Evaluate Safety and Efficacy of ABT-335 in Combination With Rosuvastatin Calcium in Subjects With Multiple Abnormal Lipid Levels in the Blood: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluate Safety and Efficacy of ABT-335 in Combination With Rosuvastatin Calcium in Subjects With Multiple Abnormal Lipid Levels in the Blood (NCT00300482), a randomized phase 3 study. The primary outcome is Mean Percent Change in Triglycerides From Baseline to Final Visit.

2. Clinical Question

3. Trial Design

This was a phase 3, randomized, double-blind, prospective, parallel-group study. The official title describes it as multicenter and identifies fenofibric acid and rosuvastatin combination therapy and monotherapy as the treatments under comparison. The registry record lists quadruple masking, which indicates that masking was recorded at that level in the trial record. The record does not specify which roles were masked or the practical masking procedures.

Six arms are listed: ABT-335 with 10 mg rosuvastatin, ABT-335 with 20 mg rosuvastatin, ABT-335 alone, and rosuvastatin alone at 10 mg, 20 mg, or 40 mg. Random allocation to parallel groups allows treatment groups to be compared while limiting systematic differences at the point of allocation. It does not guarantee that every group remains identical in all respects after allocation, nor does the design description alone explain how the analysis handled departures from assigned treatment.

The record lists enrollment of 1445 and a completed status. The study start is recorded as 2006-03 and primary completion as 2006-12. The primary triglyceride outcome is specified from baseline to 12 weeks. The outcome population required a baseline triglyceride value and at least 1 postbaseline value; last observation carried forward was used for that population. Arm-specific numbers contributing to the primary outcome are reported alongside the means, but these are not necessarily the same as enrollment or the number initially randomized to each group.

Safety and efficacy were both study objectives. The posted material includes counts of participants affected by serious and other adverse events, but the numbers at risk for those summaries are not reported in the registry record. Details of the allocation sequence, concealment, adherence, withdrawals, and the original sample-size justification are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: Six subjects were randomized but not treated: 1 was lost to follow-up, 1 was randomized but deemed ineligible, 1 was withdrawn at the investigator's discretion, and 3 withdrew consent.

Overall Study

Analysis population for the primary outcome, as defined in the registry: All randomized subjects with a baseline triglyceride value and at least 1 postbaseline triglyceride value, last observation carried forward

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.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Mean Percent Change in Triglycerides From Baseline to Final Visit, measured over Baseline to 12 Weeks.

The primary endpoint is mean percent change in triglycerides from baseline to the final visit, with the stated timeframe of baseline to 12 weeks. For each participant, percent change is defined as the week 12 triglyceride value minus the baseline value, divided by the baseline value, then multiplied by 100. The reported arm result is the mean of this continuous measure, with standard error as the dispersion measure.

Percent change puts change in relation to each participant's starting value rather than expressing it only in the original triglyceride units. This can make results easier to compare across people whose baseline measurements differ. A negative mean indicates that the measured triglyceride values were lower at the final measurement on average than at baseline. It does not mean every participant experienced a decrease, because a group mean can conceal variation in individual responses.

Triglycerides are one component of the lipid profile, so this endpoint addresses a laboratory measure relevant to the study's mixed dyslipidemia population. A 12-week endpoint captures change over the specified short-term period; it does not establish persistence, later clinical outcomes, or effects on cardiovascular events. The endpoint is an average change, not a direct measure of symptoms or quality of life. The registry record states that the analysis population included randomized subjects with a baseline measurement and at least 1 postbaseline measurement, using last observation carried forward. The rationale for that specific missing-data approach is not reported in the registry record.

7. Results Posted in the Registry

• ABT-335 + 20 mg Rosuvastatin: mean -42.90 versus -47.10 on control; difference 4.20 (95% CI -0.83 to 9.23), standardized effect d = 0.15; z-test p = 0.102.
• ABT-335: mean -32.60 versus -47.10 on control; difference 14.50 (95% CI 9.44 to 19.56), standardized effect d = 0.51; z-test p = <0.001.
• 10 mg Rosuvastatin: mean -24.40 versus -47.10 on control; difference 22.70 (95% CI 17.68 to 27.72), standardized effect d = 0.79; z-test p = <0.001.
• 20 mg Rosuvastatin: mean -25.60 versus -47.10 on control; difference 21.50 (95% CI 16.50 to 26.50), standardized effect d = 0.75; 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.

Mean Percent Change in Non-low-density Lipoprotein Cholesterol (Non-HDL-C)From Baseline to Final Visit

• 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:
  • ABT-335 + 20 mg Rosuvastatin: mean -45.30 versus -44.70 on control; difference -0.60 (95% CI -3.50 to 2.30), standardized effect d = -0.04; z-test p = 0.685.
  • ABT-335: mean -18.50 versus -44.70 on control; difference 26.20 (95% CI 23.28 to 29.12), standardized effect d = 1.67; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ABT-335 + 20 mg Rosuvastatin and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for ABT-335 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Percent Change in Very Low-density Lipoprotein Cholesterol (VLDL-C) From Baseline to Final Visit

• 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:
  • ABT-335 + 20 mg Rosuvastatin: mean -50.60 versus -55.80 on control; difference 5.20 (95% CI -3.36 to 13.76), standardized effect d = 0.11; z-test p = 0.234.
  • ABT-335: mean -31.90 versus -55.80 on control; difference 23.90 (95% CI 15.29 to 32.51), standardized effect d = 0.50; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ABT-335 + 20 mg Rosuvastatin and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for ABT-335 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Percent Change in Total Cholesterol From Baseline to Final Visit

• 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:
  • ABT-335 + 20 mg Rosuvastatin: mean -35.70 versus -34.40 on control; difference -1.30 (95% CI -3.59 to 0.99), standardized effect d = -0.10; z-test p = 0.265.
  • ABT-335: mean -13.50 versus -34.40 on control; difference 20.90 (95% CI 18.60 to 23.20), standardized effect d = 1.60; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between ABT-335 + 20 mg Rosuvastatin and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for ABT-335 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes cover several lipid measures: non-HDL cholesterol, VLDL cholesterol, total cholesterol, apolipoprotein B, and high-sensitivity C-reactive protein. The listed results include six groups for some outcomes, while the recalculated contrasts available for review cover only selected arms and outcomes. Across the posted group means, combination regimens generally show negative percent changes for the lipid measures with reported mean changes, while the size and pattern of the changes vary by measure and treatment arm. The secondary material therefore adds context about the breadth of measured lipid changes, rather than providing a single uniform pattern across every endpoint.

The direction of the posted triglyceride results is broadly consistent with the fact that lipid measures were assessed across treatment groups, but secondary outcomes answer distinct questions. A change in one lipid measure does not establish the same change in another, and findings on laboratory measures do not establish effects on cardiovascular events or other clinical outcomes. The registry lists five secondary outcomes, but only three have detailed arm results and recalculated contrasts in the posted secondary data summarized for these analyses. Detailed results for the remaining listed secondary outcomes are not reported in the registry record.

The available unadjusted secondary comparisons should carry less inferential weight than a prespecified primary analysis. Testing multiple outcomes and treatment contrasts increases the chance of a nominally small p-value arising by chance across the family of tests. The registry record does not report a multiplicity adjustment for these secondary comparisons or a complete testing hierarchy. Without such information, secondary results are best treated as supporting or exploratory evidence, not as independent confirmation of a definitive effect. Their interpretation also depends on the outcome-specific analysis populations, which vary in the posted results.

These secondary measures can help characterize the trial's laboratory profile, but they do not resolve the limitations of short follow-up, missing-data handling, or incomplete reporting of adjusted estimates. They should be interpreted alongside the primary endpoint and sponsor analyses, with attention to the fact that each endpoint has its own scale, variability, and comparison structure.

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

• Mean Percent Change in Non-low-density Lipoprotein Cholesterol (Non-HDL-C)From Baseline to Final Visit
• Mean Percent Change in Very Low-density Lipoprotein Cholesterol (VLDL-C) From Baseline to Final Visit
• Mean Percent Change in Total Cholesterol From Baseline to Final Visit
• Mean Percent Change in Lipoprotein Apo B (Apo B) From Baseline to Final Visit
• Median Percent Change in High-sensitivity C-reactive Protein (hsCRP) From Baseline to Final Visit

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary outcome is continuous: each participant has a percent change value, and the arm-level result is the mean. A mean summarizes the center of a group distribution, while the standard error describes uncertainty in the estimated mean. It is not a measure of the full spread of individual participant values. Two groups can have similar means but different individual variability, and the standard error should not be interpreted as the range in which most participants' values fall.

What does a mean difference describe?

A mean difference compares the average outcome in one arm with the average in another. Its sign depends on the order of subtraction. In the recalculated comparisons, the reference is ABT-335 plus 10 mg rosuvastatin, and each estimate describes the other arm relative to that group. Because the outcome is percent change, more negative values correspond to greater average reductions from baseline. A positive mean difference therefore represents a less negative mean change than the reference, while a negative difference represents a more negative mean change.

These recalculations use unadjusted arm summaries rather than the sponsor's adjusted model. They are useful as a transparent comparison of the posted group means, but they do not account for covariates in the manner of an adjusted analysis. The registry lists ANCOVA for two pairwise comparisons. In general, analysis of covariance models an outcome while adjusting for specified covariates, often including a baseline measure. The covariates and exact model specification are not reported in the registry record, so the degree and details of adjustment cannot be assessed.

How should a confidence interval be read?

A confidence interval expresses uncertainty around an estimated contrast under the method used to construct it. A narrower interval indicates greater precision than a wider interval on the same scale. An interval spanning zero is compatible with a range of possible contrasts that includes no mean difference, as well as values in either direction. An interval entirely on one side of zero indicates that zero is not among the values included by that interval. This does not establish clinical importance: that judgment requires a meaningful threshold, which is not reported in the registry record.

The recalculated intervals and p-values are unadjusted. A p-value describes how compatible the observed data are with a specified null hypothesis under the statistical method; it is not the probability that the null hypothesis is true, nor a measure of the size or clinical value of an effect. Repeated testing across multiple arms and outcomes creates a multiplicity issue: the more comparisons considered, the greater the chance of at least one apparently unusual result under a collection of null hypotheses. The registry record does not report a multiplicity adjustment for the posted secondary comparisons or fully describe an adjustment plan for all primary comparisons.

What role do analysis populations play?

The primary outcome population includes randomized subjects who had a baseline triglyceride value and at least 1 postbaseline value. The record says that last observation carried forward was used. In general, this method fills a missing later measurement by carrying forward an earlier observed value. Its validity depends on assumptions about why values are missing and how the outcome would otherwise have changed. If participants leave the study for reasons related to their outcomes, carried-forward values can misrepresent the trajectory. The missingness reasons and sensitivity analyses are not reported in the registry record.

The record also lists group-specific outcome counts, which are below the total enrollment for some arms and reflect the outcome analysis rather than total enrollment. The exact flow from randomized participants to those analyzed, including reasons for missing values or exclusion, is not reported in the registry record. Consequently, the posted averages should be read as results for the specified measured-outcome population, not automatically as results for every randomized participant.

Why does the sponsor analysis matter?

The sponsor reports ANCOVA p-values for two comparisons, each <0.001. The recalculated figures are derived from posted summary statistics and use the combination arm with 10 mg rosuvastatin as their reference. One recalculated comparison therefore does not match either sponsor-reported pairwise contrast. The comparison of ABT-335 plus 20 mg rosuvastatin with 20 mg rosuvastatin monotherapy is a distinct contrast from comparing the two combination arms. This distinction can account for different numerical conclusions without implying that either result is a calculation error.

When the sponsor's own reported analysis and a simple unadjusted recalculation disagree, the sponsor's analysis takes precedence for the trial's inferential result. The recalculation remains descriptive and is not a substitute for the trial's prespecified model. The registry record does not report whether the sponsor's analyses were prespecified in a protocol or statistical analysis plan, nor the full model output needed to reproduce them.

11. Confidence Intervals and Precision

The recalculated mean difference between ABT-335 plus 20 mg rosuvastatin and the 10 mg combination reference is 4.2, with a 95% confidence interval from -0.831 to 9.231. This interval spans zero and includes both a small contrast in the negative direction and a larger contrast in the positive direction. The unadjusted p-value is 0.102. The estimate is therefore not highly precise about the direction of the difference between these two combination arms.

For the comparisons of the monotherapy arms with the 10 mg combination reference, the recalculated confidence intervals are entirely positive: 9.441 to 19.56 for ABT-335, 17.68 to 27.72 for 10 mg rosuvastatin, 16.5 to 26.5 for 20 mg rosuvastatin, and 8.982 to 21.02 for 40 mg rosuvastatin. These intervals describe uncertainty in the unadjusted mean differences, not uncertainty in the sponsor's adjusted ANCOVA estimates. The registry record does not provide confidence intervals or estimates for the reported sponsor ANCOVA comparisons.

Precision is not the same as importance. An interval can be statistically separated from zero while still including effects that may be more or less meaningful to patients or clinicians. No minimally important difference is reported in the registry record. Similarly, the standard errors shown with the arm means quantify the estimated means' precision; they do not reveal the full distribution of participant responses.

Interpretation should also account for the analysis population and last observation carried forward. The precision calculations do not capture uncertainty due to possible bias from missing outcome data or the choice of analysis model. The registry record does not report sensitivity analyses to show whether conclusions change under alternative missing-data assumptions. Thus, the intervals provide a focused measure of sampling uncertainty for the recalculated contrasts, but not a complete account of every source of uncertainty.

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 contrast, not the trial's original design calculation. For these continuous outcomes, the listed values mean participants per arm, not total participants across arms or total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row considers an effect equal to 100%, 75%, or 50% of the observed effect. The trial's own original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the estimated participants per arm for a future comparison to have an 80% chance of detecting the stated effect under those assumptions, if that effect is truly present. The 90% power column applies the same idea with a higher probability of detection, which generally requires more participants. These are conditional planning quantities: they depend on the assumed effect and the statistical assumptions, and they are not guarantees about the result of a future study.

Smaller effects are harder to distinguish from random variation, so the illustration requires more participants per arm when planning for a smaller fraction of the observed effect. This relationship is particularly relevant when the observed contrast is small: planning around a small effect can yield very large estimated requirements. The figures are not observed power; they are hypothetical sample-size estimates for future planning based on the observed effect sizes and stated assumptions.

The plotted planning curve should be read in the same way as the listed values: it represents changing sample-size needs as the assumed effect changes, rather than evidence about treatment response or the number of participants actually enrolled. The listed values can differ substantially across contrasts because observed effect estimates differ. This illustration does not incorporate the sponsor's ANCOVA model or establish that the original trial was designed around these recalculated contrasts.

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 = 4.200; pooled standard deviation σ = 28.726.
2. n = 2 × (1.96 + 0.842)² × 28.726² / 4.200² = 734.3, rounded up to 735 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 <- 4.2                      # observed difference in means
sigma <- 28.7260295472085                     # 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(300482)
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 = 735 (80% power), 984 (90% power)
 75% of observed effect: n per arm = 1,306 (80% power), 1,748 (90% power)
 50% of observed effect: n per arm = 2,938 (80% power), 3,933 (90% power)
power.t.test: 735.3 
Simulated power at n = 735 per arm: 0.797 (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 sponsor analyses using ANCOVA for two primary-outcome comparisons: ABT-335 plus 10 mg rosuvastatin versus 10 mg rosuvastatin, and ABT-335 plus 20 mg rosuvastatin versus 20 mg rosuvastatin. The reported p-value is <0.001 for each. The registry supplies a 95% confidence level for the second comparison, but does not report the estimated contrast or confidence limits for either sponsor analysis.

The unadjusted recalculations use ABT-335 plus 10 mg rosuvastatin as the reference for the other arms. As a result, the recalculated comparison of the two combination arms is not the same question as comparing the 20 mg combination with 20 mg rosuvastatin alone. The remaining recalculated contrasts also use a different comparison framework from the sponsor's paired dose comparisons. Differences in the contrast being tested and in adjustment can produce different estimates and p-values.

Because the discrepancy flag is true, the sponsor's own analysis takes precedence over recalculated results when they differ. The recalculations are simple comparisons from posted summary statistics; they do not reproduce the ANCOVA. The exact model covariates, model assumptions, and adjusted estimates are not reported in the registry record, limiting independent assessment of the sponsor's results.

15. Multiplicity, Missing Data and Analysis Populations

The trial has several treatment arms, a primary outcome, and multiple secondary outcomes. Each additional comparison creates another opportunity for a low p-value, even when the corresponding null hypotheses are all true. In general, multiplicity procedures control the overall risk of false-positive conclusions across a defined family of tests. The registry record does not report a comprehensive multiplicity strategy, testing hierarchy, or adjustment for the posted unadjusted recalculations. The sponsor's reported ANCOVA p-values are available for two comparisons, but the registry does not describe how those tests relate to a broader family of comparisons.

The primary outcome population required a baseline triglyceride result and at least 1 postbaseline result. Last observation carried forward was used. In general, this approach assumes that carrying a prior observation forward provides a reasonable substitute for an unobserved later value. That assumption may be questionable if missingness is associated with treatment, tolerability, or the participant's unobserved outcome trajectory. The reasons for missing data, timing and frequency of missing measurements, and any sensitivity analyses are not reported in the registry record.

The registry record does not provide a complete participant flow accounting for all randomized participants, analysis exclusions, and withdrawals by arm. It does not report enough information to assess whether the primary outcome analysis followed an intention-to-treat approach, nor the protocol-defined analysis populations beyond the stated measurement requirement. The exact handling of intercurrent events and deviations from assigned treatment is also not reported in the registry record.

These omissions matter because p-values and confidence intervals describe uncertainty under a particular analysis; they do not automatically account for bias caused by missing outcomes, selective reporting, or multiple testing. The available results support a description of the posted analyses, but do not permit a full audit of the statistical plan or robustness to alternative assumptions.

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

The primary results describe average triglyceride percent changes from baseline to the 12-week measurement across six treatment groups. The mean change was negative in every group, with the combination of ABT-335 and 10 mg rosuvastatin showing the most negative listed mean. The recalculated unadjusted comparison of the two combination arms has a confidence interval that includes zero, while the recalculated comparisons of the monotherapy arms with the 10 mg combination reference have intervals entirely above zero. The sponsor separately reports ANCOVA p-values below 0.001 for two comparisons of combination therapy with the corresponding rosuvastatin monotherapy.

These findings are about group averages and a specified laboratory endpoint. They do not show that every participant had the same response, nor do they tell the reader how individual responses were distributed. A statistically detectable contrast is not necessarily clinically important, and the registry record does not state a threshold for clinical importance. The results do not establish effects on symptoms, quality of life, cardiovascular events, or other outcomes not reported for the primary endpoint.

The timeframe is baseline to 12 weeks. The posted results cannot establish whether observed changes persist after that period. Nor do the outcome values alone establish a complete safety profile. The adverse-event summaries include affected counts, but denominators for those summaries are not reported in the registry record; death counts and numbers at risk are also not reported.

Comparisons should be understood in light of the analysis method and population. The sponsor's adjusted comparisons address different pairs of arms than some of the unadjusted recalculations, and the sponsor's analysis takes precedence when results conflict. The last-observation-carried-forward approach and incomplete reporting of missing-data details limit what can be concluded about outcomes for all randomized participants. The findings are therefore a record of short-term measured changes and reported statistical comparisons, not a standalone basis for judgments about an individual's care.

17. Why This Trial Matters Statistically

This trial is statistically instructive because it includes multiple active regimens and component monotherapies. That structure makes the exact contrast important: comparing the combination regimens with each other asks a different question from comparing each combination with its corresponding rosuvastatin monotherapy. A result from one contrast cannot simply be transferred to another, even when the treatment arms share a component.

The record also illustrates why an arm mean and an adjusted treatment comparison are not interchangeable. Simple differences calculated from posted means provide a transparent descriptive check, but the sponsor reports ANCOVA for selected pairwise comparisons. Without the model specification and adjusted estimates, the two analyses cannot be fully reconciled. Where a discrepancy exists, the sponsor's own analysis takes precedence, while the recalculation clarifies the contrast represented by the posted summary statistics.

Precision is another central issue. The confidence interval for the recalculated comparison between the combination arms spans zero, whereas intervals for the recalculated monotherapy comparisons with the 10 mg combination reference do not. These intervals make uncertainty visible and discourage treating a point estimate as exact. At the same time, statistical separation from zero does not establish that a difference is important in practice.

Finally, this example shows how endpoint population, imputation, and multiplicity affect interpretation. The reported primary analysis uses a measurement-defined population and last observation carried forward. Multiple outcomes and pairwise contrasts raise questions about multiplicity, but the registry record does not report a complete adjustment plan. Reading the result well therefore requires attention not only to p-values, but also to the population analyzed, the contrast tested, the uncertainty, and the limits of the record.

18. 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 primary endpoint is a short-term laboratory measure, assessed through baseline to 12 weeks. The registry record does not report longer-term follow-up for this outcome or clinical endpoints that would establish whether the measured changes correspond to later health outcomes. A mean percent change also conceals the distribution of individual responses.

The primary analysis population excludes randomized subjects without the required baseline and postbaseline measurements. Last observation carried forward was used, but the record does not explain the reasons for missingness, report sensitivity analyses, or provide enough participant-flow detail to assess the potential effect of missing outcomes. The outcome counts differ by arm and do not equal the enrollment figure, but the reasons for those differences are not reported in the registry record.

The registry reports only selected sponsor ANCOVA comparisons and p-values. Estimated adjusted differences, confidence intervals, model covariates, and the full statistical specification are not reported. This limits evaluation of the sponsor analysis and prevents a direct comparison of its estimates with the unadjusted recalculations. The sponsor's reported comparisons also use different arm pairs from the recalculated contrasts based on the 10 mg combination reference.

Multiplicity procedures and the trial's original sample-size justification are not reported in the registry record. The planning illustrations based on observed effects are not a substitute for the original design rationale. Safety summaries report affected counts, but denominators and death counts are not reported; the listed serious-event terms have no accompanying arm-specific denominators in the record. Together, these reporting gaps limit assessment of uncertainty, generalizability, and the completeness of safety comparisons.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Evaluate Safety and Efficacy of ABT-335 in Combination With Rosuvastatin Calcium in Subjects With Multiple Abnormal Lipid Levels in the Blood — NCT00300482.
• Registry record last updated 2009-06-10; status last verified 2009-06.
• Results first posted 2009-03-25.
• 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.

22. Record Summary

The registry describes a completed, randomized, double-blind phase 3 parallel-group study of ABT-335 and rosuvastatin in people with mixed dyslipidemia. Six arms evaluated two combination regimens, ABT-335 monotherapy, and three rosuvastatin monotherapy doses. The primary outcome was mean percent change in triglycerides from baseline to 12 weeks, analyzed in subjects with a baseline value and at least 1 postbaseline value using last observation carried forward.

All six reported arm means were negative. The sponsor reports ANCOVA p-values <0.001 for combination therapy compared with the corresponding 10 mg and 20 mg rosuvastatin monotherapies. Unadjusted recalculations use a different comparison framework; because a discrepancy is flagged, the sponsor's own analysis takes precedence. Missing-data sensitivity analyses, the complete statistical plan, and the original sample-size justification are not reported in the registry record.