Geniculate Artery Embolization for Knee Pain Secondary to Osteoarthritis (OA): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Geniculate Artery Embolization for Knee Pain Secondary to Osteoarthritis (OA) (NCT03362957), a randomized na study. The primary outcome is Mean WOMAC Scores at Month 6.

2. Clinical Question

The central clinical question was whether embolization changed the pain and disability measured by the Western Ontario and McMaster University Osteoarthritis Index (WOMAC), compared with a placebo procedure. The posted primary outcome is the mean WOMAC score at month 6. The registry describes the measure as a questionnaire covering pain, stiffness, and physical function, with higher total scores indicating more difficulty.

This question is constrained by the trial’s crossover. The initial sham comparison was not maintained through the six-month outcome: all sham participants crossed over to receive embolization after 1 month. As a result, the reported six-month mean for the crossover arm is not a six-month sham-control result. The registry reports a mean for the GAE Procedure arm and a separate mean for the Crossover Arm; it does not report a six-month mean for the Sham Procedure arm.

The record describes a small randomized study intended to evaluate safety and efficacy, but a treatment effect cannot be established from the posted six-month means alone. The available groups differ in their treatment histories, and there is no reported six-month sham comparison or inferential analysis. These results therefore describe the observed WOMAC scores in the groups with data rather than answer a controlled six-month comparison of embolization against sham.

3. Trial Design

This was a randomized, placebo-controlled study of geniculate artery embolization for knee pain secondary to osteoarthritis. The registry lists a planned enrollment of 21 and records the study as completed. Its dates are 2018-05-25 for study start and 2020-04-24 for primary completion. The phase is listed as not applicable.

The registry identifies the model as crossover. Participants assigned to the sham procedure could cross over to embolization if they saw no improvement after 1 month. The record says all sham participants crossed over, so the sham group had no available six-month data. The Crossover Arm represents a distinct reported group; its six-month outcome should not be read as a sham outcome or as a randomized sham comparison.

There is a difference in the registry’s descriptions of masking. The study metadata lists double masking, while the brief summary calls the study single-blinded. The registry record does not explain this difference or specify who was masked. It also does not report the randomization method, allocation ratio, or how assignment was concealed.

The listed procedures included screening, enrollment, embolization or sham procedure, and follow-up assessments at 24 hours, 2 weeks, 1, 3, 6, and 12 months. The summary says an MRI may be performed after the 1-month visit. These are described procedures; the registry does not report how completely each visit or optional MRI occurred.

The registry reports 13 participants in the GAE Procedure arm and 6 in the Crossover Arm for the primary outcome, with no six-month value or sample size for the Sham Procedure arm. It separately gives safety denominators of 14, 7, and 7 for the GAE, Sham, and Crossover groups. The record does not explain the difference between those safety denominators and the primary-outcome counts. No allocation ratio or detailed participant flow is reported.

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.

Before assignment: All patients that received the sham procedure crossed over into the crossover arm.

Overall Study

Analysis population for the primary outcome, as defined in the registry: One participant was excluded from the GAE arm due to increased pain that resulted in additional medication and was excluded from the analysis. All sham participants crossed over to receive treatment after 1 month, thus no 6-month data available for this group. One crossover participant did not have WOMAC data available for the 6 month visit.

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: All participants in the sham arm crossed over into the crossover arm so the sham participants and crossover participants are the same baseline characteristics except the VAS and WOMAC where they had new baseline scores following the sham procedure.

6. Primary Endpoint

The registry lists the primary outcome as Mean WOMAC Scores at Month 6, measured over 6 months following procedure.

The primary endpoint is the mean WOMAC score at 6 months following the procedure. WOMAC is a questionnaire measure of osteoarthritis symptoms and function. The registry says it includes pain, stiffness, and physical function: pain has 5 items, stiffness has 2 items, and physical function has 17 items. Participants select a level of difficulty for tasks using the listed 5 point Likert scale, with response labels and values given as 0=None, 1=Slightly, 3=Very, and 4=Extremely. The registry describes a total score from 0 to 96 and says higher scores indicate more difficulty.

A mean summarizes the group’s average score, while the standard deviation describes how spread out individual scores are around that average. Together, these statistics describe both the central level and variability of the measured outcome. The score is relevant to the trial’s stated aim because it covers symptoms and activity-related difficulty, rather than only a single pain rating.

The reported GAE Procedure mean is 29.0 with a standard deviation of 25.0 among 13 participants. The Crossover Arm mean is 26.0 with a standard deviation of 31.0 among 6 participants. The registry gives no six-month result for the Sham Procedure arm because all sham participants crossed over after 1 month. Consequently, the endpoint as posted does not provide the intended six-month comparison against an ongoing sham group.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: each participant has a WOMAC score, and the registry reports group means and standard deviations. A mean is the arithmetic center of the observations, while the standard deviation describes the spread of individual scores around that center. Neither statistic is itself a comparison between randomized groups. A treatment-effect analysis would ordinarily estimate a between-group contrast and quantify its uncertainty, but no such analysis is posted for this endpoint.

What does a standard deviation tell the reader?

General statistical explanation: a standard deviation describes how dispersed individual measurements are within a group. It is not the same as a standard error or confidence interval. A confidence interval, when calculated from an appropriate analysis, gives a range of values compatible with the data and model assumptions for the quantity being estimated. The registry reports no confidence interval for the six-month WOMAC means and no confidence interval for a treatment contrast. The standard deviations of 25.0 and 31.0 therefore should not be read as uncertainty intervals for the group means or as evidence about statistical significance.

Why does the crossover matter?

Randomization is useful for comparing groups because it aims to make their treatment assignments independent of participant characteristics. That logic applies to the assigned groups and a defined comparison. In this trial, every sham participant crossed over to embolization after 1 month. At the six-month endpoint, there is consequently no group remaining under the sham procedure for comparison. The Crossover Arm has a treatment history that includes the initial sham period followed by embolization; the registry does not supply enough detail to treat that arm as a contemporaneous sham control. Its mean is informative as a descriptive group result, but it cannot replace a six-month placebo comparison.

The design also raises questions about how the crossover was defined and analyzed. The registry describes crossover for participants who saw no improvement after 1 month, but does not report the operational definition of improvement, the timing of individual crossovers, or a statistical model that accounts for periods and treatment sequence. These details are not reported in the registry record. Without an analysis plan and participant-level timing, the posted means cannot disentangle outcomes associated with the initial assignment, later treatment, or time in follow-up.

What can be said about statistical testing?

The registry records zero analyses for the outcome and states that values are missing or not estimable for a two-arm comparison of this continuous endpoint. This means there is no posted inferential test to interpret. In particular, there is no reported p-value, adjusted estimate, or confidence interval. A p-value, as a general concept, describes how compatible data are with a specified statistical null model; it is not a measure of the size or clinical importance of an effect. Since the registry gives no p-value, no claim about statistical significance can be made from the record.

General statistical explanation: analyses of continuous outcomes can account for baseline scores, repeated measurements, and other prespecified factors, depending on the protocol and data. Such choices can alter the estimand and its uncertainty. The registry does not report the planned statistical model, adjustment variables, baseline WOMAC results, or a prespecified estimand for the crossover. It is therefore not possible to determine whether the reported means are adjusted or whether they represent the full planned analysis; the posted results identify them as means with standard deviations.

How should planning calculations be understood?

The registry describes sample-size illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The illustrations are described as planning calculations for different fractions of the observed effect. They are not observed power and do not alter the evidence from the completed trial. No resulting required sample sizes are reported in the registry record. In addition, the assumptions describe a two-arm setting, whereas the posted six-month endpoint has no sham group value, so those illustrations should not be mistaken for a reported test of this endpoint.

General statistical explanation: planning sample sizes are conditional on assumed effect sizes, variability, allocation, and analysis methods. They do not show the probability that the observed result is true, and they cannot repair missing outcome data or recreate an absent randomized comparison. The trial’s enrollment and outcome counts are descriptive details; a planning illustration is a hypothetical design calculation rather than an additional observed result.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("GAE Procedure", "Crossover Arm")
n   <- c(13, 6)
m   <- c(29, 26)
sp  <- c(25, 31)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  GAE Procedure  13  29.000  25.000  13.893  44.107
CARM  Crossover Arm  6  26.000  31.000  -6.533  58.533

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 outcome measures and 1 posted secondary outcome, described as the number of participants with a reduction in medication. It records no analyses for the primary outcome. The full set of outcome results and any analysis hierarchy are not reported in the registry record. The record does not state whether the primary endpoint was designated in advance as the sole confirmatory analysis, whether secondary measures were exploratory, or whether any adjustment for multiple testing was planned.

General statistical explanation: when multiple outcomes or repeated tests are examined, the chance of a false-positive result can increase if each test is interpreted separately without an appropriate plan. A prespecified hierarchy or multiplicity adjustment can address that issue, but the registry does not report whether either approach was used. No conclusion about multiplicity control can be drawn from the listed outcome count alone.

For the six-month WOMAC endpoint, the Sham Procedure arm has no data because all sham participants crossed over after 1 month. The registry also says one participant was excluded from the GAE analysis after increased pain resulted in additional medication. One sentence about the crossover group is incomplete: it states that one crossover participant did not have, but does not identify what was absent. The specific missing information and its handling are not reported in the registry record.

The record does not describe an intention-to-treat analysis, a per-protocol analysis, imputation, or sensitivity analyses for missing outcomes. General statistical explanation: if people with missing outcomes differ systematically from those with observed outcomes, estimates based only on observed data can be biased. The direction and size of any such bias cannot be assessed from these registry entries. The reported group counts should therefore be read as the numbers contributing to the posted summaries, not as evidence that all randomized participants were included in the endpoint analysis.

12. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports serious adverse events for three groups. In the GAE Procedure group, 0 of 14 participants at risk had a serious adverse event; in the Sham Procedure group, 0 of 7; and in the Crossover Arm, 0 of 7. Deaths are also reported as 0 in each group. No serious adverse-event terms are listed.

Other, non-serious adverse events were reported for 8 of 14 participants at risk in the GAE Procedure group, 2 of 7 in the Sham Procedure group, and 5 of 7 in the Crossover Arm. These are the registry’s affected-participant counts, not counts of all event episodes. The record does not provide event-level detail in the listed results, so the counts do not indicate event type, severity within the non-serious category, timing, or whether a participant had more than one event.

The safety denominators differ from the primary-outcome counts: safety lists 14, 7, and 7 participants at risk, whereas the primary endpoint lists 13 for GAE and 6 for crossover and has no six-month sham result. The reason for this difference is not reported in the registry record. These safety summaries should be interpreted as reported group counts, not used to infer a comparative risk or to reconcile outcome availability. The registry does not provide a statistical comparison of adverse-event rates.

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

The registry’s six-month WOMAC results show group averages and variability for participants reported in the GAE Procedure and Crossover Arms. The score runs from 0 to 96, with higher scores indicating more difficulty. These values offer a descriptive account of measured symptoms and function among those included in the respective summaries. The registry also reports non-serious adverse events in each group and no serious adverse events or deaths.

The results do not establish the effect of embolization compared with sham at 6 months. All sham participants crossed over after 1 month, leaving no six-month sham data. The two reported means are from groups with different treatment histories, and no valid controlled contrast, p-value, or confidence interval is posted. The record therefore does not quantify the size or precision of a six-month treatment effect.

Nor do the group summaries show how individual participants changed from baseline. Baseline WOMAC values and participant-level trajectories are not reported. A group mean at a follow-up time is not equivalent to the average within-person change, and neither can be reconstructed from the posted information. The registry also does not report an analysis accounting for the crossover or the participant excluded from the GAE endpoint analysis.

The safety findings are similarly limited to counts of participants affected. The absence of reported serious events in these groups is a description of the registry’s recorded data, not a guarantee about risks beyond this study. Non-serious event counts do not establish a causal relationship to a procedure, and the registry does not supply detailed event descriptions or inferential comparisons.

In short, the posted record documents a randomized crossover study, two six-month WOMAC summaries, and group-level safety counts. It does not provide the six-month sham comparison or statistical analysis needed to support a precise estimate of comparative treatment effect. The distinction between observed summaries and comparative evidence is central to interpreting the result.

14. Why This Trial Matters Statistically

This trial illustrates how crossover can change the question that a follow-up outcome can answer. Random assignment initially establishes a sham comparison, but the registry reports that every sham participant crossed over after 1 month. By the six-month WOMAC assessment, the sham arm has no data. The result is a record with follow-up summaries in two groups but without a contemporaneous sham group at the primary time point.

That makes the distinction between a group description and a treatment-effect estimate especially important. The reported means and standard deviations describe observed scores. They do not themselves account for different treatment histories, missing outcomes, or exclusion from analysis. A statistically appropriate analysis needs to correspond to the design and the clinical question; the registry does not report such an analysis for the continuous primary outcome.

The record also demonstrates why denominators should be read alongside the outcome definition. The primary endpoint includes 13 GAE participants and 6 crossover participants, while safety denominators are 14, 7, and 7. The registry does not explain the differences. Those counts refer to separate reported summaries and should not be assumed interchangeable.

Finally, the masking descriptions differ: the metadata says double masking and the brief summary says single-blinded. This matters for a placebo-controlled procedural study because readers need to know who was masked to interpret possible influences on subjective questionnaire responses. The registry does not resolve the discrepancy. Together, the crossover, incomplete analysis reporting, and masking ambiguity show why a registry result needs to be read in context rather than reduced to the numerical means alone.

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.

The principal limitation is that the registry reports no six-month sham outcome. All sham participants crossed over after 1 month, so the primary endpoint lacks an ongoing placebo comparison at the stated time point. The reported crossover group cannot serve as a sham control because its treatment history includes embolization.

The registry gives no inferential analysis for the continuous endpoint: no estimated between-group effect, confidence interval, or p-value is reported. It also does not describe the statistical model, analysis population, treatment of baseline scores, or handling of repeated measures. The recorded exclusion of one GAE participant because increased pain led to additional medication may affect which participants are represented in the reported mean; no sensitivity analysis is reported.

Missing-data reporting is incomplete. The record contains an unfinished statement that one crossover participant did not have something, without identifying the missing item. The reason and handling of that absence are not reported. The registry also does not explain why safety denominators differ from primary-outcome counts or provide a complete participant flow.

The study is described as double masked in metadata and single-blinded in the summary, with no explanation of the discrepancy. Details of allocation concealment and who was masked are not reported. The registry lists 8 outcome measures, but does not report a multiplicity plan or clarify how outcomes were prioritized beyond identifying the primary endpoint.

These limitations concern the completeness and interpretability of the registry record; they do not establish what the full study data or any unreported analysis might show. The posted results should be read as limited descriptive summaries rather than as a complete comparative assessment.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Geniculate Artery Embolization for Knee Pain Secondary to Osteoarthritis (OA) — NCT03362957.
• Registry record last updated 2021-03-22; status last verified 2021-03.
• Results first posted 2021-03-15.
• 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

The registry describes a completed randomized, placebo-controlled crossover study of geniculate artery embolization for knee pain secondary to osteoarthritis, with planned enrollment of 21. The primary endpoint was mean WOMAC score 6 months following the procedure. The score measures pain, stiffness, and physical function, with higher scores indicating more difficulty.

At six months, the GAE Procedure group had 13 participants, a mean WOMAC score of 29.0, and a standard deviation of 25.0. The Crossover Arm had 6 participants, a mean of 26.0, and a standard deviation of 31.0. No six-month sham data are reported because all sham participants crossed over after 1 month. The registry reports no analysis for the continuous two-arm comparison.

Safety results report no serious adverse events and no deaths in the listed groups. Other adverse events affected 8 of 14 participants in the GAE Procedure group, 2 of 7 in the Sham Procedure group, and 5 of 7 in the Crossover Arm. The registry does not report a six-month controlled treatment-effect estimate or its precision.