Prospective Randomized Endovascular Therapy in Multiple Sclerosis - PREMiSE: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Prospective Randomized Endovascular Therapy in Multiple Sclerosis - PREMiSE (NCT01450072), a randomized na study. The primary outcome is Percentage of Patients With Immediate and Short-term SAE.

2. Clinical Question

The clinical question recorded for PREMiSE is whether therapeutic balloon angioplasty, compared with venography and sham angioplasty, can be evaluated for safety and effectiveness in the context of venous narrowing in multiple sclerosis. The posted primary endpoint addresses safety, not effectiveness: it counts participants with severe adverse events. A secondary outcome concerns participants with more than 75% of normal outflow at 1 year post-angioplasty, but a result for that outcome is not reported in the registry record.

These are distinct questions. The safety count describes observed events within the stated timeframe; it does not itself establish whether the intervention changes clinical outcomes or venous outflow. The record identifies randomized allocation and a parallel design, but does not report the detailed procedures for assignment, follow-up, or endpoint adjudication. Those omissions limit how specifically the registry entries can be interpreted.

3. Trial Design

PREMiSE is identified as a randomized, single-masked, parallel-group study. The registry lists an enrollment of 20 and marks the study completed. Its start is recorded as 2010-06, with primary completion recorded as 2011-12. The phase is listed as NA. The sponsor is the State University of New York at Buffalo.

The active arm is described as therapeutic balloon angioplasty, and the control arm as venography and sham angioplasty. The intervention listing names selective venography followed by therapeutic balloon angioplasty. These descriptions establish the intended contrast, but the registry record does not report procedural details such as how the sham was conducted or how treatment assignment was concealed.

Single masking is recorded, but the identity of the masked party or parties is not reported in the registry record. The method used to generate or implement random allocation is also not reported. The record does not specify whether the primary analysis followed randomized assignment regardless of treatment received, or describe any other analysis population.

The posted arm counts for the primary endpoint are 10 participants in the control arm and 9 in the active arm. The registry enrollment is 20. These entries refer to different record fields; the posted results do not explain how the arm counts relate to the enrollment figure or describe the status of any participant not represented in those arm counts. Reasons for missing or unavailable outcome data are not reported in the registry record.

The record lists two outcome measures and one secondary outcome. The primary outcome is the percentage of patients with immediate and short-term severe adverse events, with the timeframe recorded as 24 hours-1 month. The secondary outcome concerns participants with more than 75% of normal outflow at 1 year post-angioplasty. The record does not include posted analyses for the outcomes.

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

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 Percentage of Patients With Immediate and Short-term SAE, measured over 24 hours-1 month.

The primary endpoint is titled “Percentage of Patients With Immediate and Short-term SAE.” SAE refers in the description to severe adverse events. The unit is participants, and the outcome is defined as a count of participants with an event. Its timeframe is recorded as 24 hours-1 month, while the description says that short-term safety is measured at 1 month and refers separately to immediate and short-term SAE rates.

This endpoint is used to describe safety by recording whether participants experienced severe adverse events during the specified period. Such counts can identify events that occurred in the study arms, but the registry entry does not provide event definitions, grading rules, attribution procedures, or details of how events were collected. The timing language also leaves the precise distinction between the immediate and short-term assessments unclear in the posted result.

The posted primary counts are zero for the control arm, with 10 participants, and zero for the active arm, with 9 participants. The registry description says that exact confidence intervals were to be obtained for the immediate and short-term SAE rates and that the rates were to be analyzed with Fisher's exact test. Those confidence intervals and test results are not reported in the registry record.

The endpoint is a safety measure, not a direct measure of effectiveness. The listed secondary outcome about normal outflow is separate, and its result is not reported in the registry record. No inference about that outcome can be drawn from the posted severe-adverse-event counts.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary at the participant level: a participant either is counted as having a severe adverse event or is not. The registry lists the arm outcomes as counts of participants, rather than as continuous measurements. The posted values are zero events in each arm. These are descriptive results; a comparison requires an inferential method as well as clear information about which participants and time windows entered the analysis.

What does an exact confidence interval convey?

A confidence interval is a range of values produced by a statistical procedure to express uncertainty around an estimate. An exact interval is based on the discrete structure of event counts rather than relying on a large-sample approximation for the endpoint estimate. This can be useful for sparse binary outcomes, where standard approximations may be unreliable. The registry description says exact confidence intervals were planned for the immediate and short-term SAE rates. The intervals themselves are not reported in the registry record, so their width and the precision of the rate estimates cannot be assessed from the posted results.

A zero observed count does not make uncertainty disappear. When an event is uncommon, a study may record no events even though the underlying event probability is not known with certainty. The interval is important because it communicates how much uncertainty remains around an observed rate. No interval bounds are available in the registry record, and none can be inferred from the count alone without performing a calculation not posted in the record.

Why use Fisher's exact test?

Fisher's exact test is a method for comparing event proportions in a contingency table. It evaluates how compatible the observed allocation of events across arms is with a specified null hypothesis of no association, using the discrete distribution of possible tables conditional on the margins. The registry description names this test for the immediate and short-term SAE rates.

With sparse event data, including zero counts in the posted arms, a method based on exact probabilities avoids relying on the usual large-sample approximation to a test statistic. The method does not eliminate the limitations of the data: a test result still depends on the observed counts and the analysis definition, and a non-significant result would not prove equality. No Fisher test statistic or p-value is reported in the registry record.

What can be said about a recalculated comparison?

The record indicates that values are missing or not estimable for a two-arm comparison of the binary endpoint. Consequently, no recalculated comparative estimate or test should be presented as if it were a posted result. Recalculations from summary statistics, when available, are unadjusted and do not replace the sponsor's own analysis. The record does not report a sponsor comparison for this endpoint, so there is no sponsor estimate to set against a recalculation.

General statistical interpretation also distinguishes an effect estimate from a test of statistical significance. An estimate describes the direction and size of an observed contrast, while a confidence interval expresses uncertainty around that estimate. A p-value, if reported, would describe compatibility with a null model under the test assumptions; it would not measure the probability that a treatment works. None of these comparative results is included in the registry record.

How should planning illustrations be read?

Planning illustrations are based on assumptions rather than on observed trial performance. The stated assumptions for any such illustration are a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and an effect to detect defined as a stated fraction of the observed effect. For non-time-to-event outcomes, planning results are expressed as participants per arm for a specified power. These assumptions do not describe a reported trial design calculation, and the illustrations are not observed power.

Observed power is not a substitute for an effect estimate and its uncertainty. In general, power is a design-stage probability calculated under an assumed effect and analysis plan. A planning illustration based on an observed effect is especially dependent on that observed estimate and should not be read as proof that the study was adequately or inadequately powered. The registry record does not report a sample-size justification or its original planning assumptions. The posted comparison is marked as not estimable, so a planning figure based on a calculated observed effect should not be treated as a result.

Why does over-dispersion matter in other outcome types?

Over-dispersion is a general concern for count outcomes when observed variability is greater than a model assumes. If ignored, standard errors may be too small and intervals or tests may appear more precise than warranted. That issue is not the stated primary endpoint: the primary measure is a participant-level binary event count, and the registry description specifies exact intervals and Fisher's exact test. The registry does not provide a count model or a dispersion estimate for the primary outcome. The general concept should not be confused with evidence that over-dispersion was present in this study.

In short, the registry names suitable exact methods for sparse binary safety data, but it does not post the resulting uncertainty intervals or comparison test. Method names establish what was described as planned; they do not supply missing statistical results.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Control Arm", "Active Arm")
n      <- c(10, 9)
events <- c(0, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Control Arm  10  0  0.0  0.0  30.8  0.0  27.8  0.2  28.5
ARM  Active Arm  9  0  0.0  0.0  33.6  -0.0  29.9  0.3  30.8

11. Multiplicity, Missing Data and Analysis Populations

The record lists two outcome measures and identifies one secondary outcome, concerning participants with more than 75% of normal outflow at 1 year post-angioplasty. The registry record does not report a result for that secondary outcome. It also does not describe a multiplicity strategy, such as whether inference across outcomes was adjusted or which outcome comparisons were intended to be confirmatory. The order in which outcomes are listed is not enough to establish how multiple testing was handled.

The primary endpoint description refers to immediate and short-term SAE rates and says that each would receive an exact confidence interval and a Fisher's exact test. The record does not clarify whether these time periods were separate analyses, how any multiplicity across them was handled, or which result would be considered primary if they differed. A stated test plan should not be mistaken for a complete statistical analysis plan.

Information about missing data is not reported in the registry record. The available entries do not describe withdrawals, incomplete follow-up, missing endpoint status, or any imputation or sensitivity analysis. The arm counts associated with the primary outcome are 10 for control and 9 for active, while the registry enrollment is 20; the record does not explain the difference between these entries. It is not possible to assume that every enrolled participant contributed a complete endpoint assessment.

The analysis population is also not reported. The record does not say whether results were analyzed according to randomized assignment, treatment received, or another rule. It does not report exclusions, protocol deviations, or how participants with unavailable data were handled. These details matter because they define which observations support an estimate and what population that estimate describes. With no analysis details or secondary result posted, conclusions should remain limited to the recorded counts and the explicitly documented design.

12. 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 non-serious adverse event terms

The primary safety outcome records zero participants with severe adverse events in the control arm and zero in the active arm. The corresponding posted arm counts are 10 and 9 participants. These counts describe the registry entries for the stated endpoint; they are not a guarantee that severe events cannot occur, nor do they quantify the uncertainty around the event rate.

The adverse-event entries also list three other adverse events in the control arm and three in the active arm. The record distinguishes these from serious adverse events, but does not provide detailed terms for the other events in the registry record describing the results. Death counts are not reported. No serious-event terms are listed, and the primary serious-event count is zero in each arm; the available record does not provide a fuller event narrative.

The timeframe is recorded as 24 hours-1 month, while the endpoint description refers to immediate and short-term rates and measurement at 1 month. The registry does not report separate event counts for each interval or explain the precise handling of participants whose follow-up was incomplete. The listed counts should therefore be interpreted within the stated reporting boundaries. They describe observed entries in the registry, not a complete estimate of all possible safety risks or a comparison supported by reported confidence intervals and tests.

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

The result establishes a limited descriptive fact about the posted primary outcome: the control arm and active arm each have a recorded count of zero participants with severe adverse events. It also establishes that the registry lists three other adverse events in each arm. These entries are useful for understanding what the registry reports, but they do not by themselves resolve the comparative safety question.

The result does not show that the underlying risk is zero. An observed count is based on the participants and follow-up represented in the record, and the record does not report confidence intervals for the SAE rates, a comparative effect estimate, or a test result. The primary outcome description says that exact confidence intervals and Fisher's exact tests were planned, but planned analyses cannot be interpreted as completed findings.

Nor does the result establish effectiveness. The study summary describes evaluation of safety and effectiveness, but the primary endpoint is a safety outcome. The secondary outcome about more than 75% of normal outflow at 1 year is listed without a posted result in the registry record. The safety counts cannot answer that secondary question or establish a clinical benefit.

Interpretation is further limited by unreported details: the registry does not explain how enrollment relates to the posted arm counts, whether outcome data were complete, which analysis population was used, or how the immediate and short-term time windows were handled. These uncertainties do not negate the reported counts, but they constrain the conclusions that can be drawn from them.

Thus the posted information supports a cautious summary of the recorded safety observations, not a definitive judgment about the intervention. It does not show that angioplasty is effective, ineffective, safe, or unsafe. More complete outcome reporting and analysis details would be needed to assess the magnitude and precision of any between-arm difference.

14. Why This Trial Matters Statistically

PREMiSE is statistically notable because its primary endpoint is a binary safety outcome with zero severe adverse events recorded in each arm. Sparse event data make uncertainty especially important: a point count alone can look definitive while providing little information about the range of plausible underlying rates. The registry describes exact confidence intervals and Fisher's exact test, methods suited to discrete binary data, but does not report their results.

The record also illustrates why a randomized design and a posted result are not the same as a fully interpretable comparison. Random allocation is recorded, and the arms are described, but a comparative effect estimate, confidence interval, and test output are not reported. Without these, readers cannot assess the precision of the contrast or distinguish a small observed difference from a meaningful underlying difference. In this record the event counts are equal, but equal observed counts do not establish equivalence.

Safety reporting also requires clear definitions and follow-up accounting. The primary timeframe is stated as 24 hours-1 month, while the description refers to immediate and short-term rates and assessment at 1 month. The registry does not clarify how those intervals map to the posted values. The relationship between enrollment and the posted arm counts is also not explained, and missing-data handling is not reported.

These features make the record a useful example of the limits of interpreting counts without their uncertainty, analysis population, and outcome definitions. The listed secondary outflow outcome has no posted result, so the safety entries cannot stand in for effectiveness evidence. Statistically, the central lesson is that a zero count is a descriptive observation; its interpretation depends on follow-up, precision, and transparent analysis reporting.

15. 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 reporting gaps limit interpretation. The registry identifies single masking but does not specify who was masked. It records random allocation but does not describe the sequence-generation or assignment procedures. Details of the sham angioplasty and the venography procedures are not reported in the registry record, limiting assessment of how closely the comparator matched the active procedure.

The primary endpoint has potentially distinct immediate and short-term components, yet the listed timeframe is 24 hours-1 month and the description also mentions measurement at 1 month. The record does not report separate results for the time periods or explain how the entries were combined. It does not supply operational definitions for severe adverse events, event adjudication procedures, or attribution rules.

The registry enrollment is 20, while the posted primary arm counts are 10 and 9. The record does not explain the relationship between these figures. It also does not describe completeness of follow-up, missing outcome status, withdrawals, or the analysis population. This makes it difficult to determine the scope of the population represented by the posted outcome counts.

No confidence intervals, comparative estimate, Fisher test result, or result for the secondary outflow outcome is reported. Multiplicity handling, sample-size justification, and the complete statistical analysis plan are not reported in the registry record. These are limits of what can be concluded from the registry entries; they do not establish that the underlying study lacked procedures or analyses that were not recorded.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Prospective Randomized Endovascular Therapy in Multiple Sclerosis - PREMiSE — NCT01450072.
• Registry record last updated 2025-06-18; status last verified 2025-06.
• Results first posted 2025-06-18.
• 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.

19. Record Summary

PREMiSE (NCT01450072) is listed as a completed, randomized, single-masked, parallel study sponsored by the State University of New York at Buffalo. The registry records enrollment of 20, a start of 2010-06, and primary completion of 2011-12. The phase is listed as NA.

The study concerns multiple sclerosis and compares therapeutic balloon angioplasty with venography and sham angioplasty. Its primary endpoint is the count of participants with immediate and short-term severe adverse events, with the timeframe stated as 24 hours-1 month. The posted counts are zero in the control arm and zero in the active arm. Other adverse-event counts are three in each arm.

The record describes exact confidence intervals and Fisher's exact tests for the primary safety rates, but does not report their results or a comparative estimate. A secondary outcome on normal outflow at 1 year post-angioplasty is listed without a result. Information about missing data, analysis populations, multiplicity, and the relationship between enrollment and the posted arm counts is not reported in the registry record.