Adefovir Dipivoxil In Compensated Chronic Hepatitis B Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Adefovir Dipivoxil In Compensated Chronic Hepatitis B Patients (NCT00316719), a randomized phase 3 study. The primary outcome is Mean Change From Baseline in Hepatitis B Virus (HBV) DNA at Week 52.

2. Clinical Question

3. Trial Design

This was a randomized, quadruple-masked, parallel-group phase III study of adefovir dipivoxil and lamivudine in patients with compensated chronic hepatitis B. The registry lists GlaxoSmithKline as sponsor and records the study as completed. The recorded start is 2006-01, and primary completion is recorded as 2008-01. Enrollment was 105.

The study compared ADV 10 mg with LAM 100 mg. The brief summary describes an efficacy and safety comparison over 52-week periods. The primary outcome was mean change in HBV DNA from baseline to Week 52, with the primary analysis specified for the Per Protocol Set. The reported group counts for that outcome are 50 for ADV and 52 for LAM. These are outcome-analysis counts, not a substitute for the registry's enrollment figure.

Randomized allocation supports a design intended to compare groups formed by chance, while a parallel design means participants were assigned to one treatment group rather than crossing between treatments. Quadruple masking indicates masking at four roles or levels, but the registry record does not identify those roles. The exact randomization procedure, allocation concealment process, and operational details of masking are not reported in the registry record.

The registry lists secondary outcomes concerning HBV DNA loss, hepatitis B e antigen loss and seroconversion, and hepatitis B s antigen loss and seroconversion, including time-to-onset outcomes. It records 13 outcome measures and one analysis. There is also a posted Week 52 ALT outcome. The available details do not establish a complete analysis plan for all listed outcomes.

The trial's original sample-size justification, assumptions for its design, and any planned interim analyses are not reported in the registry record. The planning illustrations accompanying the posted results are based on the observed effect and should not be mistaken for the rationale used to choose the original enrollment.

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:

Analysis population for the primary outcome, as defined in the registry: Per Protocol Set (PPS): participants in the Full Analysis Set (all subjects who entered the study, received at least one dose of investigational product, and had at least one efficacy assessment after the treatment initiation) population with no major protocol violations

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 From Baseline in Hepatitis B Virus (HBV) DNA at Week 52, measured over Baseline and Week 52.

The primary endpoint is the mean change from baseline in serum HBV DNA at Week 52, measured in log10 copies/mL. The change is defined as the difference between the HBV DNA copy numbers in serum collected by blood draw at baseline and at Week 52. It is a continuous measure: each participant contributes a numerical change, and the group result is described by a mean and standard deviation.

Change from baseline focuses the comparison on how the measurement shifted over the study period rather than only on the Week 52 measurement. A negative mean change represents a lower Week 52 value than baseline on the stated scale; it does not, by itself, establish a between-group difference. The relevant comparison is the contrast between the group means, interpreted alongside its uncertainty and the analysis specified for the trial.

HBV DNA is used as a laboratory measure of viral material in serum. The registry separately lists HBV DNA loss below the stated threshold among secondary outcomes, but that is a different endpoint from the primary continuous change measure. The primary endpoint's population is also specific: it is the Per Protocol Set, which excludes participants with major protocol violations from the Full Analysis Set as defined in the registry. How the endpoint was handled for participants without a Week 52 measurement is not reported in the registry record.

7. Results Posted in the Registry

• Lamivudine (LAM): mean -3.40 versus -3.69 on control; difference 0.29 (95% CI -0.32 to 0.90), standardized effect d = 0.18; z-test p = 0.350.

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 Alanine Aminotransferase (ALT) Level 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 deviations.
• Result:
  • Lamivudine (LAM): mean 33.00 versus 32.30 on control; difference 0.70 (95% CI -8.32 to 9.72), standardized effect d = 0.03; z-test p = 0.879.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Lamivudine (LAM) and the control. The interval gives the range of effects that remain compatible with the data.

The listed secondary outcomes broaden the assessment beyond mean change in HBV DNA. They include threshold-based HBV DNA loss, hepatitis B e antigen loss and seroconversion, and hepatitis B s antigen loss and seroconversion, with some outcomes expressed as time to onset. The registry also includes a Week 52 ALT measure. These endpoints address different features of the study and are not interchangeable with the continuous primary endpoint.

The available numerical secondary results are insufficient to characterize a coherent pattern across all listed outcomes. In particular, most outcome-specific estimates and analyses are not reported in the registry record. The secondary results therefore cannot be used to fill in the uncertainty surrounding the primary comparison or to determine whether findings were consistent across biological measures. Any broader claim about agreement or disagreement among secondary outcomes would go beyond the posted information.

Secondary endpoints often involve multiple comparisons. When many outcomes are examined, the chance of at least one apparently notable result can increase unless the analysis plan addresses multiplicity. The registry does not report whether an adjustment or hierarchical testing strategy was used. Unadjusted secondary results, where available, should consequently be treated as less definitive than a prespecified primary analysis. They may be useful for context or hypothesis generation, but should not be treated as independent confirmation without details on the analysis plan and multiplicity handling.

The primary outcome and secondary outcomes also differ in scale and interpretation. A continuous change measure uses group means and standard deviations; a loss or seroconversion endpoint is categorical, while time-to-onset involves event timing. Their statistical methods and uncertainty summaries need not match. The registry record does not provide enough detail to assess all those methods or their relationship to the sponsor's primary ANCOVA analysis.

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

• Percentage of Participants With HBV DNA Loss (<400 Copies/mL) at Week 52
• Time to Onset of HBV DNA Loss (< 400 Copies/mL)
• Percentage of Participants With Hepatitis B e Antigen (HBeAg) Loss at Week 52
• Percentage of Participants With Hepatitis B e Antigen/Antibody (HBeAg/Ab) Seroconversion at Week 52
• Time to Onset of HBeAg Loss
• Time to Onset of HBeAg/Ab Seroconversion
• Percentage of Participants With Hepatitis B s Antigen (HBsAg) Loss at Week 52
• Percentage of Participants With Hepatitis B s Antigen/ Antibody (HBsAg/Ab) Seroconversion at Week 52

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean change represent?

A continuous endpoint can take many numerical values rather than placing participants into only a small number of categories. For this trial, the participant-level measure is the change in serum HBV DNA from baseline to Week 52. The group mean is the average of those changes, and the standard deviation describes how dispersed the participant-level changes are around that average. A mean is useful for comparing groups on the scale of the endpoint, but it can conceal the spread and shape of individual results.

The posted mean changes are -3.69 for ADV and -3.4 for LAM. The unadjusted recalculation expresses the LAM mean relative to the ADV mean as a difference of 0.29. A positive or negative sign depends on which group is subtracted from which; it does not mean that a treatment is beneficial or harmful by itself. The clinical meaning also depends on the endpoint, its measurement scale, and the analysis population.

What is an unadjusted recalculation?

The recalculation uses the posted group summary statistics rather than the individual participant records or the sponsor's adjusted model. It estimates a simple difference between the group means and gives an uncertainty interval and unadjusted p-value. Such a calculation can help readers understand the information contained in the summary data, but it cannot reproduce an ANCOVA analysis that adjusts for baseline or other covariates. The registry record does not give the covariates or model specifications used for the sponsor analysis.

The recalculated mean difference is 0.29, with a confidence interval from -0.319 to 0.899 and p = 0.35. Since this is an unadjusted recalculation and the record flags a discrepancy with the sponsor analysis, it is not the preferred result for representing the trial's own analysis. The sponsor's reported analysis takes precedence.

What does ANCOVA do?

ANCOVA, or analysis of covariance, compares an outcome between groups while accounting for one or more covariates. In a baseline-adjusted analysis, baseline measurement is commonly used to improve the comparison by accounting statistically for where participants started. This is a general description of the method, not a claim about the exact covariates in this trial: the adjustment variables and model details are not reported in the registry record.

The sponsor reports an ANCOVA net mean difference of -0.33, a 95% confidence interval from -0.94 to 0.28, and p < 0.001. The analysis is labeled non-inferiority or equivalence. In those frameworks, the question may be whether a difference is within a prespecified acceptable margin, rather than whether the difference equals zero. The margin and test direction are not reported in the registry record. As a result, the estimate, interval, and p-value should be read as the posted sponsor output, not reinterpreted as a standard superiority test.

Why do confidence intervals and p-values answer different questions?

A confidence interval gives a range of values compatible with the data under the assumptions of the analysis. Its width communicates precision: wider intervals indicate more uncertainty about the effect's magnitude. A p-value measures how unusual the data would be under a specified null hypothesis and statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or importance of an effect.

For the simple recalculation, the interval crosses zero and the unadjusted p-value is 0.35. For the sponsor's analysis, the reported interval also crosses zero, but the p-value is below 0.001. The different analysis and the non-inferiority/equivalence classification matter: the p-value can relate to a margin-based test rather than a test of zero difference. Without the margin and test specification, the apparent contrast cannot be resolved into a conclusion about the formal criterion.

Why can variation affect a comparison?

The standard deviation describes variability across participant measurements within a group. More variation can make a group mean less precise, all else being equal, because individual measurements provide a less uniform signal about the group average. The posted standard deviations are 1.169 for ADV and 1.896 for LAM. These values describe dispersion; they are not uncertainty intervals for the group means and should not be confused with the confidence interval for a between-group estimate.

When variation differs between groups, an analysis needs to account appropriately for the observed data structure. The recalculation uses the posted summary information, while the sponsor's ANCOVA may use additional data and covariate adjustment. The registry does not report diagnostic checks, model assumptions, or robust variance procedures, so no claim about those details is warranted.

What does the Per Protocol Set imply?

The primary endpoint is analyzed in the Per Protocol Set. The record defines this as participants in the Full Analysis Set who had no major protocol violations. The Full Analysis Set definition requires entry into the study, at least one dose, and at least one efficacy assessment after treatment initiation. A per-protocol analysis describes results among participants meeting those criteria; it is not necessarily the same estimand as an analysis of every randomized participant. The registry record does not provide a separate primary result for all randomized participants or explain how protocol violations were classified.

Because inclusion in a per-protocol population can depend on events after randomization, the population definition is important when judging how broadly an estimate applies. The recorded primary outcome counts are 50 for ADV and 52 for LAM. Reasons that particular participants were not included in those outcome counts, and the handling of missing Week 52 measurements, are not reported in the registry record.

What is the role of statistical significance?

Statistical significance is tied to a specified hypothesis, analysis, and threshold. It is not a direct measure of clinical importance, and a non-significant result does not prove there is no difference. Likewise, a small p-value does not reveal whether an effect is large or relevant. In this study, the sponsor's p-value belongs to an analysis labeled non-inferiority or equivalence, while the recalculated p-value is unadjusted. Their distinct purposes prevent treating them as two interchangeable tests of the same question.

General statistical principles can clarify how these methods work, but the registry record does not supply enough detail to reconstruct the sponsor's model or formal decision rule. Interpretation should therefore preserve the distinction between the posted sponsor analysis, the unadjusted recalculation, and information that is not reported.

11. Confidence Intervals and Precision

Precision concerns how narrowly the data estimate a quantity, rather than whether an estimate is important in clinical terms. A confidence interval gives a range of effect values compatible with the data and analysis assumptions. A narrow interval indicates greater precision; a wide interval leaves more uncertainty about the magnitude. The interval does not assign probability to each possible true value, and it is not a range containing every participant's result.

For the unadjusted comparison, the estimated mean difference is 0.29, with a confidence interval from -0.319 to 0.899. That interval includes zero and allows for differences on either side of zero. The posted standard deviations, 1.169 for ADV and 1.896 for LAM, describe participant-level variability, not the precision interval itself. The unadjusted p-value is 0.35.

The sponsor's adjusted analysis reports a 95% confidence interval from -0.94 to 0.28 around a net mean difference of -0.33. It too includes zero, but its p-value is reported as p < 0.001 because the analysis is classified as non-inferiority or equivalence. That label makes the missing margin especially important: a margin-based test and a test of zero difference can use the same estimate and interval yet address different hypotheses. The margin is not reported in the registry record.

The recalculated interval and the sponsor interval should not be compared as if they came from the same model and estimand. One is a simple unadjusted calculation from posted summaries; the other is the sponsor's ANCOVA output. Covariate adjustment and analysis specifications can change an estimate and its uncertainty. The sponsor analysis takes precedence, while the available record does not permit a full assessment of its precision or decision criterion.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed effect, not the trial's original design assumptions. For this continuous primary outcome, the listed figures are participants per arm, not total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The trial's own original sample-size assumptions are not reported in the registry record.

The columns for 80% and 90% power describe planning targets. In plain terms, a design with 80% power has an 80% chance of detecting the specified effect under the assumed model and threshold if that effect is the true one; a 90% target sets a higher chance under those same assumptions. These are prospective planning concepts, not descriptions of what happened in the completed trial.

The display considers planning for 100%, 75%, and 50% of the observed effect. As the target effect becomes smaller, more participants per arm are needed under the stated assumptions. For example, planning for 100% of the observed effect gives 468 participants per arm for 80% power and 626 for 90% power. Planning for 75% gives 831 and 1,112 participants per arm, respectively; planning for 50% gives 1,869 and 2,502. These values are illustrations tied to the observed effect and assumptions, not estimates of the trial's required enrollment under an original protocol.

This is not observed power. Observed power is not a useful substitute for reading the estimated effect and its confidence interval after a study has been completed. The displayed calculations instead show how assumed effect size and desired planning power relate to a hypothetical participant count.

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 = 0.290; pooled standard deviation σ = 1.582.
2. n = 2 × (1.96 + 0.842)² × 1.582² / 0.290² = 467.2, rounded up to 468 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 <- 0.29                      # observed difference in means
sigma <- 1.58207807961554                     # 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(316719)
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 = 468 (80% power), 626 (90% power)
 75% of observed effect: n per arm = 831 (80% power), 1,112 (90% power)
 50% of observed effect: n per arm = 1,869 (80% power), 2,502 (90% power)
power.t.test: 468.2 
Simulated power at n = 468 per arm: 0.804 (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 registry reports the sponsor's ANCOVA result as a net mean difference of -0.33, with a 95% confidence interval from -0.94 to 0.28 and p < 0.001. It classifies the analysis as non-inferiority or equivalence. The sponsor's own analysis takes precedence over the simple unadjusted recalculation, which uses the posted group means and standard deviations and gives a mean difference of 0.29, an interval from -0.319 to 0.899, and p = 0.35.

The two results differ in their estimates and methods. The recalculation is unadjusted; the sponsor reports ANCOVA. The record does not give enough detail about the covariates, estimand, direction convention, or model to reconcile the numerical difference. The displayed order of groups may also affect the sign, so the estimates should not be compared without attention to the contrast definition.

The sponsor result's p-value should be read in the context of its non-inferiority/equivalence classification, not assumed to be a standard test of a zero difference. The margin and formal decision rule are not reported in the registry record. Thus, the sponsor analysis is the principal reported analysis, but the registry entries do not permit an independent assessment of whether its stated criterion was satisfied.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study tests several outcomes, contrasts, time points, or subgroups. The more hypotheses examined, the greater the possibility of an apparently unusual result arising by chance if each is assessed separately at the same threshold. Common strategies include adjusting significance thresholds, controlling a family-wise error rate, or specifying a hierarchy in which tests proceed in a planned order. These are general methods, not claims about this trial's conduct.

The registry lists a primary outcome and multiple secondary outcomes, including several related measures of HBV DNA and hepatitis B antigen loss or seroconversion. The record does not report a multiplicity adjustment, a hierarchical testing procedure, or which outcomes were formally confirmatory. That omission makes unadjusted secondary findings less persuasive as standalone evidence. It also prevents judging whether the sponsor's p-value belongs to a broader testing sequence.

The primary analysis is defined for the Per Protocol Set, with the Full Analysis Set and eligibility conditions described in the record. The registry does not report a separate primary estimate in an intention-to-treat or all-randomized population. It also does not state how missing Week 52 observations were handled, whether any imputation was used, how participants with incomplete measurements were treated, or whether sensitivity analyses assessed missing-data assumptions. Those details matter because the primary result depends on observed change between baseline and Week 52.

The record does not report the original sample-size calculation, its target effect, planned power, or the assumptions behind the enrollment target. Although planning illustrations based on the observed effect are available, they are not evidence of what was specified before the trial began. The registry record also does not report interim analyses, stopping rules, or the timing of any review of accumulating data.

These missing details do not establish that a particular method was or was not used. They mark limits on what can be evaluated from the registry record. Interpretation can describe the posted population, estimates, and analysis label, but cannot reconstruct a full statistical analysis plan or assess every potential source of bias.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: hepatitis (1.9%), pneumonia bacterial (1.9%), lymphoma (1.9%), urinary retention (1.9%).

The safety record lists serious adverse events for 52 participants at risk in the ADV group and 53 at risk in the LAM group. The number with a serious adverse event is listed as 0 for ADV and 4 for LAM. These are the registry's counts; they should be read alongside their recorded at-risk denominators and not converted into percentages without an explicit calculation.

The record lists Hepatitis, Pneumonia bacterial, Lymphoma, and Urinary retention among the top serious terms, each with a reported value of 1.9%. The listed terms are not assigned to a particular treatment group in the information available, so those values do not establish a group-specific pattern. Death counts are not reported in the registry record.

The other adverse-event counts are 39 among 52 at risk for ADV and 47 among 53 at risk for LAM. These are separate from the serious-event counts, and the record does not provide enough detail to characterize event severity, duration, relatedness, or participant-level overlap. Safety counts are descriptive and do not by themselves establish comparative risk or causation.

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

The primary outcome compares mean change in serum HBV DNA from baseline to Week 52 in the Per Protocol Set. The posted group means are -3.69 log10 copies/mL for ADV and -3.4 for LAM. The unadjusted recalculation estimates a difference of 0.29, but its confidence interval spans zero and its p-value is 0.35. This result does not establish that the group means are identical; it reflects uncertainty in a simple comparison based on summary statistics.

The registry's sponsor analysis is an ANCOVA result labeled non-inferiority or equivalence. It reports a net mean difference of -0.33, a 95% confidence interval from -0.94 to 0.28, and p < 0.001. The sponsor's analysis takes precedence over the unadjusted recalculation, but the registry record does not report the margin or decision rule needed to independently interpret the non-inferiority/equivalence test. The p-value cannot be treated as a conventional test of a zero difference without that context.

The findings do not by themselves establish a treatment effect across outcomes, patient groups, or time periods. The primary endpoint is a laboratory measure at Week 52, and the analysis is limited to the stated Per Protocol Set. Most secondary endpoint results are not reported in enough detail to assess an overall pattern. The posted safety counts are descriptive and do not establish causation or comparative risk.

The registry record does not report key details needed for a full audit of the statistical plan, including missing-data handling, multiplicity control, the sponsor model's covariates, and the formal non-inferiority/equivalence margin. Accordingly, the posted results can be described and their uncertainty discussed, but the available information does not support reconstructing every design decision or making broader claims beyond the recorded analyses.

18. Why This Trial Matters Statistically

This study illustrates why a posted treatment comparison can have more than one statistical result attached to it. A simple unadjusted calculation from group summaries produces a different estimate and p-value from the sponsor's ANCOVA analysis. The distinction is not merely technical: the analyses may target different contrasts, use different information, and answer different hypotheses. The registry classifies the sponsor analysis as non-inferiority or equivalence, making the unreported margin central to its interpretation.

It also demonstrates why a confidence interval and p-value must be interpreted together with the analysis design. Both posted intervals span zero, while the sponsor p-value is reported 0.001. Rather than treating this as a contradiction to resolve by selecting one statistic, readers need the test hypothesis and margin. The record does not provide those details, so the correct statistical reading preserves that uncertainty.

The Per Protocol Set is another important feature. The primary endpoint is not described as an analysis of all randomized participants; it is defined using treatment exposure, post-initiation efficacy assessment, and absence of major protocol violations. Knowing the analysis population changes how one understands the scope of the estimate and what additional results would help assess robustness.

Finally, multiple listed secondary outcomes and incomplete reporting of their numerical results highlight the need to distinguish a prespecified primary analysis from exploratory or unadjusted findings. Statistical interpretation depends not only on the observed estimate, but also on the population, estimand, model, missing-data rules, multiplicity plan, and design assumptions. Where these details are absent, the limits of inference are part of the result's meaning.

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 most consequential limitation for interpreting the primary comparison is the difference between the unadjusted recalculation and the sponsor's reported ANCOVA result. The registry labels a discrepancy, and the sponsor analysis takes precedence, but the record does not provide model covariates, the exact estimand, or enough detail to reconcile the estimates. The non-inferiority/equivalence margin and formal decision rule are also not reported, limiting independent interpretation of the sponsor p-value.

The primary analysis is in the Per Protocol Set, not a separately reported all-randomized population. The record defines that set but does not explain how major protocol violations were determined or give reasons for participants not represented in the primary outcome counts. Missing Week 52 data, imputation, and sensitivity analyses are not reported in the registry record.

The registry contains a broad list of secondary outcomes, but most numerical secondary results and their analysis details are not available in the record described. It also does not report a multiplicity strategy. These gaps prevent a full assessment of consistency across endpoints or the evidential status of secondary findings.

Several design and reporting details are absent, including the original sample-size justification, interim-analysis plans, allocation concealment procedure, the roles covered by quadruple masking, and the full statistical analysis plan. The posted planning illustrations use observed effects and cannot replace original design assumptions. Safety reporting is also limited: the serious-event terms are not linked to a treatment group in the posted information, deaths are not reported, and further details such as relatedness and duration are unavailable. These limitations restrict the conclusions that can be drawn from the registry record without implying that unreported procedures were not performed.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Adefovir Dipivoxil In Compensated Chronic Hepatitis B Patients — NCT00316719.
• Registry record last updated 2009-10-06; status last verified 2009-10.
• Results first posted 2009-08-03.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The registry record describes a completed randomized, quadruple-masked, parallel-group phase III study sponsored by GlaxoSmithKline. It enrolled 105 participants with compensated chronic hepatitis B and compared ADV 10 mg with LAM 100 mg over 52-week periods. The primary endpoint was mean change in serum HBV DNA from baseline to Week 52 in the Per Protocol Set.

The sponsor reported an ANCOVA net mean difference of -0.33, a 95% confidence interval from -0.94 to 0.28, and p < 0.001, with the analysis classified as non-inferiority or equivalence. An unadjusted recalculation gives a mean difference of 0.29, an interval from -0.319 to 0.899, and p = 0.35; the sponsor analysis takes precedence. The non-inferiority/equivalence margin and several statistical-plan details are not reported in the registry record, limiting independent evaluation of the formal conclusion.