Total Knee Arthroplasty: Comparison of Fixed and Mobile Bearings: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Total Knee Arthroplasty: Comparison of Fixed and Mobile Bearings (NCT00417859), a randomized na study. The primary outcome is Gait Stride Length.

2. Clinical Question

3. Trial Design

NCT00417859 is a completed, randomized, triple-masked, parallel-group study sponsored by University of Lausanne Hospitals. Its listed phase is NA. The registry records enrollment of 64 and gives the condition as knee osteoarthritis. The study start is 2004-02 and primary completion is 2015-11. These registry dates describe the recorded study timeline; further information about recruitment, follow-up, or the timing of individual assessments is not reported in the registry record.

The treatment comparison is between mobile-bearing and fixed-bearing total knee arthroplasty. The arm descriptions identify “TKA mobile” and “TKA fix,” while the intervention list names mobile-bearing total knee arthroplasty. The allocation ratio, method for generating the random sequence, concealment procedures, and which people were masked are not reported in the registry record. Triple masking is recorded, but the record does not specify the roles to which it applied.

The primary outcome is gait stride length at five years. Its posted summaries use the labels “Preoperative” and “5 Years,” with reported sample sizes of 26 and 29. Those labels do not match the fixed- and mobile-bearing arm labels. The record does not explain whether the primary summaries pool treatment arms, how the analyses relate to randomization, or whether the same participants contributed at both times. The reported secondary outcomes are a knee score and the number of participants with dislocation. The registry records four outcome measures and one analysis, but further details about the additional outcome measures are not 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: 8 patients were excluded

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 Gait Stride Length, measured over 5years.

Gait stride length is the distance covered in a step during walking, recorded in meters. The registry describes measuring it with a data logger that uses five inertial sensors and identifies the assessment time as five years. The study’s stated rationale is to use gait parameters as an objective way to assess outcomes after total knee arthroplasty, rather than relying only on clinical scores.

As a general measurement principle, an objective instrument can make the collection of an outcome more standardized, but the measure still needs to be interpreted in context. The record does not report validation details for this device or define what size of change in stride length would be important to patients or clinicians. It also does not describe the gait protocol, such as the walking conditions or how many recordings contributed to each participant’s result. Those details are not reported in the registry record.

The posted primary summaries give a mean of 1.2 meters for both “Preoperative” and “5 Years,” with an interquartile range from 1.1 to 1.3 for each. The record lists a Wilcoxon (Mann-Whitney) test, but does not provide a treatment-group contrast, its confidence interval, or a clear mapping from the time-labelled summaries to the randomized groups. The outcome therefore indicates what was recorded and when, while the available summary does not fully establish the comparative effect of bearing type.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary analysis is identified as a Wilcoxon (Mann-Whitney) test. In general, this rank-based method compares the ordering of observations between two groups rather than directly testing a difference between their means. It can be useful when an outcome’s distribution is not well represented by assumptions required for a mean-based test. A rank-based test does not automatically estimate a clinically interpretable difference in the original measurement units. To understand magnitude, an effect estimate and its uncertainty are also useful.

For the gait outcome, the registry specifies a mean and interquartile range as the summary measures, while naming a Wilcoxon (Mann-Whitney) test for the analysis. These describe different aspects of the data: the mean is an average, the interquartile range describes the central spread, and the test uses ranks. That combination is not inherently invalid, but the record does not explain the analysis choices or provide enough detail to reconstruct the treatment contrast. In particular, the posted primary groups are labelled “Preoperative” and “5 Years,” not by randomized treatment. The analysis population, handling of any repeated measurements, and mapping between these time labels and the randomized arms are not reported in the registry record.

The registry gives p = 0.05 for that test but no test statistic, effect estimate, confidence interval, or details about whether the p-value is exact or approximate. A p-value should be read alongside an estimate of the effect and its precision. A confidence interval, as a general statistical concept, describes a range of values compatible with the data and the method’s assumptions. Its width indicates how much uncertainty remains about the estimated effect. No primary confidence interval is reported, so primary-effect precision cannot be assessed from an interval.

The knee-score comparison has a recalculated, unadjusted mean difference of -2.4 for fixed versus mobile bearing, with a 95% confidence interval from -11.51 to 6.711 and p = 0.606. These are recalculated from the posted group summaries, not a sponsor-reported analysis. The interval includes zero and also includes differences on either side of zero. Thus, this estimate is compatible with a range of possible values; it is not a demonstration of equivalence or proof of no difference. The reported group means are 51.7 for mobile bearing and 49.3 for fixed bearing, with standard deviations of 12.8 and 19.4, respectively.

In general, a standard deviation describes variation among individual observations around a group mean. The standard error describes uncertainty in an estimated quantity, such as a mean; it is not the same as the variation among participants. The recalculation includes standard errors of 2.56 and 3.88 for the two group means and a pooled standard deviation of 16.43. Its standardized effect estimate is -0.146. These summary-statistic calculations rely on the posted group summaries and do not adjust for other variables or study design features. The registry does not report a sponsor analysis for this secondary outcome, so the recalculated comparison should be understood as an unadjusted descriptive analysis.

The planning illustrations use the observed knee-score effect as their starting point, with the effect to detect set to a stated fraction of that observed effect. They are not observed power and do not show what the trial was designed to detect. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Under those assumptions, planning for 100% of the observed effect gives 737 participants per arm for 80% power and 986 for 90% power. Planning for 75% of the observed effect gives 1,309 participants per arm for 80% power and 1,752 for 90% power. Planning for 50% of the observed effect gives 2,945 participants per arm for 80% power and 3,942 for 90% power. These are illustrative sample-size requirements, not observed trial results. The registry’s own sample-size justification is not reported.

As a general concept, over-dispersion occurs when observed variability is greater than a statistical model expects. Ignoring it can make uncertainty appear smaller than it is. Whether over-dispersion was assessed or relevant to any analysis in this trial is not reported in the registry record. More broadly, the record does not provide enough information to assess model diagnostics, adjustments, or sensitivity analyses. Such details should not be inferred from the name of a test or from the posted summary statistics alone.

10. Analyses Reported by the Sponsor

The registry reports a sponsor analysis for the primary gait outcome using a Wilcoxon (Mann-Whitney) test, with p = 0.05. It does not provide the corresponding test statistic, treatment-group effect estimate, confidence interval, or a clear account of how the “Preoperative” and “5 Years” summaries relate to the randomized bearing groups. The posted primary result is therefore the sponsor-reported inferential result, but the record does not contain enough detail to quantify the primary treatment contrast.

For the knee score, the registry has no sponsor-reported analysis. The mean difference, confidence interval, standardized effect, and unadjusted p-value discussed in the statistical interpretation are recalculated from the posted group summaries. They should be distinguished from a prespecified or adjusted sponsor analysis. The registry does not describe a sponsor analysis for this endpoint, its model, or its handling of missing observations.

No discrepancy is flagged between a sponsor result and a recalculation. That does not make the recalculated result equivalent to a full trial analysis: it is a simple, unadjusted comparison derived from summary statistics. The registry’s primary p-value remains the only reported primary inferential result, and the knee-score recalculation adds a separate estimate whose limitations should be kept in view.

11. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary gait outcome and two secondary outcomes, but does not report a detailed statistical analysis plan. It is not reported whether the primary endpoint was selected before data review, whether secondary hypotheses were arranged in a testing hierarchy, or whether an adjustment for multiple testing was used. In general, evaluating several outcomes without accounting for multiplicity can increase the chance of a false-positive finding. This is a concern to consider, not evidence that any particular result is false.

The primary outcome summaries are labelled “Preoperative” and “5 Years,” while the randomized arms are labelled by bearing type. The record does not explain the relationship between these labels or whether the primary analysis compared within-person change, separate time groups, or treatment groups. The summary sample sizes are 26 and 29, but those figures do not establish whether observations are paired or how many randomized participants contributed to each analysis. The analysis population, including whether it was intention-to-treat or another population, is not reported in the registry record.

Missing outcome data and missing reporting are distinct issues. The registry does not explain whether any measurements were unavailable, why they might be unavailable, or how missing values were handled. It does not report an imputation method, sensitivity analysis, or participant flow for the primary endpoint. The absence of these details does not show that observations were missing; it means the record does not describe their status or handling.

The registry also does not report interim analyses, stopping rules, a sample-size justification, or a detailed plan for covariate adjustment. These elements should not be inferred from the randomized, triple-masked, parallel design. Without the analysis population, missing-data approach, and multiplicity strategy, the precision and robustness of the reported findings are harder to assess. The posted estimates remain interpretable as reported summaries, but their relation to the full randomized study analysis is not fully documented.

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 adverse-event summary lists 25 participants at risk in each bearing group. It records zero serious adverse events in the mobile-bearing group and zero in the fixed-bearing group. For other adverse events, it records zero in the mobile-bearing group and one in the fixed-bearing group. These are the registry’s posted counts, not estimates of comparative risk.

The event description, severity details beyond the serious/non-serious categories, timing, and assessment method are not reported in the registry record. The record does not state whether the single other event was related to the procedure or device. Death counts are also not reported in the registry record. The separately listed secondary outcome of participants with dislocation has no result posted.

As a general statistical principle, zero observed events does not establish that an event cannot occur, and small event counts provide limited information about differences between groups. The adverse-event figures should therefore be read as a brief descriptive account, not as a conclusive comparison of safety. They do not by themselves establish a causal relationship or resolve uncertainty about uncommon events.

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

The registry supports a limited description of what was recorded: a randomized comparison of fixed- and mobile-bearing total knee arthroplasty, a gait stride-length endpoint at five years, and a knee-score outcome. For gait, the posted time-labelled summaries have the same reported mean and interquartile range, alongside a Wilcoxon (Mann-Whitney) p-value of 0.05. Because the summaries are not identified by randomized treatment group and no primary effect estimate or confidence interval is reported, they do not quantify the bearing-specific difference.

The secondary knee-score recalculation provides a group contrast, but its confidence interval is wide enough to include values on either side of no difference. In general, uncertainty of this kind means the estimate does not pin down a narrow range of plausible effects. It should not be interpreted as proof that the groups are equivalent, nor as proof that a difference exists. The dislocation outcome is named but has no posted result.

The trial’s randomized design is relevant to the intended causal comparison, but the registry record does not report key details needed to evaluate how the primary analysis used that design. In particular, it does not clarify the relation between the primary time labels and treatment groups or describe the analysis population and missing-data handling. The numerical results should be interpreted within those reporting limits.

These findings describe the posted record and its statistical precision. They do not establish that either bearing type is better or worse, and they are not a basis for medical advice. A result that is imprecise or incompletely reported is not the same as evidence of no effect; it means the available information does not settle the comparative question.

14. Why This Trial Matters Statistically

This trial raises an important statistical distinction between measuring an outcome and estimating a treatment effect. An ambulatory device with five inertial sensors can yield a quantitative gait measure, but a numerical endpoint becomes informative about a randomized comparison only when the groups, analysis, and uncertainty are clearly reported. Here, the primary summaries are organized by time labels rather than bearing type, and the registry gives a test result without a treatment contrast estimate or confidence interval.

The study also illustrates why outcome summaries and hypothesis tests answer different questions. Means and interquartile ranges describe the recorded distributions in broad terms, while the Wilcoxon (Mann-Whitney) test uses ranks. Neither the summary values nor the p-value alone communicates the magnitude and precision of a fixed-versus-mobile difference. Reporting a clearly defined effect estimate with an uncertainty interval would make that comparison easier to interpret.

The secondary knee-score result shows a related issue: a recalculation from group summaries can be useful for orientation, but it is not the same as a sponsor-reported analysis or a complete account of the planned analysis. The planning illustrations likewise show that sample-size requirements depend on the effect chosen for detection and on specified assumptions. They are prospective planning illustrations, not evidence about power achieved by the completed trial.

For clinical researchers, the record is a reminder to report endpoint definitions, analysis populations, missing-data handling, multiplicity decisions, and group-specific effects alongside test results. Those details help readers distinguish an inconclusive estimate from a precisely estimated small effect, and a time comparison from a randomized treatment comparison. Their absence limits what this registry entry can establish, even though it does not determine what other study documents may contain.

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 trial-specific limitation is the mismatch between the primary outcome labels and the treatment comparison. The gait summaries are labelled “Preoperative” and “5 Years,” while the study arms are labelled by bearing type. The registry does not explain whether these are pooled summaries, separate treatment-group summaries at different times, or another analysis. It also does not post a primary treatment effect estimate or confidence interval. This makes the primary result difficult to connect directly to the randomized comparison.

The primary endpoint description gives the measurement unit, sensor setup, and assessment time, but not the gait protocol or the interpretation threshold for a meaningful change. Although the outcome is described as a mean with an interquartile range, further distributional information and the reasoning behind the named rank-based test are not reported. The registry does not report whether the five-year analysis accounts for baseline measurements or repeated observations.

The record does not describe the randomization procedure, allocation ratio, masking roles, analysis population, missing-data handling, or multiplicity approach. It also does not report the trial’s sample-size justification. These omissions limit assessment of bias, precision, and the relationship between the registry results and any planned analysis.

Reporting is incomplete across outcomes: the dislocation outcome has no result, and the registry does not identify the additional outcome measures implied by its overall outcome count. The adverse-event summary has few reported events and lacks event descriptions and death counts. Finally, the knee-score comparison is a recalculation from posted summaries rather than a sponsor-reported analysis. It is unadjusted and should not be treated as a substitute for a fuller trial analysis.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Total Knee Arthroplasty: Comparison of Fixed and Mobile Bearings — NCT00417859.
• Registry record last updated 2020-04-21; status last verified 2020-04.
• Results first posted 2020-04-07.
• 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

NCT00417859 is a completed randomized, triple-masked, parallel-group study sponsored by University of Lausanne Hospitals. It compares mobile- and fixed-bearing total knee arthroplasty in the setting of knee osteoarthritis. The registry records enrollment of 64, a start date of 2004-02, and primary completion in 2015-11.

The primary outcome is gait stride length at five years, measured in meters with a data logger using five inertial sensors. The posted summaries are labelled “Preoperative” and “5 Years,” and the reported Wilcoxon (Mann-Whitney) test has p = 0.05. A bearing-specific primary effect estimate and confidence interval are not reported. The knee-score outcome has posted group summaries and an unadjusted recalculation; the named dislocation outcome has no posted result.

The registry also reports zero serious adverse events in both groups and one other adverse event in the fixed-bearing group. The primary analysis labels, analysis population, missing-data handling, multiplicity approach, and several other details are not reported in the registry record. Consequently, the available information describes selected outcomes but does not fully quantify the primary randomized treatment contrast.