A Safety and Efficacy Study of Mepolizumab in Subjects With Severe Asthma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Safety and Efficacy Study of Mepolizumab in Subjects With Severe Asthma (NCT03562195), a randomized phase 3 study. The primary outcome is Rate of Clinically Significant Exacerbations of Asthma.

2. Clinical Question

The clinical question is whether adding mepolizumab to participants’ existing standard of care changes the rate of clinically significant asthma exacerbations, compared with adding placebo, among Chinese participants with severe asthma and eosinophilic inflammation. The registry describes a randomized, double-blind, parallel-group phase 3 study. Both groups continued baseline standard of care for asthma, and salbutamol was available as rescue medication on an as-needed basis.

The registry record reports an enrollment of 300. It does not report the distribution of participants by country or the detailed eligibility criteria beyond the stated population. The trial’s primary endpoint concerns an event rate over treatment exposure, rather than simply whether each participant experienced at least one event.

3. Trial Design

The record describes a randomized, double-blind, parallel-group phase 3 study sponsored by GlaxoSmithKline. Participants were assigned to mepolizumab 100mg or matching placebo, with both treatments added to baseline standard of care for asthma. The brief summary describes a 1:1 allocation and an intended total of 300 participants. The registry lists the trial as completed and records enrollment of 300.

The treatment descriptions specify subcutaneous administration every 4 weeks for a total treatment period of 52 weeks. Salbutamol was available as rescue medication on an as-needed basis. The brief summary gives a maximum study duration of 56 weeks, while the primary endpoint timeframe is through Week 52. The registry record does not describe the timing or purpose of the additional period beyond the treatment timeframe.

The primary analysis population was the Modified Intent-to-Treat Set, defined in the record as randomized participants who received at least one dose of trial medication. Participants who were randomized but did not receive study treatment were excluded. The posted primary group summaries include 149 participants in the mepolizumab group and 151 in the placebo group. These are the analysis-group counts reported for the primary outcome, not a description of every secondary endpoint’s available data.

The registry lists Asthma as the condition and specifies severe asthma with eosinophilic inflammation in the study summaries and official title. The record’s primary outcome is the rate of clinically significant exacerbations. It also lists secondary outcomes involving exacerbation probability and healthcare use, respiratory questionnaire scores, lung function, adverse events, and laboratory measures. A full protocol, detailed eligibility criteria, recruitment locations, and the original sample-size justification are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Approximately 371 participants with severe asthma and eosinophilic inflammation who meet the protocol defined inclusion criteria were screened. Off these, 300 participants were randomized in 1:1 ratio to mepolizumab: placebo arms. 149 participants in the mepolizumab treatment group and 151 participants in the placebo treatment group.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis was performed on the Modified Intent-to-Treat (MITT) Set that included all randomized participants who received at least one dose of trial medication. 'Modified' implies that participants who were randomized but did not receive study treatment were excluded.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Rate of Clinically Significant Exacerbations of Asthma, measured over Up to Week 52.

The primary endpoint is the rate of clinically significant asthma exacerbations, expressed as exacerbations per year, through Week 52. The registry defines a clinically significant exacerbation as worsening asthma requiring systemic corticosteroids and/or hospitalization and/or an emergency department visit. The outcome therefore counts events and relates them to time on treatment; it is not simply the proportion of participants who had at least one exacerbation.

Event rates are useful when participants contribute different amounts of observation time or when some participants experience more than one event. A rate can retain information about repeated events, whereas a yes-or-no outcome generally does not distinguish one event from multiple events. The registry’s model used logarithm of time on treatment as an offset, which is a standard way to account for exposure time when modeling event counts.

In the posted group summaries, the annualized rates were 0.45 in the mepolizumab group and 1.31 in the placebo group. These are descriptive rates for the primary analysis population. The sponsor’s reported comparison is a model-based rate ratio, not a direct difference between the two rates. The primary endpoint focuses on exacerbations judged clinically significant by the specified treatment or healthcare-use criteria; it does not capture every change in asthma symptoms or all aspects of asthma control.

7. Results Posted in the Registry

• Mepolizumab 100mg: 0.45 versus 1.31 (Exacerbations per year) on control; rate ratio 0.34 (95% CI 0.26–0.45, approximate, Poisson model); 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.

Mean Change From Baseline in St. George's Respiratory Questionnaire (SGRQ) at Week 52

• 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:
  • Mepolizumab 100mg: mean -14.29 versus -7.20 on control; difference -7.09 (95% CI -11.39 to -2.79), standardized effect d = -0.40; z-test p = 0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Mepolizumab 100mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Change From Baseline in Clinic Prebronchodilator Forced Expiratory Volume in 1 Second (FEV1) at Week 52

• 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:
  • Mepolizumab 100mg: mean 262.79 versus 125.67 on control; difference 137.12 (95% CI 40.33 to 233.91), standardized effect d = 0.36; z-test p = 0.005.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Mepolizumab 100mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Platelets Count

• 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 deviations.
• Result:
  • Mepolizumab 100mg: mean 0.60 versus -3.50 on control; difference 4.10 (95% CI -5.67 to 13.87), standardized effect d = 0.10; z-test p = 0.411.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mepolizumab 100mg and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a broad range of secondary outcomes, including exacerbation timing and healthcare use, respiratory questionnaire scores, lung function, adverse events, and laboratory counts. The posted numerical summaries include selected continuous outcomes. Across those results, the estimated change in SGRQ was more negative for mepolizumab than placebo, the estimated change in FEV1 was more positive, and the platelet-count difference was small relative to its interval of uncertainty. These measures describe different aspects of respiratory status and safety; they do not measure the same event process as the primary exacerbation rate.

The secondary findings can provide context for the primary comparison, but they should not be combined into a single claim about benefit or harm. A questionnaire score, a lung-function measure, and an event rate have different scales, meanings, and sources of variability. The registry record does not supply enough information to assess every secondary outcome listed, including detailed results for all posted outcome measures.

Unadjusted secondary results generally carry less inferential weight when many outcomes are examined. In general, testing multiple outcomes increases the chance that some results will appear noteworthy by chance, even when no corresponding underlying effect exists. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the secondary outcomes. Therefore, isolated secondary p-values should be treated as exploratory evidence rather than as independent confirmation of the primary result. The pattern may help frame questions for further study, but it does not replace the sponsor’s primary analysis.

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

• Percent Probability of First Clinically Significant Exacerbations at Week 16, Week 32, and Week 52
• Mean Change From Baseline in St. George's Respiratory Questionnaire (SGRQ) at Week 52
• Number of Participants With Clinically Significant Exacerbations Requiring Hospitalization (Including Intubation and Admittance to an Intensive Care Unit [ICU]) or Emergency Department (ED) Visits
• Number of Participants With Clinically Significant Exacerbations Requiring Hospitalization
• Mean Change From Baseline in Clinic Prebronchodilator Forced Expiratory Volume in 1 Second (FEV1) at Week 52
• Number of Participants With Adverse Events (AEs) Including Systemic (i.e., Allergic [Type I Hypersensitivity] and Other Systemic) and Injection Site Reactions
• Change From Baseline in Platelets Count
• Change From Baseline in Erythrocytes Count

9. Statistical Interpretation

10. Statistical Methods Explained

The sponsor’s primary comparison used a negative binomial model. This model is designed for count outcomes, such as the number of exacerbations, and can accommodate variability in counts beyond what a simple Poisson model assumes. In general, over-dispersion means that counts vary more than a Poisson model would expect; ignoring it can make uncertainty estimates too optimistic. The registry identifies the negative binomial model but does not report a dispersion estimate.

The modeled outcome was the number of clinically significant exacerbations. The model included treatment group, baseline maintenance oral corticosteroid therapy, region, exacerbations in the year before the study as an ordinal variable, and baseline percent-predicted FEV1. These are the covariates listed in the primary endpoint description. Adjustment means the reported treatment comparison comes from a model that accounts for these listed characteristics, rather than from a comparison of the two raw event totals alone.

The model used the logarithm of time on treatment as an offset variable. In general, an offset allows a count model to estimate a rate while accounting for the amount of time each participant contributes. Participants observed for more time can have more opportunity to experience events; the offset incorporates exposure time into the rate comparison. The primary endpoint is therefore appropriately interpreted as an exacerbation rate over time, not as a simple count.

What does the rate ratio express?

A rate ratio compares the modeled event rate in one group with the rate in the comparison group. A value of 1 corresponds to equal modeled rates; values below 1 correspond to a lower modeled rate in the group in the numerator. The sponsor reports a rate ratio of 0.35 for mepolizumab versus placebo. This is a relative comparison under the fitted model. It is not the number of events prevented per participant, and it does not state how many participants experienced any event.

The unadjusted re-calculation uses the posted group rates and the exposure-time information encoded in the posted summary statistics. It is a simple check on the direction and approximate size of the comparison. It does not reproduce the sponsor’s covariate adjustment or all features of the trial analysis. The sponsor’s reported negative binomial analysis remains the primary statistical result.

How should the p-value be read?

The sponsor reports p < 0.001 for the negative binomial comparison. In general, a p-value describes how compatible the observed data are with a specified null hypothesis under the model and its assumptions. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The p-value should be considered alongside the rate ratio, its confidence interval, the endpoint definition, and the study population.

The record separately reports a Bayesian Dynamic Borrowing result as a posterior probability of 1.000. A posterior probability is interpreted within a Bayesian framework, conditional on the model and assumptions used. The record does not report the details needed to evaluate how information was borrowed, what prior distributions were used, or how the posterior probability was calculated. It should not be treated as the same statistic as the negative binomial p-value.

What does the analysis population include?

The primary analysis used the Modified Intent-to-Treat Set. It included randomized participants who received at least one dose of trial medication and excluded randomized participants who did not receive treatment. This definition is reported in the registry record. It is narrower than a population including every randomized participant, so the primary estimate describes the defined treated, randomized analysis set.

The registry identifies the primary model, covariates, and offset, but does not report several other details that can affect interpretation: how missing outcome data were handled, whether any interim analyses were planned or conducted, the original sample-size justification, or the full statistical analysis plan. Their absence in the record does not establish that such procedures were absent from the trial; those details are not reported in the registry record.

General statistical interpretation also depends on whether the model is suitable for the data and whether its assumptions are reasonable. A negative binomial model allows extra variability in count outcomes, but no model removes uncertainty or guarantees that all relevant differences between groups have been addressed. Randomization and blinding help support a fair comparison, while the model’s covariate adjustment and exposure-time offset specify how the primary rate comparison was estimated.

11. Confidence Intervals and Precision

The sponsor reports a 95% confidence interval from 0.24 to 0.50 for the rate ratio of 0.35. In general, a confidence interval describes the range of estimates compatible with the observed data under the analysis method and its assumptions. It is not a guarantee that the true value lies in the interval, nor does it describe the range of individual participants’ outcomes. Here, the reported interval is entirely below 1, the value corresponding to equal rates, which is consistent with a lower modeled rate in the mepolizumab group under the sponsor’s analysis.

The unadjusted recalculation gives a rate ratio of 0.344 with an interval from 0.260 to 0.453. This interval is based on a simple re-calculation from posted summary statistics. It should not be confused with the sponsor’s adjusted model interval, which is the primary reported analysis. The two intervals are similar in location, but they arise from different calculations and should be interpreted according to their respective methods.

For the secondary continuous outcomes, the registry posts confidence intervals for the mean differences in SGRQ change and FEV1 change. The SGRQ mean difference is -7.09, with an interval from -11.39 to -2.789; the FEV1 mean difference is 137.1 mL, with an interval from 40.33 to 233.9. The interval for platelet-count change is 4.1, from -5.673 to 13.87. These intervals communicate uncertainty in the estimated group differences, but the registry lists no sponsor-reported analysis for these outcomes in the results.

Precision depends on the amount and variability of the available data, as well as the analysis method. A confidence interval does not address bias from missing data, limitations in the analysis population, or applicability beyond the enrolled study population. The record does not report every detail needed to assess those issues.

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 planning figures are illustrations based on the observed effect, not the trial’s original design calculations. For the primary rate endpoint, the figures are participants per arm. The planning assumptions listed in the registry-derived information are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set at the stated fraction of the observed effect. For time-to-event outcomes, the convention would be total events, but the primary endpoint in this trial is a rate outcome; the planning figures for it are participants per arm.

The 80% power column gives the number of participants per arm estimated to be needed for an 80% chance of detecting the specified effect under those planning assumptions, if that effect were the true one and the assumptions held. The 90% power column gives the corresponding planning number for a 90% chance. These are design-planning concepts, not guarantees about a future study’s result.

The rows describe planning for the full observed effect, 75% of the observed effect, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so the illustration requires more participants per arm as the assumed effect becomes smaller. The listed counts are conditional on the observed effect from this trial and on the stated simplified assumptions; they are not the sample-size justification used to design this trial. The original design assumptions and justification are not reported in the registry record.

These figures are not observed power. They do not tell how likely the completed trial was to detect its observed result after the data were collected. The useful reading is prospective: under the stated assumptions, they illustrate how planning requirements change as the target effect changes.

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, λ₀ and λ₁ are event rates per patient-year, and T is follow-up per participant in years. This assumes Poisson counts and equal follow-up.

Worked calculation for this trial

1. Control rate λ₀ = 1.3100; treatment rate λ₁ = 0.4500 events per patient-year; follow-up T = 1.00 years.
2. n = (1.96 + 0.842)² × (1/1.3100 + 1/0.4500) / (1.00 × [ln(0.344)]²) = 20.6, rounded up to 21 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
lambda0 <- 1.31                      # control rate, events per patient-year
lambda1 <- 0.45                      # treatment rate
tt      <- 0.996550402453047                       # follow-up per participant, years
n_per_arm <- function(power, frac = 1) {
  l1 <- lambda0 * (lambda1 / lambda0)^frac   # smaller effect = fraction of the observed log rate ratio
  ceiling(round((z_a + qnorm(power))^2 * (1 / lambda0 + 1 / l1) / (tt * log(l1 / lambda0)^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 = ",")))

# Simulation 1: Poisson counts. Simulation 2: over-dispersed counts (variance = 1.5 x mean), analysed with the matching variance.
set.seed(3562195)
n    <- n_per_arm(0.80)
reps <- 20000
m0 <- lambda0 * n * tt; m1 <- lambda1 * n * tt
x0 <- rpois(reps, m0); x1 <- rpois(reps, m1)
z  <- log(x1 / x0) / sqrt(1 / x0 + 1 / x1)
cat(sprintf("Simulated power, Poisson counts, n = %s per arm: %.3f\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE)))
phi <- 1.5
y0 <- rnbinom(reps, mu = m0, size = m0 / (phi - 1)); y1 <- rnbinom(reps, mu = m1, size = m1 / (phi - 1))
z2 <- log(y1 / y0) / sqrt(phi * (1 / y0 + 1 / y1))
cat(sprintf("Simulated power, over-dispersed counts (phi = 1.5), same n: %.3f\n", mean(abs(z2) > z_a, na.rm = TRUE)))

Output from running the code

100% of observed effect: n per arm = 21 (80% power), 28 (90% power)
 75% of observed effect: n per arm = 31 (80% power), 41 (90% power)
 50% of observed effect: n per arm = 58 (80% power), 77 (90% power)
Simulated power, Poisson counts, n = 21 per arm: 0.858
Simulated power, over-dispersed counts (phi = 1.5), same n: 0.671

The first simulation draws Poisson counts and should land close to the target power. The second allows extra variability (variance 1.5 times the mean, a purely illustrative value), which lowers the power at the same sample size, so rate trials usually need more participants than the Poisson formula suggests.

14. Analyses Reported by the Sponsor

The sponsor reports two analyses of the primary comparison. The negative binomial model gives a rate ratio of 0.35 with a 95% confidence interval from 0.24 to 0.50 and p < 0.001. The registry also lists a Bayesian Dynamic Borrowing analysis with a posterior probability of 1.000 for superiority. These results use different statistical approaches and their reported quantities should not be treated as interchangeable.

The simple unadjusted re-calculation from posted summary statistics gives a rate ratio of 0.344, with an interval from 0.260 to 0.453 and p < 0.001. It is close to the sponsor’s reported rate ratio and points in the same direction. The recalculation is not a reconstruction of the sponsor’s covariate-adjusted negative binomial model, so its interval and p-value are best treated as a check rather than as the primary inferential result.

The registry does not flag a discrepancy between the sponsor analysis and the re-calculation. The sponsor’s reported model is therefore the central result for interpreting the trial, while the recalculation offers a straightforward comparison based on posted summaries. Further details about the Bayesian borrowing method are not reported in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies the Modified Intent-to-Treat Set for the primary analysis: randomized participants who received at least one dose of trial medication. It explicitly excludes randomized participants who did not receive study treatment. This definition matters because the estimated primary comparison applies to that analysis population. The record does not report the analysis population used for every secondary endpoint or explain whether outcome-specific exclusions were made.

Missing outcome data can affect estimates if participants with unavailable observations differ from those whose data are observed. The registry record does not report the amount or pattern of missing primary or secondary outcome data, the method used to handle missing outcomes, or sensitivity analyses addressing missingness. No particular missing-data procedure should therefore be assumed from the registry information alone.

The study reports multiple outcomes, including the primary exacerbation rate and secondary measures of exacerbation timing, healthcare use, respiratory status, safety, and laboratory counts. The registry record does not report a multiplicity adjustment, a testing hierarchy, or whether any secondary outcomes were designated as confirmatory. Without that information, interpretation of secondary p-values should account for the possibility of chance findings across multiple comparisons. This caution concerns the strength of inference, not the direction of any particular estimate.

The record also does not report whether interim analyses were planned or conducted, the original sample-size calculation, or the full statistical analysis plan. Those details would help clarify how analysis choices and multiple testing were handled. Their status is not reported in the registry record; absence of reporting should not be taken as evidence that no such procedures existed.

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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: asthma (7.9%), pneumonia (3.3%), nasal polyps (2.0%), cataract (1.3%), bronchitis (1.3%).

The registry reports serious adverse events for 18 participants in the mepolizumab group and 25 in the placebo group. The corresponding groups at risk are reported as 149 and 151. Other adverse events are reported for 134 participants in the mepolizumab group and 146 in the placebo group. The record also reports 0 deaths in the mepolizumab group and 2 in the placebo group. These are descriptive counts; they do not by themselves establish that a treatment caused or prevented an event.

The listed serious adverse event terms include Asthma at 7.9%, Pneumonia at 3.3%, Nasal polyps at 2.0%, Cataract at 1.3%, Bronchitis at 1.3%, and Nasal septum deviation at 0.7%. The record does not make clear in this summary whether those term percentages are presented across groups or for a particular group, so they should not be assigned to an arm. Event counts and percentages also do not convey severity, timing, duration, or relatedness to trial treatment.

Comparisons of safety counts require attention to exposure, follow-up, event definitions, and participant characteristics. The registry record does not provide a complete account of these details in the reported safety summary. The figures describe events recorded during the trial, not a prediction of an individual participant’s risk.

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

The primary result describes a lower modeled rate of clinically significant asthma exacerbations in the mepolizumab group than in the placebo group over the reported treatment timeframe, under the sponsor’s negative binomial analysis. The estimate is a group-level rate comparison in the Modified Intent-to-Treat Set, which includes randomized participants who received at least one dose. It does not describe every randomized participant or predict the outcome for any one person.

The endpoint is restricted to exacerbations requiring systemic corticosteroids and/or hospitalization and/or an emergency department visit. It does not encompass every symptom change or all possible definitions of worsening asthma. The rate ratio is a relative measure and should be considered with the group rates, confidence interval, endpoint definition, and analysis population.

The result does not establish that the same estimate would apply in populations with different characteristics, in settings with different standard care, or over a different duration. The registry record concerns Chinese participants with severe asthma and eosinophilic inflammation, receiving either study treatment alongside baseline standard of care. The record does not report detailed eligibility criteria or recruitment locations that would further define generalizability.

The secondary outcomes offer additional measurements, but the registry does not report a multiplicity strategy, complete missing-data details, or the full analysis plan. Those gaps limit how firmly secondary results can be interpreted. The safety counts are descriptive and do not alone establish causation. Statistical evidence from this trial should therefore be read as evidence about the recorded comparison under the reported analysis, not as a complete account of all possible outcomes or treatment effects.

18. Why This Trial Matters Statistically

This trial illustrates the distinction between an event count and an event rate. Participants may contribute different amounts of time, and some may experience repeated exacerbations. Modeling a rate with time on treatment as an offset addresses exposure time and allows the analysis to use information about recurrent events rather than reducing each participant’s experience to a yes-or-no indicator.

It also shows why the model and its adjustment matter. The sponsor used a negative binomial model and specified covariates for treatment group, baseline maintenance oral corticosteroid therapy, region, prior exacerbations, and baseline percent-predicted FEV1. The reported estimate is consequently a model-based rate ratio, not merely a comparison of the raw group rates. A simple unadjusted recalculation can serve as a check, but it does not supersede the sponsor’s model.

The registry reports both frequentist and Bayesian analyses of the primary comparison. Their distinct outputs—a rate ratio and p-value for the negative binomial model, and a posterior probability for Bayesian Dynamic Borrowing—illustrate that statistical conclusions depend on the framework and assumptions used. The record does not supply enough detail to fully assess the borrowing method.

Finally, the study’s numerous secondary outcomes make multiplicity and transparent reporting important. A clear primary endpoint helps anchor interpretation, while secondary endpoints can add context without each being treated as confirmatory. The registry record does not report the complete multiplicity plan, missing-data methods, interim analyses, or original sample-size justification, underscoring why a posted result should be read alongside the scope 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.
• Rate comparisons use a Poisson approximation unless the sponsor posted confidence limits; real event counts are often over-dispersed.

The registry record gives the primary model and its covariates, but does not report a full statistical analysis plan, model diagnostics, or a dispersion estimate. The negative binomial approach allows extra variability in counts in general, but the information available does not show how well the model fit this outcome or whether alternative model specifications were assessed.

The primary analysis population excludes randomized participants who did not receive treatment. That Modified Intent-to-Treat definition is explicit, but the record does not report how many randomized participants were excluded for that reason or describe the effect of the exclusion on the comparison. Missing outcome data and the methods used to address them are also not reported in the registry record.

The record does not report the original sample-size justification, the assumptions behind it, or whether interim analyses were planned or conducted. The planning illustrations based on observed effects cannot fill that gap. Nor does the record report a multiplicity adjustment or testing hierarchy for the many secondary outcomes.

The primary endpoint depends on the registry’s definition of clinically significant exacerbation, which combines worsening asthma with specified medication use or healthcare use. Results may not describe other dimensions of asthma control. The available summary also does not provide detailed eligibility criteria, recruitment locations, or enough information to assess the applicability of the results beyond the stated population. Safety findings are reported as counts and selected terms, without enough detail to assess event timing, duration, treatment relatedness, or exposure-adjusted risk.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Safety and Efficacy Study of Mepolizumab in Subjects With Severe Asthma — NCT03562195.
• Registry record last updated 2024-09-19; status last verified 2024-04.
• Results first posted 2024-09-19.
• 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

NCT03562195 is a completed, randomized, double-blind, parallel-group phase 3 study of mepolizumab added to standard asthma care versus matching placebo added to standard care. The registry reports enrollment of 300 and describes a population of Chinese participants with severe asthma and eosinophilic inflammation. Both groups could use salbutamol as rescue medication as needed.

The primary endpoint was the rate of clinically significant exacerbations through Week 52. The sponsor’s negative binomial model reported a rate ratio of 0.35, a 95% confidence interval from 0.24 to 0.50, and p < 0.001. The registry also reports a Bayesian Dynamic Borrowing posterior probability of 1.000 for superiority. The primary analysis population comprised randomized participants who received at least one dose of trial medication.

The record lists secondary respiratory, healthcare-use, safety, and laboratory outcomes. The full statistical analysis plan, original sample-size justification, multiplicity approach, and missing-data methods are not reported in the registry record.