A Placebo-controlled Study to Investigate Safety and Efficacy of BIA 2-093: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Placebo-controlled Study to Investigate Safety and Efficacy of BIA 2-093 (NCT02170077), a randomized phase 2 study. The primary outcome is The Percentage of Participants With a 50% or Greater Reduction in Seizure Frequency (Further Referred to as "Responders") in a Treatment Period Compared to the Baseline Period.

2. Clinical Question

3. Trial Design

This was a randomized, triple-masked, parallel-group phase 2 study. Its title identifies it as a placebo-controlled investigation of BIA 2-093, and the registry lists epilepsy as the condition. The stated purpose was to investigate safety and efficacy for controlling refractory partial seizures when the study treatment was added to ongoing therapy. Overall enrollment is listed as 144. The record gives a start date of 2002-04, primary completion in 2002-11, and a completed status.

The three groups were the once-daily BIA 2-093 group, the twice-daily BIA 2-093 group, and the placebo group. In each BIA 2-093 group, daily doses increased at four-weekly periods through 400 mg, 800 mg and 1200 mg. The placebo group received placebo. The registry does not report further details about the dose escalation procedures, adherence, or the ongoing therapies participants used.

Randomized allocation means participants were assigned to study groups by a random process, in general. Triple masking indicates that the record describes the study as masked at three levels, but it does not identify those roles in the registry record. A parallel-group design assigns participants to a group for comparison rather than describing a crossover between treatments. These design labels alone do not establish how allocation was implemented or how masking was maintained.

The arm-specific denominators for the primary outcome were 50 for once-daily BIA 2-093, 46 for twice-daily BIA 2-093, and 47 for placebo. These are the denominators attached to the posted outcome results; they should not be treated as a description of all participant flow. Reasons for the difference between these denominators and the overall enrollment, and details of exclusions or withdrawals, are not reported in the registry record. The trial’s original design assumptions and sample-size justification are also 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.

Overall Study

Reasons for not completing this period:

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 The Percentage of Participants With a 50% or Greater Reduction in Seizure Frequency (Further Referred to as "Responders") in a Treatment Period Compared to the Baseline Period, measured over baseline, week 12.

The primary endpoint was the percentage of participants with a 50% or greater reduction in seizure frequency during a treatment period compared with the baseline period. The registry refers to people meeting this threshold as responders and gives the timeframe as baseline and week 12. It is a binary outcome at the participant level: each person either meets the stated threshold or does not, and the group result is expressed as a percentage.

A threshold-based responder outcome offers a clinically interpretable way to describe how many participants reached a prespecified degree of reduction, rather than describing only an average change in seizure frequency. General statistical context: turning a continuous or count-based change into a responder category can make group proportions easy to compare, but it also discards information about how much participants’ seizure frequency changed beyond the threshold. The registry record does not provide the individual seizure counts, the distribution of changes, or further details about how baseline and treatment-period frequencies were measured.

The endpoint compares each treatment-period frequency with that participant’s baseline period. The posted percentages are 54.0 for the once-daily group, 41.0 for the twice-daily group, and 28.0 for placebo. These are group proportions meeting the threshold, not the average reduction in seizure frequency. They should not be read as the percentage reduction experienced by every participant or as a measure of seizure frequency for the group as a whole.

The endpoint is useful for asking whether the proportion reaching the stated response threshold differed between groups. It does not by itself describe other outcomes, the size of each participant’s change, or the balance of benefits and harms. No secondary outcome results are posted in the registry record.

7. Results Posted in the Registry

• ODG - Once Daily Group: 54.0% (27/50) versus 27.7% (13/47) on control; risk difference 26.3 percentage points (95% CI 7.5 to 45.2), risk ratio 1.95 (1.15–3.31); unadjusted two-proportion z-test p = 0.008.
• TDG - Twice Daily Group: 41.3% (19/46) versus 27.7% (13/47) on control; risk difference 13.6 percentage points (95% CI -5.5 to 32.8), risk ratio 1.49 (0.84–2.66); unadjusted two-proportion z-test p = 0.166.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a binary participant-level outcome. Each participant is classified according to whether seizure frequency was reduced by 50% or more during the treatment period compared with baseline. The group-level measure is the percentage meeting that definition. This approach turns each person’s result into a responder or non-responder classification; it does not retain the full amount of change for those who meet or fall short of the threshold.

What does the risk difference represent?

The registry reports risk differences for each BIA 2-093 group compared with placebo. In general, a risk difference is the difference between the two groups’ proportions with the outcome, expressed on the proportion scale. A positive estimate in these comparisons means that the responder proportion was larger in the BIA 2-093 group than in placebo. The posted estimates are 0.263 for once-daily dosing and 0.136 for twice-daily dosing. They are unadjusted comparisons; the registry does not report a separate adjusted estimate.

For context, the registry also reports risk ratios and odds ratios with confidence intervals. For once-daily BIA 2-093 versus placebo, the risk ratio is 1.952, with a confidence interval from 1.151 to 3.312; the odds ratio is 3.07, with a confidence interval from 1.316 to 7.163. For twice-daily BIA 2-093 versus placebo, the risk ratio is 1.493, with a confidence interval from 0.839 to 2.658; the odds ratio is 1.84, with a confidence interval from 0.773 to 4.383. These are relative measures, unlike the risk difference, and they answer a different scale of comparison. They should not be conflated with the endpoint percentages themselves.

How should the confidence intervals be understood?

General statistical explanation: a confidence interval is a range of values compatible with the observed data and the method used to calculate the interval. It conveys uncertainty around an estimate; it does not identify the exact underlying difference. Wider intervals indicate less precision than narrower intervals when the same quantity and method are being considered. For the once-daily comparison, the reported risk-difference interval is 0.075 to 0.452. For the twice-daily comparison, it is -0.055 to 0.328. The latter includes zero, while the former does not.

A confidence interval is not a range containing every participant’s outcome, nor does it show the spread of individual seizure-frequency changes. It pertains to the group comparison. Its interpretation also depends on the analysis method and assumptions. The registry record does not provide enough detail to assess all calculation choices or the treatment of missing endpoint data.

What does an unadjusted p-value say?

General statistical explanation: a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that null hypothesis and the analysis assumptions held. It is not the probability that the null hypothesis is true, and it is not a measure of clinical importance. The posted unadjusted p-values are 0.008 for the once-daily comparison and 0.166 for the twice-daily comparison. The registry does not report a multiplicity adjustment, so these values do not account for the fact that two BIA 2-093 groups are each compared with placebo.

The p-values and confidence intervals are related ways of describing the comparisons, but neither replaces attention to the estimated difference, its uncertainty, the endpoint definition, or the design. The registry’s unadjusted comparisons should not be interpreted as adjusted results or as proof that the same estimates would be obtained elsewhere.

Why can variation matter for a binary endpoint?

General statistical explanation: binary outcomes depend on the number of participants with and without the event and on the denominators used. When event proportions are estimated from finite groups, they vary from sample to sample. This is one reason interval estimates are useful. The reported arm-level event counts are 27 of 50 for once-daily BIA 2-093, 19 of 46 for twice-daily BIA 2-093, and 13 of 47 for placebo. The record does not describe how missing outcome observations, if any, were handled or whether all randomized participants were included in the endpoint analysis.

Over-dispersion is a general concern when observed variability exceeds what a chosen statistical model expects; it can make model-based uncertainty too small if it is ignored. The posted primary outcome is binary, and the registry does not describe a model-based analysis or report an over-dispersion assessment. There is therefore no basis in the registry record for asserting that over-dispersion was present or that a particular adjustment was used.

The trial’s randomization, masking, and parallel-group labels describe important design features, but do not specify the full statistical analysis plan. Details such as allocation implementation, analysis population, handling of missing data, model specification, and prespecified multiplicity procedures are not reported in the registry record. The posted comparison estimates can be described, but those unreported details limit how fully the analysis can be reconstructed.

10. Confidence Intervals and Precision

Precision is about how narrowly the data locate a group difference, not whether that difference is important in practice. The once-daily comparison has a reported risk difference of 0.263 and a 95% confidence interval from 0.075 to 0.452. The interval is entirely above zero, while still spanning a range of possible positive differences. The estimate should therefore be read together with the breadth of that interval, rather than as a single exact value.

For twice-daily BIA 2-093 versus placebo, the risk difference is 0.136, with a 95% confidence interval from -0.055 to 0.328. This interval includes zero and extends to values on either side of zero. The data are consequently compatible with a range of differences, including no difference, under the reported analysis. The point estimate alone would not convey this uncertainty.

General statistical explanation: confidence intervals tend to be more informative than a point estimate alone because they show a range of values consistent with the data and the method. They are not guarantees about the true value, and their coverage depends on the analysis assumptions. Nor do they describe the range of outcomes among individual participants. In this trial, the outcome denominators were 50, 46, and 47 for the three groups, and the posted uncertainty reflects those group comparisons.

The registry record does not report additional precision measures, adjusted intervals, or a direct interval for the difference between the two active dosing groups. The reported confidence intervals therefore support interpretation of each treatment-versus-placebo comparison only. They do not establish the precision of a once-daily versus twice-daily contrast.

11. 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 participants per arm for the two-group comparisons, not total events. They are illustrations based on the observed effect and are not the trial’s own design justification. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each planning scenario, the effect to detect is the stated fraction of the observed effect. The trial’s original sample-size assumptions are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the specified effect under those planning assumptions, if that effect and the assumptions held. The 90% power column gives the corresponding per-arm number for a design with a 90% chance. These are design calculations, not results observed in the trial. They do not say what power the completed trial had after observing its data; this is not observed power.

The rows describing 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect ask what planning sample size would be needed for progressively smaller target differences. Smaller effects are harder to distinguish from random variation, so they require more participants under the same assumptions. For example, the once-daily comparison’s illustration lists 54 participants per arm for 80% power at 100% of the observed effect, and 203 participants per arm at 50% of the observed effect. The twice-daily comparison lists 190 and 722 participants per arm for those corresponding scenarios.

These calculations are conditional on the stated assumptions and the observed effect used as their starting point. If the true effect, outcome rate, analysis method, allocation, or other design features differed, the required numbers could also differ. The registry record does not report that these planning calculations were used when the trial was designed.

12. 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 = 0.277; treatment proportion p₁ = 0.540; average p̄ = 0.408.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6951 = 1.3624.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6697 = 0.5636.
4. n = (1.3624 + 0.5636)² / (0.2634)² = 53.5, rounded up to 54 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 <- 0.276595744680851                          # control proportion
p1 <- 0.54                          # 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(2170077)
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 = 54 (80% power), 72 (90% power)
 75% of observed effect: n per arm = 94 (80% power), 125 (90% power)
 50% of observed effect: n per arm = 203 (80% power), 271 (90% power)
power.prop.test: 53.5 
Simulated power at n = 54 per arm: 0.800 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry posts two primary treatment-versus-placebo comparisons: once-daily BIA 2-093 versus placebo and twice-daily BIA 2-093 versus placebo. Their unadjusted p-values are 0.008 and 0.166, respectively. The registry record does not report whether these comparisons were prespecified as a family, whether a multiplicity procedure was planned, or whether any other multiplicity strategy was used. The p-values should therefore be understood as unadjusted values for the individual comparisons, not as values adjusted for considering both treatment groups.

Multiplicity matters in general because making several statistical comparisons creates more opportunities for an apparently notable result to arise by chance. A specified adjustment can alter how evidence is assessed across a group of tests. The record does not give a statistical analysis plan or an adjustment method, so no adjusted interpretation can be reconstructed from the registry information. There is also no reported direct comparison between once-daily and twice-daily dosing.

The registry lists one outcome measure and no posted secondary outcome results. It does not report an analysis population definition, an intention-to-treat designation, the handling of missing primary endpoint data, or participant-level exclusions from the outcome analysis. The arm denominators are 50, 46, and 47, but the record does not explain how those denominators relate to the overall enrollment of 144 or describe withdrawals and missing observations. It is not possible to infer from the posted denominators alone whether missing data occurred or how they were treated.

Interim analyses, stopping rules, and the trial’s original statistical analysis plan are not reported in the registry record. Their absence from the record should not be taken as evidence that such procedures were or were not used. Without those details, the posted estimates and unadjusted p-values can be described, but a full audit of prespecification, missing-data handling, and the overall testing strategy is not possible.

14. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: viral gastroenteritis (2.1%), bacterial infection of intestines (2.0%), ischaemic stroke (2.0%).

The registry reports serious adverse events by group, with the number at risk and the number affected. The once-daily BIA 2-093 group had 2 serious adverse events among 50 participants at risk; the twice-daily group had 0 among 46; and the placebo group had 1 among 47. These are reported event counts and denominators, not adjusted comparisons. The record does not post a statistical comparison of serious adverse events between groups.

Other adverse events are also counted separately: 18 participants in the once-daily group, 19 in the twice-daily group, and 21 in the placebo group were affected, using the same group denominators as the serious-event summaries. The registry lists viral gastroenteritis at 2.1%, bacterial infection of intestines at 2.0%, and ischaemic stroke at 2.0% among the top serious adverse-event terms. The record does not identify these terms as counts specific to any one treatment group.

Deaths are not reported in the registry record. Adverse-event counts alone do not establish causation or fully describe severity, timing, duration, or exposure. The posted information should be read as descriptive safety reporting rather than as a comparative safety analysis.

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

The primary outcome reports the proportion of participants who met a threshold of at least a 50% reduction in seizure frequency during treatment compared with baseline. In the registry, the responder percentages were 54.0 for once-daily BIA 2-093, 41.0 for twice-daily BIA 2-093, and 28.0 for placebo. These are the observed group percentages for the posted endpoint, not a statement that every participant in a group experienced the same change.

The unadjusted once-daily versus placebo estimate was 0.263, with a 95% confidence interval from 0.075 to 0.452 and an unadjusted p-value of 0.008. The twice-daily versus placebo estimate was 0.136, with a 95% confidence interval from -0.055 to 0.328 and an unadjusted p-value of 0.166. The intervals indicate different levels of uncertainty in the two comparisons; in particular, the twice-daily interval includes zero. These results describe the posted analyses and do not establish a direct difference between dosing schedules.

The endpoint does not describe the magnitude of change for each participant, seizure frequency after the stated assessment, or outcomes that were not posted. It also cannot by itself summarize the safety profile. The serious adverse-event counts were 2 in the once-daily group, 0 in the twice-daily group, and 1 in the placebo group, while deaths are not reported in the registry record. Those counts are descriptive and are not evidence of causation.

The findings apply to the trial groups and its recorded endpoint. The registry record does not provide enough detail to assess all aspects of the analysis, participant flow, missing data, or multiplicity. It also does not supply a direct comparison between once-daily and twice-daily dosing or secondary outcome results. The posted data should therefore be interpreted as estimates with uncertainty in a particular randomized, triple-masked phase 2 trial, rather than as definitive evidence for other populations or settings.

16. Why This Trial Matters Statistically

This study is statistically informative because it reports a binary responder endpoint for three randomized, parallel groups and provides separate comparisons of each active dosing schedule with placebo. The group percentages and arm denominators allow the outcome to be understood as counts of participants reaching a defined threshold, rather than as an abstract measure alone. The primary endpoint’s threshold is explicit: a 50% or greater reduction in seizure frequency compared with baseline during the treatment period.

The two placebo comparisons illustrate why point estimates should be considered alongside uncertainty. The once-daily comparison has a risk-difference confidence interval that excludes zero, while the twice-daily comparison’s interval includes zero and spans values on both sides. The p-values also differ, but they are unadjusted. This provides a practical example of how two estimates pointing in the same direction can carry different precision and how conclusions should not rely on the point estimates alone.

The record also makes visible the boundary between a reported endpoint result and a full statistical account. It lists a single primary outcome and no secondary outcome results, but does not report the analysis population, missing-data approach, multiplicity procedure, or original sample-size justification. Such information matters when assessing how securely the posted estimates reflect the planned analysis.

Finally, the planning illustrations demonstrate how assumptions about the effect to detect influence design requirements: smaller target effects call for more participants under the stated planning approach. These calculations are distinct from the completed trial’s observed results and do not establish its observed power. Taken together, the record supports careful interpretation of the endpoint estimates while showing why complete analysis details are essential for reproducibility.

17. 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 registry record gives limited information for reconstructing the trial’s conduct and analysis. It identifies the study as randomized, triple-masked, and parallel, but does not describe the randomization procedure, allocation concealment, or which roles were masked. The population is described as people with refractory partial seizures, including simple or complex partial seizures with or without secondary generalization, but detailed eligibility criteria and baseline participant characteristics are not reported in the registry record.

The outcome is defined by a responder threshold and a baseline-to-treatment comparison, with week 12 listed as the timeframe. The record does not describe how seizure frequency was collected, how baseline values were established, or how the endpoint was assessed when data were unavailable. The arm-specific denominators are reported, but participant flow, withdrawals, reasons for missing observations, and analysis-population rules are not.

Only one outcome measure is listed, and no secondary results are posted. The registry provides unadjusted comparisons for each active group versus placebo, but no direct comparison of the two dosing schedules. It does not report a multiplicity procedure, an adjusted analysis, interim analyses, or the trial’s original statistical analysis plan. The original sample-size justification is also not reported in the registry record; the listed planning illustrations use observed effects and are not a substitute for that justification.

Safety information is presented as group counts, without a posted statistical comparison or details about event timing, duration, severity beyond the serious-event classification, or attribution. Deaths are not reported in the registry record. These limitations restrict what can be concluded from the posted information and prevent a complete independent reconstruction of the study’s statistical and safety analyses.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Placebo-controlled Study to Investigate Safety and Efficacy of BIA 2-093 — NCT02170077.
• Registry record last updated 2017-08-18; status last verified 2017-07.
• Results first posted 2014-08-12.
• 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.

21. Record Summary

The registry record describes a completed, randomized, triple-masked, parallel-group phase 2 placebo-controlled study of BIA 2-093 for refractory partial seizures. Overall enrollment is listed as 144. The study began in 2002-04 and reached primary completion in 2002-11.

The primary outcome was the percentage of participants meeting a 50% or greater reduction in seizure frequency compared with baseline, with week 12 specified for the treatment period. The posted responder percentages were 54.0 for once-daily BIA 2-093, 41.0 for twice-daily BIA 2-093, and 28.0 for placebo. Unadjusted risk-difference comparisons with placebo were reported for both dosing groups.

The registry lists serious adverse events and other adverse events by group, but deaths are not reported in the registry record. It lists one outcome measure, no secondary outcome results, and does not report the original sample-size justification or several details needed to reconstruct the full analysis plan.