Peripheral Nerve Blocks vs Periarticular Local Anesthetic Injection for Total Knee Arthroplasty (TKA): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Peripheral Nerve Blocks vs Periarticular Local Anesthetic Injection for Total Knee Arthroplasty (TKA) (NCT02223364), a randomized phase 4 study. The primary outcome is Maximum Pain Post-Operative Day (POD) 1 (Morning).

2. Clinical Question

The pain rating used a numeric scale from 0, meaning no pain, to 10, meaning the worst pain imaginable. The primary measurement window was postoperative day one, approximately 6 am to 12:00 pm. The registry reports three pairwise Wilcoxon (Mann-Whitney) tests: peripheral nerve block versus ropivacaine, peripheral nerve block versus liposomal bupivacaine, and ropivacaine versus liposomal bupivacaine.

This is a comparison of three strategies for providing pain relief after surgery, not a comparison of surgery with no surgery or of pain relief with no pain relief. The results describe the recorded pain outcome during the stated window. They do not, on their own, establish how the approaches compare on other outcomes, in other settings, or over other follow-up periods.

3. Trial Design

The study was a randomized, parallel-group, phase 4 trial comparing three analgesia approaches for patients undergoing total knee arthroplasty. The registry lists the study as completed, with a start date of 2014-07 and primary completion of 2016-07. Mayo Clinic is named as the sponsor. The registry reports enrollment of 165.

The three groups were Peripheral Nerve Block (PNB), Ropivacaine (PAI-R), and Liposomal Bupivacaine (PAI-L). The peripheral nerve block group received a continuous femoral nerve block and a single-injection sciatic nerve block consisting of bupivacaine. The two periarticular injection groups received injection into the surgical joint area using ropivacaine or liposomal bupivacaine. The registry describes the ropivacaine and liposomal bupivacaine injection volumes as 120 milliliters, but the detailed arm descriptions are truncated in the registry record.

Masking is listed as none, so this is an open-label comparison. The registry does not report further details about how randomization was generated or concealed, nor does it describe additional steps to limit bias in outcome assessment. Those details are not reported in the registry record.

The primary outcome population is described as intent-to-treat. Posted group sizes for the primary outcome are 50.0 for PNB, 55.0 for PAI-R, and 52.0 for PAI-L. The primary outcome is the maximum pain rating on postoperative day one in the morning. Secondary outcomes include pain measures at other postoperative time points and opioid-use measures, but their results are not part of the primary comparison. The registry record does not report a rationale for the enrollment target or a detailed account of the trial’s planned statistical analysis.

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: Intent-to-Treat Analysis

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 Maximum Pain Post-Operative Day (POD) 1 (Morning), measured over Post-Operative Day 1, approximately 6 am to 12:00 pm.

The primary endpoint is maximum pain on postoperative day one in the morning, during a window described as approximately 6 am to 12:00 pm. Pain was measured using a numeric rating scale from 0, meaning no pain, to 10, meaning the worst pain imaginable. A maximum rating captures the highest reported pain during the specified measurement period, rather than an average across that period.

The registry reports the endpoint as a median and interquartile range. The median describes the middle of the ordered observations, while the interquartile range describes the middle half of the observed values. These summaries can be useful when a measure is skewed or when a few relatively high or low observations make the mean less representative of a typical value. They do not show every participant’s score or the full shape of the distribution.

Postoperative pain is relevant to the trial’s stated comparison because the study sought to compare approaches intended to provide comfort after knee surgery. The primary endpoint focuses on one defined time window and one feature of pain: the maximum rating on the morning of postoperative day one. It does not, by itself, summarize pain over the whole recovery period, opioid use, or other consequences of the analgesia approaches. Those are distinct outcomes, and the registry lists several of them as secondary outcomes.

The posted medians were 3.0 (interquartile range 1.0 to 6.0) for PNB, 4.0 (2.0 to 6.0) for PAI-R, and 4.5 (3.0 to 6.0) for PAI-L. These are descriptive summaries; interpretation of the group comparisons also depends on the reported pairwise tests and on the uncertainty and limitations of the record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the Wilcoxon (Mann-Whitney) test compare?

The registry reports Wilcoxon (Mann-Whitney) tests for each pair of groups. In general, this rank-based test compares the distributions of observations in two independent groups by considering the ordering of values across the groups. It can be useful when a continuous or ordinal outcome is not well described by a normal distribution, or when an analysis is intended to rely less on distributional assumptions than a conventional two-sample mean comparison.

A rank-based test is not automatically a test of medians alone. If the groups have different distribution shapes or spreads, a test result can reflect more than a difference between their medians. The posted medians and interquartile ranges help describe the pain scores, but the registry does not provide the raw distributions or a detailed analysis plan that would clarify the exact interpretation intended by the investigators.

How should the medians and interquartile ranges be read?

The median is the central observation after values are ordered. The interquartile range marks the interval from the lower quartile to the upper quartile and describes the central portion of the observed distribution. In the posted results, PNB has a median of 3.0 and an interquartile range from 1.0 to 6.0; PAI-R has a median of 4.0 and an interquartile range from 2.0 to 6.0; PAI-L has a median of 4.5 and an interquartile range from 3.0 to 6.0.

These summaries do not give the standard error of a group median or an interval estimate for a between-group contrast. Interquartile ranges are measures of observed variability, not confidence intervals. In general, a confidence interval for an effect estimate communicates a range of values compatible with the data and analysis assumptions. No such interval for the pairwise effects is reported in the registry record.

What do the p-values tell the reader?

The p-values reported are 0.144 for PNB versus PAI-R, 0.016 for PNB versus PAI-L, and 0.196 for PAI-R versus PAI-L. In general, a p-value is calculated under a null hypothesis and measures how unusual a test statistic at least as extreme as the observed statistic would be if that null model and the test assumptions held. It is not an effect size, a measure of clinical importance, or a probability that the null hypothesis is true.

A smaller p-value does not show how large a difference is, and a larger p-value does not establish that groups are equivalent. The lack of a reported effect estimate and confidence interval limits interpretation of the magnitude and precision of the pairwise contrasts. The p-values must be considered alongside the descriptive summaries, the design, and the number of outcomes and analyses.

Why can over-dispersion matter?

For count outcomes, over-dispersion means that observed variability is greater than a statistical model expects. In general, a model that ignores over-dispersion can understate uncertainty and make results appear more precise than they are. This is a general consideration for count-data analysis, not a reported feature of the primary pain endpoint in this trial. The registry does not provide enough detail about models for the other outcomes to assess whether over-dispersion was relevant to those analyses.

What can planning calculations establish?

The sample-size material is described as a planning illustration based on the observed effect, not the trial’s own design. It uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is defined as a stated fraction of the observed effect. In general, such calculations illustrate how a future study’s planning target could depend on an assumed effect and statistical assumptions. They are not observed power and do not establish that the completed trial was adequately powered. The registry does not report the trial’s sample-size justification or the assumptions used to set its enrollment target.

Planning illustrations should also be interpreted cautiously when they use an effect estimated from the same trial. An observed effect can be imprecise, and a future study’s required enrollment depends on the chosen outcome, target effect, design, analysis, and assumptions. The record does not provide enough information to treat the planning illustration as a replacement for the original design rationale.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparisons using Wilcoxon (Mann-Whitney) tests, with p-values of 0.144 for PNB versus PAI-R, 0.016 for PNB versus PAI-L, and 0.196 for PAI-R versus PAI-L. These are the posted inferential results for the primary endpoint. No numerical effect estimate or confidence interval accompanies these comparisons in the registry record.

No discrepancy is flagged between the posted analysis and a recalculation. Where a discrepancy is present, the sponsor’s own analysis takes precedence; that situation does not arise in this record. The available primary endpoint summaries are medians and interquartile ranges, rather than the information needed to reproduce the reported rank-based tests independently. The record does not report participant-level data or additional test details that would allow a reader to reconstruct those p-values from the summaries.

Accordingly, the sponsor’s reported tests should be read alongside the descriptive medians and their spread, without treating the p-values as effect estimates. Any separate planning calculations based on observed effects are planning illustrations, not observed power and not evidence about the analysis actually used for the trial’s primary result.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 15 outcome measures and 40 analyses, while the primary endpoint results include three pairwise comparisons among the three groups. Considering multiple outcomes or comparisons raises the possibility that some apparently noteworthy findings may occur by chance. In general, multiplicity procedures can control the overall false-positive risk or otherwise define how a family of tests is interpreted. The registry record does not report whether such a procedure was planned or applied to these analyses.

The record does not describe a hierarchy for primary and secondary hypotheses beyond identifying the primary endpoint. It also does not state whether the three pairwise tests were prespecified, whether the p-values are adjusted, or how other analyses contributed to interpretation. These omissions limit the ability to interpret the reported p-values as a coordinated family of tests. They do not establish that multiplicity was ignored; the approach is not reported in the registry record.

The primary outcome population is labeled intent-to-treat, which generally refers to analyzing participants according to their randomized group regardless of treatment received. The registry does not provide further details about the population definition, treatment deviations, exclusions, or how missing primary outcome measurements were handled. The posted primary results include group sizes of 50.0, 55.0, and 52.0, but the registry does not state whether these represent all randomized participants or only those with usable primary outcome data.

Missing-data amounts, reasons for missingness, and any imputation or sensitivity analyses are not reported in the registry record. Without that information, readers cannot assess whether missing outcomes could have affected the comparisons. The secondary outcomes include additional pain and opioid-use measures, but their results and handling of missing observations are not detailed in the primary endpoint record. These gaps call for caution when generalizing from the primary result to the broader set of outcomes.

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 zero serious adverse events in each group and zero deaths in each group. It reports other adverse events affecting 2 participants in PNB, 3 in PAI-R, and 3 in PAI-L. The numbers at risk for the adverse-event summaries are 50, 55, and 52, respectively. The record does not list serious adverse event terms.

These figures are descriptive counts, not a complete assessment of comparative safety. The registry record does not report definitions for the serious and other adverse-event categories, the collection and attribution procedures, the observation period for these summaries, or whether events were actively solicited. It also does not provide uncertainty estimates for between-group differences in adverse events.

Zero reported serious events or deaths in these groups should not be interpreted as proof that such events cannot occur, or that the approaches have equivalent safety. The event counts are limited to what is recorded for this trial and its stated reporting. They should be read with the study design and the absence of further safety details, rather than as a general safety conclusion.

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

The primary results describe maximum pain ratings on the morning of postoperative day one among patients in three randomized groups. The observed medians were 3.0 for peripheral nerve block, 4.0 for periarticular ropivacaine, and 4.5 for periarticular liposomal bupivacaine. The reported pairwise tests give p-values of 0.144, 0.016, and 0.196 for the comparisons PNB versus PAI-R, PNB versus PAI-L, and PAI-R versus PAI-L, respectively.

These results do not establish the size or precision of the group differences because the registry does not report effect estimates or confidence intervals. The rank-based tests assess the pairwise data under their assumptions, but their p-values do not indicate whether a difference is important to patients. The medians and interquartile ranges describe the observed ratings, not the full distribution or the uncertainty in the contrasts.

The findings are restricted to the specified pain outcome and measurement window. They do not answer whether the groups differed in pain at every other time point, in opioid consumption, or in other aspects of recovery. Although those outcomes are listed in the registry, the associated results are not included in this primary endpoint record.

The results also do not establish that one approach is generally preferable, that the groups are equivalent, or that the observed pattern will apply to other patient populations or care settings. The study was randomized, but the registry offers limited detail about randomization procedures, missing outcomes, multiplicity, and the full analysis plan. The findings are best understood as a set of recorded comparisons within this particular trial, with unresolved uncertainty about effect magnitude and some aspects of analysis and reporting.

14. Why This Trial Matters Statistically

This trial is statistically informative because it compares three approaches within a randomized, parallel-group design and reports a defined patient-reported pain outcome. The primary endpoint is specific: maximum pain on postoperative day one in the morning, measured on a numeric rating scale. A clearly defined outcome and time window make the scope of the result easier to identify and help prevent broad claims based on a narrower measurement.

The result also illustrates why descriptive summaries and inferential tests answer different questions. The medians and interquartile ranges show the observed central pain ratings and their spread. The Wilcoxon (Mann-Whitney) p-values address pairwise evidence under the test’s assumptions. Neither component alone supplies a complete account: the summaries do not quantify uncertainty in a group contrast, and the p-values do not state the contrast’s magnitude or clinical importance.

There are three reported pairwise comparisons, alongside a registry listing of 15 outcome measures and 40 analyses. This makes transparency about prespecified hypotheses and multiplicity important to interpretation. Those details are not reported in the registry record. Similarly, the primary population is labeled intent-to-treat, but the registry does not explain missing-data handling or provide the full analysis plan.

The posted result therefore offers a useful example of how to read trial summaries without overextending them. It identifies the measured outcome, the observed group summaries, and the sponsor-reported tests, while leaving important questions about precision and analysis choices unresolved. Keeping those boundaries explicit is central to interpreting clinical trial statistics accurately.

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 registry record does not report an effect estimate or confidence interval for the pairwise primary comparisons. The medians and interquartile ranges describe the observed groups, but they do not quantify the precision of a contrast. This limits assessment of which sizes of difference remain compatible with the data.

The analysis details are incomplete. The record identifies the primary outcome population as intent-to-treat and names the Wilcoxon (Mann-Whitney) test, but does not report the full statistical analysis plan, randomization method, allocation concealment, or details of outcome assessment. Masking is listed as none. Because the primary endpoint is a reported pain rating, awareness of group assignment may be relevant to how the outcome was assessed, although the record does not describe its practical influence.

Missing-data counts, reasons, handling methods, and sensitivity analyses are not reported in the registry record. Nor does the record state whether the three pairwise tests or the larger set of outcomes and analyses were adjusted for multiplicity. The number of outcomes and analyses therefore makes the absence of a stated testing strategy an important interpretive limitation.

The detailed periarticular arm descriptions are truncated in the registry record. Although the record gives the injection volume, the complete procedural descriptions are not available in this discussion used for these results. The registry also does not provide numerical results for the secondary outcomes in this primary endpoint record, including opioid-use outcomes relevant to the stated hypothesis.

Safety reporting is limited to counts of serious events, other events, and deaths, without event terms, definitions, collection details, or uncertainty estimates. These limitations constrain interpretation of both effectiveness-related outcomes and safety. They do not negate the reported findings, but they define how narrowly those findings should be read.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Peripheral Nerve Blocks vs Periarticular Local Anesthetic Injection for Total Knee Arthroplasty (TKA) — NCT02223364.
• Registry record last updated 2017-08-31; status last verified 2017-08.
• Results first posted 2017-08-31.
• 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

This completed phase 4 randomized, open-label, parallel-group trial compared peripheral nerve blocks with periarticular injection of ropivacaine or liposomal bupivacaine for patients undergoing total knee arthroplasty. Mayo Clinic is listed as the sponsor. The registry reports enrollment of 165 and identifies the primary endpoint population as intent-to-treat.

The primary endpoint was maximum pain on the morning of postoperative day one, measured on a 0-to-10 numeric rating scale. Median pain ratings were 3.0 for PNB, 4.0 for PAI-R, and 4.5 for PAI-L. Reported Wilcoxon (Mann-Whitney) p-values were 0.144 for PNB versus PAI-R, 0.016 for PNB versus PAI-L, and 0.196 for PAI-R versus PAI-L.

The record does not report effect estimates or confidence intervals for these comparisons. It also does not provide full details on multiplicity, missing-data handling, or the statistical analysis plan. The posted results therefore describe the primary pain outcome and pairwise tests, while leaving uncertainty about the precision and broader interpretation of the comparisons.