A Study of Lasmiditan (LY573144) Over Four Migraine Attacks: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Lasmiditan (LY573144) Over Four Migraine Attacks (NCT03670810), a randomized phase 3 study. The primary outcome is Percentage of Participants That Are Pain Free 2 Hours Postdose During the First Attack.

2. Clinical Question

The trial focused on acute treatment of migraine attacks, including attacks with or without aura. Its primary endpoint was narrower than the overall four-attack study framework: it assessed pain freedom after the first attack. Other posted outcomes considered pain relief, sustained pain freedom, and outcomes across repeated attacks. Those additional measures help describe different aspects of the observed response, but they do not change the definition of the primary question.

The registry record describes the primary analysis population as randomized participants who used at least 1 dose for an Intent-to-Treat evaluable attack. An attack qualified when pain was at least mild and a postdose pain-severity assessment was available at or before 2 hours. This definition identifies which participants contributed to the primary endpoint; it does not mean that every randomized participant necessarily contributed an endpoint observation.

3. Trial Design

This was a randomized, double-masked, parallel phase 3 trial of lasmiditan for migraine. The registry record lists enrollment as 1633 and status as completed. The study began on 2019-06-24, with primary completion on 2020-06-12. Eli Lilly and Company is listed as sponsor.

The main posted efficacy groups were 100 mg lasmiditan, 200 mg lasmiditan, and placebo. The primary comparison was made separately for each lasmiditan group against placebo. The intervention descriptions state that study tablets were administered orally within 4 hours of onset of an attack, with treatment available for up to 4 attacks. Matching placebo tablets were included in the descriptions to maintain masking. The registry also lists maximum extended enrollment groups and an open-label extension; their presence should not be confused with the main randomized efficacy comparisons.

The trial design therefore combines a parallel randomized comparison with an outcome framework that includes repeated attacks. The primary endpoint is specifically tied to the first attack, while several secondary outcomes concern responses over subsequent attacks or sustained outcomes. This distinction matters: results for a first attack and results requiring responses across multiple attacks address related but different questions.

The registry record does not report the original sample-size justification, the detailed randomization procedure, or the full protocol and statistical analysis plan. The available descriptions do not establish how masking success was assessed or how every treatment sequence was assigned. Those details should not be inferred from the basic design labels or the posted group results.

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: An ITT evaluable attack is defined as a treated attack of least mild pain severity with any postdose pain severity assessments at or before 2 hours postdose.

Only the first 6 of 9 groups are shown.

Main Study

Reasons for not completing this period:

Open Label Extension

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants who used at least 1 dose of study drug for an Intent-to-Treat (ITT) evaluable attack, defined as a treated attack of at least mild pain severity with any postdose pain severity assessments at or before 2 hours postdose. As pre-specified in the analysis plan, outcome measures will not be reported for the Maximum Extended Enrollment (MEE) arms/groups but only for the main global study arms/groups.

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.

Population: All randomized participants who were treated with at least 1 dose of study drug, regardless of whether or not they undergo any study assessments.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants That Are Pain Free 2 Hours Postdose During the First Attack, measured over 2 Hours Postdose.

The primary endpoint is the percentage of participants who were pain free 2 hours after dosing during the first attack. The registry definition describes pain-free status as a headache that was mild, moderate, or severe becoming none at that assessment. The endpoint is binary at the participant level: each eligible participant either met the definition at the specified assessment or did not.

This measure addresses an acute, patient-relevant symptom state at a defined time after treatment. A fixed assessment time makes the outcome interpretable across randomized groups and permits comparison of the observed percentages. It is different from pain relief, which can include improvement without complete absence of pain, and from sustained pain freedom, which requires the state to persist over a longer period. Those distinctions explain why the registry lists these as separate outcomes.

The analysis population was not simply everyone enrolled. The registry specifies randomized participants who used at least 1 dose for an Intent-to-Treat evaluable attack: an attack with at least mild pain and a postdose pain assessment at or before 2 hours. The record does not report all details of how unavailable assessments were handled. Thus the endpoint describes the posted analysis population and definition, not every possible outcome for every randomized participant.

7. Results Posted in the Registry

• 100 mg Lasmiditan: 25.8% (108/419) versus 8.4% (37/443) on control; risk difference 17.4 percentage points (95% CI 12.5 to 22.3), risk ratio 3.09 (2.18–4.37); unadjusted two-proportion z-test p = <0.001.
• 200 mg Lasmiditan: 29.3% (127/434) versus 8.4% (37/443) on control; risk difference 20.9 percentage points (95% CI 15.9 to 25.9), risk ratio 3.50 (2.49–4.93); 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.

Percentage of Participants With Pain Relief at 2 Hours Post Dose During the First Attack

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • 100 mg Lasmiditan: 65.4% (274/419) versus 41.3% (183/443) on control; risk difference 24.1 percentage points (95% CI 17.6 to 30.5), risk ratio 1.58 (1.39–1.80); unadjusted two-proportion z-test p = <0.001.
  • 200 mg Lasmiditan: 65.2% (283/434) versus 41.3% (183/443) on control; risk difference 23.9 percentage points (95% CI 17.5 to 30.3), risk ratio 1.58 (1.39–1.80); 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 100 mg Lasmiditan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for 200 mg Lasmiditan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With Pain Relief at 2 Hours Postdose in at Least 2 Out of 3 Attacks

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • 100 mg Lasmiditan: 62.3% (207/332) versus 36.9% (118/320) on control; risk difference 25.5 percentage points (95% CI 18.1 to 32.9), risk ratio 1.69 (1.43–2.00); unadjusted two-proportion z-test p = <0.001.
  • 200 mg Lasmiditan: 66.7% (222/333) versus 36.9% (118/320) on control; risk difference 29.8 percentage points (95% CI 22.5 to 37.1), risk ratio 1.81 (1.54–2.13); 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 100 mg Lasmiditan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for 200 mg Lasmiditan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With 24-Hour Sustained Pain Freedom During the First Attack

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • 100 mg Lasmiditan: 13.6% (57/419) versus 4.3% (19/443) on control; risk difference 9.3 percentage points (95% CI 5.5 to 13.1), risk ratio 3.17 (1.92–5.24); unadjusted two-proportion z-test p = <0.001.
  • 200 mg Lasmiditan: 17.3% (75/434) versus 4.3% (19/443) on control; risk difference 13.0 percentage points (95% CI 9.0 to 17.0), risk ratio 4.03 (2.48–6.55); 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 100 mg Lasmiditan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for 200 mg Lasmiditan than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Across the posted secondary outcomes, the overall pattern is directionally consistent with the primary result: the reported lasmiditan-group percentages are generally higher than the placebo percentages for pain relief, repeated-attack pain relief, and sustained pain freedom. These outcomes extend the description beyond pain freedom at 2 hours during the first attack. They consider improvement without requiring complete pain freedom, persistence over a longer interval, or response across multiple attacks.

This broad pattern provides supporting context for the primary comparison, but each endpoint measures a different aspect of response. A secondary result cannot be treated as a replication of the primary endpoint simply because it points in a similar direction. The endpoint definition, time frame, and analysis population matter, and the numbers of participants contributing to the repeated-attack outcomes differ from those for the first-attack outcomes.

The registry lists 20 posted secondary outcomes, alongside 22 outcome measures and 67 analyses overall. Testing many outcomes increases the chance of observing apparently strong results even when no underlying effect exists for a particular outcome. General statistical principle: multiplicity adjustment or a prespecified hierarchy can help control this issue, but the registry record does not report whether such a strategy was used for these secondary analyses.

For that reason, unadjusted secondary p-values carry less inferential weight than a clearly designated primary analysis, particularly when many endpoints and comparisons are considered. The secondary results are useful for describing the range of outcomes recorded and whether their broad direction aligns with the primary finding. They should not be treated as independent proof, and their pattern does not resolve missing details about multiplicity or analysis methods.

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

• Percentage of Participants With Pain Relief at 2 Hours Post Dose During the First Attack
• Percentage of Participants With Pain Relief at 2 Hours Postdose in at Least 2 Out of 3 Attacks
• Percentage of Participants With 24-Hour Sustained Pain Freedom During the First Attack
• Percentage of Participants With 48-Hour Sustained Pain Freedom During First Attack
• Percentage of Participants That Are Pain Free 2 Hours Postdose During the First Attack in Triptan Insufficient Responders.
• Percentage of Participants With no Disability as Measured by the Disability Item, at 2 Hours Postdose During the First Attack
• Percentage of Participants That Are Pain Free at 2 Hours Postdose in at Least 2 Out of 3 Attacks in Triptan Insufficient Responders
• Percentage of Participants Free of Most Bothersome Symptom (MBS) Associated With Migraine at 2 Hours Postdose During the First Attack

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary: participants are classified according to whether they met the pain-free definition at the 2-hour assessment. For a binary outcome, a percentage is the proportion of analyzed participants with the event of interest. Here, the event is pain freedom, not pain remaining. The posted percentages are descriptive summaries; a comparison between groups adds an estimate of how those proportions differ or relate.

What does a risk difference express?

A risk difference compares the event proportions by subtraction, on the original probability scale. It answers an absolute question: how far apart are the observed group proportions? A risk difference is often easier to relate to event frequency than a relative measure, but it depends on the underlying event rates. In these results, the recalculated risk differences compare each lasmiditan group with placebo. The estimates and their confidence intervals are unadjusted recalculations from the posted summary statistics, rather than the sponsor's logistic regression estimates.

What does an odds ratio express?

Odds are the probability of an event divided by the probability of its non-occurrence. An odds ratio compares those odds between groups. It is not the same as a risk ratio, and it should not be read as a direct multiplication of the probability or percentage. When an outcome is not rare, an odds ratio can appear farther from the null value than a risk ratio for the same underlying data. The sponsor reported odds ratios from logistic regression. The registry record identifies the method as logistic regression but does not report the model's covariates or all other modeling details.

Why can analysis methods give slightly different estimates?

A direct recalculation from group summaries uses the posted event counts and denominators without adjustment. A regression analysis can use a model and analysis details that are not reproduced in the basic summaries. Differences can therefore arise from modeling choices, analysis populations, or rounding. The registry record does not flag a discrepancy in this case, and the sponsor's reported odds ratios are close to the recalculated odds ratios. The sponsor analysis is the principal reported analysis; the recalculation is a supplementary check, not a replacement.

What does a p-value tell a reader?

A p-value describes how compatible the observed data are with a specified null hypothesis under the analysis assumptions. A small p-value is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The sponsor reports p < 0.001 for both primary comparisons. The unadjusted recalculations also report p < 0.001. These values do not address multiplicity across the full set of outcomes unless an adjustment or testing strategy is applied and documented.

How should repeated attacks be understood?

The study allowed treatment for up to 4 attacks, while the primary endpoint is based on the first attack. Some secondary outcomes evaluate whether a response occurred in at least 2 out of 3 attacks. Such a repeated-attack outcome summarizes consistency across attacks differently from a first-attack outcome. It can also involve a different analysis population, as reflected by different denominators in the posted results. General statistical principle: repeated observations from the same participant may be related, so an analysis of multiple attacks may need to account for within-participant dependence. The exact modeling approach for each repeated-attack endpoint is not reported in the registry record.

The registry lists 22 outcome measures and 67 analyses. Multiple endpoints and comparisons create multiple opportunities for an apparently notable result to occur by chance. General statistical principle: multiplicity procedures can control the overall error rate when testing a family of hypotheses, but they must be specified and interpreted in context. The registry record does not report whether or how the secondary outcomes were adjusted for multiplicity, so unadjusted secondary p-values should be treated as less definitive than the primary reported analysis.

Finally, statistical significance and precision answer different questions. A p-value summarizes evidence relative to a null hypothesis; a confidence interval displays a range of estimates compatible with the data and model. Neither alone answers whether a difference is important to patients. The endpoint definition, outcome timing, population, effect scale, and uncertainty all inform interpretation.

11. Confidence Intervals and Precision

The sponsor reports a 95% confidence interval for each primary odds ratio. For 100 mg lasmiditan versus placebo, the odds ratio is 3.83, with an interval from 2.56 to 5.73. For 200 mg lasmiditan versus placebo, the odds ratio is 4.56, with an interval from 3.07 to 6.77. These intervals describe uncertainty around the estimated odds ratios under the reported logistic regression analyses.

In general, a confidence interval is a range generated by a statistical procedure that, under its assumptions, has a specified long-run coverage rate. It is not a range containing a known proportion of individual participant outcomes, nor is it a guarantee that the true value lies inside a particular calculated interval. A narrower interval generally indicates greater precision on that effect scale; a wider interval indicates more uncertainty.

The recalculated risk-difference intervals use an absolute scale and should not be compared directly with the sponsor's odds-ratio intervals. The lower and upper endpoints of an interval depend on the measure and its calculation. For the secondary comparisons, the reported intervals also vary in width, reflecting differing event frequencies and analysis denominators. The registry record does not give confidence intervals for the arm percentages themselves.

The primary intervals do not include the null odds ratio of 1, and the sponsor reports p < 0.001 for each comparison. That is consistent with the reported superiority analyses, but it does not make the estimates exact. The intervals leave uncertainty about the size of the association, and they do not establish an individual-level outcome or clinical importance. Precision should therefore be read together with the observed percentages, the endpoint definition, and the limitations of the analysis population and model reporting.

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 estimates are planning illustrations based on the observed effects, not the trial's original design assumptions. The registry record does not report the original sample-size justification. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. These figures are participants per arm for the binary outcomes, rather than total events.

The 80% power column gives the estimated participants per arm under those assumptions for an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column gives the corresponding planning number for a 90% chance. For example, planning for 100% of the observed effect in the primary 100 mg comparison gives 72 participants per arm for 80% power and 96 participants per arm for 90% power. For the 200 mg comparison, the corresponding figures are 54 and 72 participants per arm.

Rows that plan for 75% or 50% of the observed effect represent smaller target differences. Smaller effects are harder to distinguish from random variation, so they require more participants under the stated assumptions. The values are illustrative rather than a reconstruction of the trial's recruitment plan, and the planning assumptions may not match the sponsor's actual design decisions.

This is not observed power. The estimates do not measure the probability that the completed trial detected an effect, and they do not change the interpretation of its results. The sample-size curve should be read as a sensitivity display: as the assumed effect becomes smaller, the estimated participants per arm increase. The figures apply to the specified comparison and assumptions, not automatically to every secondary endpoint or to a study with a different allocation or analysis plan.

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 = 0.084; treatment proportion p₁ = 0.258; average p̄ = 0.171.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5320 = 1.0427.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5176 = 0.4356.
4. n = (1.0427 + 0.4356)² / (0.1742)² = 72.0, rounded up to 72 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 <- 0.0835214446952596                          # control proportion
p1 <- 0.257756563245823                          # 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(3670810)
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 = 72 (80% power), 96 (90% power)
 75% of observed effect: n per arm = 116 (80% power), 154 (90% power)
 50% of observed effect: n per arm = 229 (80% power), 306 (90% power)
power.prop.test: 72 
Simulated power at n = 72 per arm: 0.820 (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 sponsor's reported primary analyses use logistic regression and express the comparison as an odds ratio. The reported estimate for 100 mg lasmiditan versus placebo is 3.83, with a 95% confidence interval from 2.56 to 5.73; for 200 mg versus placebo it is 4.56, with a 95% confidence interval from 3.07 to 6.77. Both reported p-values are p < 0.001.

The recalculated comparisons use posted event counts and denominators to obtain unadjusted estimates. Their odds ratios are 3.811 for 100 mg and 4.539 for 200 mg, with confidence intervals of 2.55 to 5.693 and 3.058 to 6.738, respectively. The recalculations are close to the sponsor estimates. The registry record does not flag a discrepancy; accordingly, the sponsor's own analysis remains the main inferential result, and the unadjusted calculations serve as a check from the summary data.

The two approaches are not identical: the sponsor's method is logistic regression, while the recalculation is based directly on the posted group summaries. The registry record does not report enough modeling detail to reproduce every aspect of the sponsor analysis. Estimates can also differ slightly because of rounding. The recalculated risk differences answer an absolute-scale question and should not be confused with the sponsor's odds ratios.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary endpoint and lists multiple secondary outcomes. It also records 22 outcome measures and 67 analyses. This breadth creates a multiplicity issue: with many statistical tests, the chance of at least one small p-value can increase even when the tests do not represent true effects. General statistical principle: a planned testing sequence or adjustment may address this, but the registry record does not report whether a multiplicity procedure was used for the secondary analyses.

The primary endpoint is defined for randomized participants who used at least 1 dose for an Intent-to-Treat evaluable attack. The attack definition requires at least mild pain and a postdose pain assessment at or before 2 hours. This provides a stated basis for the primary analysis population, but does not fully describe participant flow or the disposition of every missing assessment. Detailed missing-data methods, including any imputation or sensitivity analyses, are not reported in the registry record.

The record identifies logistic regression for the sponsor's primary comparisons, but does not report the model covariates, handling of missing covariates, or full model specification. It also does not report the complete statistical analysis plan, the original sample-size justification, or interim analyses. Those omissions limit independent assessment of whether every analysis followed a prespecified plan.

Different outcomes have different analysis denominators, particularly outcomes defined across multiple attacks. The registry's posted results should therefore be interpreted outcome by outcome rather than assuming one common population for all analyses. The record does not give enough detail to determine how all exclusions, unavailable assessments, or repeated measurements were handled. These are reporting limitations, not evidence that a particular method was or was not used.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: migraine (1.8%), abortion spontaneous (0.5%), appendicitis (0.4%), intermenstrual bleeding (0.2%), uterine leiomyoma (0.2%).

The registry reports serious adverse events by safety group, with both the number at risk and the number affected. In the main groups, serious adverse events affected 7 of 485 participants in the 100 mg group, 8 of 486 in the 200 mg group, and 7 of 500 in the main control group. The maximum extended enrollment groups list 1 of 55 for 100 mg, 0 of 53 for 200 mg, and 0 of 54 for control. The open-label extension lists 19 of 477.

These are group-specific counts, not pooled estimates. The registry also lists serious-event terms, including Migraine at 1.8%, Abortion spontaneous at 0.5%, and Appendicitis at 0.4%; the listed terms are not an assessment of causation. The record lists 0 deaths in the randomized and maximum extended enrollment groups and 1 in the open-label extension. These counts alone do not establish whether an event was caused by study treatment, and they should not be used to compare safety without considering study exposure, group definitions, and reporting methods.

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

The posted primary results show that pain freedom at 2 hours during the first attack was observed more often in each lasmiditan group than in the placebo group. The sponsor's logistic regression analyses also report odds ratios above 1, with confidence intervals that do not include 1 and p-values of p < 0.001. Together, these describe the direction and statistical uncertainty of the recorded group comparison.

The result does not mean that every participant in a lasmiditan group became pain free, or that an individual can be assured of the same outcome. The reported percentages are group summaries. An odds ratio is a comparison of odds, not an individual probability, and a p-value is not the probability that a treatment effect is real. Statistical evidence does not by itself establish the importance of the observed difference to patients.

The primary finding is specifically about the first attack and a 2-hour assessment in the defined analysis population. It does not directly answer whether an outcome persisted, whether pain relief occurred in other attacks, or whether every participant responded consistently across attacks. Those questions are addressed by separate secondary outcomes, which have their own populations and multiplicity considerations.

The registry record also does not provide all details needed to assess missing-data handling, the full regression model, or adjustment across secondary analyses. The safety counts are descriptive and do not establish causation. Therefore, the results are best understood as a randomized group comparison for a defined endpoint, with statistical uncertainty and reporting limitations, rather than as a guarantee, a comprehensive account of long-term outcomes, or medical advice.

18. Why This Trial Matters Statistically

This trial illustrates why an endpoint's definition and time frame matter. A study can include treatment across multiple migraine attacks while its primary analysis focuses on pain freedom after the first attack. Keeping those questions separate prevents a first-attack result from being misread as evidence about every later attack or sustained response.

It also shows how effect measures answer different questions. The posted arm percentages give an intuitive description of event frequency. A risk difference expresses the comparison in absolute terms, while the sponsor's logistic regression reports odds ratios. These measures are related but not interchangeable. Reading both the observed percentages and the model-based estimate helps avoid treating an odds ratio as a direct change in probability.

The registry record provides a useful contrast between a primary analysis and numerous secondary analyses. With 22 outcome measures and 67 analyses recorded, multiplicity and outcome selection are important to interpretation. A consistent direction across several secondary endpoints may add descriptive context, but without reported multiplicity methods it should not carry the same weight as the designated primary analysis.

Finally, the case demonstrates the value of reporting confidence intervals and analysis populations alongside p-values. The interval conveys uncertainty in the estimated comparison, and the population definition clarifies who contributed data. Missing methodological detail limits reproducibility, even when the main outcome and sponsor estimates are clearly reported. These distinctions are central to careful reading of clinical trial results.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The primary result applies to pain freedom 2 hours after dosing during the first attack, in the registry-defined evaluable population. It does not summarize all attacks treated in the study. Although treatment was available for up to 4 attacks, first-attack and repeated-attack outcomes address different questions and have different denominators.

The primary outcome population excludes randomized participants who did not meet the stated treated-attack and assessment criteria. The record does not provide a full accounting of missing assessments or detailed missing-data procedures, so the consequences of unavailable outcome measurements cannot be assessed from the posted information alone.

The registry identifies logistic regression for the sponsor analyses but does not report the full model specification or covariates. The record also does not report whether or how multiplicity was handled across the many secondary analyses. These gaps limit independent reproduction and make secondary p-values harder to interpret.

The safety results are reported as counts within multiple groups, including maximum extended enrollment and open-label extension groups. These groups should not be treated as a single randomized comparison, and the reported counts do not establish treatment causation. The record does not report the original sample-size rationale or detailed protocol information, so the planning illustrations cannot be taken as the trial's design justification.

Finally, the posted summary concerns specific measured outcomes and follow-up periods. It does not establish effects on outcomes that were not reported in the registry record or provide a complete account of all possible benefits and harms.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Lasmiditan (LY573144) Over Four Migraine Attacks — NCT03670810.
• Registry record last updated 2022-07-29; status last verified 2022-06.
• Results first posted 2021-07-02.
• 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 registry record describes a completed randomized, double-masked, parallel phase 3 trial of lasmiditan for migraine, with enrollment of 1633. The primary endpoint was the percentage of participants pain free 2 hours after dosing during the first attack, among participants meeting the stated evaluable-attack criteria.

The posted percentages were 25.8% for 100 mg lasmiditan, 29.3% for 200 mg lasmiditan, and 8.4% for placebo. The sponsor reported logistic-regression odds ratios of 3.83 and 4.56 for the respective lasmiditan comparisons with placebo, both with p < 0.001. The registry also records multiple secondary outcomes and safety groups.

Important details, including the original sample-size justification, full missing-data methods, and multiplicity approach, are not reported in the registry record. The findings describe the posted outcomes and analyses, not guaranteed individual results.