Personalised HTO Versus Generic HTO Virtual Clinical Trial: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Personalised HTO Versus Generic HTO Virtual Clinical Trial (NCT03419598), a registered phase not stated study. The primary outcome is Mechanical Stress in Plate During Gait.

2. Clinical Question

The comparison is virtual: the registry describes a cohort of models receiving both procedures, an arrangement it says is only possible in a virtual scenario. This design focuses the question on mechanical behaviour under simulated functional loading rather than on clinical outcomes after surgery. The primary measure is mechanical stress in the plate during gait at 6 weeks simulated post-operation. Mechanical strain in bone around screws during gait is listed as a secondary outcome.

The clinical question is framed as safety equivalence, but a statement of that question is not evidence that equivalence was demonstrated. Assessing equivalence requires results and a defined acceptable boundary for the difference in outcomes. Such a boundary and the primary outcome results are not reported in the registry record. The record therefore describes the comparison of interest, but does not establish whether the procedures met an equivalence criterion.

3. Trial Design

This is a completed virtual clinical trial sponsored by the University of Bath. The registry gives a start date of 2017-01-03 and a primary completion date of 2019-06-28. Enrollment is listed as 30. The condition is knee osteoarthritis. Rather than testing surgical procedures in people, the study uses existing anonymised 3D imaging data to create virtual tibia models.

The registry describes intact, un-operated tibia models based on each subject’s CT geometry. The virtual cohort is described as receiving both procedures: an opening wedge high tibial osteotomy with a personalised subject-specific custom plate and one with a generic HTO plate. The summary identifies the generic plate as the Tomofix plate. This within-model feature is central to the virtual setup, but the record also cautions that the arms are not a like-for-like control comparison. Allocation is not stated.

The primary endpoint is mechanical stress in the plate during gait at 6 weeks simulated post-operation. It is calculated as Von Mises stress under functional loading using finite element analysis. Mechanical strain in bone around screws during gait is listed as a secondary outcome. The registry lists two outcome measures, but reports zero analyses. Numerical outcome estimates, uncertainty measures, and a comparison result are not reported in the registry record.

The study status is COMPLETED. Completion status indicates the recorded trial status; by itself, it does not indicate that the primary question was answered or that the results support equivalence. The record’s stated purpose is to examine safety in terms of mechanical loads on the tibia and support plate, not to report clinical events following actual operations.

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:

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 Mechanical Stress in Plate During Gait, measured over 6 weeks simulated post-operation.

The primary endpoint is mechanical stress in the plate during gait, measured at 6 weeks simulated post-operation. The registry describes the measure as Von Mises stress, in MPa, calculated for functional loading using finite element analysis. It is reported as a mean, with standard error specified as the dispersion measure.

In general, mechanical stress describes the load experienced within a material under force. A plate’s stress under simulated gait loading is relevant to this study’s stated mechanical-safety question because the registry focuses on loads placed on the support plate and tibia. The endpoint is a model-based mechanical measure, not a direct measure of pain, function, complications, or other clinical experiences.

In general, the mean represents the central value of measurements across the analysed units. The standard error describes the precision of an estimated mean, rather than the spread of individual observations. It is therefore useful to distinguish standard error from a standard deviation: the latter describes variability among observations, while the former concerns uncertainty in the estimated mean. The registry’s endpoint description specifies standard error, but does not report the observed mean or its standard error.

Because the measure comes from finite element analysis and simulated gait, interpretation depends on the modelled geometry, loading conditions, and analytic choices. The registry identifies the general method and timeframe, but details of those modelling choices and the numerical results are not reported in the registry record. A measured or estimated stress value alone would not establish clinical safety; the registry’s safety question requires an interpretable comparison and a criterion for equivalence.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary measure represent?

The primary measure is the mean mechanical stress in the plate during simulated gait, expressed in MPa. The registry describes it as Von Mises stress calculated using finite element analysis under functional loading. In general, finite element analysis represents a structure as a mathematical model and estimates how it responds to specified forces and material properties. The resulting value is a model-based estimate; it is not a direct measurement of a patient’s clinical experience.

The registry specifies the mean as the summary measure and standard error as the dispersion measure. In general, the mean describes the centre of the analysed values, while the standard error indicates uncertainty in the estimated mean. Standard error is not interchangeable with standard deviation: standard deviation describes variation among observations, whereas standard error describes the precision of an estimate. The observed mean and standard error are not reported in the registry record.

How should uncertainty be assessed?

In general, an estimate should be interpreted together with an interval describing its uncertainty. A confidence interval gives a range of values compatible with the data and the method used to construct it. A narrow interval indicates greater precision than a wide interval, although precision alone does not establish clinical importance or equivalence. No confidence interval for the primary outcome or for a comparison is reported in the registry record.

A p-value, when reported, addresses how compatible data are with a specified null hypothesis under the assumptions of a statistical test. It does not measure the size or importance of an effect and does not, on its own, demonstrate equivalence. No p-value or inferential analysis is reported in the registry record. The record’s stated equivalence question would require a suitable equivalence framework and a prespecified margin; the margin and analysis plan are not reported.

What does the virtual comparison imply for analysis?

The registry says the virtual cohort receives both procedures. In general, when each model contributes results under both procedures, the comparison may need to account for the relationship between measurements from the same model. Treating such measurements as independent when they are related can misstate uncertainty. The record does not report the statistical method used to handle this feature, and it cautions that the arms are not a like-for-like control comparison. Allocation is not stated.

Finite element results also depend on the model assumptions and the conditions used to represent functional loading. In general, changes in geometry, material properties, constraints, or applied forces can affect model outputs. The registry identifies CT-based tibia geometry and functional loading but does not report further modelling assumptions, validation details, or sensitivity analyses. These details are important for assessing how far the mechanical estimates can be generalised beyond the modelled setup.

What do the planning calculations mean?

The registry describes sample-size assumptions as a planning illustration computed from the observed effect, rather than from the trial’s own design. It specifies a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. In general, planning calculations estimate the sample size that might be needed under chosen assumptions. They are not a substitute for reporting the actual outcome estimates, and they are not observed power. The registry does not report a sample-size justification for the study’s original design.

For non-time-to-event calculations, the registry’s planning convention is participants per arm for 80% power or another stated power level. Time-to-event calculations use a total number of events. These conventions describe how planning illustrations are expressed; they do not show that the trial achieved a particular power or that the comparison was adequately precise. No new sample-size calculation is made in this explanation.

What statistical information is missing?

The registry lists two outcome measures and zero analyses. It does not report an analysis population, endpoint-specific counts, outcome values, treatment contrasts, confidence intervals, p-values, or a statistical model. It also does not report an interim analysis plan or adjustment for multiple testing. These omissions prevent an independent assessment of the primary comparison, the uncertainty of its estimate, and the relationship between the primary and secondary outcomes. The study status of COMPLETED does not fill those reporting gaps.

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("Personalised HTO", "Generic HTO")
n   <- c(28, 28)
m   <- c(108.3, 90.6)
sp  <- c(3.81, 3.56)
sd  <- sp * sqrt(n)      # standard errors converted to SDs
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  Personalised HTO  28  108.300  20.161  100.483  116.117
CARM  Generic HTO  28  90.600  18.838  83.295  97.905

11. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures: mechanical stress in the plate during gait as the primary outcome, and mechanical strain in bone around screws during gait as a secondary outcome. It reports zero analyses. The record does not describe whether multiple outcomes were tested, whether a testing sequence was specified, or whether any adjustment for multiple comparisons was planned. These details are not reported in the registry record.

In general, when several outcomes or comparisons are examined, the chance of at least one apparently notable result can increase. A prespecified approach to multiplicity helps clarify which findings are confirmatory and how uncertainty should be interpreted across outcomes. Without the analysis plan and results, it is not possible to assess how the primary and secondary outcomes were handled or to distinguish planned from exploratory analyses.

The registry does not report missing-data rules, exclusions, imputation methods, or the number of models contributing data to each endpoint. It also does not define an analysis population. Enrollment is listed as 30, but enrollment alone does not establish the number of usable observations for a particular outcome or comparison. Reasons for any missing or excluded model data are not reported in the registry record.

Allocation is not stated. The registry describes the virtual cohort as receiving both procedures, while also cautioning that the arms are not a like-for-like control comparison. The analysis population and method for handling within-model comparisons are not reported. These missing details limit evaluation of the analysis, including whether its uncertainty estimates appropriately reflect the virtual design.

The record does not report an interim analysis plan or a prespecified equivalence margin. Both are relevant to interpreting a claim of equivalence, although the registry’s stated question alone does not show that such a claim was tested. Without a defined margin, analysis results, and uncertainty estimates, equivalence cannot be evaluated from the registry record.

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

The record establishes that the study was designed as a virtual comparison of personalised and generic HTO plates, focused on mechanical loads in the tibia and support plate. It specifies plate stress during gait at 6 weeks simulated post-operation as the primary outcome, and bone strain around screws during gait as a secondary outcome. It also identifies finite element analysis as the method for calculating the primary measure.

The record does not report numerical results for either outcome, a comparison estimate, or uncertainty around a comparison. It therefore does not show whether stress or strain was higher under either procedure, how large any difference was, or how precisely a difference was estimated. It does not establish equivalence or non-equivalence. The objective of assessing equivalence should not be mistaken for a result.

In general, model-based mechanical measures can help address a specific question about simulated forces and structural response. They do not, by themselves, answer questions about clinical outcomes after an operation. The registry describes virtual interventions using anonymised imaging data rather than procedures performed on participants, so the record does not provide clinical outcome evidence from actual surgery.

The registry’s adverse-event fields list zero serious adverse events, other adverse events, and deaths, but also list zero participants at risk. In this virtual study context, these entries should not be interpreted as evidence that actual patients experienced no adverse events. The record does not report a clinical safety follow-up population or event rates.

In general, a result from a finite element model applies to the model and assumptions used to generate it. Judging wider relevance requires information about model construction, loading assumptions, and validation, as well as outcome results. The registry record does not report those details or the numerical findings. Its contribution in the available record is therefore a defined virtual mechanical question and outcome framework, rather than evidence that one procedure is safer, less safe, or equivalent to the other.

13. Why This Trial Matters Statistically

This trial is statistically distinctive because it addresses a mechanical question through a virtual cohort rather than procedures performed on people. The registry describes CT-based tibia geometry and finite element calculations under functional loading. That changes what the outcome means: the primary endpoint is a model-derived estimate of plate stress, not a clinical event or patient-reported outcome.

The virtual cohort is described as receiving both procedures, potentially allowing each model’s geometry to inform results under each procedure. In general, such designs can support efficient within-model comparisons, but the analysis must account for the dependence between measurements from the same model. The registry does not report how this dependence was handled, and it cautions that the arms are not a like-for-like control comparison. Those design features make transparent methods and clear limits on interpretation especially important.

The question is framed as equivalence. Statistically, equivalence is not simply a lack of evidence for a difference. It requires an acceptable difference to be defined in advance and the uncertainty in the estimated comparison to be sufficiently constrained relative to that boundary. The registry record does not report an equivalence margin or the results needed to judge one. This distinction helps prevent a study objective from being read as a demonstrated finding.

The record also illustrates why enrollment and completion status cannot replace outcome reporting. Enrollment is listed as 30 and the study status is COMPLETED, but the record reports zero analyses and no numerical outcome results. Those details do not indicate the precision of the comparison or whether the data support the intended question. For model-based studies, reporting the endpoint estimates, uncertainty, analysis population, and modelling assumptions is central to evaluating what the computation can establish.

14. 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 primary or secondary outcome values, comparison estimates, confidence intervals, or p-values. It lists zero analyses. This is the main limitation for interpreting the study statistically: the stated question cannot be answered from the recorded results.

The record does not report an equivalence margin or a statistical analysis plan. It also does not report how measurements from models receiving both procedures were compared, or how the caution that the arms are not a like-for-like control comparison was addressed. Allocation is not stated. Without these details, it is not possible to assess whether an analysis would estimate the intended contrast or how its uncertainty should be interpreted.

The primary endpoint is derived from finite element analysis under functional loading. The record identifies CT-based geometry and the broad calculation method, but does not report further assumptions about model construction, material properties, loading conditions, validation, or sensitivity analyses. These factors may affect model-based estimates in general, and their omission limits evaluation of the robustness and scope of any potential result.

The record does not report endpoint-specific analysis counts, missing-data procedures, exclusions, or an analysis population. Enrollment of 30 does not establish how many models contributed to an endpoint. It is also not clear from the record how the secondary outcome or multiplicity was handled. Finally, because the study is virtual, its mechanical outputs do not directly describe clinical outcomes or adverse events following actual surgery. The adverse-event fields list zero at-risk participants, so the event counts should not be treated as clinical event-rate evidence.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Personalised HTO Versus Generic HTO Virtual Clinical Trial — NCT03419598.
• Registry record last updated 2020-12-07; status last verified 2020-11.
• Results first posted 2020-12-07.
• 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.

18. Record Summary

NCT03419598 is a completed virtual clinical trial titled “Personalised HTO Versus Generic HTO Virtual Clinical Trial,” sponsored by the University of Bath. The registry lists knee osteoarthritis as the condition, enrollment as 30, a start date of 2017-01-03, and a primary completion date of 2019-06-28.

The study uses anonymised 3D imaging data to create CT-based tibia models and compares virtual opening wedge high tibial osteotomy procedures with personalised subject-specific and generic plates. The primary endpoint is mean mechanical stress in the plate during gait at 6 weeks simulated post-operation, calculated using finite element analysis. Mechanical strain in bone around screws during gait is listed as a secondary outcome.

The registry describes an equivalence question concerning mechanical loads on the tibia and support plate, but reports zero analyses and no numerical outcome results. It does not report an equivalence margin, analysis plan, or comparison uncertainty. The available record therefore describes the study’s design and measures but does not establish whether the procedures were mechanically equivalent.