Safety and Efficacy of Gabapentin in Postherpetic Neuralgia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Gabapentin in Postherpetic Neuralgia (NCT00636636), a randomized phase 3 study. The primary outcome is Mean Change in Baseline Observation Carried Forward (BOCF) Average Daily Pain Score.

2. Clinical Question

The registry frames the study as an evaluation of safety and efficacy of extended-release gabapentin for postherpetic neuralgia. The statistical question can be stated more narrowly: whether the groups differed in their mean change in pain score by the final week of the efficacy treatment period. Because the endpoint is change from baseline, a negative change indicates a lower pain score relative to baseline. The posted group estimates are negative in both arms, and the mean change is more negative in the G-ER arm. The analysis should be read as a comparison of group averages, not as a description of how every participant's pain changed.

The record also includes patient and clinical global impressions of change and average daily sleep interference as secondary outcomes. Those measures broaden the set of outcomes described, but they do not replace the primary question. The registry record does not report the trial's original sample-size justification or its planned approach to interpreting multiple outcomes.

3. Trial Design

This was a Phase 3, randomized, double-blind, parallel-group, placebo-controlled study. It compared gabapentin extended-release tablets with placebo in people with postherpetic neuralgia. The sponsor is listed as Depomed. The registry record gives status as completed, trial start as 2008-03, and primary completion as 2009-08.

Enrollment was 452. For the primary outcome, the posted arm summaries report 220 participants in the G-ER group and 230 in the placebo group. The primary analysis population is described as ITT, BOCF, and the outcome was assessed through the final week of a 10-week efficacy treatment period. The record does not explain the relationship between overall enrollment and the numbers included in the primary outcome summaries.

Randomization indicates that treatment allocation was assigned by chance, while double masking indicates that the study was designed so that participants and at least one other party were unaware of assignment. The registry record does not identify all masked roles or describe allocation concealment, the randomization sequence, or procedures for maintaining masking. It identifies a parallel model, meaning the treatment groups were followed as separate arms rather than participants crossing between treatments.

The record names five outcome measures and five analyses, and results are posted for the primary outcome and three secondary outcomes. The original protocol's detailed statistical plan, the planned sample-size justification, and whether interim analyses were intended are not reported in the registry record. Those omissions limit what can be concluded about how the reported analyses compare with the trial's prespecified plans.

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: Study included screening visit, wash-out from all PHN medications as necessary, followed by a week of baseline period. Patients who successfully completed the baseline week, if still continues to meet entry criteria, then get randomized to Active or Placebo groups.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT, BOCF

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 Mean Change in Baseline Observation Carried Forward (BOCF) Average Daily Pain Score, measured over 10 weeks.

The primary endpoint is mean change from baseline to the final week of the 10-week efficacy treatment period in average daily pain score. Pain was rated on an 11-point numerical rating scale, where 0 means no pain and 10 means the worst possible pain. The outcome is reported as a least-squares mean change using baseline observation carried forward (BOCF), in the ITT population.

Change from baseline puts the endpoint on a scale of improvement or worsening relative to each participant's starting score. A negative change represents a lower average daily pain score at follow-up than at baseline. The reported least-squares means are -2.12 for G-ER and -1.63 for placebo. The sponsor's reported least-squares mean difference is -0.49, with a 95% confidence interval from -0.88 to -0.11.

General: a continuous endpoint such as a pain score retains information about the size of each recorded value, rather than reducing the outcome to a yes-or-no event. Averaging can describe the overall direction and magnitude of change between groups, while concealing variation among individuals. The least-squares mean is an adjusted model-based mean; the registry record identifies the method as ANCOVA for the between-group comparison, but does not give the full model specification.

This endpoint is used to address the trial's central comparative question with a defined measure and assessment time. The record does not report a threshold for a clinically meaningful difference, so statistical evidence about a group-average difference cannot by itself establish how important that difference is to participants.

7. Results Posted in the Registry

• G-ER: mean -2.12 versus -1.63 on control; difference -0.49 (95% CI -0.95 to -0.03), standardized effect d = -0.20; z-test p = 0.037.

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.

Patient Global Impression of Change (PGIC)

• 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:
  • G-ER: 42.7% (94/220) versus 33.5% (77/230) on control; risk difference 9.2 percentage points (95% CI 0.3 to 18.2), risk ratio 1.28 (1.01–1.62); unadjusted two-proportion z-test p = 0.043.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for G-ER than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Clinical Global Impression of Change (CGIC)

• 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:
  • G-ER: 44.1% (97/220) versus 33.9% (78/230) on control; risk difference 10.2 percentage points (95% CI 1.2 to 19.1), risk ratio 1.30 (1.03–1.64); unadjusted two-proportion z-test p = 0.027.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for G-ER than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Average Daily Sleep Interference Score

• 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:
  • G-ER: mean -2.30 versus -1.59 on control; difference -0.71 (95% CI -1.14 to -0.28), standardized effect d = -0.31; z-test p = 0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for G-ER than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The three secondary outcomes with posted results cover participant global impression of change, clinical global impression of change, and average daily sleep interference. Across these measures, the point estimates are directionally consistent with the primary result: the global-impression outcomes have more participants counted in the G-ER arm, and the sleep-interference score has a more negative mean change in that arm. This is a broad pattern across different types of outcomes, not a single combined measure.

These outcomes complement the primary pain score by reflecting global impressions and sleep interference. They may help describe whether the groups' patterns extend beyond the primary measure, but the registry record does not report a formal analysis combining them with the primary endpoint. Nor does it state that each secondary result was intended to confirm the primary finding independently.

The record lists five outcome measures and five analyses, while results are posted for three secondary outcomes. The two other outcome measures are not identified in the posted secondary results. The reported secondary tests are unadjusted in the record as presented, and a multiplicity correction or testing hierarchy is not reported. When several outcomes are examined, some apparently noteworthy findings can occur by chance even if there is no underlying difference for those outcomes. General: adjustment or a prespecified hierarchy can help control the interpretation of repeated testing, but whether either approach was used in this trial is not reported in the registry record.

For that reason, the secondary findings carry less weight than the primary comparison when describing the trial's central result. Their direction and estimates add context, but the nominal p-values should not be read as if each were the only hypothesis tested. The record does not provide enough information to assess whether the secondary measures were prespecified, how their results were prioritized, or whether analyses were selected for posting.

The posted pattern is therefore best treated as supportive descriptive context, with uncertainty about multiplicity and the broader set of outcomes. It does not, by itself, establish that all aspects of pain, sleep, or participants' overall experience changed in the same way.

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

• Patient Global Impression of Change (PGIC)
• Clinical Global Impression of Change (CGIC)
• Average Daily Sleep Interference Score

9. Statistical Interpretation

10. Statistical Methods Explained

The primary comparison is reported as an ANCOVA least-squares mean difference. ANCOVA is a regression-based method that can compare outcome means while adjusting for covariates included in the model. The registry record identifies the method but does not report the complete model specification, including which covariates were used. The reported estimate is -0.49, with a 95% confidence interval from -0.88 to -0.11 and p = 0.013.

What does a least-squares mean represent?

General: a least-squares mean is a model-based estimate of a group mean, often adjusted for variables in the analysis. It is not necessarily identical to a raw arithmetic average. The registry provides least-squares mean changes for each arm and a least-squares mean difference for the comparison, but does not give enough detail to reconstruct the adjustment model. The comparison should therefore be interpreted as the sponsor-reported ANCOVA result, not as a fully specified model that can be independently reproduced from the registry entry.

What does BOCF mean?

Baseline observation carried forward means that a participant's baseline observation is carried forward for the outcome when a follow-up value is unavailable, according to the stated approach. General: this is one method for handling missing follow-up measurements; it assigns a particular value rather than estimating an unobserved outcome from a model. The record identifies BOCF and the ITT population for the primary endpoint, but does not report the amount or pattern of missing data, the detailed implementation, or sensitivity analyses using alternative assumptions. Therefore, the effect of missing-data assumptions on the result cannot be assessed from the registry record alone.

How does the simple recalculation differ?

The supplementary unadjusted calculation uses posted arm-level summary statistics to estimate a simple difference between the groups. It does not reproduce the sponsor's ANCOVA or its possible covariate adjustments. For the primary endpoint, the recalculated mean difference is -0.49 and the unadjusted p-value is 0.037. Similar point estimates do not make the methods identical: the model, standard error, confidence interval, and p-value can depend on the analysis method. The registry does not identify a discrepancy flag, but the recalculation remains a secondary check rather than a substitute for the trial's reported analysis.

What can a p-value say?

General: a p-value measures how unusual results at least as incompatible with the null hypothesis as the observed result would be, if the null hypothesis and the test's assumptions held. It is not the probability that a treatment has no effect, and it does not quantify the size or importance of a difference. The primary p-value is interpreted together with the estimated difference and confidence interval, not on its own.

Why do the confidence intervals matter?

General: a confidence interval expresses statistical uncertainty around an estimate under the method's assumptions. A narrower interval means greater precision for that estimate; a wider interval leaves a broader range of values compatible with the analysis. The primary sponsor-reported interval ranges from -0.88 to -0.11. Its endpoints are not guaranteed bounds on the true effect, and the interval does not account for limitations not represented in the analysis, such as unreported model choices or missing-data assumptions.

What does over-dispersion mean in general?

General: over-dispersion occurs when observed variability is greater than a statistical model assumes. If a model underestimates variability, standard errors can be too small and confidence intervals or tests can appear more decisive than warranted. The primary outcome is analyzed as a continuous measure using ANCOVA, and the registry record does not report an over-dispersion assessment. This concept is not a specific finding about this trial.

The registry also reports Z tests for the binary global-impression outcomes and ANCOVA for sleep interference. The posted analyses give estimates, confidence intervals, and p-values for those endpoints. The record does not report whether these tests were part of a prespecified hierarchy, how multiplicity was handled, or whether any interim analyses were conducted. Those details matter when interpreting several outcomes together, but they should not be inferred from the methods labels alone.

11. Confidence Intervals and Precision

The sponsor-reported primary estimate is a least-squares mean difference of -0.49, with a 95% confidence interval from -0.88 to -0.11. This interval describes uncertainty in the estimated average difference under the ANCOVA analysis. It includes values representing a relatively smaller negative difference as well as a larger one; it does not establish that the true difference is exactly -0.49.

Because zero is outside the reported interval, the interval is consistent with a group difference in the negative direction under the stated analysis. That statistical observation does not establish clinical importance. The registry record does not give a threshold for a clinically meaningful change, and the confidence interval does not describe the range of responses among individual participants.

General: precision and statistical significance are related but not identical. An estimate can be statistically distinguishable from zero while still having uncertainty about its magnitude. Conversely, an interval can be broad enough to include both small and larger effects. The width of a confidence interval reflects the information and variability available to the analysis, together with its model assumptions.

The primary group summaries have 220 participants for G-ER and 230 for placebo. Those outcome counts provide context for the estimate, but do not by themselves describe missingness or explain the difference from overall enrollment of 452. The registry record does not report a confidence interval for every arm-level mean, nor does it provide the full covariance and model details needed to reconstruct the sponsor's ANCOVA interval.

The simple unadjusted recalculation has a mean difference of -0.49 and an unadjusted p-value of 0.037. Its confidence interval is not repeated because the sponsor's reported ANCOVA interval is the more relevant summary of the trial's own analysis. The two approaches can have different uncertainty estimates even when their rounded point estimates agree.

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 calculation. For the continuous primary outcome, the figures are participants per arm. The planning assumptions 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. The original trial's sample-size assumptions are not reported in the registry record.

The 80% power column gives the illustrative participants per arm for a design with an 80% chance of detecting the specified effect under those assumptions, if that effect and the assumptions were correct. The 90% power column describes the corresponding target for a 90% chance. These are design-planning quantities, not properties measured after the trial. This is not observed power.

Rows for 100%, 75%, and 50% of the observed effect illustrate how the planning requirement changes when the target effect is smaller than the observed estimate. The smaller-effect scenarios require more participants per arm in the posted planning figures. The observed estimate itself is uncertain, so using it as the basis for a future design can produce an optimistic planning target if the effect was overestimated in this trial.

The figures do not report how many participants were enrolled in this trial, and they do not establish that the original trial was adequately or inadequately sized. They answer a conditional planning question: how many participants per arm would be indicated by the listed assumptions and target effect. The registry record does not report the original sample-size justification or whether it used these assumptions.

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 = -0.490; pooled standard deviation σ = 2.488.
2. n = 2 × (1.96 + 0.842)² × 2.488² / -0.490² = 404.7, rounded up to 405 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 <- -0.49                      # observed difference in means
sigma <- 2.48790741357793                     # 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(636636)
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 = 405 (80% power), 542 (90% power)
 75% of observed effect: n per arm = 720 (80% power), 964 (90% power)
 50% of observed effect: n per arm = 1,619 (80% power), 2,168 (90% power)
power.t.test: 405.6 
Simulated power at n = 405 per arm: 0.802 (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 mean change. It gives a mean difference of -0.49, a 95% confidence interval from -0.88 to -0.11, and p = 0.013. This is the result to prioritize when describing the trial's own primary analysis.

The unadjusted recalculation from posted summary statistics also gives a rounded mean difference of -0.49, with an unadjusted p-value of 0.037. It uses an elementary comparison of the arm summaries rather than the sponsor's ANCOVA. The agreement in the rounded point estimate does not mean that the analyses are interchangeable; their uncertainty estimates and p-values arise from different methods.

The recalculation is a check based on summary statistics, not a reconstruction of the sponsor's adjusted model. The registry record does not report enough model detail to independently reproduce that ANCOVA. Thus, the sponsor's reported estimate, interval, and p-value should be distinguished from the unadjusted calculation rather than blended into a single analysis.

For the primary result, the registry does not flag a discrepancy between the reported analysis and recalculation. Even so, the unadjusted result is supplementary context. It should not displace the analysis reported for the trial, particularly when that analysis is model-based and the full model specification is unavailable.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies five outcome measures and five analyses, and posts results for the primary outcome and three secondary outcomes. The original statistical plan is not reported in enough detail to establish the intended hierarchy of endpoints, whether secondary outcomes were confirmatory, or how the family of tests was to be interpreted. A multiplicity adjustment is not reported in the registry record.

General: testing several outcomes increases the chance that at least one result will appear statistically notable by chance. This does not make every secondary finding false, but it means nominal p-values across a set of outcomes do not have the same interpretation as one isolated test. Without a stated correction or testing sequence, the secondary results are appropriately treated as less definitive than the primary analysis.

The primary endpoint is described as ITT, BOCF. The record does not report the number of participants with missing endpoint data, the reasons for missingness, or how many observations were carried forward. It also does not describe sensitivity analyses using alternative missing-data assumptions. BOCF is a specified approach, but its presence alone does not show how robust the finding is to other plausible assumptions about unavailable outcomes.

The record does not report the full ANCOVA covariate specification, details of allocation concealment, or whether interim analyses were conducted. It also does not provide the trial's original sample-size justification. These omissions make it difficult to compare the posted analyses with a complete prespecified plan or assess how each design choice might affect the inference.

These are limits of the registry entry, not proof that the trial lacked a protocol or that analyses were necessarily inappropriate. The record may not contain all documentation available elsewhere, but additional details cannot be inferred from the results listed. Interpretation should therefore distinguish what is explicitly described—the endpoint, population, BOCF approach, and reported analyses—from planning and analytic details that are not reported in the registry record.

16. 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 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: left arm fracture (0.5%), osteochondrosis (0.5%), pancoast tumor (0.5%), pancreatitis chronic (0.5%), cardiac failure congestion (0.4%).

For serious adverse events, the registry reports 4 participants affected among 221 at risk in the G-ER group and 6 among 231 at risk in the placebo group. The most frequent serious-event terms listed include Left arm fracture, Osteochondrosis, Pancoast tumor, and Pancreatitis chronic, each at 0.5%, and Cardiac failure congestion and Cellulitis, each at 0.4%. These term percentages are the registry's reported values.

The record also reports other adverse events for 92 participants among 221 at risk in the G-ER group and 63 among 231 at risk in the placebo group. The registry record does not give further detail about event severity categories, duration, relatedness to study treatment, or whether participants could contribute more than one event. Death counts are not reported in the registry record.

These counts describe events recorded in the study; they are not a complete account of each participant's clinical course and do not establish causation. Differences in counts should not be interpreted as evidence of a treatment effect without the relevant definitions, follow-up details, and formal safety analysis. The serious-event term percentages are small reported values, but the record does not provide the basis or precision for those percentages beyond the values listed.

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

The posted primary result describes a between-group difference in mean change in average daily pain score at the end of the 10-week efficacy treatment period. Both arms had negative mean changes, and the G-ER arm had the more negative least-squares mean change. The sponsor-reported ANCOVA estimate is -0.49, with a 95% confidence interval from -0.88 to -0.11. These values characterize a group-average comparison in the stated analysis population.

The result does not mean that every participant's score changed by the group mean, or that all participants experienced improvement. An average can combine a range of individual outcomes, and the registry record does not provide the distribution of participant-level changes. The confidence interval describes uncertainty in the estimated group difference under the analysis, not the range of individual responses.

The statistical result also does not, on its own, establish the practical importance of the difference. No threshold for a clinically meaningful change is reported in the registry record. The p-value is evidence about compatibility with a null hypothesis under the test assumptions; it is not the probability that the result is due to chance, nor a measure of how important the effect is.

The secondary outcomes offer additional context, but their interpretation is limited by the unreported multiplicity approach and incomplete posting of the full set of outcome results. The safety counts are descriptive and do not establish that an event was caused by treatment. The trial findings should be understood in relation to its population, endpoint, time frame, and analysis method rather than generalized beyond those details.

In short, the posted analysis supports a difference in the average primary endpoint between the arms under the reported ANCOVA model. It does not resolve uncertainty about the model details, missing-data assumptions, clinical importance, or how results might differ in other settings or populations. Those questions are not answered by the registry results alone.

18. Why This Trial Matters Statistically

This trial illustrates why a reported treatment comparison should be read through its endpoint, analysis population, and statistical method rather than through a p-value alone. The primary outcome is a continuous change score, summarized with least-squares means and compared using ANCOVA. The result includes both an effect estimate and a confidence interval, allowing readers to consider magnitude and precision alongside the test result.

It also illustrates the distinction between a sponsor-reported adjusted analysis and a simple recalculation from posted summaries. The recalculation is useful as a check, but it does not reproduce the trial's ANCOVA. A rounded point estimate can agree while the inferential methods differ. Keeping those analyses separate helps prevent a summary-statistic calculation from being mistaken for the original model.

The registry record names BOCF and ITT for the primary outcome, making the missing-data approach part of the interpretation. At the same time, details needed to evaluate its implementation and robustness are not reported. This demonstrates why identifying an analysis label is useful but not equivalent to having a complete statistical plan.

Finally, the three posted secondary outcomes show why multiple endpoints require careful framing. Their direction is broadly consistent with the primary result, but the registry does not report a multiplicity procedure or the complete outcome set. Results across endpoints can add context without carrying the same inferential weight as a clearly identified primary comparison.

These statistical distinctions matter because conclusions depend not only on the numerical estimate but also on what was measured, how it was analyzed, and what remains uncertain. The record supports a focused interpretation of the 10-week primary comparison while leaving important planning and analytic details unreported.

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 registry entry does not report the full ANCOVA model specification. It therefore is not possible to determine which covariates were included or to reproduce the sponsor's adjusted estimate from the information in the record. The unadjusted recalculation is not a substitute for those details.

The primary endpoint uses BOCF in the ITT population, but the record does not state how many outcome values were missing, why they were missing, or how results changed under alternative missing-data assumptions. The sensitivity of the primary estimate to these choices cannot be assessed from the registry record.

The registry does not report the original sample-size justification, the planned multiplicity strategy, or whether interim analyses occurred. It lists five outcome measures and five analyses, while only three secondary outcomes are posted. This leaves uncertainty about how the full set of outcomes and tests was intended to be interpreted.

The outcome counts for the primary analysis are 220 in the G-ER group and 230 in the placebo group, while overall enrollment is reported as 452. The registry record does not explain this difference, so its reason should not be inferred. The record also does not provide the participant-level distribution of changes or a threshold for clinically meaningful change.

Safety reporting is limited in the information listed: death counts are not reported in the registry record, and further detail on event relatedness and the structure of the other adverse-event counts is not given. These limits restrict how fully safety and generalizability can be assessed using this registry record alone.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy of Gabapentin in Postherpetic Neuralgia — NCT00636636.
• Registry record last updated 2012-02-22; status last verified 2012-02.
• Results first posted 2012-01-13.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed Phase 3, randomized, double-blind, parallel-group, placebo-controlled study sponsored by Depomed. It enrolled 452 participants and compared gabapentin extended-release tablets with placebo for postherpetic neuralgia.

The primary outcome was ITT, BOCF mean change in average daily pain score at 10 weeks. The sponsor-reported ANCOVA difference was -0.49, with a 95% confidence interval from -0.88 to -0.11 and p = 0.013. Three secondary outcomes have posted results, alongside descriptive adverse-event counts.

The record does not report the original sample-size justification, complete model details, multiplicity approach, or missing-data sensitivity analyses. These omissions limit assessment beyond the stated outcomes and analyses.