis, so their estimates and statistical significance cannot be assessed. The registry record also does not report a complete protocol-based multiplicity strategy spanning primary and secondary outcomes, or whether interim analyses were planned or conducted.

12. 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 non-serious adverse event terms

The safety figures report serious adverse events affecting 0 participants in each group at Month 5. The corresponding numbers at risk are 55 for OAT plus placebo, 53 for OAT plus beta blocker, 55 for OAT plus BMM and placebo, and 69 for OAT plus BMM and beta blocker. These figures describe the registry’s serious-event counts for the listed groups and time point; they do not establish that no serious events occurred outside the recorded scope or follow-up.

For other adverse events, the registry lists affected counts of 3, 7, 2, and 2 across those groups. The numbers at risk for these other events are not reported in the registry record. Death counts are also not reported in the registry record. The serious-event terms are not listed, and the record does not supply event narratives, severity details, relatedness assessments, or exposure-adjusted summaries.

In general, an event count is not the same as an event rate, and counts alone do not account for differences in follow-up or exposure. These safety entries should therefore be read as the reported counts, not as a complete comparison of risk between groups. The registry record does not provide enough detail to determine the clinical nature or comparative importance of the events.

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

The primary endpoint results describe average changes in migraine episodes per 30 days from Month 1 to Month 10. Each group’s mean change was negative, and the largest reported decrease was in the group receiving OAT, BMM, and beta blocker. The sponsor’s reported pairwise analyses found p < .001 for comparisons of that combined group with each of the other listed groups. The remaining pairwise comparisons have reported p-values above .25, .98, and .29, as specified in the registry record.

These findings concern the recorded endpoint, time interval, treatment combinations, and analysis reported in the registry. They do not show whether the observed changes were important to participants, because a threshold for meaningful change is not reported in the registry record. Nor do the results establish the reason for the group differences, the role of adherence or other care, or whether the same findings would occur in a different population or setting.

A statistically significant comparison is not equivalent to certainty or proof. Its interpretation depends on the study design, outcome data, model assumptions, missingness, and handling of multiple comparisons. Several of these details are not reported in the registry record. Likewise, a comparison with a larger p-value does not prove that two treatment groups have identical effects; it means the reported test did not provide strong evidence against its null hypothesis under the stated method.

The primary outcome should also be kept distinct from the secondary outcomes. Changes in migraine days, quality of life, and later episode counts address related but different questions. Their results are not included in the endpoint estimates, so conclusions about the primary episode-count measure should not be extended to those outcomes. The registry record supports a statistical description of the posted Month 10 endpoint and its reported comparisons, not a broader claim about every outcome or every possible use of the treatment combinations.

14. Why This Trial Matters Statistically

This trial is statistically informative because the treatment groups share OAT while varying the addition of a beta blocker, BMM, or both. The four group summaries make it possible to examine the reported changes under each combination and to compare the combined approach with groups receiving one or neither of the additions. The registry’s reported comparisons show why considering contrasts separately matters: the combined group’s reported net differences and p-values differ from those for beta blocker alone or BMM with placebo.

The endpoint also illustrates the importance of defining a count outcome. The registry specifies a 24-hour pain-free interval between episodes, a monthly time scale, a Month 1 reference, and electronic diary recording. These definitions help establish what the mean change represents. Without them, the same numerical count could be interpreted differently depending on episode rules, observation period, or data source.

The record highlights the distinction between group-level estimates and inferential comparisons. The means and their confidence intervals convey the estimated average changes within groups; the mixed models and t-tests address comparisons across groups. Pairwise interval limits, missing-data methods, and multiplicity handling are not reported, so the reported p-values do not provide a complete account of uncertainty or analysis choices.

Finally, the stated intent-to-treat population gives an important description of the efficacy analysis, but it does not replace information on outcome availability and missing-data handling. Taken together, the trial offers a concrete example of how a registry can report a randomized design, defined endpoint, group estimates, and statistical tests while leaving important aspects of analytic implementation and interpretation undocumented.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several trial-specific details needed to assess the results are not reported in the registry record. These include the randomization sequence and allocation concealment methods, who was masked, and how masking was maintained. The record states that the trial was quadruple-masked but does not explain the masked roles or procedures. It also does not provide the full protocol, the sponsor’s sample-size justification, or details about any planned or conducted interim analyses.

The primary outcome is based on daily electronic diary entries, but diary completion, missing entries, and the handling of incomplete records are not described. The record states that efficacy analyses were intent-to-treat and included all randomized participants, yet it does not explain how missing Month 10 outcomes were addressed or whether sensitivity analyses were conducted. The model specification for the mixed models analysis is also not reported, limiting assessment of its assumptions and the exact meaning of the overall test.

The six reported pairwise comparisons are not accompanied by confidence-interval limits, despite a stated 95% confidence level. Adjustment for multiple testing is not reported. The registry also does not identify a prespecified hierarchy or explain how the primary comparison was chosen. These gaps make it difficult to assess the precision and family-wise interpretation of the comparisons from the record alone.

The registry provides limited safety detail. Serious-event counts are reported, but serious-event terms are absent; other-event denominators and death counts are not reported in the registry record. Follow-up details for these safety summaries are not given beyond Month 5. Finally, the outcome material described for the primary endpoint cannot answer questions about the listed secondary outcomes, whose results and analysis details are not included in the registry record for this analysis.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Drug and Non-Drug Treatment Of Severe Migraine — NCT00910689.
• Registry record last updated 2016-11-22; status last verified 2016-10.
• Results first posted 2009-06-01.
• 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.

19. Record Summary

The registry identifies NCT00910689 as a completed Phase 4, randomized, quadruple-masked, parallel trial sponsored by Ohio University. Enrollment was 232. The recorded start was 2001-07 and primary completion was 2005-11. The study compared combinations of Optimal Acute Therapy, beta-blocker medication or placebo, and Behavioral Migraine Management.

The primary endpoint was change in migraine episodes per 30 days from Month 1 to Month 10, using daily electronic diary data and a 24-hour pain-free interval to distinguish episodes. The registry reports mean changes and 95% confidence intervals for all four groups, an overall mixed models p-value of p < .01, and six pairwise t-tests. The combined BMM and beta-blocker group had the largest reported decrease and pairwise p-values of p < .001 against each other group.

The record identifies the efficacy population as intent-to-treat, including all randomized participants. It does not report key details including missing-data methods, pairwise interval limits, multiplicity handling, full model specifications, or detailed safety-event terms.