This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
SPARTAN was a randomized, double-blind, phase 3 trial evaluating three doses of lasmiditan against placebo in the acute treatment of migraine with or without aura. The registry reports 3,005 enrolled participants, four treatment arms, two primary binary endpoints, and logistic-regression analyses expressed as odds ratios.
| Feature | SPARTAN |
|---|---|
| Trial name | SPARTAN |
| Phase | Phase 3 |
| Condition | Migraine With or Without Aura |
| Design | Randomized, double-blind, sequential |
| Allocation | Randomized |
| Primary purpose | Treatment |
| Enrollment | 3,005 |
| Arms | 4 |
| Primary endpoints | Headache pain free at 2 hours; most bothersome symptom free at 2 hours |
| Primary endpoint type | Binary |
| Primary statistical method | Logistic regression |
| Primary effect measure | Odds ratio |
| Hypothesis type | Superiority |
| Status | Completed |
| Start | 2016-05 |
| Primary completion | 2017-06-30 |
| Lead sponsor | Eli Lilly and Company |
2. Clinical Question
The central question was whether acute treatment with lasmiditan at 50 mg, 100 mg, or 200 mg increased the odds of achieving headache pain freedom and being free of the most bothersome migraine-associated symptom at 2 hours post dose compared with placebo.
Population
Participants with migraine with or without aura enrolled in the phase 3 SPARTAN trial.
Intervention
Lasmiditan at 50 mg, 100 mg, or 200 mg.
Comparator
Placebo.
Primary question
For each lasmiditan dose, are the odds of headache pain freedom and MBS freedom at 2 hours greater than with placebo?
3. Trial Design
Lasmiditan 50 mg
- Lasmiditan 50 mg
- Compared with placebo in the registry's reported analyses
- Primary endpoints assessed at 2 hours post dose
Lasmiditan 100 mg
- Lasmiditan 100 mg
- Compared with placebo in the registry's reported analyses
- Primary endpoints assessed at 2 hours post dose
Lasmiditan 200 mg
- Lasmiditan 200 mg
- Compared with placebo in the registry's reported analyses
- Primary endpoints assessed at 2 hours post dose
Placebo
- Placebo
- Comparator for each reported lasmiditan dose comparison
- Primary endpoints assessed at 2 hours post dose
The registry identifies the overall design model as sequential and the masking as double. The reported primary comparisons are dose-specific lasmiditan-versus-placebo comparisons.
4. Analysis Population
The primary analyses were based on all randomized participants who received at least one dose of study drug and had evaluable endpoint data. The exact evaluable population differed by endpoint because each analysis required evaluable data for the specific outcome.
| Endpoint | Analysis population |
|---|---|
| Headache pain free | All randomized participants who received at least one dose of study drug and had evaluable headache pain free data. |
| Most bothersome symptom free | All randomized participants who received at least one dose of study drug and had evaluable MBS data. |
| Headache relief | All randomized participants who received at least one dose of study drug and had evaluable headache relief data. |
| Rescue medication | All randomized participants who received at least one dose of study drug and had evaluable use of rescue medication data. |
| Nausea free | All randomized participants who received at least one dose of study drug and had evaluable nausea free data. |
| Phonophobia free | All randomized participants who received at least one dose of study drug and had evaluable phonophobia free data. |
| Photophobia free | All randomized participants who received at least one dose of study drug and had evaluable photophobia free data. |
This distinction is important because the analysis population is not simply synonymous with the enrollment count. A participant contributes to a particular logistic-regression analysis only when the registry identifies that participant as having the required evaluable data for that endpoint.
5. Primary Endpoints
| Endpoint | Registry definition | Time frame | Type |
|---|---|---|---|
| Percentage of Participants Headache Pain Free at 2 Hours Post Dose | The percentage of participants defined as mild, moderate, or severe headache pain becoming none. | 2 hours post dose | Binary |
| Percentage of Participants Who Are Most Bothersome Symptom (MBS) Free | The percentage of participants defined as the associated symptom present and identified as MBS (nausea, photophobia, or phonophobia) prior to dosing being absent. | 2 hours post dose | Binary |
Both primary endpoints are binary outcomes. That structure naturally supports a comparison of the probability of success between each lasmiditan dose and placebo. The registry reports logistic regression with odds ratios and two-sided 95% confidence intervals.
6. Statistical Methodology
Logistic regression for binary endpoints
The registry reports logistic regression as the method for the primary analyses. Logistic regression models the probability of a binary outcome through the log odds of the event. For a treatment indicator, the exponentiated treatment coefficient is an odds ratio.
For a treatment indicator X, exp(β1) represents the estimated odds ratio comparing the treatment group with its comparator.
Odds ratios
An odds ratio compares the odds of the specified binary outcome between two groups. An odds ratio above 1 indicates higher odds of the outcome in the lasmiditan group relative to placebo; an odds ratio below 1 indicates lower odds.
For example, an OR of 1.7 means the estimated odds of the specified outcome are 1.7 times the odds under placebo. It does not mean that the probability of the outcome is 70% higher.
Two-sided 95% confidence intervals
Each posted primary analysis reports a two-sided 95% confidence interval for the odds ratio. The interval describes statistical uncertainty around the estimated treatment effect under the model and sampling framework. It does not describe the range of individual patient responses.
Superiority testing
The registry identifies the hypothesis type for the primary analyses as superiority. The dose-specific comparisons therefore ask whether the observed evidence is compatible with an odds ratio of 1 versus an effect in either direction, using the reported two-sided p-values.
Multiple dose comparisons
Each of the three lasmiditan doses was compared with placebo for each primary endpoint. This produces six primary endpoint analyses: three dose comparisons for headache pain freedom and three for MBS freedom. The registry reports the individual odds ratios, confidence intervals, and p-values, but the ClinicalTrials.gov record does not report a multiplicity-adjustment procedure for these six comparisons.
7. Primary Results: Headache Pain Free at 2 Hours
The first primary endpoint was the percentage of participants whose headache pain became none at 2 hours post dose. The registry reports logistic-regression analyses for each lasmiditan dose versus placebo.
Lasmiditan 50 mg vs placebo
95% CI: 1.1–1.9 · P = 0.003
Analysis population: all randomized participants who received at least one dose of study drug and had evaluable headache pain free data.
An odds ratio of 1.5 means that the estimated odds of being headache pain free at 2 hours were 1.5 times the corresponding odds with placebo. Expressed as a relative comparison of odds, this is 50% higher odds; it is not equivalent to saying that the probability of being pain free was 50% higher.
The 95% CI of 1.1–1.9 indicates uncertainty around the estimated odds ratio. Because the interval is entirely above 1, the reported two-sided p-value of 0.003 is consistent with evidence against an odds ratio of 1 under the stated model.
The p-value measures evidence against the null hypothesis; it does not measure the magnitude or clinical importance of the effect. The analysis is also conditional on the specified evaluable population and logistic-regression model.
Lasmiditan 100 mg vs placebo
95% CI: 1.3–2.2 · P < 0.001
Analysis population: all randomized participants who received at least one dose of study drug and had evaluable headache pain free data.
An odds ratio of 1.7 indicates estimated odds of headache pain freedom that were 1.7 times the placebo odds at 2 hours. The odds scale is important: converting an odds ratio into an absolute probability difference requires the underlying event probabilities, which are not reported in the posted analysis data here.
The 95% CI of 1.3–2.2 provides a relatively constrained range around the point estimate and remains above 1. The p-value of <0.001 indicates strong evidence against the null odds ratio of 1 under the reported analysis.
The confidence interval addresses precision around the estimated association; the p-value does not indicate that the effect is necessarily clinically large, nor does it quantify the probability that the treatment works.
Lasmiditan 200 mg vs placebo
95% CI: 1.8–3.1 · P < 0.001
Analysis population: all randomized participants who received at least one dose of study drug and had evaluable headache pain free data.
An odds ratio of 2.3 means the estimated odds of headache pain freedom at 2 hours were 2.3 times the placebo odds. This is a comparison of odds, not a 2.3-fold increase in the probability of becoming pain free.
The 95% CI of 1.8–3.1 remains above 1 and indicates that the estimate is compatible with a range of positive associations rather than a single exact treatment effect. The p-value of <0.001 indicates strong evidence against an odds ratio of 1 under the reported model.
The result should still be interpreted within the endpoint-specific evaluable population and the broader set of dose and endpoint comparisons, because multiple statistical tests are part of the trial's primary analysis structure.
| Headache pain freedom at 2 hours | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 1.5 | 1.1–1.9 | 0.003 |
| Lasmiditan 100 mg vs placebo | 1.7 | 1.3–2.2 | <0.001 |
| Lasmiditan 200 mg vs placebo | 2.3 | 1.8–3.1 | <0.001 |
The three reported odds ratios are all above 1, with the point estimate increasing from 1.5 at 50 mg to 2.3 at 200 mg. This describes the pattern of the observed estimates; it does not by itself establish a statistically significant dose-response trend because the ClinicalTrials.gov record does not report a formal trend test.
8. Primary Results: Most Bothersome Symptom Free
The second primary endpoint assessed whether the participant was free at 2 hours of the symptom identified before dosing as the most bothersome among nausea, photophobia, or phonophobia.
Lasmiditan 50 mg vs placebo
95% CI: 1.1–1.8 · P = 0.009
Analysis population: all randomized participants who received at least one dose of study drug and had evaluable MBS data.
An odds ratio of 1.4 indicates estimated odds of being MBS free at 2 hours that were 1.4 times the placebo odds. The 95% CI of 1.1–1.8 remains above 1, while the p-value of 0.009 provides evidence against an odds ratio of 1 under the reported two-sided logistic-regression analysis.
The result does not tell us the absolute percentage of participants who became MBS free, because an odds ratio alone does not identify the underlying probabilities. The estimate also depends on the endpoint-specific evaluable population.
Lasmiditan 100 mg vs placebo
95% CI: 1.2–2.0 · P < 0.001
Analysis population: all randomized participants who received at least one dose of study drug and had evaluable MBS data.
An odds ratio of 1.6 corresponds to estimated odds of MBS freedom that were 1.6 times the placebo odds at 2 hours. The 95% CI of 1.2–2.0 indicates uncertainty around that estimate while remaining above 1.
The p-value of <0.001 provides strong evidence against an odds ratio of 1 under the reported model. It should not be interpreted as a measure of the size, durability, or clinical importance of the treatment effect.
Lasmiditan 200 mg vs placebo
95% CI: 1.4–2.4 · P < 0.001
Analysis population: all randomized participants who received at least one dose of study drug and had evaluable MBS data.
An odds ratio of 1.9 means the estimated odds of being free of the most bothersome symptom at 2 hours were 1.9 times the placebo odds. The 95% CI of 1.4–2.4 indicates uncertainty around the point estimate while remaining above the null value of 1.
The p-value of <0.001 indicates strong evidence against an odds ratio of 1 under the stated analysis. As with the other primary results, this is not a direct estimate of the absolute probability difference and should be interpreted alongside the multiple primary comparisons.
| MBS freedom at 2 hours | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 1.4 | 1.1–1.8 | 0.009 |
| Lasmiditan 100 mg vs placebo | 1.6 | 1.2–2.0 | <0.001 |
| Lasmiditan 200 mg vs placebo | 1.9 | 1.4–2.4 | <0.001 |
The pattern of point estimates is again upward across the three lasmiditan doses, from 1.4 to 1.9. As with headache pain freedom, that descriptive pattern should not be confused with a formal statistical test for dose-response because no such test is reported here.
9. Secondary and Other Pre-Specified Results
The registry also posts logistic-regression analyses for headache relief, use of rescue medication, nausea freedom, phonophobia freedom, and photophobia freedom. These are binary outcomes analyzed with odds ratios in the same general framework.
Headache Relief at 2 Hours
| Comparison | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 1.7 | 1.3–2.2 | <0.001 |
| Lasmiditan 100 mg vs placebo | 2.3 | 1.7–2.9 | <0.001 |
| Lasmiditan 200 mg vs placebo | 2.4 | 1.8–3.1 | <0.001 |
For this endpoint, all three reported odds ratios exceed 1. The 95% confidence intervals remain above 1, and each reported p-value is below 0.001. These results describe the odds of the registry-defined headache relief outcome at 2 hours, rather than headache pain freedom specifically.
Use of Rescue Medication at 2 Hours
| Comparison | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 0.7 | 0.5–0.9 | 0.002 |
| Lasmiditan 100 mg vs placebo | 0.5 | 0.4–0.7 | <0.001 |
| Lasmiditan 200 mg vs placebo | 0.3 | 0.3–0.4 | <0.001 |
Here the direction is reversed because the event is use of rescue medication rather than freedom from a symptom. An odds ratio below 1 indicates lower estimated odds of rescue-medication use in the lasmiditan group than in the placebo group. The 200 mg comparison has an estimated odds ratio of 0.3 with a 95% CI of 0.3–0.4.
Use of Rescue Medication From 2 to 24 Hours
| Comparison | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 1 | 0.7–1.5 | 0.917 |
| Lasmiditan 100 mg vs placebo | 0.7 | 0.5–1.1 | 0.129 |
| Lasmiditan 200 mg vs placebo | 0.8 | 0.6–1.3 | 0.456 |
These estimates are different from the 2-hour rescue-medication results because they address the period from 2 hours post dose up to 24 hours. The confidence intervals for all three comparisons include 1, and the reported p-values are 0.917, 0.129, and 0.456.
Nausea Free at 2 Hours
| Comparison | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 0.9 | 0.7–1.2 | 0.522 |
| Lasmiditan 100 mg vs placebo | 1.1 | 0.8–1.4 | 0.622 |
| Lasmiditan 200 mg vs placebo | 1 | 0.8–1.3 | 0.992 |
The registry-reported odds ratios for nausea freedom are close to 1, and all three 95% confidence intervals include 1. The corresponding p-values are 0.522, 0.622, and 0.992.
Phonophobia Free at 2 Hours
| Comparison | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 1.4 | 1.1–1.7 | 0.008 |
| Lasmiditan 100 mg vs placebo | 1.6 | 1.3–2.0 | <0.001 |
| Lasmiditan 200 mg vs placebo | 1.6 | 1.3–2.1 | <0.001 |
Photophobia Free at 2 Hours
| Comparison | Odds ratio | 95% CI | P-value |
|---|---|---|---|
| Lasmiditan 50 mg vs placebo | 1.4 | 1.1–1.7 | 0.007 |
| Lasmiditan 100 mg vs placebo | 1.7 | 1.3–2.1 | <0.001 |
| Lasmiditan 200 mg vs placebo | 1.8 | 1.4–2.3 | <0.001 |
The phonophobia and photophobia results show odds ratios above 1 for all three dose comparisons. These outcomes are distinct from the MBS endpoint: MBS is defined by the symptom identified as most bothersome before dosing, whereas these analyses separately evaluate phonophobia and photophobia freedom.
10. Safety Results
The registry reports serious adverse events by treatment sequence/arm as affected participants divided by participants at risk. Because the design is identified as sequential, the reported arm labels distinguish lasmiditan-to-lasmiditan, lasmiditan-to-placebo, and placebo-to-placebo sequences.
| Reported arm | Serious adverse events | Participants at risk | Affected / at risk |
|---|---|---|---|
| Lasmiditan 50 mg / Lasmiditan 50 mg | 1 | 429 | 1/429 |
| Lasmiditan 50 mg / Placebo | 0 | 225 | 0/225 |
| Lasmiditan 100 mg / Lasmiditan 100 mg | 1 | 423 | 1/423 |
| Lasmiditan 100 mg / Placebo | 1 | 212 | 1/212 |
| Lasmiditan 200 mg / Lasmiditan 200 mg | 3 | 435 | 3/435 |
| Lasmiditan 200 mg / Placebo | 0 | 217 | 0/217 |
| Placebo / Placebo | 2 | 646 | 2/646 |
The serious-adverse-event data should be interpreted as counts within the reported sequential arms rather than as a single simple lasmiditan-versus-placebo comparison. The denominators differ substantially across the reported arms, and the registry presents the sequence-specific affected/at-risk values.
11. Statistical Methods Explained
Why was logistic regression used?
The primary endpoints are binary: a participant is either headache pain free at 2 hours or not, and either MBS free at 2 hours or not. Logistic regression is designed for this type of outcome because it models the probability of the event through its odds and provides a treatment-effect estimate on the odds-ratio scale.
What does an odds ratio of 2.3 mean?
An OR of 2.3 means that the estimated odds of the outcome are 2.3 times the odds in the comparator group. It does not mean that 2.3 times as many participants achieved the outcome. Odds and probabilities are related but are not interchangeable, particularly when the outcome is common.
Why is an odds ratio below 1 appropriate for rescue medication?
The direction depends on what is defined as the event. For headache pain freedom, an OR above 1 favors the treatment because the desired outcome is more frequent in the treatment group. For rescue-medication use, an OR below 1 corresponds to lower odds of the event being analyzed.
What does the confidence interval tell us?
The 95% confidence interval describes uncertainty around the estimated odds ratio. A narrow interval generally indicates greater statistical precision than a wide interval, although precision is not the same thing as clinical importance. When the interval includes 1, the data are compatible with both lower and higher odds under the stated model.
Why does the p-value not measure effect size?
The p-value addresses the compatibility of the observed data with a specified null hypothesis. It does not tell us how large the treatment effect is. The odds ratio describes the estimated magnitude, while the confidence interval describes uncertainty around that magnitude.
Why does multiplicity matter here?
The primary analysis contains six dose-specific comparisons across two primary endpoints. When multiple hypotheses are tested, the chance of observing at least one small p-value can increase if no multiplicity adjustment is used. The registry supplies the individual p-values and identifies the analyses as superiority tests, but the ClinicalTrials.gov record does not report a multiplicity-adjustment procedure for these six primary comparisons.
Does an increasing OR across doses prove a dose-response relationship?
No. The observed point estimates increase across the three doses for both primary endpoints, but a dose-response claim requires an appropriate statistical comparison or trend model. The posted analyses identify pairwise dose-versus-placebo logistic regressions rather than a formal dose-response test.
12. Understanding the Primary Odds Ratios
| Primary endpoint | 50 mg OR | 100 mg OR | 200 mg OR |
|---|---|---|---|
| Headache pain free at 2 hours | 1.5 | 1.7 | 2.3 |
| MBS free at 2 hours | 1.4 | 1.6 | 1.9 |
The two primary endpoints show the same broad direction in the posted estimates: each lasmiditan dose has an odds ratio above 1 relative to placebo. The magnitude differs by endpoint and dose. The headache pain-free estimates are 1.5, 1.7, and 2.3, while the MBS-free estimates are 1.4, 1.6, and 1.9.
The graphic is a visual display of the reported point estimates, not a substitute for the confidence intervals. Statistical interpretation should rely on the full estimate, interval, p-value, endpoint definition, and analysis population.
13. Results Across the Posted Binary Endpoints
| Outcome | 50 mg OR (95% CI) | 100 mg OR (95% CI) | 200 mg OR (95% CI) |
|---|---|---|---|
| Headache pain free, 2 hours | 1.5 (1.1–1.9) | 1.7 (1.3–2.2) | 2.3 (1.8–3.1) |
| MBS free, 2 hours | 1.4 (1.1–1.8) | 1.6 (1.2–2.0) | 1.9 (1.4–2.4) |
| Headache relief, 2 hours | 1.7 (1.3–2.2) | 2.3 (1.7–2.9) | 2.4 (1.8–3.1) |
| Rescue medication, 2 hours | 0.7 (0.5–0.9) | 0.5 (0.4–0.7) | 0.3 (0.3–0.4) |
| Rescue medication, 2–24 hours | 1.0 (0.7–1.5) | 0.7 (0.5–1.1) | 0.8 (0.6–1.3) |
| Nausea free, 2 hours | 0.9 (0.7–1.2) | 1.1 (0.8–1.4) | 1.0 (0.8–1.3) |
| Phonophobia free, 2 hours | 1.4 (1.1–1.7) | 1.6 (1.3–2.0) | 1.6 (1.3–2.1) |
| Photophobia free, 2 hours | 1.4 (1.1–1.7) | 1.7 (1.3–2.1) | 1.8 (1.4–2.3) |
This table illustrates why endpoint definition matters. An odds ratio of 0.3 for rescue-medication use and an odds ratio of 2.3 for headache pain freedom both represent treatment-versus-placebo effects, but they describe different events and therefore have different favorable directions.
14. P-Values and Confidence Intervals Together
The primary analyses illustrate why p-values and confidence intervals should be considered together. The p-value provides a hypothesis-testing summary, while the confidence interval places the estimated effect on a scale that conveys both magnitude and uncertainty.
Headache pain freedom
The 200 mg comparison has OR 2.3 with 95% CI 1.8–3.1 and P < 0.001. The interval communicates substantially more information about the possible size of the odds ratio than the p-value alone.
MBS freedom
The 50 mg comparison has OR 1.4 with 95% CI 1.1–1.8 and P = 0.009. The estimate is above 1, but the interval shows meaningful uncertainty around its magnitude.
Neither confidence interval should be interpreted as a range in which an individual patient's treatment effect must fall. These are intervals for the population-level parameter estimated by the statistical model.
15. What an Odds Ratio Does — and Does Not — Mean
An odds ratio of 1.7 means that the estimated odds of the specified binary outcome are 1.7 times the odds in the comparator group.
It does not mean that 70% of participants benefited, that the probability increased by 70 percentage points, or that every participant experienced a 70% improvement.
A 95% CI such as 1.3–2.2 describes uncertainty around the estimated odds ratio. The width of the interval provides information about statistical precision, while the location of the interval relative to 1 informs interpretation of the null odds-ratio hypothesis.
Odds ratios are relative measures. Without the underlying event rates, an odds ratio cannot be translated into an absolute probability increase. Two studies with the same odds ratio can have very different absolute treatment effects when baseline event probabilities differ.
16. Limitations and Interpretation Issues
- Multiple comparisons: the primary analysis includes three dose comparisons for each of two primary endpoints. The ClinicalTrials.gov record reports individual p-values but does not report a multiplicity-adjustment procedure for these six comparisons.
- Odds ratio interpretation: odds ratios should not be read as relative risks or percentage-point differences in response probability.
- Endpoint-specific populations: each analysis requires evaluable data for the particular outcome, so the analysis population can differ across endpoints.
- Sequential design: the registry identifies the design model as sequential, and the safety data are reported using sequence-specific arms. This structure should be considered when interpreting arm-level safety counts.
- No dose-response test reported: increasing point estimates across doses describe the observed pattern but do not constitute a formal statistical test for trend.
- Small serious-event counts: serious adverse events are uncommon in the reported arms, so the observed counts provide limited precision for comparing safety event frequencies.
- Binary endpoint timing: the primary endpoints are assessed at a specific time point, 2 hours post dose. They therefore answer an acute treatment question at that time rather than describing all subsequent migraine outcomes.
- Model dependence: the odds ratios are model-based estimates from logistic regression and therefore depend on the statistical model and analysis population.
17. Why This Trial Matters Statistically
SPARTAN is a useful teaching example because it shows how a randomized, double-blind phase 3 trial can use logistic regression to analyze multiple binary clinical outcomes across several treatment doses.
| Concept | How it appears in SPARTAN |
|---|---|
| Randomization | The trial uses randomized allocation across four arms. |
| Blinding | The registry identifies the trial as double-blind. |
| Binary endpoints | Primary outcomes classify participants as meeting or not meeting the specified acute treatment outcome. |
| Logistic regression | The posted primary and other pre-specified analyses use logistic regression. |
| Odds ratio | Treatment effects are reported as odds ratios. |
| Confidence intervals | Each posted statistical analysis with an estimate reports a two-sided 95% CI. |
| P-values | Each posted analysis includes a p-value, allowing hypothesis-test interpretation alongside the effect estimate. |
| Multiple comparisons | Three lasmiditan doses are compared with placebo across two primary endpoints. |
| Sequential design | The registry identifies the design model as sequential and reports sequence-specific safety arms. |
| Safety analysis | Serious adverse events are reported as affected participants divided by participants at risk for each arm. |
18. Trial Timeline
Trial begins
The registry lists the study start as 2016-05.
Randomized double-blind evaluation
The trial evaluates lasmiditan 50 mg, 100 mg, and 200 mg against placebo in acute migraine treatment.
Primary completion
The registry lists the primary completion date as 2017-06-30.
Results posted
The registry reports posted outcome measures and statistical analyses, including six primary-endpoint analyses with estimates and confidence intervals.
19. Interpreting the Six Primary Comparisons
The primary statistical structure is straightforward but important: there are two binary primary endpoints and three active-dose-versus-placebo comparisons for each endpoint.
| Endpoint | 50 mg | 100 mg | 200 mg |
|---|---|---|---|
| Headache pain free | OR 1.5 95% CI 1.1–1.9 P = 0.003 | OR 1.7 95% CI 1.3–2.2 P < 0.001 | OR 2.3 95% CI 1.8–3.1 P < 0.001 |
| MBS free | OR 1.4 95% CI 1.1–1.8 P = 0.009 | OR 1.6 95% CI 1.2–2.0 P < 0.001 | OR 1.9 95% CI 1.4–2.4 P < 0.001 |
All six confidence intervals lie above the null odds ratio of 1. The point estimates also show higher odds ratios at each successive dose for both primary endpoints. Those are descriptive properties of the posted analyses; they should be kept separate from a formal claim that the dose-response relationship itself was statistically tested.
20. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The six primary logistic-regression analyses report odds ratios above 1, with two-sided 95% confidence intervals above 1 and reported p-values of 0.009 or smaller. The estimates quantify the relative odds of the specified binary outcomes at 2 hours.
Clinical interpretation
The endpoints concern acute migraine outcomes at a specific 2-hour time point. Interpreting their practical importance requires considering the endpoint definitions, absolute response probabilities, the broader efficacy profile, and safety rather than relying on odds ratios alone.
21. Statistical Takeaways
Several principles from SPARTAN generalize to the analysis of randomized clinical trials with binary outcomes:
1. Match the model to the endpoint
Binary endpoints naturally support logistic-regression methods, with treatment effects commonly expressed as odds ratios.
2. Define the event carefully
The direction of the odds ratio depends on whether the event is desirable, such as pain freedom, or undesirable, such as rescue-medication use.
3. Read OR and CI together
The odds ratio describes the estimated magnitude, while the confidence interval communicates uncertainty around that estimate.
4. Do not confuse OR with probability
An odds ratio of 2 does not generally mean that the probability of success is twice as high. The baseline probability is needed for that conversion.
5. Account for multiplicity
Six primary comparisons create a broader hypothesis-testing framework than a single treatment comparison.
6. Keep efficacy and safety distinct
Binary efficacy endpoints and serious-adverse-event counts answer different questions and should not be collapsed into one statistical conclusion.
22. Related Tutorials
Learn more about the methods used in this trial:
23. Related Calculators
24. Sources
- ClinicalTrials.gov: NCT02605174 — SPARTAN.
- PubMed: 36125279.
- PubMed: 35779194.
- PubMed: 35471625.
- PubMed: 35350937.
- PubMed: 34644189.
Continue through the Clinical Biostats knowledge graph
Connect the statistical methods used in SPARTAN with deeper tutorials and practical statistical calculators.
25. Record Summary
SPARTAN provides a clear example of randomized clinical-trial analysis for binary outcomes. Three lasmiditan doses were compared with placebo across two primary endpoints at 2 hours post dose, with logistic regression used to estimate odds ratios and two-sided 95% confidence intervals. The primary headache pain-free analyses reported odds ratios of 1.5, 1.7, and 2.3 for 50 mg, 100 mg, and 200 mg, respectively, while the MBS-free analyses reported odds ratios of 1.4, 1.6, and 1.9. The registry also reports analyses of headache relief, rescue-medication use, nausea freedom, phonophobia freedom, and photophobia freedom, together with sequence-specific serious-adverse-event counts.