Does Rosuvastatin Delay Progression of Atherosclerosis in HIV: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Does Rosuvastatin Delay Progression of Atherosclerosis in HIV (NCT01813357), a randomized phase 4 study. The primary outcome is Progression of Carotid Intima Media Thickness.

2. Clinical Question

3. Trial Design

The registry describes a completed, randomized, parallel-group, placebo-controlled phase 4 trial. Its official title identifies a multicentre, double-blind design, while the masking entry specifies quadruple masking. These entries indicate that the study was designed to conceal treatment assignment, though the record does not identify each role covered by the masking label or describe how masking was maintained.

The registry lists enrollment as 84. The posted arm summaries show 40 participants in the placebo group and 44 in the rosuvastatin group. Participants assigned to rosuvastatin received 20mg daily; the placebo was described as an encapsulated sugar pill designed to look like the active treatment. The record lists Bayside Health as sponsor.

The primary outcome was progression of carotid intima media thickness, measured by ultrasonography. The stated endpoint was change from baseline to week 96, with the intention-to-treat population specified. The trial also listed rates of adverse events as a secondary outcome. The results record includes serious and other adverse-event counts, but does not provide a detailed account of how events were collected or classified.

The registry lists the trial as completed, with a start date of 2013-07-02 and primary completion date of 2018-05-17. These dates and design features describe the registered study; they do not by themselves establish how closely every aspect of conduct followed the registered plan. The registry record does not report the original sample-size justification, a detailed statistical analysis plan, or the schedule for any interim analyses. It also does not explain the specific allocation procedure beyond identifying the study as randomized.

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: Intention to treat

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 Progression of Carotid Intima Media Thickness, measured over Baseline to week 96.

Carotid intima media thickness is a measurement of the thickness of the inner layers of the carotid artery, assessed in this trial by ultrasonography. The primary endpoint was defined as the change from baseline to week 96. A change measure aims to describe how the observed arterial measurement shifted over the specified follow-up period, rather than relying only on a measurement at one time point.

The registry describes the outcome as a continuous measure summarized by a mean, with standard error as its measure of dispersion. In general, a continuous endpoint retains information about the degree of change across participants, rather than reducing the outcome to a yes-or-no event. The standard error indicates the precision of an estimated mean; it is not a description of the spread of individual participant measurements.

Carotid intima media thickness was used as the trial’s measure of atherosclerosis progression. The endpoint is therefore a measured arterial characteristic, not a direct count of clinical cardiovascular events. The registry’s primary endpoint description specifies change from baseline to week 96, while its posted sponsor analysis is labeled as a mixed-model analysis of final values. The record does not explain the relationship between that final-values analysis and the stated change endpoint. That distinction should be kept in view when reading the arm summaries and model result.

7. Results Posted in the Registry

• Rosuvastatin: mean 0.00 versus 0.01 on control; difference -0.00 (95% CI -0.01 to 0.01), standardized effect d = -0.09; z-test p = 0.679.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference represent?

The endpoint is continuous, so each arm can be summarized by its average change in carotid intima media thickness. A mean difference compares the average in one group with the average in the other. The recalculation uses the posted arm summaries and expresses the contrast as rosuvastatin minus placebo. A negative estimate therefore corresponds to a smaller average change in the rosuvastatin arm under that orientation. The estimate is not a statement about every participant; individual changes can vary around each group mean.

The registry reports a mean and standard error for each arm. In general, the mean describes the center of the observed values, while the standard error describes how precisely the sample mean estimates its underlying group mean. A standard error is not the standard deviation of individual responses. The posted summaries do not include the raw participant-level measurements, so features such as skewness, outliers, or variation in response cannot be assessed directly from the arm means and standard errors.

What does the confidence interval add?

In general, a confidence interval gives a range of values that are reasonably compatible with the data and the assumptions used to calculate it. It conveys uncertainty about the estimated mean difference, not the range in which most individual participant measurements lie. The recalculated interval runs from a negative value to a positive value, so it includes no difference as well as possible differences in either direction. Its width indicates that the posted summaries do not determine the contrast precisely.

The recalculated interval is based on unadjusted summary-statistic calculations. The sponsor’s reported result comes from a mixed-model analysis of final values. Such analyses can use model structure and assumptions that are not captured by a simple comparison of arm summaries. The registry record does not give enough detail to reproduce the sponsor’s model or to determine all of its assumptions. The two analyses should therefore not be treated as identical estimators merely because both compare the randomized groups.

How should the p-values be read?

The recalculated unadjusted p-value is 0.679, and the sponsor’s reported p-value is 0.684. In general, a p-value measures how unusual data at least as incompatible with the null hypothesis as those observed would be if the null hypothesis and analysis assumptions were correct. It is not a measure of the probability that the treatment has no effect, and it is not a measure of the size of the observed contrast. A p-value alone also does not show whether an estimate is precise; the confidence interval is needed to understand the range of uncertainty.

The registry labels the sponsor’s analysis as a superiority analysis. A superiority test asks whether the data support a difference in a specified direction or contrast under its testing framework. The record does not report a prespecified equivalence margin, so a result that does not establish superiority cannot be interpreted as evidence that the treatments are equivalent. The record also does not state whether the sponsor’s reported p-value was adjusted for other analyses or outcomes.

Why does the analysis population matter?

The primary outcome population is identified as intention to treat. In general, an intention-to-treat analysis seeks to retain participants in the groups to which they were randomized, regardless of later treatment adherence, according to the analysis rules. The registry does not describe the implementation details for this outcome, including how missing measurements were handled, whether any participants were excluded from the analysis, or how the model incorporated incomplete follow-up. The label alone does not answer those questions.

The primary endpoint is described as a change from baseline to week 96, but the sponsor’s analysis entry is labeled as a mixed-model analysis of final values. A mixed model can account for a defined structure of repeated or otherwise related observations, but the registry record does not provide the model specification, covariates, covariance structure, or measurement schedule used. It is therefore not possible to reconstruct the model’s estimand or assess its assumptions from the posted description alone. This is a limitation of the available reporting, not evidence that the model was inappropriate.

What does dispersion mean for this endpoint?

In general, dispersion describes how much observations vary. For continuous outcomes, using a mean and standard error summarizes the estimated group average and its precision, but does not show the full distribution of individual values. The recalculation uses the reported standard errors and sample sizes to form an unadjusted comparison. It cannot account for unreported covariates or participant-level patterns. If observations are more variable than a model assumes, or if the data depart substantially from the model’s distributional assumptions, standard errors and tests may be affected. The registry does not report diagnostic checks that would allow those assumptions to be evaluated.

The listed outcome measure count is 2, while the record lists one secondary outcome and one reported analysis. The registry does not provide a complete analysis plan identifying all planned comparisons, the order of testing, or any multiplicity procedure. General statistical practice recognizes that testing multiple outcomes or contrasts can increase the chance of at least one nominally unusual result. Whether and how that issue was addressed in this trial is not reported in the registry record.

10. Confidence Intervals and Precision

The recalculated mean difference has a confidence interval that extends from negative to positive values. It therefore includes no difference, while also allowing differences in either direction. This interval is the main guide to the precision of the unadjusted summary-statistic comparison. It does not establish that the true difference is exactly zero, and it does not show that the groups are equivalent.

In general, a confidence interval combines an estimate with its uncertainty under specified assumptions. A narrow interval would indicate that the estimate is more tightly constrained; a wide interval leaves more plausible values in play. Here, the interval does not tightly locate the contrast around one direction. The posted arm standard errors and the recalculated interval both indicate that uncertainty should be considered alongside the observed means.

The sponsor’s mixed-model result reports a mean difference of 0.002 and p-value of 0.684, but the registry does not report a confidence interval for that model estimate. Its precision cannot therefore be assessed from a model-based interval in the posted results. The recalculated confidence interval is based on arm-level summaries and is not a replacement for the sponsor model’s uncertainty estimate.

Precision is also distinct from clinical importance. The registry does not identify a threshold for a clinically meaningful change in carotid intima media thickness. Without such a threshold, the interval cannot be used to classify the range of differences as clinically important or unimportant. The endpoint is a surrogate measurement, so precision about this measurement would not by itself quantify precision about clinical cardiovascular events.

11. Planning a Similar Trial: Sample Size

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

The sample-size figures are planning illustrations based on the observed effect, not the trial’s original design calculations. The original design assumptions and sample-size justification are not reported in the registry record. These illustrations use 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 figures are participants per arm, not total event counts.

For planning at 100% of the observed effect, the display gives 1,909 participants per arm for 80% power and 2,555 participants per arm for 90% power. At 75% of the observed effect, it gives 3,393 participants per arm for 80% power and 4,542 participants per arm for 90% power. At 50% of the observed effect, it gives 7,634 participants per arm for 80% power and 10,219 participants per arm for 90% power.

In plain language, 80% power is the probability, under the planning assumptions and if the specified effect were the true effect, of detecting an effect by the chosen test. The 90% power column represents a higher chance of detection under those same assumptions. Increasing the target power generally requires more participants. The illustrations also show that planning for a smaller effect requires more participants than planning for a larger effect.

These figures are not observed power. They are hypothetical planning counts derived from the observed effect and the stated assumptions. They do not show what the trial was designed to achieve, and they do not change the interpretation of the trial’s observed estimates, confidence interval, or p-values.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.002; pooled standard deviation σ = 0.024.
2. n = 2 × (1.96 + 0.842)² × 0.024² / -0.002² = 1,908.3, rounded up to 1,909 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 <- -0.0022                      # observed difference in means
sigma <- 0.0242567365548765                     # 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(1813357)
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 = 1,909 (80% power), 2,555 (90% power)
 75% of observed effect: n per arm = 3,393 (80% power), 4,542 (90% power)
 50% of observed effect: n per arm = 7,634 (80% power), 10,219 (90% power)
power.t.test: 1909.3 
Simulated power at n = 1,909 per arm: 0.798 (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.

13. Analyses Reported by the Sponsor

The sponsor’s posted analysis is a mixed-model analysis of final values. It reports a mean difference of 0.002 and a p-value of 0.684. The registry does not state the direction of the reported contrast or provide a confidence interval for this model estimate. The primary endpoint description, by contrast, defines the outcome as change from baseline to week 96. The record does not explain why the analysis entry is described in terms of final values or how the model’s target quantity relates to the stated change endpoint.

The recalculation uses the posted arm means, standard errors, and sample sizes to compare change summaries without adjustment. It reports an unadjusted p-value of 0.679, and its confidence interval includes values on either side of no difference. The small difference between that p-value and the sponsor’s reported p-value does not make the two analyses interchangeable: they are based on different reported descriptions and potentially different analytical procedures. The model’s details are not reported, so the sponsor’s calculation cannot be independently reconstructed from this record.

The record does not flag a discrepancy between the posted sponsor analysis and recalculation. The sponsor’s analysis remains the trial’s reported analysis; the recalculation is a secondary, unadjusted comparison of the summary statistics and should be interpreted as such. Neither result alone establishes equivalence or resolves the uncertainty about the size and direction of the contrast.

14. Multiplicity, Missing Data and Analysis Populations

The registry record does not provide enough information to determine how multiplicity was handled. It lists a primary outcome, rates of adverse events as a secondary outcome, an outcome measure count of 2, and one reported analysis. It does not give the complete set of planned endpoints and contrasts, identify which tests were prespecified, or state whether any adjustment was used for multiple testing. Without that information, the reported p-values should be read as the values listed, not assumed to be adjusted or unadjusted beyond the specific recalculation being discussed.

In general, multiplicity arises when several outcomes, treatment contrasts, time points, or subgroups are tested. Conducting more tests can raise the chance of at least one result appearing unusual under a null model. Prespecifying a primary outcome and an analysis hierarchy can help readers understand which result is central, but the registry record does not report a detailed testing hierarchy or statistical analysis plan. It is not possible to infer from the number of posted outcomes alone how many analyses were planned or performed.

The primary population is identified as intention to treat. The registry does not report how many participants contributed primary outcome data at week 96, how missing baseline or follow-up measurements were handled, or whether the model included all randomized participants. It also does not describe the assumptions made about missingness or any sensitivity analyses. These details matter because incomplete follow-up can affect both the estimated contrast and its uncertainty, particularly when the availability of measurements is related to participant characteristics or outcomes.

The posted arm summaries do not include participant-level data, so missingness patterns and the distribution of individual changes cannot be examined. The record also does not report protocol deviations, analysis exclusions, or whether the sponsor’s model used adjustments beyond its mixed-model structure. These aspects are not reported in the registry record and should not be filled in by assumption.

The absence of these details limits independent assessment of the analysis, but it does not establish that the sponsor failed to use appropriate procedures. It means that the public registry description is insufficient to verify the full statistical handling of multiplicity and missing data.

15. Safety Results

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

Deaths

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: acute myocardial infarction (4.5%), type two diabetes (4.5%), elevated alanine aminotransferase (alt) (2.5%), acute mesentric ischaemia (2.5%), oesophageal malignancy (2.5%).

The registry reports serious adverse events affecting 6 participants in the placebo arm and 7 in the rosuvastatin arm. The recorded at-risk counts are 40 for placebo and 44 for rosuvastatin. The record also reports other adverse events affecting 29 participants in the placebo arm and 36 in the rosuvastatin arm. Deaths are listed as 0 in the placebo arm and 1 in the rosuvastatin arm. These are observed counts as recorded, not estimates of a causal treatment effect.

The serious-event terms listed include acute myocardial infarction and type two diabetes at 4.5% each, and elevated alanine aminotransferase (ALT), acute mesenteric ischaemia, oesophageal malignancy, and lumbar vertebral disc herniation at 2.5% each. The registry does not provide enough detail in this record to establish the event-level relationship between these terms and the participant counts, or to determine whether an event was judged related to treatment.

In general, adverse-event counts describe events observed and recorded during a study; they do not by themselves establish that a treatment caused an event. Interpretation depends on definitions, follow-up, severity, timing, and the method of event collection. The registry record does not provide a full safety analysis or comparative uncertainty estimates. The reported counts should therefore be read descriptively, without treating differences in counts as proof of a difference in risk.

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

The results show that the mean change in carotid intima media thickness was numerically smaller in the rosuvastatin group than in the placebo group. The recalculated unadjusted confidence interval includes no difference and extends in both directions. The recalculated p-value is 0.679. The sponsor’s mixed-model analysis of final values reports a mean difference of 0.002 and p-value of 0.684. Together, these results do not provide a precise estimate that establishes a difference between the randomized groups on the posted measurement.

This is not the same as demonstrating that rosuvastatin and placebo have identical effects. A result that does not distinguish a difference from no difference can reflect uncertainty, and the confidence interval indicates that more than one direction remains compatible with the summary data. The registry does not report an equivalence margin or an equivalence analysis. It therefore does not support an equivalence conclusion.

The primary endpoint is a measure of carotid artery structure over the stated follow-up period. It is not itself a count of heart attacks, strokes, or other clinical events. Although the registry’s study description refers to cardiovascular risk and atherosclerosis progression, the posted primary result should be interpreted as a comparison of the specified carotid measurement, not as a complete account of clinical cardiovascular outcomes.

The safety record contains serious and other adverse-event counts, as well as a death count for each group. These counts provide descriptive information but do not establish causation or comparative risk. The available record does not give the detail needed to fully assess event definitions, treatment attribution, missing safety follow-up, or uncertainty around group differences.

Overall, the posted data leave uncertainty about the size and direction of any between-group difference in the primary measurement. They do not justify a claim that treatment works or does not work, and they should not be used alone to make a treatment decision.

17. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the distinction between an observed arm difference and a well-determined treatment contrast. The rosuvastatin arm’s mean change is numerically smaller than the placebo arm’s, but the recalculated interval includes no difference and values in both directions. A point estimate alone can look directional; the interval shows why uncertainty must be considered before drawing a conclusion.

It also illustrates how the endpoint definition and analysis description matter. The primary outcome is described as change from baseline to week 96, while the sponsor’s posted analysis is labeled as a mixed-model analysis of final values. These summaries may address related questions, but the registry does not explain their relationship. Readers need enough information about the estimand, model, and analysis population to know exactly what a reported contrast represents.

The comparison between the recalculated p-value and the sponsor’s p-value highlights the limits of summary-statistic re-analysis. The values are close, but the recalculation is unadjusted and based on posted arm summaries, whereas the sponsor reports a mixed model. Similar p-values do not prove that the analyses use the same method or answer the same estimand.

Finally, the planning illustrations show why sample-size interpretation depends on the effect selected for planning. The listed counts increase as the assumed target effect becomes smaller, and higher target power also calls for more participants. These are hypothetical planning calculations based on the observed effect, not a record of the trial’s original design. Distinguishing prospective planning from observed results helps prevent power calculations from being mistaken for evidence about what the completed trial established.

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 registry record leaves important details unavailable for independent evaluation. Although the primary endpoint is defined as change from baseline to week 96, the sponsor’s analysis is described as a mixed-model analysis of final values. The record does not explain that distinction, state the exact model specification, or provide a model-based confidence interval. The reported sponsor estimate’s direction is also not specified.

The summary-statistic recalculation is unadjusted and cannot reproduce the sponsor’s model. It relies on arm-level means, standard errors, and sample sizes rather than participant-level observations. The registry does not report model covariates, the handling of repeated measurements, diagnostic checks, or sensitivity analyses. These omissions restrict assessment of the assumptions and robustness of the analysis.

The record identifies intention to treat as the primary outcome population but does not report the number with observed week 96 measurements, the handling of missing data, or any exclusions from the analysis. It also does not provide a full multiplicity plan, the original sample-size justification, or details of interim analyses. These items are not reported in the registry record.

The primary endpoint is a surrogate measurement, and the posted result does not provide a direct comparison of clinical cardiovascular events. The safety information is presented as counts and selected serious-event terms, without detailed event attribution, timing, or comparative uncertainty estimates. The record’s adverse-event summary therefore supports descriptive reading but not a comprehensive assessment of comparative safety.

These reporting limits do not establish that the trial’s conduct or analysis was flawed. They mark the boundary of what can be evaluated from the registry record and counsel against drawing conclusions beyond the reported endpoint, analysis, and follow-up.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Does Rosuvastatin Delay Progression of Atherosclerosis in HIV — NCT01813357.
• Registry record last updated 2020-08-28; status last verified 2020-08.
• Results first posted 2020-08-28.
• 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 identifies NCT01813357 as a completed, randomized, parallel-group, phase 4, placebo-controlled trial sponsored by Bayside Health. The study enrolled 84 participants and compared rosuvastatin 20mg daily with an encapsulated placebo in people with HIV at moderate cardiovascular risk. The registry lists a start date of 2013-07-02 and a primary completion date of 2018-05-17.

The primary endpoint was change in carotid intima media thickness from baseline to week 96, measured by ultrasonography, with an intention-to-treat population specified. The sponsor reported a mixed-model analysis of final values, with mean difference 0.002 and p-value 0.684. The unadjusted recalculation of the posted change summaries has p-value 0.679 and a confidence interval spanning both sides of no difference. The registry also reports adverse-event counts, but does not provide sufficient detail for a full comparative safety assessment.