Nerve Ablation by Cooled Radiofrequency Compared to Corticosteroid Injection for Management of Knee Pain: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Nerve Ablation by Cooled Radiofrequency Compared to Corticosteroid Injection for Management of Knee Pain (NCT02343003), a randomized na study. The primary outcome is Numeric Rating Scale (NRS).

2. Clinical Question

The primary endpoint is a participant count, not a mean pain-score change. In the cooled radiofrequency group, 43 of the 58 participants who completed the 6 month outcome met the threshold. In the corticosteroid injection group, 11 of 68 completers met it. The posted comparison describes the risk difference for corticosteroid injection relative to cooled radiofrequency as negative, with a confidence interval that remains below zero and an unadjusted p < 0.001.

The endpoint addresses whether each participant crossed a prespecified threshold for pain reduction, rather than the average amount of change across all participants. The registry record does not report the threshold's rationale or how the primary comparison was specified in the statistical plan. The question is therefore best read as a comparison of observed proportions meeting the threshold among participants with outcome data, alongside the registry's reported contrast—not as a complete account of every aspect of pain or function.

3. Trial Design

The registry describes a prospective, multicenter, randomized controlled trial with a parallel-group model. It lists no masking. Enrollment was 151, and the study status is completed. The listed phase is NA. The study began in 2015-01, and primary completion was 2016-08.

Participants were assigned to cooled radiofrequency or corticosteroid injection. The registry records 76 originally randomized participants in the cooled radiofrequency group and 75 in the corticosteroid injection group. The primary outcome population counts differ from those original group sizes: 58 cooled radiofrequency participants and 68 corticosteroid injection participants completed the 6 month outcome. The registry record does not explain the reasons for these differences.

The device intervention was described as delivering cooled radiofrequency energy to the knee to ablate sensory nerves. The comparator was corticosteroid injections administered to the knee. The registry names Numeric Rating Scale and Oxford Knee Score as secondary outcomes and lists 6 outcome measures in total. It does not report further detail about the other outcome measures in the material available in the registry record.

Random allocation supports a planned comparison between groups, but the record lists no masking. The posted endpoint counts are based on those who completed the outcome, not the full originally randomized groups. That distinction matters when interpreting the result: the displayed counts describe observed participants with six-month data, while the record does not specify whether a separate analysis included all randomized participants or addressed missing outcomes. The protocol's original sample-size justification and other design assumptions are not reported in the registry record.

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: 58 of the 76 originally-randomized study subjects in the Cooled radiofrequency group completed this 6 month outcome. 68 of the originally-randomized 75 study subjects in the Corticosteroid injection group completed this 6 month outcome.

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 Numeric Rating Scale (NRS), measured over 6 months.

The primary endpoint counts participants whose knee pain was reduced by at least 50% on the Numeric Rating Scale at 6 months. The NRS is described as an 11-point scale running from 0 points, meaning no pain, to 10 points, meaning the worst pain. The endpoint converts each participant's score change into a yes-or-no threshold outcome, then counts those meeting the threshold in each group.

A threshold outcome can be easier to interpret as a count of participants reaching a defined degree of improvement than as a group average. It also discards information: participants just below the threshold and participants with little change are both counted as not meeting it, while participants meeting it are counted together even if their score changes differ. The registry record does not report the rationale for selecting this threshold or provide the underlying individual score changes.

The posted counts are 43 of 58 cooled radiofrequency completers and 11 of 68 corticosteroid injection completers. The descriptions state that these are participants who completed the six-month outcome; they are not the original randomized denominators. The endpoint therefore measures threshold attainment among participants with recorded outcome data. It does not, on its own, describe pain at other time points, average score change, or the separate secondary outcome of Oxford Knee Score.

7. Results Posted in the Registry

• Corticosteroid Injection: 16.2% (11/68) versus 74.1% (43/58) on control; risk difference -58.0 percentage points (95% CI -72.2 to -43.7), risk ratio 0.22 (0.12–0.38); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: a participant either met the specified reduction threshold or did not. The registry reports counts for participants meeting the threshold and the number completing the outcome in each group. A comparison of proportions is therefore a natural way to summarize the endpoint, although the exact statistical test or model used is not reported in the registry record.

What does a risk difference describe?

A risk difference subtracts one group's event proportion from the other group's event proportion. Here, the registry defines the contrast as corticosteroid injection relative to cooled radiofrequency. The estimate is negative, which corresponds to a lower observed proportion meeting the threshold in the corticosteroid injection group. A confidence interval around a risk difference gives a range of values compatible with the data under the method's assumptions. The registry reports an interval entirely below zero, but its calculation method is not reported in the registry record.

Other ways of comparing binary outcomes include a risk ratio and an odds ratio. A risk ratio compares proportions directly, while an odds ratio compares odds. These measures answer related but distinct questions and can have different numerical values. The registry record includes risk-ratio and odds-ratio comparisons as well, but the interpretation focuses on the posted risk difference and the endpoint counts. The calculations should not be treated as interchangeable measures.

What does the p-value contribute?

The posted p-value is unadjusted and is reported as < 0.001. In general, a p-value measures how unusual results at least as incompatible with a specified null hypothesis would be if that hypothesis and the test assumptions held. It does not give the probability that the null hypothesis is true, the probability that the observed result occurred by chance in an everyday sense, or the probability that an individual participant will meet the endpoint. The test used to produce the posted p-value is not reported in the registry record.

Statistical significance and effect size are different concepts. A small p-value does not, by itself, communicate the magnitude or precision of the group contrast. The risk-difference estimate and its confidence interval address the scale and uncertainty of the contrast; the counts show the observed outcome data behind the comparison. All should be read together, with attention to the denominators.

Why do the denominators matter?

The original randomized groups contained 76 and 75 participants, but the endpoint counts use 58 and 68 participants who completed the six-month outcome. General statistical practice distinguishes an analysis of observed completers from an analysis that includes all randomized participants under a specified missing-data approach. The registry record does not say whether the primary result is a completer analysis by design, whether other analyses were performed, or whether any missing values were imputed.

When outcomes are unavailable, a complete-case comparison can be affected if outcome availability is associated with group assignment or the unobserved outcome. This is a general concern, not a claim that such bias occurred in this trial. Without the reasons for noncompletion and a reported missing-data strategy, the extent and direction of any effect cannot be assessed from the registry record.

What is not known about the analysis?

The registry lists no analysis records. It nevertheless posts a group comparison, confidence interval, and p-value for the primary result. The exact test or model, covariate adjustment, analysis population, and handling of missing data are not reported in the registry record. The record also does not report whether interim analyses were planned or conducted, or whether any statistical adjustment was made for multiple outcomes or comparisons.

The listed total of 6 outcome measures and 2 posted secondary outcomes does not by itself establish how many statistical hypotheses were tested or whether they were treated as confirmatory. Multiple testing can increase the chance of at least one apparently notable finding when many hypotheses are tested. General statistical practice may address this with a prespecified hierarchy or an adjustment, but the registry record does not report such a plan. The posted primary p-value is explicitly unadjusted, so it should be read as such.

The trial's original sample-size design and justification are also not reported in the registry record. A separate planning illustration uses the observed effect and specified assumptions; it is not evidence of the trial's original design process. The distinction between a planning calculation and an observed analysis is essential: a sample-size illustration asks how many participants might be needed under chosen assumptions, whereas the endpoint comparison describes the trial's observed data.

10. Confidence Intervals and Precision

The risk-difference confidence interval is reported entirely below zero for the contrast of corticosteroid injection relative to cooled radiofrequency. This indicates that, under the interval method's assumptions, the range of values compatible with the data does not include no difference. The exact interval endpoints are not reproduced because the registry's reported values have more than three decimal places.

A confidence interval is not a guarantee that the true effect lies within its limits, nor is it a probability statement about the particular interval after it has been calculated. In general, it describes the performance of a procedure across repeated samples under specified assumptions. A narrower interval indicates greater precision; a wider interval indicates greater uncertainty about the effect's magnitude. The method used to calculate this interval is not reported in the registry record.

Precision also depends on the endpoint data available. The group counts are based on 58 and 68 participants who completed the six-month outcome, rather than all 76 and 75 originally randomized. The registry record does not explain the missing outcomes or report how they affect uncertainty. The observed counts and the confidence interval should therefore be interpreted with that denominator distinction in mind.

The interval supports a direction for the posted contrast under its assumptions, but it does not settle questions about the clinical importance of the difference, possible reporting effects in an unmasked trial, or generalizability beyond the study participants. Those judgments require context that is not reported in the registry record. The unadjusted p-value provides a separate summary of compatibility with a null hypothesis; it is not a substitute for the interval's information about the range of effect sizes.

11. 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 illustration reports participants per arm for each assumed fraction of the observed effect. For planning at 100% of the observed effect, it gives 11 participants per arm for 80% power and 14 participants per arm for 90% power. For 75% of the observed effect, it gives 20 participants per arm for 80% power and 26 participants per arm for 90% power. For 50% of the observed effect, it gives 44 participants per arm for 80% power and 59 participants per arm for 90% power.

In plain terms, the 80% power column gives the planned number per group under which the calculation has an 80% chance of detecting the assumed effect at the stated significance level, if the planning assumptions and model apply. The 90% power column represents a higher chance of detecting that assumed effect, and therefore requires more participants per group in these illustrations. The rows assume progressively smaller effects; smaller effects are harder to distinguish from random variation and require more participants.

The calculation assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The assumed effect is the stated fraction of the observed effect. These figures are participants per arm, not total events. This is a planning illustration based on the observed effect, not observed power. The trial's own original sample-size assumptions and justification are not reported in the registry record.

12. 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 = 0.741; treatment proportion p₁ = 0.162; average p̄ = 0.452.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7038 = 1.3794.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5721 = 0.4815.
4. n = (1.3794 + 0.4815)² / (-0.5796)² = 10.3, rounded up to 11 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 <- 0.741379310344828                          # control proportion
p1 <- 0.161764705882353                          # 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(2343003)
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 = 11 (80% power), 14 (90% power)
 75% of observed effect: n per arm = 20 (80% power), 26 (90% power)
 50% of observed effect: n per arm = 44 (80% power), 59 (90% power)
power.prop.test: 10.3 
Simulated power at n = 11 per arm: 0.841 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 6 outcome measures in total and names Numeric Rating Scale and Oxford Knee Score as secondary outcomes; 2 secondary outcomes are posted. The primary endpoint is the six-month count of participants meeting the NRS pain-reduction threshold. The record does not report a hierarchy for the outcomes, a family of hypotheses, or a multiplicity-adjustment plan. The posted p-value is unadjusted. Without a stated analysis plan, it is not possible to determine how the trial treated the possibility of multiple outcome or time-point comparisons.

The registry lists no analysis records, and the primary comparison's statistical test, analysis population, covariate adjustment, and missing-data method are not reported in the registry record. The outcome description states that 58 of 76 originally randomized participants in the cooled radiofrequency group and 68 of 75 in the corticosteroid injection group completed the six-month outcome. The registry does not state why outcomes were unavailable for the other randomized participants, whether they were excluded from the primary analysis, or whether any method was used to account for missing outcomes.

In general, an analysis including only participants with observed outcomes can differ from an analysis of all randomized participants if outcome availability is related to prognosis, outcome, or treatment assignment. This is a methodological possibility, not evidence that the posted result is biased. The registry record does not identify the primary analysis population as intention-to-treat, per-protocol, or another defined population. It also does not report sensitivity analyses addressing missing outcomes.

These reporting gaps do not alter the posted counts or comparison, but they limit assessment of whether the analysis followed a prespecified plan and how robust the result is to missingness or multiple testing. Details of interim analyses and any statistical monitoring plan are not reported in the registry record.

14. 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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: gastrointestinal (4.0%), other (2.7%), cardiovascular (2.7%), urogenital (1.3%), procedure (1.3%).

The registry reports serious adverse events affecting 3 of 76 participants at risk in the cooled radiofrequency group and 9 of 75 participants at risk in the corticosteroid injection group. These are participant counts, not a comparison adjusted for follow-up time, baseline risk, or other factors. The record also reports other adverse events affecting 43 of 76 and 25 of 75 participants, respectively.

The listed serious-event terms and percentages are gastrointestinal, 4.0%; other, 2.7%; cardiovascular, 2.7%; urogenital, 1.3%; procedure, 1.3%; and respiratory, 1.3%. The registry record does not identify these terms by treatment group in the information available, so they should not be attributed to one group from these figures alone. The meaning of the term other is not further described.

These counts describe events reported in the registry and do not establish causation. The record does not provide enough detail to compare event severity, timing, duration, or relatedness to the intervention. Death counts are not reported in the registry record. Safety counts should be read separately from the primary pain-threshold outcome.

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

The registry result describes the proportion of six-month completers who reached a threshold of at least 50% pain reduction on the NRS. There were 43 such participants among 58 cooled radiofrequency completers and 11 among 68 corticosteroid injection completers. The reported risk difference for corticosteroid injection relative to cooled radiofrequency is negative, its confidence interval is below zero, and the unadjusted p-value is < 0.001. In the posted endpoint data, meeting the threshold was more frequent among cooled radiofrequency completers.

This finding does not establish the full pattern of pain change. A threshold outcome does not show how much scores changed for each participant, whether participants just below the cutoff had similar outcomes to those above it, or how scores changed at other times. It also does not describe the Oxford Knee Score or other outcome measures. The registry record does not report results for those measures in the registry record for this result.

The result does not resolve the impact of incomplete six-month outcome data. The original randomized groups were 76 and 75, while the primary endpoint counts include 58 and 68 completers. Reasons for noncompletion and methods for handling missing outcomes are not reported in the registry record. The study is listed as unmasked, and the primary measure is participant-reported; the record does not describe how expectations or reporting effects were addressed.

Nor does the posted comparison alone establish how applicable the result is to people beyond the enrolled participants, the durability of outcomes after 6 months, or the balance of all benefits and harms. Serious and other adverse-event counts are separately reported, but the registry information does not provide a full clinical interpretation of these events. These data support a statistical description of the recorded endpoint, not medical advice or a definitive conclusion about treatment for an individual.

16. Why This Trial Matters Statistically

This trial illustrates how a randomized comparison can produce a clear statistical contrast while still leaving important interpretation questions. The primary outcome is a binary threshold, so the result can be expressed as participant counts and a risk difference. That makes the endpoint straightforward to describe, but it also compresses a continuous pain scale into whether a participant crossed one cutoff. The counts do not preserve the amount of score change within either category.

Denominator choice is especially important. The trial originally randomized 76 participants to cooled radiofrequency and 75 to corticosteroid injection, while the six-month outcome was completed by 58 and 68, respectively. A comparison of completers is not automatically equivalent to a comparison of all randomized participants. Without reported missing-data methods, the reader cannot determine how the unobserved outcomes were incorporated or whether the analysis was sensitive to them.

The registry also lists no masking. Because the primary endpoint is based on a participant-reported pain scale, awareness of treatment assignment is a relevant design feature when evaluating possible measurement and expectation effects. This does not establish that such effects occurred, but it is part of the context for interpreting a patient-reported outcome.

Finally, the posted p-value is unadjusted, and the registry record does not report a multiplicity plan or the statistical test used. The interval and counts help convey the contrast and its precision, but missing analysis details limit assessment of the analysis plan. Together, these features demonstrate why statistical reading requires attention not just to whether a result crosses a significance threshold, but also to endpoint definition, analysis population, missingness, masking, and precision.

17. 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 primary endpoint is reported among participants who completed the six-month assessment, not all participants originally randomized. The registry provides the original group sizes and the completer denominators but does not report reasons for noncompletion, the timing of missing outcomes, or a missing-data strategy. This prevents a complete assessment of how missing data may affect the comparison.

The trial is listed as having no masking. The primary outcome is based on participant-reported pain, so awareness of treatment assignment is relevant when considering expectations and reporting. The registry record does not describe measures intended to address those influences. The record also does not report the precise statistical test or model underlying the posted comparison and confidence interval, or define the analysis population.

The primary outcome reduces an 11-point pain scale to a binary threshold of at least 50% reduction. That choice allows a count of participants reaching the threshold, but does not show the full distribution of score changes. The record does not report the rationale for the cutoff in the registry record for this result. Although Numeric Rating Scale and Oxford Knee Score are named as secondary outcomes, detailed secondary results are not included.

The registry lists 6 outcome measures and provides an unadjusted p-value, but does not report a multiplicity strategy or interim-analysis plan. It also does not report the original sample-size justification. The safety summary gives participant counts but limited detail on event timing, relatedness, or group attribution for the listed serious-event terms. These are limitations of the available registry reporting; they do not establish that the corresponding methods or details were absent from the full study record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Nerve Ablation by Cooled Radiofrequency Compared to Corticosteroid Injection for Management of Knee Pain — NCT02343003.
• Registry record last updated 2019-05-13; status last verified 2017-02.
• Results first posted 2017-10-09.
• 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.

21. Record Summary

NCT02343003 is a completed, prospective, multicenter, randomized parallel-group trial with no masking listed. It enrolled 151 participants with osteoarthritis of the knee and compared cooled radiofrequency with corticosteroid injection. The primary endpoint counted participants with at least 50% pain reduction on the NRS at 6 months.

Among participants completing that outcome, 43 of 58 in the cooled radiofrequency group and 11 of 68 in the corticosteroid injection group met the threshold. The posted risk-difference comparison for corticosteroid injection relative to cooled radiofrequency is negative, with a confidence interval below zero and an unadjusted p < 0.001. Original design assumptions, the primary analysis method, missing-data handling, and a multiplicity plan are not reported in the registry record.