A Study to Test if Fremanezumab Reduces Headache in Participants With Posttraumatic Headache (PTH): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Test if Fremanezumab Reduces Headache in Participants With Posttraumatic Headache (PTH) (NCT03347188), a randomized phase 2 study. The primary outcome is DB Period: Mean Change From Baseline in Monthly Average Number of Headache Days of at Least Moderate Severity During the 12-Week Treatment Period After the First Dose of Fremanezumab.

2. Clinical Question

3. Trial Design

NCT03347188 is a phase 2, multicenter, randomized, quadruple-masked, parallel-group, placebo-controlled study of fremanezumab for posttraumatic headache. The registry record reports 87 enrolled participants and a completed study. The recorded start date is 2017-12-18, and the primary completion date is 2020-03-13.

The study had a 12-week double-blind treatment period followed by a 12-week open-label treatment period. During the double-blind period, participants were assigned to fremanezumab or matching placebo. The fremanezumab regimen was 675 milligrams given as 3 subcutaneous injections at randomization, Week 4, and Week 8. The registry description for placebo likewise specifies 3 matching injections at those time points. It also describes participants who completed the double-blind period and continued into the open-label period receiving fremanezumab, but the arm descriptions in the registry record are incomplete.

Randomization and masking support a comparison between the assigned groups during the double-blind period. The registry record identifies the study as quadruple-masked but does not specify which four roles were masked. Details of the randomization procedure, allocation concealment, and the planned sample-size justification are not reported in the registry record.

The primary analysis population was the full analysis set: randomized participants who received at least 1 dose of study drug and had at least 1 post-baseline efficacy assessment for the primary endpoint. The endpoint group counts were 42 for fremanezumab and 43 for placebo. The registry record does not explain why these counts differ from total enrollment or provide a full accounting of exclusions and missing endpoint assessments.

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.

Double-blind (DB) Period (12 Weeks)

Reasons for not completing this period:

Open-label (OL) Period (12 Weeks)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set (FAS) included all randomized participants who received at least 1 dose of the study drug and had at least 1 post-baseline efficacy assessment on the primary endpoint.

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: Intent-to-treat (ITT) analysis set included all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as DB Period: Mean Change From Baseline in Monthly Average Number of Headache Days of at Least Moderate Severity During the 12-Week Treatment Period After the First Dose of Fremanezumab, measured over Baseline (Day -28 to Day -1), up to Week 12.

The primary endpoint was the mean change from baseline in monthly average headache days of at least moderate severity during the 12-week double-blind treatment period. The baseline window was Day -28 to Day -1, and the primary timeframe extended up to Week 12. A headache day counted when a participant reported a headache of at least moderate severity.

Monthly averages were normalized to a 28-day equivalent. The registry description specifies dividing the number of days with the efficacy variable by the number of days with assessments recorded in the electronic diary over the relevant period, then multiplying by 28. Change was defined as the post-baseline value minus the baseline value. Consequently, a negative change represents fewer such headache days on average than at baseline.

This endpoint is useful for comparing changes in a recurring symptom over a defined period, while accounting for differing numbers of diary assessment days. It summarizes frequency rather than intensity on each headache day, duration, or an individual participant’s overall experience. The least-squares means were estimated with an analysis of covariance model that included treatment and duration of posttraumatic headache history as fixed effects, with baseline monthly average headache days as a covariate. The result is therefore an adjusted group comparison, not simply a comparison of raw participant-level changes.

7. Results Posted in the Registry

• Fremanezumab: mean -3.60 versus -5.10 on control; difference 1.50 (95% CI -1.27 to 4.27), standardized effect d = 0.23; z-test p = 0.289.

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.

DB Period: Number of Participants Reaching at Least 50% Reduction From Baseline in Monthly Average Number of Headache Days of Any Severity During 12-Week Treatment With Fremanezumab

• 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:
  • Fremanezumab: 11.9% (5/42) versus 9.3% (4/43) on control; risk difference 2.6 percentage points (95% CI -10.5 to 15.7), risk ratio 1.28 (0.37–4.44); unadjusted two-proportion z-test p = 0.697.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Fremanezumab and the control. The interval gives the range of effects that remain compatible with the data.

DB Period: Number of Participants Reaching at Least 50% Reduction From Baseline in the Monthly Average Number of Headache Days of at Least Moderate Severity During 12-Week Treatment With Fremanezumab

• 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:
  • Fremanezumab: 21.4% (9/42) versus 25.6% (11/43) on control; risk difference -4.2 percentage points (95% CI -22.2 to 13.8), risk ratio 0.84 (0.39–1.81); unadjusted two-proportion z-test p = 0.652.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Fremanezumab and the control. The interval gives the range of effects that remain compatible with the data.

DB Period: Change From Baseline in Disability Score, as Measured by the 6-Item Headache Impact Test (HIT-6) Total Score at Week 12 After the First Dose of Fremanezumab

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Fremanezumab: mean -6.90 versus -10.80 on control; difference 3.90 (95% CI -8.84 to 16.64), standardized effect d = 0.14; z-test p = 0.548.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Fremanezumab and the control. The interval gives the range of effects that remain compatible with the data.

The registry record lists 14 secondary outcome measures, spanning headache-day response thresholds, changes at different intervals, disability, patient global impression of change, and treatment-emergent adverse events. The numerical secondary comparisons available for interpretation cover responder outcomes for headache days and the change in HIT-6 disability score. Across these outcomes, the estimated contrasts do not form a consistent pattern that clearly reinforces one direction for the primary comparison.

These secondary measures address related but distinct questions. Responder outcomes classify participants according to whether they reach a threshold reduction, while the primary endpoint compares average changes in headache days. HIT-6 assesses disability rather than headache-day frequency. In general, agreement across different measures can add context, but disagreement or uncertainty is not unexpected when measures capture different aspects of the condition and are based on limited participant counts.

The available secondary comparisons are not a substitute for the primary endpoint. Their confidence intervals are broad, and the registry record does not report whether the secondary analyses were adjusted for multiple comparisons. When many outcomes are examined, some results can appear notable by chance even if no underlying differences exist. Unadjusted secondary results therefore carry less weight than a prespecified primary analysis and are most appropriately treated as exploratory evidence requiring cautious interpretation.

The registry record does not report numerical results for every listed secondary measure in the information summarized for these comparisons. That means the overall secondary-outcome picture cannot be reconstructed solely from the listed measures and the detailed comparisons. The secondary findings should be considered alongside the primary estimate, its confidence interval, the study’s phase 2 status, and the limits of the reported analysis details—not as independent confirmation of a definitive treatment effect.

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

• DB Period: Number of Participants Reaching at Least 50% Reduction From Baseline in Monthly Average Number of Headache Days of Any Severity During 12-Week Treatment With Fremanezumab
• DB Period: Number of Participants Reaching at Least 50% Reduction From Baseline in the Monthly Average Number of Headache Days of at Least Moderate Severity During 12-Week Treatment With Fremanezumab
• DB Period: Number of Participants Reaching at Least 50% Reduction From Baseline in the Monthly Average Number of Headache Days of at Least Moderate Severity During First 4-Week (Month 1), 5- to 8-Week (Month 2), and 9- to 12-Week (Month 3)
• DB Period: Change From Baseline in the Number of Headache Days of At Least Moderate Severity During the First 4-Week (Month 1), 5- to 8-Week (Month 2), and 9- to 12-Week (Month 3)
• DB Period: Change From Baseline in Disability Score, as Measured by the 6-Item Headache Impact Test (HIT-6) Total Score at Week 12 After the First Dose of Fremanezumab
• DB Period: Number of Participants (Responder and Non-Responder) With the Patient Global Impression of Change (PGIC) Scale at Weeks 4, 8, and 12 After the First Dose of Fremanezumab
• DB Period: Number of Participants With Treatment-Emergent Adverse Events (TEAEs)
• OL Period: Number of Participants With TEAEs

9. Statistical Interpretation

10. Statistical Methods Explained

The primary comparison is an adjusted comparison of average changes in a continuous outcome. The registry reports least-squares means from an analysis of covariance model, or ANCOVA. In general, ANCOVA estimates group differences while accounting for specified covariates. Here the model included treatment and duration of posttraumatic headache history as fixed effects, and baseline monthly average headache days of at least moderate severity as a covariate. The duration categories were less than 12 month since the brain injury or greater or equal to 12 month since the brain injury, as described in the registry record.

What does a least-squares mean represent?

In general, a least-squares mean is a model-based estimate of a group mean after accounting for the covariates included in the model. It is not necessarily identical to the simple arithmetic average of the observed participant values. The primary group estimates were -3.6 days for fremanezumab and -5.1 days for placebo. The reported contrast of 1.5 days is expressed as fremanezumab minus placebo. For a change score defined as post-baseline minus baseline, negative group estimates indicate decreases from baseline; a positive between-group contrast means the estimate for fremanezumab is less negative than the estimate for placebo.

Why adjust for baseline and headache history?

In general, baseline adjustment can improve the comparison by accounting for initial differences in the measured outcome and can reduce unexplained variation when baseline values are associated with follow-up values. A history-duration term allows the model to account for the specified categories of headache history. Adjustment does not eliminate uncertainty, and the estimated contrast remains dependent on the model and the participants with usable outcome data. The registry record does not report further details of model diagnostics, missing-data handling, or sensitivity analyses.

How should the diary-based measure be understood?

The primary measure used monthly averages normalized to a 28-day equivalent. In general, diary normalization can make periods with differing numbers of recorded assessment days more comparable, but it relies on the available recorded days adequately representing the period. The registry defines a headache day as a day with a reported headache of at least moderate severity. Its change score is post-baseline minus baseline, so a reduction is represented by a negative value. The record does not describe diary completion rates or analyses testing the effect of incomplete diary data.

What is the role of a standard error?

The registry gives standard errors alongside the arm-level least-squares means: 0.99 for fremanezumab and 1.01 for placebo. In general, a standard error describes the uncertainty in an estimated mean, not the spread of individual participant outcomes. It is used in statistical inference, including construction of confidence intervals. The recalculation uses the posted arm summary statistics and an approximate pooled standard deviation; it is an unadjusted analysis and should not be confused with the sponsor’s ANCOVA model.

Why can the adjusted and unadjusted comparisons differ?

In general, an adjusted analysis and an unadjusted comparison can yield different uncertainty estimates because they use different information and assumptions. ANCOVA accounts for the stated covariates, while the recalculation is based on posted summary statistics. The recalculated confidence interval from -1.272 to 4.272 and p = 0.289 are therefore supplementary. The registry-reported adjusted analysis has a confidence interval from -0.73 to 3.67 and p = 0.188. The registry record does not identify a discrepancy between these analyses, but their distinct methods remain important to interpretation.

For binary secondary outcomes, a responder is counted if the specified reduction threshold is reached. In general, a risk difference expresses the difference between the proportions meeting that criterion in the two groups. The posted secondary comparisons include confidence intervals that can be broad when counts are small. Such intervals describe uncertainty in the group contrast; they do not imply that a participant’s result is predictable from the group average.

Interpretation also depends on the analysis population and outcome availability. The primary full analysis set required randomization, receipt of at least 1 dose, and at least 1 post-baseline primary efficacy assessment. That definition does not mean that all randomized participants necessarily contributed every secondary endpoint. The registry record does not report a comprehensive participant flow, endpoint-specific missingness, or the method used to handle missing data. Those omissions limit assessment of how robust the estimates are to incomplete outcome information.

11. Confidence Intervals and Precision

The sponsor-reported primary estimate is a 1.5-day adjusted difference between groups, with a 95% confidence interval from -0.73 to 3.67. In general, a confidence interval describes a range of estimates reasonably compatible with the data and analysis under the method’s assumptions. It is not a range that contains a known true value with certainty, and it is not a prediction interval for individual participants.

The primary interval includes zero and reaches values on both sides of it. Thus the posted estimate does not precisely distinguish between a difference favoring fremanezumab, little or no difference, or a difference favoring placebo. The interval’s width matters: even though the point estimate is a single number, it should not be read without the uncertainty around it. The reported p-value, 0.188, is consistent with the interval’s inclusion of zero under the sponsor’s analysis; it does not measure the size or clinical importance of the difference.

The unadjusted recalculation gives a 95% confidence interval from -1.272 to 4.272. It is based on summary statistics and an approximate pooled standard deviation, rather than the sponsor’s ANCOVA. It is useful as a separate check on the posted arm summaries, but it is not the preferred estimate of the adjusted primary comparison. The sponsor-reported analysis should take precedence for the trial’s primary result.

In general, greater precision can come from more informative observations, less variability, or a design and analysis that efficiently use relevant information. The primary endpoint groups contain 42 and 43 participants, and the standard errors for the arm-level least-squares means are 0.99 and 1.01. The registry record does not report the original sample-size justification, so it is not possible to determine from this record whether the achieved precision matched the trial’s design expectations.

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 trial’s original design assumptions. For this continuous primary endpoint, the figures refer to participants per arm. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each illustration uses the stated fraction of the observed effect as the effect to detect. The original trial’s sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the number of participants per arm that the planning calculation associates with an 80% chance of detecting the specified effect under its assumptions, if that effect were the true effect. The 90% power column uses a higher target chance of detection and therefore requires more participants per arm for the same effect. For the primary endpoint, the illustration gives 297 participants per arm for 80% power and 398 participants per arm for 90% power when planning for 100% of the observed effect. Planning for 75% of the observed effect gives 528 and 707 participants per arm; planning for 50% gives 1,187 and 1,589 participants per arm.

Smaller effects are harder to distinguish from random variation, so the planning calculation requires more participants when the effect to detect is reduced. These figures depend on the stated assumptions and on using the observed effect as the reference point. They are not observed power, do not show the probability that the trial’s result is correct, and do not change the interpretation of the confidence interval from the completed trial. The registry record does not report the sponsor’s original planning effect or intended statistical power.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 1.500; pooled standard deviation σ = 6.522.
2. n = 2 × (1.96 + 0.842)² × 6.522² / 1.500² = 296.7, rounded up to 297 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 1.5                      # observed difference in means
sigma <- 6.52154242972095                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3347188)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 297 (80% power), 398 (90% power)
 75% of observed effect: n per arm = 528 (80% power), 707 (90% power)
 50% of observed effect: n per arm = 1,187 (80% power), 1,589 (90% power)
power.t.test: 297.7 
Simulated power at n = 297 per arm: 0.798 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is an ANCOVA comparison of least-squares means. It reports an adjusted difference of 1.5 days, a 95% confidence interval from -0.73 to 3.67, and p = 0.188. This is the primary analysis to use when describing the trial’s reported result because it reflects the analysis method stated in the registry record.

The recalculated comparison uses the posted arm summaries and is unadjusted. It gives the same point estimate of 1.5 days but a confidence interval from -1.272 to 4.272 and p = 0.289. These differences in interval and p-value reflect that the recalculation is not the same model-based analysis as the sponsor’s ANCOVA. It does not reproduce the covariate adjustment, and its standard deviation is approximate.

The recalculation is best read as a secondary statistical check, not as a replacement or correction of the reported analysis. Both intervals include zero, but the sponsor’s estimate and uncertainty should anchor the interpretation of the primary endpoint. The registry record does not report a discrepancy flag, and neither analysis should be converted into a claim of equivalence or proof that there is no difference.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple secondary outcomes in addition to the primary endpoint. When investigators test several outcomes, the chance of obtaining at least one apparently notable result by chance can increase. In general, multiplicity procedures can control that risk by adjusting the testing strategy or by defining a hierarchy for interpreting outcomes. The registry record does not report whether multiplicity was controlled, whether outcomes were ranked in a testing sequence, or which comparisons were considered confirmatory.

The primary endpoint description identifies the outcome, timeframe, analysis population, and ANCOVA covariates. However, the record does not report the full statistical analysis plan, the original sample-size justification, interim analyses, or additional model assumptions. It also does not report whether sensitivity analyses were planned or conducted. Without those details, the reported primary comparison can be described, but its place in a broader prespecified testing strategy cannot be assessed.

The primary full analysis set included randomized participants who received at least 1 dose and had at least 1 post-baseline assessment on the primary endpoint. This definition explains who qualified for the primary efficacy analysis, but it does not itself describe how missing data were handled. The registry record does not report the amount or pattern of missing diary data, the imputation method, or analyses evaluating alternative assumptions about missing outcomes. It also does not provide a detailed account of participants excluded from the endpoint analysis.

Different outcomes can have different analysis populations. The HIT-6 comparison, for example, has fewer participants in its endpoint groups than the primary comparison, but the record does not explain the reasons for that difference. The secondary endpoint analysis populations and missing-data procedures are not reported in full. These gaps matter because selective availability of outcomes can affect the interpretation of estimates, particularly when group sizes are limited.

Overall, the primary result is interpretable as the registry-reported ANCOVA estimate for the defined analysis population. The absence of reported details on multiplicity, missing-data methods, and the complete analysis plan limits how confidently the secondary findings and the robustness of the primary estimate can be assessed.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: gastrointestinal haemorrhage (2.3%), non-cardiac chest pain (2.3%), cholelithiasis (2.3%), lung neoplasm malignant (2.3%), hypotension (2.3%).

During the double-blind period, serious adverse events affected 2 of 43 participants at risk in the fremanezumab group and 1 of 43 participants at risk in the placebo group. The registry record lists no deaths in either group. Other adverse events affected 28 of 43 participants in the fremanezumab group and 29 of 43 in the placebo group. These are participant counts, not estimates of whether an event was caused by study treatment.

The serious-event terms listed in the registry each have a reported frequency of 2.3%: gastrointestinal haemorrhage, non-cardiac chest pain, cholelithiasis, lung neoplasm malignant, and hypotension. The registry does not provide enough detail in these figures to infer timing, severity beyond the serious-event classification, relatedness, or whether the same participant experienced more than one event. The listed terms should therefore be read as descriptive safety information rather than as evidence establishing a causal relationship.

Safety counts are informative but limited by the number of participants and the follow-up described. The study included a 12-week double-blind period and a 12-week open-label period; the serious-event counts above refer to the posted group figures and should not be generalized beyond their stated period or population. The registry record does not report a detailed safety analysis plan or a full account of event-related withdrawals.

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

The primary result describes an adjusted average difference in change in monthly headache days of at least moderate severity during the 12-week double-blind period. The estimated difference was 1.5 days, with a sponsor-reported 95% confidence interval from -0.73 to 3.67. This interval includes zero and allows for estimates in either direction. The result therefore does not establish a clear between-group difference, but it also does not establish that the groups are equivalent.

The result is specific to the endpoint definition, timeframe, analysis model, and participants included in the primary analysis. It concerns average changes in a diary-derived count normalized to a 28-day equivalent. It is not a direct statement about headache intensity on every day, the experience of each participant, longer-term outcomes, or outcomes outside the study population. The registry record describes adults aged 18 to 70 years, inclusive, with posttraumatic headache.

The trial is identified as a phase 2 proof-of-concept study. Its results can be understood as estimates in a randomized comparison, with uncertainty that remains substantial for the primary contrast. The reported p-value does not show the probability that the treatment works, and a confidence interval including zero is not proof that there is no effect. Likewise, a numerical difference between the group estimates is not by itself evidence of a clinically important difference.

The secondary results do not resolve these limitations. They address other outcomes and include uncertainty, while the registry record does not report whether multiplicity was accounted for. Safety figures are also descriptive and should not be read as proving or disproving causation. Taken together, the posted results describe what was observed and estimated in this trial, while leaving meaningful uncertainty about the size and direction of the primary treatment contrast.

18. Why This Trial Matters Statistically

This trial illustrates why the estimated difference between groups should be read together with its uncertainty. The primary contrast is 1.5 days, but the sponsor-reported confidence interval ranges from -0.73 to 3.67. A point estimate alone can make a comparison appear more definite than the data justify; the interval shows that several interpretations remain compatible with the analysis.

It also demonstrates the distinction between an adjusted model-based result and a simple recalculation from group summaries. ANCOVA included baseline headache-day averages and duration of posttraumatic headache history, whereas the recalculation is unadjusted and uses an approximate pooled standard deviation. The two analyses answer related questions, but they are not interchangeable. The registry-reported analysis is the appropriate primary reference.

The endpoint construction is another important statistical feature. Diary observations were normalized to a 28-day equivalent, and the outcome was a change from baseline. Understanding the direction of that change is essential: negative values indicate decreases, and the positive between-group contrast reflects a less negative estimate in the fremanezumab group than in the placebo group.

Finally, the study shows why secondary outcomes, multiplicity, and missing-data reporting matter. A range of symptom, disability, and safety measures can provide useful context, but their interpretation depends on the analysis plan and the completeness of outcome reporting. The registry record leaves several of those details unreported. Statistical reading therefore requires both attention to the numerical estimates and awareness of the limits of the available methods information.

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 principal limitation for interpreting this trial’s efficacy results is the imprecision of the primary between-group estimate. The sponsor-reported confidence interval includes zero and extends in both directions. The enrolled and analyzed groups are relatively small in the context of the planning illustrations, which are based on the observed effect and are not the original trial design assumptions.

The registry record does not report the original sample-size rationale, the full statistical analysis plan, or whether interim analyses occurred. It also does not report multiplicity procedures for the many secondary outcomes. This makes it difficult to determine which secondary comparisons were intended as confirmatory and how their results should be weighted against one another.

Missing-data information is limited. The primary analysis population required at least 1 post-baseline efficacy assessment, but the record does not describe diary completion, the amount or pattern of missing data, or the method used to handle missing values. The endpoint counts do not by themselves explain all differences between enrollment and analysis populations. The record also does not provide the reasons for the smaller groups on some secondary outcomes.

The primary endpoint depends on electronic diary assessments and a normalization procedure based on days with recorded assessments. The registry record does not report analyses testing how incomplete diary recording might affect the estimates. The sponsor’s ANCOVA and the unadjusted recalculation also rely on different methods; the latter uses an approximate pooled standard deviation and should not be interpreted as an independent reproduction of the adjusted model.

Finally, the registry’s open-label period followed the 12-week double-blind period, but the reported primary comparison is for the double-blind period. The arm descriptions for the open-label period are incomplete, and the information summarized for the primary endpoint does not establish longer-term comparative effects. These limitations constrain interpretation to the posted short-term randomized comparison and its stated analysis population.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Test if Fremanezumab Reduces Headache in Participants With Posttraumatic Headache (PTH) — NCT03347188.
• Registry record last updated 2022-12-13; status last verified 2022-12.
• Results first posted 2021-03-26.
• 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

NCT03347188 is a completed phase 2 randomized, quadruple-masked, parallel-group study of fremanezumab versus placebo for posttraumatic headache. The registry reports enrollment of 87 participants and a 12-week double-blind period followed by a 12-week open-label period.

The primary endpoint was change from baseline to Week 12 in monthly average headache days of at least moderate severity. The sponsor-reported ANCOVA estimated a between-group difference of 1.5 days, with a 95% confidence interval from -0.73 to 3.67 and p = 0.188. The interval includes zero, leaving uncertainty about the direction and size of the difference.

The record lists multiple secondary outcomes and descriptive safety counts. Important details—including the original sample-size justification, multiplicity strategy, and missing-data methods—are not reported in the registry record.