Carotid Occlusion Surgery Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Carotid Occlusion Surgery Study (NCT00029146), a randomized phase 3 study. The primary outcome is Surgical Group:Ipsilateral Ischemic Stroke in 2 Yrs From Randomization and All Stroke & Death Through 30d Post-surgery; Non-surgical Group:Ipsilateral Ischemic Stroke in 2 Yrs From Randomization and All Stroke & Death Through 30d Post-randomization.

2. Clinical Question

3. Trial Design

The Carotid Occlusion Surgery Study is listed as a phase 3, randomized, single-masked, parallel-group trial sponsored by the University of North Carolina, Chapel Hill. The registry lists enrollment as 700. The listed study start is 2002-07, primary completion is 2010-06, and the status is terminated. A reason for termination is not reported in the registry record.

Participants were assigned to a Surgical Group or a Non-surgical Group. The Surgical Group was assigned to undergo extracranial-intracranial arterial bypass in addition to best current practice medical therapy. The Non-surgical Group received best current practice medical therapy. The registry’s masking description does not identify who was masked, and further detail about the allocation process is not reported in the registry record.

The primary outcome record uses estimated two-year rates and reports an intention-to-treat population. Its arm entries use 97 participants for the Surgical Group and 98 for the Non-surgical Group. These outcome entries do not match the registry’s enrollment figure, and the reason for that difference is not reported in the registry record. The record lists 12 outcome measures and 11 analyses; it also lists 9 posted secondary outcomes. Further detail on how those outcome counts relate to each other is not reported in the registry record.

The study’s focus was a specific population with recently symptomatic carotid occlusion and increased oxygen extraction fraction on PET, rather than people with carotid disease generally. The registry record does not report the trial’s original sample-size justification, its detailed statistical analysis plan, or further operational details about conduct and follow-up.

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 principle.All participants analyzed in the group to which they were originally randomized.

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 Surgical Group:Ipsilateral Ischemic Stroke in 2 Yrs From Randomization and All Stroke & Death Through 30d Post-surgery; Non-surgical Group:Ipsilateral Ischemic Stroke in 2 Yrs From Randomization and All Stroke & Death Through 30d Post-randomization, measured over within 2 yrs of randomization.

The primary endpoint is an estimated proportion of participants experiencing the specified event by two years from randomization. The registry describes the values as two-year Kaplan-Meier estimates. Ipsilateral ischemic stroke is defined as a clinically diagnosed focal neurological deficit due to cerebral ischemia, localizable to the internal carotid artery territory distal to the symptomatic occluded artery, lasting more than 24 hours. The definition of all stroke includes a focal deficit due to ischemia or hemorrhage, localizable to the brain and lasting more than 24 hours; death is from any cause.

The outcome title specifies additional short-term criteria: for the Surgical Group, all stroke and death through 30 days after surgery; for the Non-surgical Group, all stroke and death through 30 days after randomization. Thus, the endpoint combines a two-year ipsilateral ischemic stroke estimate with group-specific short-term stroke and death criteria. That distinction matters when interpreting what counts as an event in each arm.

Kaplan-Meier estimation is commonly used for time-to-event outcomes because it can account for different lengths of observed follow-up and censoring. In general, censoring means that a participant’s event status is not observed through the entire period of interest. The registry describes the endpoint as a Kaplan-Meier estimate, but further details about censoring and the underlying analysis are not reported in the registry record. The endpoint is useful for assessing the cumulative occurrence of clinically important events over a stated period, rather than only comparing event counts without regard to event timing.

7. Results Posted in the Registry

• Non-surgical Group: 22.4% (22/98) versus 20.6% (20/97) on control; risk difference 1.8 percentage points (95% CI -9.7 to 13.4), risk ratio 1.09 (0.64–1.86); unadjusted two-proportion z-test p = 0.756.

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.

All Stroke

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Non-surgical Group: 26.5% (26/98) versus 23.7% (23/97) on control; risk difference 2.8 percentage points (95% CI -9.3 to 15.0), risk ratio 1.12 (0.69–1.82); unadjusted two-proportion z-test p = 0.650.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Non-surgical Group and the control. The interval gives the range of effects that remain compatible with the data.

Disabling Stroke

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Non-surgical Group: 2.0% (2/98) versus 6.2% (6/97) on control; risk difference -4.1 percentage points (95% CI -9.7 to 1.4), risk ratio 0.33 (0.07–1.59); unadjusted two-proportion z-test p = 0.145.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Non-surgical Group and the control. The interval gives the range of effects that remain compatible with the data.

Fatal Stroke

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Non-surgical Group: 2.0% (2/98) versus 1.0% (1/97) on control; risk difference 1.0 percentage points (95% CI -2.4 to 4.5), risk ratio 1.98 (0.18–21.47); unadjusted two-proportion z-test p = 0.567.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Non-surgical Group and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results add context, but they do not overturn the primary finding. Across the secondary outcomes with numerical results available, the estimated directions are not uniform: some estimates are higher in the Surgical Group and others are higher in the Non-surgical Group. Their reported confidence intervals include no difference, so these analyses do not provide a clear, precise separation between the groups.

The registry lists secondary outcomes covering all stroke, disabling stroke, fatal stroke, death, modified Rankin measures, a modified Barthel measure and a Summary SS-QOL Score. Numerical results for the remainder of these outcomes are not reported in the registry record. The visible pattern among the outcomes with reported comparisons is therefore limited: it should not be treated as a complete assessment of every listed secondary measure.

In general, secondary endpoints are useful for describing other dimensions of outcome and for identifying questions that may warrant further study. However, examining multiple outcomes increases the chance that at least one result will appear notable by chance. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting the secondary results. Consequently, unadjusted secondary comparisons carry less inferential weight than the primary analysis and should not be treated as independent confirmation of an effect.

Secondary measures may also differ in clinical meaning, event frequency and precision. A less frequent outcome can have a wide interval even when its point estimate appears notably different. Conversely, a result that is not statistically clear does not establish that the groups have identical outcomes. The secondary findings should be viewed as supplementary descriptions alongside the primary estimate, with attention to their uncertainty and the incomplete numerical reporting in the registry record.

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

• All Stroke
• Disabling Stroke
• Fatal Stroke
• Death
• Modified Rankin 0-1
• Modified Rankin 0-2
• Modified Barthel Index 19-20
• Summary SS-QOL Score

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is reported as a percentage based on a two-year Kaplan-Meier estimate, with a 95% confidence interval for each arm. In general, an estimate is the result calculated from the observed study data; it is not the true event rate known without uncertainty. A confidence interval gives a range of values that are reasonably compatible with the data and the method used to calculate it. It is not a guarantee that the true value lies in the interval, and it is not a probability statement about the treatment effect after the study has been completed.

What does the Kaplan-Meier estimate add?

In general, Kaplan-Meier methods estimate the probability of remaining free of an event over time while allowing participants to have different observed follow-up periods. Participants whose observation ends before an event is seen can be censored, rather than treated as if they experienced the event at the end of their observed follow-up. This approach depends on assumptions about censoring and follow-up. The registry identifies the estimates as Kaplan-Meier estimates but does not report further details of the censoring approach or its assumptions.

Because the primary outcome concerns event occurrence over time, an estimated rate can differ from a simple proportion calculated by dividing an event count by an arm denominator. A direct re-calculation from posted counts is therefore a useful descriptive check, but it does not replace the time-to-event analysis. In this record, the primary arm estimates are 21.0% and 22.7%, whereas the unadjusted calculation from posted counts gives a risk difference of 0.018. Those figures come from different summaries and should not be treated as interchangeable estimates.

How should the group difference and p-value be read?

The sponsor reports a difference in estimated two-year rates of 1.7, a 95% confidence interval from -10.4 to 13.8, and p = 0.78, using a two-sided z-statistic. A difference expresses a contrast between the group estimates. Its sign depends on the order in which groups are subtracted, so the direction should be interpreted using the stated group order and endpoint definitions. The sponsor-reported interval crosses zero, the value corresponding to no difference.

In general, a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the test assumptions were correct. It does not measure the probability that the null hypothesis is true, the probability that the result is due to chance, or the size or clinical importance of an effect. A p-value of 0.78 is not evidence that the groups are identical; it indicates that the analysis did not show a statistically clear difference under the reported test.

The posted unadjusted re-calculation has risk difference 0.018, 95% confidence interval -0.097 to 0.134 and p = 0.756. It is based on the posted counts and denominators, not the sponsor’s comparison of estimated two-year rates. This is a simple unadjusted calculation, so it does not capture the full time-to-event method. The registry record reports no discrepancy flag, but the sponsor’s specified trial analysis is still the analysis that should anchor interpretation of the primary outcome.

Why do confidence intervals matter alongside significance tests?

In general, an interval shows both the direction and the precision of an estimate. An interval that includes zero for a difference means that no difference is among the values compatible with the data under that interval method. The interval can also contain differences in either direction that could matter scientifically or clinically. A non-significant test alone does not distinguish between a genuinely small effect and an estimate too imprecise to resolve a meaningful effect.

Intervals for arm-specific estimates and intervals for a between-group difference answer different questions. The first describe uncertainty around each arm’s event estimate; the latter describe uncertainty around the contrast. Neither can be used alone to establish a causal explanation beyond the randomized comparison and its analysis. The record’s stated intention-to-treat population keeps participants classified by original assignment, which generally preserves the randomized comparison. Details of how missing outcome data were handled are not reported in the registry record.

Secondary outcomes provide additional context, but multiple outcomes create more opportunities for chance findings. Unless an analysis plan specifies how those tests are to be interpreted or adjusted, individual secondary p-values should be regarded as exploratory rather than as independent confirmation of the primary result. The registry record does not report a multiplicity adjustment or a detailed plan for interpreting secondary analyses.

11. Confidence Intervals and Precision

The sponsor-reported primary difference is 1.7, with a 95% confidence interval from -10.4 to 13.8. The interval crosses zero and spans values favoring either group. Its breadth means the estimate does not pin down the size or direction of a between-group difference precisely. The reported p-value, 0.78, is consistent with the interval’s inclusion of no difference, but the interval is more informative about the range of estimates compatible with the data.

The arm-specific primary estimates are also uncertain: 21.0% for the Surgical Group, with a 95% confidence interval from 12.8 to 29.2, and 22.7% for the Non-surgical Group, with a 95% confidence interval from 13.9 to 31.6. These are estimates of event occurrence under the registry’s endpoint definitions, not exact rates known without uncertainty. Their intervals overlap, but overlap between separate arm intervals is not a substitute for the confidence interval on the treatment contrast.

The unadjusted re-calculation likewise has an interval that includes zero: risk difference 0.018, with a 95% confidence interval from -0.097 to 0.134. That interval is based on posted summary counts rather than the sponsor’s two-year estimated-rate analysis. It should be interpreted as a separate descriptive calculation, not as a more precise version of the primary analysis.

In general, confidence intervals tend to be wider when the available information is limited or event estimates are uncertain. The registry’s different enrollment and primary arm denominator figures make it difficult to assess the completeness of the analyzed population; the reason for the difference is not reported in the registry record. The intervals quantify statistical uncertainty under their respective methods, but they do not account for every possible limitation, such as missing data or issues in how the study was conducted.

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 planning figures are expressed as participants per arm for 80% power and participants per arm for 90% power. They are not total event counts. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is set as a stated fraction of the observed effect. For time-to-event outcomes, the planning convention would instead use total events, but the primary outcome figures in this analysis are for a binary percentage endpoint.

The 80% power column describes the number of participants per arm a hypothetical study would need, under those assumptions, to have an 80% chance of detecting the chosen effect if that effect were the true one. The 90% power column uses a higher chance of detection and therefore generally requires more participants per arm for the same target effect. These are planning illustrations based on the observed effect, not the trial’s own original design assumptions; those original assumptions are not reported in the registry record.

Rows that plan for 75% or 50% of the observed effect ask what sample size would be needed to detect a smaller effect than the observed one. Smaller effects are harder to distinguish from random variation, so they require more participants per arm under the same planning assumptions. The values should not be read as evidence that the trial had a particular chance of finding its observed result. This is not observed power.

The estimates are sensitive to the effect size chosen for planning. An observed difference from a trial can be uncertain, so using it as a planning target does not guarantee that a future study will observe that effect. The figures are best read as conditional illustrations: if the assumptions and target effect were appropriate, the indicated number per arm would correspond to the stated power level. They do not establish the right design size for a future trial, and the registry record does not report the original sample-size rationale.

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 = 0.206; treatment proportion p₁ = 0.224; average p̄ = 0.215.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5813 = 1.1394.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5812 = 0.4891.
4. n = (1.1394 + 0.4891)² / (0.0183)² = 7,915.4, rounded up to 7,916 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 <- 0.206185567010309                          # control proportion
p1 <- 0.224489795918367                          # 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(29146)
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 = 7,916 (80% power), 10,596 (90% power)
 75% of observed effect: n per arm = 13,964 (80% power), 18,694 (90% power)
 50% of observed effect: n per arm = 31,173 (80% power), 41,732 (90% power)
power.prop.test: 7915.4 
Simulated power at n = 7,916 per arm: 0.799 (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-reported primary analysis compares estimated two-year rates using a two-sided z-statistic. It gives a difference of 1.7, a 95% confidence interval from -10.4 to 13.8, and p = 0.78. This is the trial’s reported analysis of the primary endpoint and should be the main reference for interpreting the study’s result.

The unadjusted re-calculation uses the posted event counts and denominators. It gives a risk difference of 0.018, a 95% confidence interval from -0.097 to 0.134, and p = 0.756. It is a simple calculation from summary data rather than a reproduction of the sponsor’s Kaplan-Meier comparison. Differences between these summaries can arise because an event-count calculation and an estimated time-to-event rate are not the same statistical quantity.

The registry record does not flag a discrepancy between the sponsor analysis and the re-calculation. The two results are broadly consistent in not showing a clear difference, but the sponsor analysis remains the appropriate estimate for the endpoint as defined and analyzed in the trial. The re-calculation is a check on the posted summary counts, not a replacement for the trial’s reported method.

15. Multiplicity, Missing Data and Analysis Populations

The registry record identifies a primary outcome and multiple secondary outcomes, but a detailed statistical analysis plan is not reported in the registry record. It also does not report whether the secondary comparisons were organized into a formal hierarchy or whether their p-values were adjusted for multiple testing. When many outcomes are examined, the chance of at least one apparently unusual result can rise. In general, multiplicity procedures control that issue by changing how evidence across several tests is evaluated. Without a reported procedure, secondary findings should be interpreted cautiously and not as confirmatory tests.

The primary population is described using the intention-to-treat principle: participants are analyzed in the group to which they were originally randomized. This preserves the randomized assignment in the analysis as described. However, the registry does not report how missing outcome data were handled, whether any participants had incomplete follow-up, or whether sensitivity analyses were used to assess missingness. Those details are important because missing information can affect estimates if it is related to outcomes or differs between groups.

The registry lists enrollment as 700, while the primary arm entries use denominators of 97 and 98. The registry record does not explain the relationship between these figures. It therefore does not allow a reader to determine from the posted summary alone why the primary analysis uses those denominators or how outcome availability relates to the listed enrollment. No explanation should be inferred from the difference.

Further details about interim analyses, stopping rules, protocol deviations and the original statistical analysis plan are not reported in the registry record. The study status is terminated, but the reason for termination is not reported in the registry record. These omissions limit assessment of how analyses were selected, how follow-up was completed and whether the posted results reflect all planned analyses.

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 0% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: non-fatal stroke (13.4%), transient ischemic attack (4.1%), respiratory distress (2.1%), epidural/subdural hematoma (2.1%), seizure (2.1%).

The serious adverse event entries report 14 affected participants among 97 at risk in the Surgical Group and 2 among 98 at risk in the Non-surgical Group. These are serious adverse event summaries, which are distinct from the primary endpoint estimates. They should not be substituted for the primary outcome or interpreted as a complete comparison of every possible harm.

The listed serious event terms include Non-fatal Stroke at 13.4%, Transient ischemic attack at 4.1%, Respiratory distress at 2.1%, Epidural/subdural hematoma at 2.1%, Seizure at 2.1%, and Fatal stroke at 1.0%. The registry’s term-level entries do not identify group-specific percentages for these terms, so they cannot be assigned to one group from the listed information.

The adverse event record lists no affected participants for other adverse events in either group. Death counts in the adverse event record are not reported in the registry record. The listed event terms and affected-participant summaries provide useful context, but the record does not give enough detail to establish the timing, duration, severity grading or causal attribution of each event.

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

The result is an estimate of the specified two-year outcome in the randomized comparison of bypass surgery plus best medical therapy with best medical therapy alone among the selected high-risk population. The sponsor-reported difference is uncertain: its confidence interval includes no difference and extends in both directions. The posted results therefore do not establish a clear difference between the groups on the primary endpoint.

This does not mean the groups have been shown to be equivalent, nor does it prove that the intervention has no effect. A result that does not show a statistically clear difference can reflect an effect that is small, an estimate with substantial uncertainty, or both. The registry does not report an equivalence margin or a non-inferiority analysis, so those conclusions cannot be drawn from the primary comparison.

The finding applies to the endpoint as defined in the registry, including its two-year timing and the differing short-term stroke and death windows in the two groups. It should not be generalized automatically to other populations, outcome definitions, or treatment settings. The record describes people with recently symptomatic carotid occlusion and increased oxygen extraction fraction on PET; it does not establish outcomes for populations outside that description.

The outcome result also does not provide a complete picture of safety, function or quality of life. Some serious adverse event information is listed, but group-specific event-term details and death counts are not reported in the registry record. Numerical results for some listed secondary outcomes are also not reported. Finally, the enrollment figure and primary outcome denominators differ, without an explanation in the registry record. These limits mean the posted summary is informative about the reported analysis but does not answer every question about benefit, harm, completeness or generalizability.

18. Why This Trial Matters Statistically

This trial is statistically informative because it tests a randomized comparison of adding a procedure to medical therapy in a specifically described high-risk population. Random allocation supports a comparison between assigned strategies, while the intention-to-treat description indicates that participants were analyzed according to their original assignment. These features help frame the result as a comparison of treatment strategies rather than as an uncontrolled association.

The outcome also illustrates why point estimates and uncertainty must be read together. The arm estimates are close, but the interval for the reported difference is broad enough to include values in either direction. Statistical interpretation is not simply a judgment about whether a p-value crosses a threshold: the size and precision of the estimate, the endpoint definition and the analysis method all matter.

The record further highlights the distinction between time-to-event estimates and direct calculations from event counts. Kaplan-Meier estimates account for the timing and censoring structure described for the primary outcome, while a simple unadjusted calculation from counts is a different summary. Comparing both can help readers understand the data, but the re-calculation cannot replace the trial’s reported analysis.

Finally, the study underscores the importance of complete reporting. The registry gives an enrollment figure, outcome denominators, a terminated status, and selected primary and secondary results, but does not explain the difference between enrollment and primary denominators or report several analysis details. These gaps matter statistically because they affect how confidently readers can assess precision, missing data, multiplicity and the scope of the conclusions.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves several trial-specific questions unresolved. It lists enrollment as 700, but the primary outcome arm entries use 97 and 98 participants; the reason for that difference and its relationship to outcome availability are not reported in the registry record. Missing-data methods, loss to follow-up and sensitivity analyses are also not reported, limiting assessment of whether incomplete outcome information could affect the estimates.

The primary endpoint has different short-term event windows by group: through 30 days after surgery for the Surgical Group and through 30 days after randomization for the Non-surgical Group, alongside the two-year ipsilateral ischemic stroke outcome. This is part of the registry’s stated endpoint definition and should be kept in view when interpreting the contrast. Further detail on how these components were operationalized in the analysis is not reported in the registry record.

The status is terminated, but a reason is not reported in the registry record. The record also does not report the original sample-size justification, interim analyses, stopping rules, or a detailed statistical analysis plan. The identity of the party masked in this single-masked trial is not reported. These omissions limit evaluation of the design assumptions, conduct and analysis choices.

Secondary results are only partially available as numerical comparisons, and multiplicity handling is not reported. Serious adverse event terms are listed without group-specific term rates, while adverse-record death counts are not reported. These limitations do not alter the posted primary estimate, but they restrict how fully the totality of efficacy and safety outcomes can be assessed from the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Carotid Occlusion Surgery Study — NCT00029146.
• Registry record last updated 2012-03-27; status last verified 2012-03.
• Results first posted 2012-03-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

The Carotid Occlusion Surgery Study was a phase 3 randomized, single-masked, parallel-group trial sponsored by the University of North Carolina, Chapel Hill. It compared extracranial-intracranial bypass surgery plus best medical therapy with best medical therapy alone in people with recently symptomatic carotid occlusion and increased cerebral oxygen extraction fraction measured by PET. The registry lists enrollment as 700, a study start of 2002-07, primary completion of 2010-06, and status as terminated.

The primary outcome was an estimated two-year rate based on the registry’s specified stroke and death criteria. The posted estimates were 21.0% in the Surgical Group and 22.7% in the Non-surgical Group. The sponsor-reported difference was 1.7, with a 95% confidence interval from -10.4 to 13.8 and p = 0.78; this does not show a clear difference between groups. The registry describes an intention-to-treat analysis. The reason for termination, missing-data methods and original sample-size justification are not reported in the registry record.