Eltrombopag To Reduce The Need For Platelet Transfusion In Subjects With Chronic Liver Disease And Thrombocytopenia Undergoing Elective Invasive Procedures: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Eltrombopag To Reduce The Need For Platelet Transfusion In Subjects With Chronic Liver Disease And Thrombocytopenia Undergoing Elective Invasive Procedures (NCT00678587), a randomized phase 3 study. The primary outcome is Number of Participants With Chronic Liver Disease and Thrombocytopenia (Platelets <50 Gi/L) Who do Not Require a Platelet Transfusion Prior to, During, and up to 7 Days Following Elective Invasive Procedures.

2. Clinical Question

The trial focused on a procedure-related clinical decision: whether a participant required a platelet transfusion in the period surrounding an elective invasive procedure. Transfusion decisions were tied to platelet counts and, for participants in an intermediate range, to the judgment of the investigator and the physician performing the procedure. The registry describes the endpoint as the number of participants who did not require transfusion, rather than as a continuous platelet-count outcome.

The primary comparison therefore concerns the proportion of randomized participants meeting the no-transfusion outcome in each group. Other outcomes included bleeding events, platelet counts, transfusion counts, adverse events, and medical resource use. The registry record does not report additional eligibility details beyond the chronic liver disease and thrombocytopenia population described in the trial record.

3. Trial Design

This was a Phase 3, randomized, double-blind, placebo-controlled, multicentre study with a parallel-group design. The registry lists triple masking. The two groups received once-daily oral eltrombopag 75 mg or once-daily oral placebo. The registry reports enrollment of 292 participants, with 147 assigned to placebo and 145 assigned to eltrombopag in the primary endpoint results.

The primary outcome was assessed in the Intent-to-Treat population, defined in the registry as all participants randomized to treatment. Its timeframe covered the period before, during, and up to seven days following the elective procedure, described as Study Days 16–19 and as covering Baseline to Day 26. The record lists 16 outcome measures and 15 posted secondary outcomes. These included bleeding, platelet measures, transfusions, safety outcomes, and resource use.

The registry lists the study status as terminated. The reason for termination is not reported in the registry record. The trial start is recorded as 2008-06 and primary completion as 2009-10. Details of the original sample-size justification, planned interim analyses, and the full statistical analysis plan are not reported in the registry record. Those omissions limit what can be said about the intended operating characteristics and analysis decisions before results were known.

The group sizes in the primary endpoint analysis are close, but the registry does not state the original allocation ratio or give a detailed account of enrollment and follow-up by randomized group beyond the posted results. The protocol’s intended allocation, recruitment targets, and procedures for handling participants with incomplete endpoint information are not reported in the registry record.

4. Participant Flow and Accounting

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

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) Population: all participants who were randomized to treatment

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 Number of Participants With Chronic Liver Disease and Thrombocytopenia (Platelets <50 Gi/L) Who do Not Require a Platelet Transfusion Prior to, During, and up to 7 Days Following Elective Invasive Procedures, measured over Prior to, during, and up to seven days following elective invasive procedures (Study Days 16-19); therefore, this covers a time period from Baseline to Day 26.

The primary endpoint counts participants with chronic liver disease and thrombocytopenia who did not require a platelet transfusion before, during, and up to seven days after an elective invasive procedure. It is a binary outcome at the participant level: each participant is counted according to whether a transfusion was avoided during the stated period. The registry describes the analysis population as all randomized participants.

The transfusion criteria incorporated platelet counts and clinical judgment. A platelet transfusion was given if the platelet count was <50 Gi/L before the procedure. It was not given if the platelet count was >80 Gi/L, consistent with the endpoint’s stated success criterion. For counts between 50 Gi/L and 80 Gi/L, transfusion was at the discretion of the investigator and the physician performing the procedure.

This endpoint is relevant to the trial’s stated aim because it directly records whether a transfusion was needed around the procedure. It is not, by itself, a complete measure of bleeding, all procedural outcomes, or longer-term clinical benefit. The registry separately lists bleeding events and other outcomes, which address different aspects of the clinical experience. The primary endpoint timeframe extends from Baseline to Day 26, while the stated transfusion-avoidance window is tied to the procedure and the seven days following it.

7. Results Posted in the Registry

• Eltrombopag 75 mg: 71.7% (104/145) versus 19.0% (28/147) on control; risk difference 52.7 percentage points (95% CI 43.0 to 62.4), risk ratio 3.77 (2.66–5.34); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants With a World Health Organization (WHO) Bleeding Score >=2 During and up to 7 Days Following Elective Invasive Procedures

• 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:
  • Eltrombopag 75 mg: 17.2% (25/145) versus 23.1% (34/147) on control; risk difference -5.9 percentage points (95% CI -15.1 to 3.3), risk ratio 0.75 (0.47–1.18); unadjusted two-proportion z-test p = 0.210.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Eltrombopag 75 mg and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With Renal Function Abnormality

• 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:
  • Eltrombopag 75 mg: 19.6% (28/143) versus 18.6% (27/145) on control; risk difference 1.0 percentage points (95% CI -8.1 to 10.0), risk ratio 1.05 (0.65–1.69); unadjusted two-proportion z-test p = 0.836.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Eltrombopag 75 mg and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With a Clinically Significant Change in Electrocardiogram (ECG) Results

• 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:
  • Eltrombopag 75 mg: 0.8% (1/130) versus 0.0% (0/128) on control; risk difference 0.8 percentage points (95% CI -0.7 to 2.3); unadjusted two-proportion z-test p = 0.320.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Eltrombopag 75 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several distinct domains: bleeding during the procedure period, platelet measures, transfusions, adverse events, renal function, vision, and ECG findings. These outcomes complement the primary transfusion-avoidance endpoint but do not measure the same thing. A change in the primary endpoint does not by itself establish a corresponding difference in bleeding, organ function, or other safety outcomes.

The posted secondary results include comparisons with uncertainty that spans zero, including the bleeding and renal function abnormality estimates. The clinically significant ECG outcome is sparse, with zero participants recorded in the placebo group and one in the eltrombopag group. These results are not a uniform pattern of statistically clear differences across outcomes, and the precision varies with the outcome and event counts. The detailed results should be interpreted on their own definitions and timeframes rather than treated as direct confirmation or refutation of the primary endpoint.

Secondary outcomes commonly involve multiple comparisons. When many outcomes are tested, the chance of obtaining at least one apparently notable result by chance can increase. The registry record does not report a multiplicity adjustment or a hierarchy that would establish which secondary analyses were confirmatory. Unadjusted secondary results therefore carry less weight as stand-alone evidence than a prespecified primary analysis, especially when the full analysis plan is unavailable. This does not make secondary outcomes uninformative; it means they are best treated as supporting or exploratory evidence unless the record establishes otherwise.

The registry lists 15 posted secondary outcomes and 16 outcome measures overall. It does not report a complete analysis plan for those measures. Differences in outcome definitions, observation windows, and available denominators also matter when comparing them. The secondary results are most useful as context for the primary finding and as descriptions of other recorded outcomes, not as a single combined test of overall benefit or harm.

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

• Number of Participants With a World Health Organization (WHO) Bleeding Score >=2 During and up to 7 Days Following Elective Invasive Procedures
• Number of Participants With the Indicated Number of Platelet Transfusions Administered
• Median Platelet Count at Screening; Days 1, 8, 15, 16-19; Procedure + 7, 14, 21, 30 Day Follow-up; Early Withdrawal; and Maximum Post-baseline
• Number of Participants With the Indicated Platelet Count at Screening; Days 8 and 15; Procedure + 7, 14, 21, 30 Day Follow-up (FU); and Maximum Post-baseline
• Number of Participants Experiencing an Adverse Event (AEs) and Serious Adverse Event (SAEs) Within the Indicated Category
• Number of Participants With a Serious Adverse Event That Occurred in Greater Than One Participant
• Number of Participants With the Indicated Event Relating to Vision
• Number of Participants With Renal Function Abnormality

9. Statistical Interpretation

10. Statistical Methods Explained

What does a risk difference describe?

For a binary endpoint, a risk difference compares the proportion with the event in one group with the proportion in the other group. Here the event is avoiding platelet transfusion. A positive difference means the no-transfusion outcome was more frequent in the eltrombopag group than in the placebo group; a negative difference means it was less frequent. The posted primary risk difference is 0.527, while the sponsor reports an absolute difference of 52.8. These values express the same type of contrast on proportion and percentage-point scales, respectively.

A risk difference is an absolute comparison. It keeps attention on how far apart the group proportions are, rather than describing their relative size. It is useful for an outcome such as transfusion avoidance because the result can be interpreted against the observed proportion in each randomized group. The counts and group sizes are also important context: the reported outcome was 28 of 147 participants in the placebo group and 104 of 145 in the eltrombopag group.

What does the confidence interval add?

A confidence interval gives a range of values compatible with the data and the statistical procedure used to construct it. It communicates uncertainty around an estimate rather than only presenting a single point estimate. The primary recalculated interval is 0.430 to 0.624; the sponsor-reported 95% interval for the absolute difference is 43.2 to 62.4. The lower and upper limits describe the range of effect sizes supported with the stated confidence level under the analysis assumptions. They are not a guarantee that the true effect lies within those particular limits.

When an interval for a difference includes zero, the data are compatible with no difference under that interval procedure. When it stays on one side of zero, the estimated direction is more clearly distinguished from no difference at the corresponding confidence level. This statistical interpretation does not settle clinical importance, generalizability, or the possibility of bias.

How should the p-value be read?

A p-value evaluates how unusual the observed data, or data more extreme, would be under a specified null hypothesis and statistical model. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. For the primary outcome, the registry reports p < 0.001, and the unadjusted recalculation reports p < 0.001. The results indicate strong statistical evidence against a null hypothesis of no difference under their respective analyses. The registry’s own result remains the sponsor-reported analysis.

The recalculated comparison is an unadjusted calculation from posted summary statistics. It does not reproduce every possible feature of a protocol analysis, such as adjustment variables or stratification, unless those are explicitly incorporated. The registry identifies a Cochran-Mantel-Haenszel method for the sponsor’s primary analysis and gives its absolute difference estimate. Details beyond that method and estimate are not reported in the registry record.

Why distinguish the sponsor analysis from a recalculation?

The sponsor-reported analysis is the trial’s posted statistical analysis. A recalculation from group counts and totals can provide a transparent, simple comparison, but it may not match methods that use stratification, covariate adjustment, or other design features. In this record, the sponsor’s reported estimate and the unadjusted recalculation are concordant in direction and close in scale when expressed on their respective scales. The registry does not flag a discrepancy. The sponsor result should therefore be read as the posted trial analysis, with the recalculation serving as a simple check based on the posted counts.

For the bleeding secondary endpoint, the sponsor reports a risk difference of -5.9 with a 95% confidence interval from -15.1 to 3.3. The unadjusted recalculation reports -0.059 with an interval from -0.151 to 0.033 and p = 0.21. The registry does not provide a sponsor p-value for that endpoint. The renal function and clinically significant ECG comparisons have recalculated estimates, but no sponsor-reported analysis is listed for those outcomes.

Why can over-dispersion matter in other kinds of outcomes?

General statistical explanation: some outcomes are counts rather than yes-or-no events. A basic count model may assume that variability is determined by the average count. If the observed variation is greater than that assumption allows, the data are over-dispersed. A model that ignores over-dispersion can make uncertainty appear too small and can overstate how precise an estimate is. The primary endpoint in this trial is reported as participants who did or did not avoid transfusion, so this count-model issue is not the central method for interpreting that endpoint. The registry’s listed secondary measures include counts and repeated platelet measurements, but detailed modeling approaches for those outcomes are not reported in the registry record.

General statistical explanation: categorical outcomes can also be analyzed with methods that account for stratification or grouping. The sponsor lists a Cochran-Mantel-Haenszel method for the primary comparison. The registry record does not describe the strata, any covariate adjustments, or further implementation details. Those elements should not be inferred from the method name alone.

11. Confidence Intervals and Precision

Precision is reflected in the width of an interval around an estimate. A narrower confidence interval indicates less statistical uncertainty under the analysis assumptions; a wider interval indicates more uncertainty. For the primary endpoint, the sponsor reports an absolute difference of 52.8 with a 95% confidence interval from 43.2 to 62.4. The recalculated risk difference is 0.527, with a 95% confidence interval from 0.430 to 0.624. Both intervals remain on the positive side of zero, so the posted primary comparison has a clear estimated direction under these procedures.

The interval does not capture every possible source of uncertainty. It represents sampling uncertainty as modeled by the analysis and does not account automatically for missing data, measurement or decision differences, deviations from the planned analysis, or limited generalizability. The registry does not report enough detail to assess all of those issues. The use of investigator and physician discretion for platelet counts between 50 Gi/L and 80 Gi/L is also relevant to understanding what the endpoint represents.

Secondary estimates are less precise in some cases. The bleeding risk-difference interval extends from -0.151 to 0.033, and the renal function abnormality interval extends from -0.081 to 0.100. Both include zero and cover effects in either direction. This uncertainty means the posted results do not sharply distinguish a difference from no difference for those outcomes. The ECG estimate is based on one participant in the eltrombopag group and none in the placebo group, with a risk-difference interval from -0.007 to 0.023. Such sparse event counts warrant particular caution in interpreting the estimate.

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 illustrations describe hypothetical future planning based on the observed effect, not the trial’s own design calculation. For binary outcomes, the displayed quantities are participants per arm; they are 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 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 represents the number of participants per arm in a hypothetical study intended to have an 80% chance of detecting the specified effect under the planning assumptions, if that effect and those assumptions held. The 90% power column represents the corresponding number per arm for a higher chance of detection. These are planning targets, not guarantees about any particular study. They are not observed power.

Rows describing a smaller fraction of the observed effect require more participants per arm. This follows because a smaller difference is harder to distinguish from random variation while keeping the same significance level and desired power. The illustrations use the effect estimated in this trial, so they should not be interpreted as definitive requirements for another study. In particular, the observed estimate may not be the effect that a future study should expect or be designed to detect.

The figures should be read alongside uncertainty in the effect estimate and the limitations of the registry record. The original justification, target effect, planned analysis details, and recruitment assumptions are not reported. The planning illustrations therefore demonstrate how sample requirements change under stated hypothetical assumptions, rather than reconstructing the trial’s original design choices.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.190; treatment proportion p₁ = 0.717; average p̄ = 0.454.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7041 = 1.3800.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5975 = 0.5029.
4. n = (1.3800 + 0.5029)² / (0.5268)² = 12.8, rounded up to 13 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 <- 0.19047619047619                          # control proportion
p1 <- 0.717241379310345                          # 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(678587)
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 = 13 (80% power), 17 (90% power)
 75% of observed effect: n per arm = 23 (80% power), 30 (90% power)
 50% of observed effect: n per arm = 49 (80% power), 65 (90% power)
power.prop.test: 12.8 
Simulated power at n = 13 per arm: 0.862 (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.

14. Analyses Reported by the Sponsor

The sponsor’s primary analysis reports an absolute difference in proportions of 52.8, with a 95% confidence interval from 43.2 to 62.4 and p < 0.001, using a Cochran-Mantel-Haenszel method. The unadjusted recalculation from the posted group counts gives a risk difference of 0.527, with a 95% confidence interval from 0.430 to 0.624 and p < 0.001. The sponsor result and recalculation agree in direction and are close when expressed on their respective scales. The registry does not flag a discrepancy.

The recalculation is a simple comparison based on the posted event counts and group totals. It is not a replacement for the sponsor’s analysis, which is the analysis recorded for the trial. The registry does not provide details of any strata, covariate adjustments, or other implementation choices beyond naming the Cochran-Mantel-Haenszel method. That limits the extent to which the sponsor analysis can be independently reconstructed from the record.

For the bleeding endpoint, the sponsor reports a risk difference of -5.9 and a 95% confidence interval from -15.1 to 3.3, while the unadjusted calculation gives -0.059 and an interval from -0.151 to 0.033. No sponsor p-value is listed for this outcome. For renal function abnormality and clinically significant ECG results, the registry lists no sponsor-reported comparison analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when many statistical tests or outcomes are examined. Even if every null hypothesis were true, some results may appear unusual by chance when enough comparisons are made. Adjustment procedures or a prespecified testing hierarchy can control how such results are interpreted. The registry lists multiple secondary outcomes but does not report whether multiplicity was controlled, how outcomes were prioritized, or whether any secondary findings were intended to be confirmatory. Consequently, secondary p-values should be read as unadjusted unless the record states otherwise.

The primary endpoint analysis population is identified as the Intent-to-Treat population, defined as all randomized participants. This is useful because it specifies the intended population for the posted primary outcome. However, the record does not provide a detailed account of missing primary endpoint data, exclusions from the analysis, imputation methods, or sensitivity analyses for missing outcomes. The way incomplete outcome information was handled is therefore not reported in the registry record.

The registry also does not report the original sample-size justification, the full statistical analysis plan, or whether interim analyses were conducted. The reason for the terminated status is not reported in the registry record. These omissions prevent a complete assessment of whether analysis choices followed a prespecified plan or whether the study’s design and conduct changed over time. They do not erase the posted results, but they narrow the conclusions that can be drawn about the trial’s planning and the status of secondary findings.

For the primary endpoint, the posted sponsor analysis and recalculation are concordant, and the trial identifies the analysis population. For a broader appraisal, readers would still need details about missingness, protocol deviations, outcome ascertainment, and multiplicity. Those details are not reported in the registry record and should not be assumed.

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: portal vein thrombosis (2.8%), mesenteric vein thrombosis (2.1%), cataract (2.1%), hepatic encephalopathy (2.1%), sepsis (1.4%).

The registry reports serious adverse events in 17 of 145 participants at risk in the placebo group and 19 of 143 at risk in the eltrombopag 75 mg group. Other adverse events are reported for 35 of 145 participants at risk in the placebo group and 40 of 143 at risk in the eltrombopag group. These are participant counts within the listed safety denominators; they are descriptive summaries and do not by themselves establish whether the groups differ because of treatment.

The registry lists serious adverse event terms including portal vein thrombosis at 2.8%, mesenteric vein thrombosis at 2.1%, cataract at 2.1%, hepatic encephalopathy at 2.1%, sepsis at 1.4%, and rectal haemorrhage at 1.4%. These entries describe terms recorded in the registry and should not be interpreted as causal attribution. The record does not provide a comparative statistical analysis for these terms in the information summarized.

Deaths are not reported in the registry record. The safety figures should be read with the listed denominators and the trial’s limited duration and endpoint windows in mind. The registry does not provide enough information to assess event timing, exposure-adjusted rates, or a full causal evaluation of individual events.

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

The primary result describes a difference in the recorded proportion of participants who avoided platelet transfusion around an elective invasive procedure. The sponsor’s analysis reports an absolute difference of 52.8, with a 95% confidence interval from 43.2 to 62.4 and p < 0.001. The direction of the result is toward more participants avoiding transfusion in the eltrombopag group than in the placebo group. The posted counts are 104 of 145 and 28 of 147, respectively.

This result does not establish that every participant avoided transfusion, that transfusions were unnecessary, or that bleeding outcomes were improved. Avoiding transfusion is the primary endpoint; bleeding was separately recorded, and its posted estimate is less conclusive. Nor does the primary result demonstrate a difference in every safety outcome. The secondary results have their own definitions, event counts, and precision, and the registry does not report a multiplicity plan.

The finding applies directly to the population and procedural setting represented in this randomized trial: participants with chronic liver disease and thrombocytopenia undergoing elective invasive procedures. The registry does not report all eligibility details needed to define the population more narrowly, and the result should not automatically be extended to other patient groups, procedure types, or clinical settings.

The endpoint also incorporates clinical decision-making. For platelet counts between 50 Gi/L and 80 Gi/L, transfusion was at the discretion of the investigator and the physician performing the procedure. The observed result therefore reflects the trial’s transfusion criteria and clinical practices as well as the recorded outcome. The registry does not report enough information to evaluate how those decisions varied across sites.

Finally, statistical evidence and clinical importance are different questions. The confidence interval and p-value describe the statistical comparison under the analysis assumptions; they do not determine whether the observed difference is worthwhile in practice or settle the balance of outcomes. No treatment recommendation follows from these registry results alone.

18. Why This Trial Matters Statistically

This trial illustrates why an endpoint should be interpreted in terms of the decision it records. The primary outcome is not simply a platelet count: it is whether a participant avoided platelet transfusion around a procedure. That makes the effect estimate directly interpretable as a between-group comparison of a clinical event, while also requiring attention to the transfusion rules and the role of clinical discretion.

The randomized, placebo-controlled design and the identified Intent-to-Treat population provide important context for the primary comparison. The sponsor’s analysis reports a large positive absolute difference and a confidence interval that remains on the positive side of zero. At the same time, the secondary outcomes show why one endpoint cannot stand in for every aspect of a trial. Bleeding and safety outcomes answer different questions, and their estimates carry their own uncertainty.

The record also demonstrates the value of distinguishing a sponsor-reported analysis from an unadjusted recalculation. The two primary estimates agree in direction and are close on their respective scales, but the sponsor analysis uses a named method and remains the trial’s posted analysis. Recalculation is a useful transparent check, not a substitute for a full protocol and analysis plan.

For biostatistical interpretation, the missing information matters as much as the reported estimates: the original sample-size rationale, missing-data handling, multiplicity approach, and reason for termination are not reported in the registry record. Recognizing these gaps helps keep conclusions appropriately bounded. Statistical significance for the primary endpoint should not be mistaken for certainty about all outcomes, all populations, or all aspects of clinical value.

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 record does not report the reason the study was terminated. Without that context, it is not possible to determine from the record whether termination was related to recruitment, operational issues, safety, efficacy, or another reason. The original sample-size justification and the trial’s own design assumptions are also not reported, so the planning illustrations cannot be used to reconstruct the intended power or target effect.

The primary endpoint depends in part on clinical decisions. For platelet counts between 50 Gi/L and 80 Gi/L, transfusions were left to the discretion of the investigator and the physician performing the procedure. The registry does not report the degree of variation in those decisions across sites or how that discretion was applied. This limits interpretation of the endpoint as a purely mechanical measure.

The record identifies an Intent-to-Treat population but does not detail missing outcome data, participant exclusions, imputation, sensitivity analyses, or protocol deviations. The full statistical analysis plan and any multiplicity strategy are not reported. Consequently, the registry does not establish whether all secondary analyses were prespecified or how the risk of chance findings across multiple outcomes was addressed.

Some secondary results involve small event counts, including the clinically significant ECG outcome. Such estimates can be imprecise, and the registry provides no sponsor-reported analysis for several secondary comparisons. The safety summaries are descriptive and do not provide a full analysis of event timing or causality. Finally, the registry’s listed condition terms are broad, and additional eligibility details are not reported; the range of patients and procedures to which the findings apply cannot be fully assessed from this record alone.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Eltrombopag To Reduce The Need For Platelet Transfusion In Subjects With Chronic Liver Disease And Thrombocytopenia Undergoing Elective Invasive Procedures — NCT00678587.
• Registry record last updated 2018-10-12; status last verified 2013-01.
• Results first posted 2010-11-08.
• 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

NCT00678587 is a randomized, triple-masked, parallel-group Phase 3 study of oral eltrombopag 75 mg once daily versus oral placebo in participants with chronic liver disease and thrombocytopenia undergoing elective invasive procedures. The registry reports enrollment of 292 and lists the study status as terminated.

The primary endpoint was the number of participants who avoided platelet transfusion before, during, and up to 7 days after the procedure. The sponsor reports an absolute difference of 52.8, with a 95% confidence interval from 43.2 to 62.4 and p < 0.001. The corresponding posted counts are 104 of 145 in the eltrombopag group and 28 of 147 in the placebo group.

The record also lists bleeding, platelet, transfusion, and safety outcomes. The original sample-size rationale, reason for termination, detailed missing-data methods, and multiplicity plan are not reported in the registry record. Secondary and safety results therefore require interpretation within their outcome definitions and stated uncertainty.