Acute Venous Thrombosis: Thrombus Removal With Adjunctive Catheter-Directed Thrombolysis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Acute Venous Thrombosis: Thrombus Removal With Adjunctive Catheter-Directed Thrombolysis (NCT00790335), a randomized phase 3 study. The primary outcome is Cumulative Incidence of Post-Thrombotic Syndrome (Villalta Scale).

2. Clinical Question

The clinical question concerns whether adding a catheter-directed procedure and thrombolytic drug to the treatment approach changes the later occurrence of post-thrombotic syndrome compared with standard anticoagulant therapy alone. The registry describes the target population as patients with symptomatic proximal deep vein thrombosis. The primary endpoint includes a Villalta score of 5 or greater, a leg ulcer, or a late endovascular procedure to treat severe venous disease in the index leg during the specified follow-up interval.

The posted primary results count participants with the endpoint in each randomized group. The intervention group had 157 participants with the endpoint among 336, and the control group had 171 among 355. The trial was randomized, but masking was none, so the study was open label. The registry identifies a modified full analysis set with intention-to-treat as the population for the primary outcome. The record does not report further details about how that modified set was defined or how any departures from the intended analysis population were handled.

3. Trial Design

ATTRACT was a phase 3, randomized, parallel-group trial with no masking. The registry records enrollment of 692 and status as completed. The study start is recorded as 2009-11, and primary completion as 2017-01. Washington University School of Medicine is listed as sponsor. The registry lists deep vein thrombosis and venous thrombosis among the conditions, as well as postphlebitic syndrome, venous thromboembolism, and post-thrombotic syndrome.

The intervention was pharmacomechanical catheter-directed thrombolysis with intrathrombus delivery of rt-PA. The intervention description gives a maximum allowable total dose of 35 mg and states that delivery could occur over a period of up to 24 hours. The registry also describes three methods of initial rt-PA delivery, but the posted arm description is incomplete after its first method. Further procedural details are not reported in the registry record.

The control arm received initial anticoagulant therapy with unfractionated heparin, enoxaparin, dalteparin, or tinzaparin for at least 5 days, overlapped with long-term oral warfarin targeting an international normalized ratio of 2.0 - 3.0. Elastic compression stockings were to be prescribed. The registry lists 336 participants in the intervention group and 355 in the control group for the primary endpoint. The record does not report the original sample-size justification, the randomization procedure beyond identifying the study as randomized, or further details of treatment adherence and crossover.

The primary analysis population is described as a modified full analysis set; intention-to-treat. The primary endpoint was evaluated between 6 and 24 months after randomization. The registry lists 27 outcome measures and 26 analyses. Those counts indicate a broad set of reported outcomes, but do not by themselves describe which analyses were planned in advance or how they were prioritized beyond the designated primary endpoint.

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: Modified full analysis set; 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.

Population: One patient was found not to meet the study inclusion criterion, and was excluded from all analyses by the adjudication committee.

6. Primary Endpoint

The registry lists the primary outcome as Cumulative Incidence of Post-Thrombotic Syndrome (Villalta Scale), measured over Between 6 and 24 months after randomization.

The primary endpoint is the cumulative incidence of post-thrombotic syndrome (PTS), measured in participants between 6 and 24 months after randomization. A participant met the endpoint if the index leg had a Villalta score of 5 or greater, a leg ulcer, or a late endovascular procedure to treat severe venous disease during the follow-up visits in that interval. The Villalta scale ranges from 0-33 points, with higher scores indicating worse symptoms or signs.

This composite definition captures more than a single examination score. It includes a threshold on the Villalta scale as well as clinically consequential events specified in the registry. Counting whether participants experienced any of these events gives a binary outcome: each participant either met the definition during the interval or did not. The registry describes the endpoint as a count of participants, rather than a mean score or a time-to-event measure.

As a general statistical concept, a prespecified endpoint gives a trial a primary outcome around which the main comparison is organized. A composite endpoint can capture several related ways an outcome may present, but its interpretation depends on the components included and on how those components contribute to the overall count. The registry definition identifies the components, but does not report component-specific primary results. The posted primary result therefore speaks to the combined endpoint as defined, not to each component separately.

7. Results Posted in the Registry

• A-Intervention: 46.7% (157/336) versus 48.2% (171/355) on control; risk difference -1.4 percentage points (95% CI -8.9 to 6.0), risk ratio 0.97 (0.83–1.14); unadjusted two-proportion z-test p = 0.704.

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.

Major Non-post-thrombotic Syndrome Treatment Failure

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • A-Intervention: 1.2% (4/336) versus 2.0% (7/355) on control; risk difference -0.8 percentage points (95% CI -2.6 to 1.1), risk ratio 0.60 (0.18–2.04); unadjusted two-proportion z-test p = 0.412.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between A-Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Any Treatment Failure

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • A-Intervention: 47.0% (158/336) versus 49.6% (176/355) on control; risk difference -2.6 percentage points (95% CI -10.0 to 4.9), risk ratio 0.95 (0.81–1.11); unadjusted two-proportion z-test p = 0.502.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between A-Intervention and the control. The interval gives the range of effects that remain compatible with the data.

Moderate-to-severe Post-thrombotic Syndrome

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • A-Intervention: 17.9% (60/336) versus 23.7% (84/355) on control; risk difference -5.8 percentage points (95% CI -11.8 to 0.2), risk ratio 0.75 (0.56–1.01); unadjusted two-proportion z-test p = 0.060.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between A-Intervention and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes provide a broader picture of the trial results, but they do not change the primary endpoint's central interpretation. Across the outcomes with posted comparisons, most analyses do not show clear statistical evidence of a difference under their reported tests. One secondary outcome, moderate-to-severe post-thrombotic syndrome, has a reported interval that excludes the null value for its risk ratio and a nominal p-value below 0.05. This finding stands out from the primary result, but it remains a secondary result rather than a substitute for the designated primary endpoint.

Other secondary comparisons have confidence intervals that include the null value and p-values that do not indicate a statistically significant difference under conventional testing. These outcomes cover treatment failure, bleeding, and recurrent venous thromboembolism, among other listed measures. Their direction and precision vary, and the record includes many outcome measures and analyses. The general pattern is therefore one of a largely inconclusive set of secondary comparisons, alongside the more favorable nominal result for moderate-to-severe post-thrombotic syndrome.

Multiplicity matters because testing many outcomes increases the chance that at least one result appears statistically significant by chance, even if no multiplicity adjustment is made. The registry record does not report a multiplicity-control procedure or a hierarchy for interpreting the secondary endpoints. The nominal secondary p-values should therefore be treated as exploratory or supportive evidence unless additional prespecified analysis details are available. This does not make a secondary finding meaningless; it changes how much weight it should carry and how confidently it can be generalized.

For interpretation, the primary endpoint remains the main result against which the trial question was organized. Secondary outcomes can add context, identify patterns, and motivate further investigation, but they generally have less confirmatory force when there are multiple comparisons and no reported adjustment. The contrast between the primary outcome and the moderate-to-severe secondary outcome is worth noting, while avoiding the claim that the secondary finding overturns or resolves the primary result.

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

• Major Non-post-thrombotic Syndrome Treatment Failure
• Any Treatment Failure
• Moderate-to-severe Post-thrombotic Syndrome
• Major Bleeding
• Major Bleeding
• Any (Minor + Major) Bleeding
• Any (Major + Minor) Bleeding
• Recurrent Venous Thromboembolism

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary at the participant level: the endpoint definition was either met or not met during the follow-up interval. The arm counts describe how many participants met the definition, while the arm denominators describe the numbers included in the posted endpoint results. The primary outcome population is identified as a modified full analysis set; intention-to-treat. The registry record does not provide additional details about the membership of that set.

What does a risk difference describe?

A risk difference compares the observed event proportions between groups by subtracting the control-group proportion from the intervention-group proportion. A value of zero represents no absolute difference; negative values indicate fewer observed events in the intervention group under this orientation. The recalculated primary risk difference is -0.014, with a 95% confidence interval from -0.089 to 0.06. The point estimate is close to zero, and the interval includes zero. The result therefore does not rule out either an absolute reduction or an absolute increase in the endpoint.

How does a risk ratio differ?

A risk ratio compares the event proportion in one group with the event proportion in the other. A value of 1 represents equal proportions; values below 1 indicate a lower observed event proportion in the intervention group when it is the numerator group. The registry-reported primary risk ratio is 0.96, with a 95% confidence interval from 0.82 to 1.11. The interval includes 1, so the data are compatible with no relative difference as well as with effects in either direction. Risk differences and risk ratios answer related but distinct questions: one describes an absolute contrast, while the other describes a relative contrast.

What does the reported test add?

The registry identifies the Cochran-Mantel-Haenszel method for the reported risk-ratio analysis. The corresponding p-value is 0.56. As a general statistical concept, a p-value measures how unusual the observed data, or data more extreme, would be under a specified null model and analysis. It is not the probability that the null hypothesis is true, and it does not quantify the clinical importance of an effect. The registry record does not report additional details about any stratification variables or other implementation choices for this method.

The recalculated primary comparison is an unadjusted calculation from the posted arm summaries. It gives a risk difference and associated interval and p-value; the registry-reported analysis uses a risk ratio and the Cochran-Mantel-Haenszel method. These figures are not identical estimands or necessarily identical analyses. They provide complementary descriptions, with the sponsor's posted analysis serving as the primary reported inferential result. No discrepancy is flagged in the posted results.

Why can interval estimates be more informative than a test alone?

A test result compresses a comparison into a statistic such as a p-value, whereas a confidence interval displays a range of effect estimates compatible with the data and model at its stated confidence level. In general, a narrow interval indicates greater precision than a wide interval, but precision is not the same as importance or freedom from bias. For the primary comparison, the reported interval includes 1 and spans both lower and higher relative risks. That range makes clear why the point estimate should not be interpreted in isolation.

The secondary outcomes are also binary participant counts. The registry reports risk ratios with confidence intervals and p-values for the listed secondary analyses. The recalculated summaries also express comparisons as risk differences, but these are unadjusted calculations from the posted summaries. The registry's primary method and reported analysis should be distinguished from these recalculations rather than treated as interchangeable.

The registry record labels the primary population modified full analysis set; intention-to-treat. In general, an intention-to-treat approach seeks to preserve randomized group assignment for analysis, while a modified analysis set may apply specified eligibility or data-availability rules. The details of any such modification, handling of missing primary outcomes, and sensitivity analyses are not reported in the registry record. Those omissions limit how fully the analysis population and robustness of the primary result can be assessed from the record alone.

The trial was open label, with masking listed as none. As a general consideration, lack of masking can affect behavior, care decisions, or subjective assessments, depending on the outcome and safeguards used. The registry record does not report additional information about blinded outcome assessment or procedures intended to limit this risk. The primary endpoint includes a scale threshold and events such as procedures, so interpretation should consider that endpoint assessment occurred in an unmasked trial, without assuming that bias did or did not occur.

11. Confidence Intervals and Precision

The registry-reported primary risk ratio is 0.96, with a 95% confidence interval from 0.82 to 1.11. The interval includes 1, the value corresponding to no relative difference. Its lower and upper limits allow for effects in either direction, so the estimate is not sufficiently precise to establish a single direction of effect from this analysis alone.

The unadjusted recalculated risk difference is -0.014, with a 95% confidence interval from -0.089 to 0.06. Zero is within this interval. In absolute terms, the observed estimate is near no difference, but the range extends across both negative and positive differences. The recalculated result is based on posted summary counts and should be read as a complementary, unadjusted summary rather than a replacement for the registry-reported analysis.

In general, a confidence interval reflects sampling uncertainty under the assumptions of the analysis. It does not account automatically for every possible source of bias, nor does it assign a probability to the true effect being inside the particular interval. A confidence interval also should not be used alone to judge clinical importance: that assessment requires context about what differences would matter, and such a threshold is not reported in the registry record.

The main precision message is that the primary estimate does not distinguish clearly between no difference and effects in either direction. A nonsignificant p-value does not mean the estimate is exact, and the interval conveys the uncertainty more directly than a binary significant/not-significant label. The registry record does not report a prespecified equivalence or noninferiority margin that would permit the interval to be judged against such a claim.

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 the binary outcomes in this trial. They are not total-event counts. The planning illustration uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It uses the observed effect as its starting point and examines planning for 100%, 75%, and 50% of that observed effect. The original trial's own design assumptions and sample-size justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the illustration estimates would be needed for an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding planning amount for a 90% chance. These are prospective planning quantities based on assumed effect size and analysis conditions, not descriptions of what the completed trial achieved. The planning figures are not observed power.

Reading across the effect assumptions shows why smaller target effects require more participants. Planning to detect 75% or 50% of the observed effect entails a larger per-arm requirement than planning for the full observed effect, because smaller contrasts are harder to distinguish from random variation under the same general framework. The figures are illustrations rather than evidence that a trial with those sizes would necessarily reproduce the observed outcome.

The curve should be interpreted in the same planning context: it relates the assumed effect size to the amount of information needed under the stated assumptions. It does not show treatment efficacy or establish that the observed effect is stable. In general, planning calculations depend on assumptions about the outcome rate, allocation, analysis, and target effect; results can differ if those assumptions do not hold. The registry record does not report the original design calculation needed to compare these illustrations with the trial's initial planning.

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 = 0.482; treatment proportion p₁ = 0.467; average p̄ = 0.474.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7062 = 1.3841.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7061 = 0.5943.
4. n = (1.3841 + 0.5943)² / (-0.0144)² = 18,801.4, rounded up to 18,802 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 <- 0.48169014084507                          # control proportion
p1 <- 0.467261904761905                          # 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(790335)
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 = 18,802 (80% power), 25,170 (90% power)
 75% of observed effect: n per arm = 33,438 (80% power), 44,763 (90% power)
 50% of observed effect: n per arm = 75,261 (80% power), 100,753 (90% power)
power.prop.test: 18801.4 
Simulated power at n = 18,802 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor's posted primary analysis is a Cochran-Mantel-Haenszel risk ratio of 0.96, with a 95% confidence interval from 0.82 to 1.11 and p = 0.56. The unadjusted recalculation from the arm summaries expresses the result as a risk difference of -0.014, with a 95% confidence interval from -0.089 to 0.06 and p = 0.704. The recalculation is a simple summary based on the posted counts; it is not the same analysis scale or method as the registry-reported result.

The posted results do not flag a discrepancy. The sponsor's analysis therefore remains the primary reported inferential result, while the recalculation can help readers see the absolute contrast and how it relates to the event counts. Both descriptions are consistent in the broad statistical message: the estimates are close to no difference, and their intervals include values corresponding to no difference. Neither should be read as proof that the groups are equivalent.

The registry record does not report enough detail to reconstruct all features of the sponsor's model or its implementation. For that reason, the recalculation should not be used to suggest that the sponsor's analysis was incorrect or to substitute a different result for the reported one. Its role is limited to an unadjusted perspective on the posted summary statistics.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 27 outcome measures and 26 analyses, including a designated primary endpoint and multiple secondary outcomes. The record does not report whether the secondary analyses were ordered in a prespecified hierarchy, whether family-wise error or another multiplicity procedure was used, or whether any outcomes were intended to be exploratory. Without that information, secondary p-values should be interpreted as nominal results rather than as evidence that necessarily accounts for the full set of comparisons.

As a general statistical concept, multiplicity arises when several hypotheses are tested. Even when each individual test uses a conventional significance threshold, conducting many tests increases the chance of at least one apparently positive result under a setting where all null hypotheses are true. Adjustments or a prespecified testing strategy can address this issue, but the registry record does not report such a strategy. This is particularly relevant to the secondary moderate-to-severe post-thrombotic syndrome result, which has a nominal p-value of 0.04.

The primary population is described as a modified full analysis set; intention-to-treat. The record does not define the modification, explain how participants with missing outcome data were handled, or report sensitivity analyses for missingness. It also does not report the amount or pattern of missing primary outcome data. These omissions make it difficult to assess whether missingness could have affected the estimates or whether conclusions are robust to alternative assumptions.

The registry identifies a randomized trial but does not report details of the randomization sequence, allocation concealment, or any stratification used. It identifies the study as unmasked, but does not report whether outcome assessors were masked or how endpoint adjudication was conducted. These details can matter when assessing risk of bias, particularly for outcomes involving clinical assessments or procedures. Their absence from the registry record does not establish that safeguards were absent; it means the record does not permit their evaluation.

The record also does not report interim analyses, stopping rules, or the original sample-size justification. No conclusions about those design features should be inferred from the completed-trial results. Together, the missing information limits assessment of multiplicity, missing-data robustness, and some features of trial conduct, while leaving the posted primary result interpretable as reported.

16. 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

Serious events with the highest arm-level rate: deep vein thrombosis (7.4%), pulmonary embolism (3.0%), chest pain (2.8%), gastrointestinal bleed (2.4%), cellulitis (1.8%).

The registry reports serious adverse events in 109 of 336 participants at risk in the intervention group and 86 of 355 participants at risk in the control group. Deaths are listed for 7 participants in the intervention group and 8 in the control group. The registry also reports zero for other adverse events in the corresponding arm fields. These figures describe the posted safety counts; they do not establish that every event was caused by the assigned treatment.

The listed serious-event terms include deep vein thrombosis at 7.4%, pulmonary embolism at 3.0%, chest pain at 2.8%, gastrointestinal bleed at 2.4%, cellulitis at 1.8%, and stroke at 1.8%. These are registry-reported percentages, and the record does not give further detail in these summary fields about event severity beyond the serious-event classification, timing, recurrence, or causal attribution.

Safety counts should be read in relation to the numbers at risk and the way events were collected. The registry record does not report a comparative statistical analysis for these serious adverse events, nor does it provide enough detail to interpret differences as causal. A descriptive difference in counts alone cannot establish the relative safety of the treatment approaches.

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

The primary result shows no statistically significant difference in the registry-reported comparison of cumulative post-thrombotic syndrome between 6 and 24 months after randomization. The estimated risk ratio is 0.96, and its 95% confidence interval includes 1. The unadjusted risk-difference estimate is -0.014, with an interval that includes zero. Taken together, these results do not provide clear evidence of a difference on the primary endpoint in this analysis.

That conclusion is narrower than saying the two treatments are the same. A failure to demonstrate superiority is not proof of equivalence, and the registry record does not report an equivalence or noninferiority margin. The confidence intervals also include effects on either side of no difference, so the result does not eliminate the possibility of a meaningful difference in either direction. Whether any range of effects would be clinically important cannot be determined from the registry summary alone.

The primary endpoint combines a Villalta score threshold with leg ulcer and late endovascular procedure events. Its result pertains to that composite definition and follow-up interval. It does not, by itself, show the effect on every component separately, on outcomes outside the interval, or on patient experiences not captured by the endpoint. The registry record does not provide component-specific primary results.

A secondary analysis of moderate-to-severe post-thrombotic syndrome has a nominally significant reported result, but it should be interpreted in the context of multiple outcomes and the absence of a reported multiplicity procedure. It does not replace the primary analysis. Likewise, descriptive adverse-event counts are important context but do not establish comparative causal safety from the summary counts alone.

In general, trial results are estimates with uncertainty, not guarantees about an individual participant's outcome. These findings describe group-level comparisons in the analyzed participants and do not provide medical advice. Further judgments about clinical importance, applicability to other populations, or individual treatment choices require information beyond the registry results.

18. Why This Trial Matters Statistically

This trial is statistically informative because it distinguishes a point estimate from the uncertainty around that estimate. The primary risk ratio is close to 1, but the 95% confidence interval extends on both sides of the no-difference value. Reporting the interval alongside the p-value makes clear that the evidence does not sharply identify one direction of effect, even though the point estimate favors a lower observed endpoint proportion in the intervention group.

It also illustrates why absolute and relative measures can complement each other. The registry's principal analysis reports a risk ratio, while an unadjusted recalculation expresses the observed contrast as a risk difference. The relative measure places the group comparison on a ratio scale; the absolute measure describes the difference in event proportions. Both are useful descriptions, but they answer different questions and should not be conflated.

The secondary results demonstrate a further statistical issue: an apparently positive secondary comparison can coexist with a primary result that is not statistically significant. With multiple outcomes, interpretation depends on prespecification, multiplicity control, and the relative priority assigned to each endpoint. The registry record does not report a strategy for controlling multiplicity, making careful separation of primary and secondary evidence especially important.

The trial also provides a practical example of how to read registry results without treating a nonsignificant test as proof of no effect. The primary confidence interval includes a range of possible effects, and the registry does not report an equivalence margin. At the same time, the endpoint's definition, analysis population, open-label status, and missing details about data handling all inform how confidently the result can be generalized. Statistical interpretation requires the estimate, its precision, the endpoint, and the design context together.

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.

Several trial-specific limitations affect interpretation of the registry results. The study was unmasked, and the record does not report whether endpoint assessors were blinded or whether endpoint events were adjudicated independently. The primary endpoint includes clinical assessment and a late procedure component, so missing information about assessment safeguards limits evaluation of possible ascertainment effects without establishing that such effects occurred.

The primary population is called a modified full analysis set; intention-to-treat, but the registry does not define the modification. It also does not report the extent or pattern of missing primary outcome data, the methods used to handle missingness, or sensitivity analyses. As a result, the record alone does not show how robust the primary estimate is to alternative analysis-population or missing-data assumptions.

The registry lists numerous outcomes and analyses, yet does not report a multiplicity strategy or a hierarchy for secondary outcomes. This makes the nominal secondary result for moderate-to-severe post-thrombotic syndrome less definitive than a primary, multiplicity-controlled result would be. The record also does not report the original sample-size justification, interim analyses, stopping rules, or further detail about the randomization and allocation procedures.

The intervention description is incomplete in the registry summary after beginning to describe the methods of initial rt-PA delivery. Further details about protocol implementation, adherence, and treatment crossovers are not reported in the record. The primary endpoint is a composite, and component-specific primary results are not reported. These limits constrain assessment of the treatment contrast and the contribution of individual endpoint components.

These are limitations of what can be assessed from the registry record, not proof that the trial lacked additional procedures or information. They warrant caution when generalizing beyond the posted outcomes and analyses.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Acute Venous Thrombosis: Thrombus Removal With Adjunctive Catheter-Directed Thrombolysis — NCT00790335.
• Registry record last updated 2018-03-29; status last verified 2018-02.
• Results first posted 2018-03-29.
• 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

ATTRACT was a phase 3, randomized, parallel-group trial with no masking, sponsored by Washington University School of Medicine. The registry records enrollment of 692 and completed status. The intervention was PCDT with intrathrombus rt-PA; the comparator was anticoagulant therapy with compression stockings prescribed.

The primary endpoint was cumulative incidence of post-thrombotic syndrome between 6 and 24 months after randomization. The registry-reported risk ratio was 0.96, with a 95% confidence interval from 0.82 to 1.11 and p = 0.56. The result did not show a statistically significant difference on the primary endpoint. A secondary moderate-to-severe post-thrombotic syndrome analysis had a nominally significant result, but multiplicity handling is not reported.

The posted results support an interpretation centered on uncertainty: the primary interval includes no relative difference and effects in either direction. The registry record does not establish equivalence, and several details about missing data, analysis populations, and multiplicity are not reported.