Inflammation and Vascular Function in Atherosclerosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Inflammation and Vascular Function in Atherosclerosis (NCT00760019), a randomized phase 2, phase 3 study. The primary outcome is Flow-mediated, Endothelium-dependent Vasodilation.

2. Clinical Question

This is a question about a measured vascular response under the two treatment conditions, not a direct assessment of clinical events or symptoms. Because the design is crossover, each sequence includes both salsalate and placebo, with treatment order differing between groups. The comparison therefore concerns measurements after each participant's two treatment periods, although the registry record does not report the statistical procedure used to account for the paired observations or treatment order.

The recorded outcome values are medians with interquartile ranges. The median was 6.8 percentage vasodilation after salsalate, with an interquartile range of 5.2 to 12.0. After placebo, the median was 8.7, with an interquartile range of 5.4 to 13.5. These are descriptive summaries. The registry record does not report an inferential test, confidence interval, or estimate of the within-participant contrast, so the precision and statistical evidence for a difference cannot be determined from the posted results alone.

3. Trial Design

The registry identifies this study as randomized, quadruple-masked, and crossover, and lists phase 2 and phase 3. It enrolled 58 participants with atherosclerosis. Brigham and Women's Hospital is listed as the sponsor. The recorded start was 2005-08, primary completion was 2009-10, and study status is completed.

Participants were allocated to one of two treatment sequences. One sequence received salsalate first, then had a 4 week washout, then received placebo for 4 weeks. The other sequence received placebo first, had a 4 week washout, and then received salsalate for 4 weeks. Both periods were described as double-blinded. The registry also identifies quadruple masking overall, but does not explain which roles were masked or how masking was maintained.

In a crossover design, participants serve as their own control by receiving both study conditions at different times. This can make within-participant comparisons informative because stable characteristics of an individual are shared across the two periods. At the same time, the interpretation can depend on treatment order, changes over time, and whether effects carry into a later period. The washout separates the periods, but the registry record does not describe how its duration was chosen or report analyses of period, sequence, or carryover effects.

The primary outcome was assessed upon completion of 4 weeks of salsalate and placebo treatment. The registry reports 56 participants in each treatment condition for the primary outcome, consistent with the outcome description that two participants' ultrasound data were discarded before unblinding. The reason for their exclusion was inadequate ultrasound data. The record does not state the trial's sample-size justification, recruitment target rationale, or any prespecified statistical analysis plan.

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

Analysis population for the primary outcome, as defined in the registry: The ultrasound data for two subjects was inadequate for analysis. This determination was made prior to unblinding. The data for these two subjects were discarded, leaving 56 subjects for analysis.

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 Flow-mediated, Endothelium-dependent Vasodilation, measured over Upon completion of 4 weeks of salsalate and placebo treatment.

The primary endpoint, flow-mediated, endothelium-dependent vasodilation, measures the percentage increase in brachial artery diameter after a 5 minute ischemic stimulus. The measurement was made at the end of the placebo treatment period and at the end of the salsalate treatment period. The registry describes the outcome in percentage vasodilation and summarizes it using the median and interquartile range.

As general background, flow-mediated vasodilation is used as a physiological measure of blood-vessel response to a stimulus. It reflects a vascular function rather than a clinical event such as a heart attack, hospitalization, or death. The endpoint is relevant to the trial's stated question because the study aimed to determine whether reducing inflammation in blood vessels with salsalate would improve blood vessel function.

The registry reports a median of 6.8 after salsalate, with an interquartile range of 5.2 to 12.0, and a median of 8.7 after placebo, with an interquartile range of 5.4 to 13.5. The values describe the observed distributions in the analyzed participants. They do not, by themselves, quantify the average within-person change or establish how uncertain the comparison is. No confidence interval or inferential test for the primary outcome is reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What do the median and interquartile range describe?

The registry identifies the primary outcome parameter as the median and the dispersion measure as the interquartile range. A median is the midpoint of an ordered set of observations: half the observations are at or below it and half are at or above it, subject to ties. The interquartile range describes the central portion of the distribution, from the lower quartile to the upper quartile. It gives a sense of how dispersed the measurements are without requiring the values to follow a symmetric distribution.

Here, the salsalate-period median is 6.8 percentage vasodilation and the reported interquartile range is 5.2 to 12.0. The placebo-period median is 8.7 and the reported interquartile range is 5.4 to 13.5. These summaries describe each condition separately. They do not show how much an individual participant's measurement changed between periods. The registry does not report the distribution of paired differences, which is especially relevant in a crossover comparison.

Why does the crossover structure matter?

In general, a crossover study measures participants under more than one condition, so observations from the same person are related rather than independent. A paired analysis uses that relationship and can account for the fact that participants may differ from one another in their underlying measurements. The sequence groups in this trial received salsalate and placebo in opposite orders, with a 4 week washout between treatment periods.

General statistical considerations for crossover designs include treatment effects, period effects, sequence effects, and possible carryover from one period into the next. A washout is intended to separate periods, but its presence alone does not show whether carryover was assessed or ruled out. The registry record does not report an analysis of these design features, nor does it identify a specific paired test or model. No particular procedure can therefore be attributed to the trial from the registry information.

What can be concluded without an inferential analysis?

An inferential analysis usually estimates a contrast and quantifies uncertainty around it, or tests a specified hypothesis. A confidence interval, as a general concept, gives a range of values compatible with the data and the assumptions of the analysis; it is not a guarantee about the true value. A p-value, in general, describes how unusual data at least as incompatible with a specified null hypothesis would be under that hypothesis and the model. Neither type of result is reported for this primary outcome in the registry record.

Consequently, the medians and interquartile ranges allow a descriptive comparison but do not establish the statistical precision of a treatment contrast. The visible difference in the two medians is not itself a paired effect estimate. Nor can the interquartile ranges be treated as confidence intervals for that effect. Without participant-level paired data or a reported model result, the size and uncertainty of the within-person contrast cannot be recovered from the posted summaries.

What do the design and planning details say?

The registry describes random allocation to treatment sequence, quadruple masking, and a crossover model. Randomization applies to the order of treatment rather than creating a parallel comparison in which separate groups remain on different conditions throughout. Masking can reduce some forms of bias in measurement and conduct, but the record does not explain the masked roles or give operational details.

The registry lists a planning-illustration approach based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the target effect is expressed as a stated fraction of the observed effect. These assumptions describe a planning illustration, not the trial's original design justification or an observed analysis. No calculated participant counts for those illustrations are reported in the record. Such calculations, even when available, are planning quantities rather than observed power and do not replace the trial's outcome analysis. The record does not report the trial's own sample-size justification.

How should the analysis population be understood?

The trial enrolled 58 participants. The primary outcome description says ultrasound data for two participants were inadequate and were discarded after this determination was made before unblinding, leaving 56 for analysis. The outcome summaries list 56 for both salsalate and placebo. The registry does not report whether any other analysis populations were defined, whether all randomized participants were included in other analyses, or how missing measurements beyond these exclusions were handled.

Pre-unblinding assessment of data adequacy is useful context because the decision was made without knowledge of treatment assignment, according to the record. Still, exclusion can affect interpretation if excluded participants differ in relevant ways, and the registry does not provide details that would allow such an assessment. No sensitivity analysis or alternative missing-data analysis is reported.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and no analyses. It does not report whether the outcome was designated as primary in a statistical analysis plan beyond its listing as the primary outcome, or whether there were additional unreported hypotheses considered during the study. A multiplicity adjustment or strategy is not reported in the registry record. Without information about the planned family of tests, it is not possible to assess how multiplicity was addressed.

For the primary outcome, the record states that ultrasound data for two participants were inadequate for analysis. The determination was made before unblinding, and the data were discarded, leaving 56 participants for analysis. The treatment summaries each list 56. This is the available account of the missing primary outcome data. The registry does not report an imputation method, sensitivity analysis, reasons for inadequacy beyond the ultrasound data description, or the distribution of missingness by treatment sequence.

In a crossover trial, missing data in either period can complicate a paired comparison because both measurements may be needed to calculate a participant's within-person contrast. The record does not explain whether the two exclusions lacked data from one period or both, or whether another analysis population was used. This is not reported in the registry record.

The randomized population was 58, while the primary outcome analysis used 56. Safety figures are reported for 58 at risk in each treatment condition. The record does not give an intention-to-treat analysis, a per-protocol definition, or separate results for alternative populations. These distinctions matter because an analysis restricted to participants with usable outcome measurements may not answer precisely the same question as an analysis retaining all randomized participants.

11. 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 non-serious adverse event terms

The adverse-event record lists 58 participants at risk for serious adverse events in each treatment condition. It reports zero serious adverse events in the salsalate group and zero in the placebo group. Deaths are also listed as zero in both groups. The registry lists no top serious adverse-event terms.

For other adverse events, the record lists 6 affected participants among 58 at risk for salsalate and 0 among 58 at risk for placebo. These are counts as entered in the registry. They are not a statistical comparison, and the record does not report a test, confidence interval, severity breakdown for these events, or causal assessment. The term “other” should not be read as a specific event type.

In a crossover study, participants receive both treatments in different periods. Safety counts organized by treatment condition do not necessarily represent separate groups of people, and the registry does not explain how repeat events or participants exposed to both conditions were handled in these summaries. Zero reported serious events in the listed groups does not demonstrate that risk is absent; it describes what the registry records for this study.

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

The posted primary outcome data describe flow-mediated vasodilation measured after the salsalate and placebo treatment periods. The median was 6.8 after salsalate and 8.7 after placebo, with interquartile ranges of 5.2 to 12.0 and 5.4 to 13.5, respectively. The numerical ordering indicates that the reported median was lower after salsalate than after placebo. The registry record does not report a formal comparison or quantify uncertainty around a treatment contrast.

That distinction is important. A difference between two descriptive medians is not necessarily the same as a participant-level treatment effect in a crossover trial. Since the same participants were intended to receive both conditions, the association between their measurements affects the variability of a paired contrast. The summaries do not show that association, and the registry does not report a paired estimate, confidence interval, or test result. Statistical significance cannot be determined from the posted results.

The endpoint is a physiological measure of vascular function, not a direct measure of a clinical outcome. The study's stated purpose was to determine whether salsalate would improve blood vessel function by reducing inflammation in blood vessels. The posted values alone do not establish an improvement, a harmful effect, or a clinically meaningful change. They also do not show whether any observed measurement pattern corresponds to effects on symptoms or health events.

The adverse-event record lists zero serious adverse events and zero deaths in both treatment conditions, alongside 6 participants with other adverse events in the salsalate condition and 0 in the placebo condition. These event counts are descriptive and do not establish comparative safety. The registry does not provide event details or an inferential safety analysis.

The most defensible conclusion is limited: the registry reports outcome distributions for 56 analyzed participants and adverse-event counts for 58 at risk per treatment condition, but it does not report the statistical analysis needed to judge the precision of the primary comparison. The results should not be extended beyond these reported measurements and counts.

13. Why This Trial Matters Statistically

This trial illustrates why study design and analysis have to be read together. The crossover structure means that participants received both salsalate and placebo in separate periods, rather than remaining in distinct treatment groups. For statistical interpretation, the relationship between a participant's measurements across periods can be central. Separate medians summarize each treatment condition but do not reveal the distribution of within-person changes.

The study also shows why endpoint summaries are not interchangeable with treatment-effect estimates. A median and interquartile range give useful information about the observed measurements and their spread. They do not quantify uncertainty in a paired contrast, and they do not supply a formal test of the primary question. Here, the registry does not report an inferential analysis, confidence interval, or p-value. The descriptive ordering of the medians therefore cannot answer whether the observed contrast is statistically distinguishable from chance under a specified model.

Missing outcome data are another important part of the statistical picture. Two participants were removed from the primary outcome analysis because ultrasound data were inadequate, with the determination made before unblinding. That timing is a relevant design detail, but readers still need to know how missingness was addressed and whether conclusions change under alternative assumptions. Those analyses are not reported in the registry record.

Finally, the registry's planning-illustration assumptions should be distinguished from the trial's own design rationale and from observed evidence. A planning calculation is conditional on assumptions about the target effect, significance level, allocation, and statistical approximation. It is not observed power and cannot compensate for an absent analysis result. The trial is statistically informative as an example of how a completed randomized crossover study can have descriptive endpoint data while leaving key inferential details unreported.

14. 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 central limitation of the registry results is that no inferential analysis is listed for the primary outcome. The record reports medians and interquartile ranges but does not report a paired treatment contrast, confidence interval, p-value, or statistical model. The precision of the comparison and its compatibility with a range of possible effects therefore cannot be assessed from the registry record.

The outcome analysis excluded two participants whose ultrasound data were inadequate. The record says this was determined before unblinding, but does not explain whether data were unavailable for one or both treatment periods, whether the excluded participants differed from those analyzed, or whether sensitivity analyses were conducted. The analysis population is thus smaller than the randomized population, with no reported alternative analysis to show the effect of that choice.

Important crossover details are not reported, including analyses of treatment order, period effects, sequence effects, and possible carryover. The study included a 4 week washout, but the registry does not describe the rationale for that duration or report evidence about carryover. It also does not identify the statistical test or model used, or provide a prespecified analysis plan.

The record does not report a sample-size justification, multiplicity strategy, interim-analysis plan, or additional analysis populations. It lists one outcome measure and no analyses. The outcome is a physiological measure rather than a clinical event, and the registry does not connect the observed measurement summaries to patient-important outcomes. Safety reporting is limited to counts, without event descriptions for the other adverse events or an inferential comparison. These limitations constrain interpretation to the descriptive results recorded.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Inflammation and Vascular Function in Atherosclerosis — NCT00760019.
• Registry record last updated 2017-05-16; status last verified 2017-04.
• Results first posted 2017-04-06.
• 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.

18. Record Summary

NCT00760019, titled “Inflammation and Vascular Function in Atherosclerosis,” is a completed randomized, quadruple-masked crossover study sponsored by Brigham and Women's Hospital. The registry lists phase 2 and phase 3. Participants with atherosclerosis received 4 weeks of salsalate and 4 weeks of placebo in randomized order, separated by a 4 week washout.

The primary outcome was flow-mediated, endothelium-dependent vasodilation at the end of each treatment period. Ultrasound data for two participants were judged inadequate before unblinding, leaving 56 participants in the outcome analysis. The reported median was 6.8 after salsalate and 8.7 after placebo; the registry does not list an inferential analysis of the comparison.

For safety, the record lists zero serious adverse events and zero deaths in each treatment condition, with 6 participants reported with other adverse events for salsalate and 0 for placebo. These are descriptive counts. The registry does not report a statistical analysis plan, sample-size justification, or multiplicity procedure.