Virtual vs. Traditional Physical Therapy Following Total Knee Replacement: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Virtual vs. Traditional Physical Therapy Following Total Knee Replacement (NCT02914210), a randomized na study. The primary outcome is Cost Difference in US Dollars.

2. Clinical Question

3. Trial Design

The registry record describes a randomized, parallel-group trial with no masking. The study title identifies it as Virtual Exercise Rehabilitation In-home Therapy: A Randomized Study (VERITAS), sponsored by Duke University. The condition is total knee replacement. The registry lists enrollment as 306 and the study status as completed. The recorded study start is 2016-10, with primary completion in 2018-03.

Participants were assigned to virtual physical therapy rehabilitation or traditional physical therapy. The virtual arm used an in-home program with care planning and remote support and monitoring by physical therapists. The traditional arm received standard home health physical therapy and/or outpatient clinic physical therapy as prescribed. Thus, the contrast is between two ways of delivering rehabilitation, not between rehabilitation and no care.

The primary endpoint is defined for patients randomized and receiving eligible surgery who completed the study through 12 weeks after surgery. The registry lists 143 participants in the virtual arm and 144 in the traditional arm for this endpoint. It does not report whether the posted outcome population follows an intention-to-treat approach, how participants were handled if they did not complete follow-up, or whether the analysis included any other participants.

The registry record does not report a masking procedure, and the trial is explicitly listed as having none. It also does not report the detailed allocation sequence or concealment procedures. These omissions limit assessment of how assignment was implemented, but do not change the basic design description: randomized allocation to parallel groups with a comparison of cost and other outcomes after surgery.

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

Analysis population for the primary outcome, as defined in the registry: Patients randomized and receiving eligible surgery, and completing the study through 12 weeks post surgery.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Enrolled patients excluding those found to be ineligible after enrollment and prior to surgery

6. Primary Endpoint

The registry lists the primary outcome as Cost Difference in US Dollars, measured over 12 weeks.

The primary endpoint is the difference in total health service use costs at 12 weeks after total knee replacement for patients receiving tele-rehabilitation-supported physical therapy versus traditional home and/or clinic-based physical therapy. The registry reports this outcome in US dollars, using a median and an inter-quartile range for each arm. The virtual arm has a reported median of 1050.0, with the reported interval from 900.0 to 1200.0. The traditional arm has a reported median of 2805.0, with the reported interval from 1644.0 to 4505.0.

Health service use costs are relevant to the study’s stated aim of comparing the resource costs associated with the two rehabilitation approaches. A median describes the middle of the observed cost distribution and is less influenced by unusually high values than an arithmetic mean. The inter-quartile range describes the central spread of the distribution. In general, cost data may be unevenly distributed, so a median and inter-quartile range can be useful descriptive summaries.

The registry presents arm-level summaries rather than a reported between-arm cost estimate with an uncertainty interval or hypothesis test. The difference in the endpoint’s title should not be mistaken for a posted estimate of the contrast: the listed numbers are the separate arm medians and their ranges. The registry record does not report the cost components included, the method used to value services, or a detailed analysis of the cost distribution.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the randomized, parallel design mean?

In a randomized study, assignment to study groups is determined by a chance-based process rather than by a participant’s or clinician’s choice. This helps reduce systematic differences between groups at the start of the comparison, although randomization does not ensure that every measured or unmeasured characteristic will be exactly balanced. A parallel design means participants are assigned to one group and outcomes in the groups are compared. The registry lists this study as randomized and parallel, but the sequence-generation and allocation-concealment procedures are not reported in the registry record.

The study is listed as having no masking. In general, masking means keeping participants, clinicians, outcome assessors, or analysts unaware of group assignment. It can reduce the influence of expectations on care, reporting, or assessment when it is feasible. The registry record does not describe which roles, if any, were unaware of assignment, and its masking classification is none. For outcomes that depend on health service use, the record also does not explain whether collection or classification was conducted by assessors unaware of group assignment.

Why report a median and inter-quartile range for cost?

The median is the central value when observations are ordered. The inter-quartile range describes the middle portion of the observed distribution, bounded by the lower and upper quartiles. These summaries can be informative when observations are skewed or include unusually high values, a possibility often considered for costs in general. This is a general statistical point, not a claim about the shape of the cost distribution in this trial. The registry does not provide a histogram, a formal distribution assessment, or further distributional details for this endpoint.

The reported cost summaries are not themselves a test of a difference between groups. A comparison would require an estimate of the contrast and a measure of its uncertainty, with an analysis method appropriate to the endpoint and study design. The registry record does not report those primary comparison details. The inter-quartile ranges describe within-arm spread; they should not be read as confidence intervals for a between-arm contrast.

How should the secondary summaries be read?

The registry posts continuous summaries for gait speed and pain as means and standard deviations. A mean describes the arithmetic average, while a standard deviation describes the spread of observations around that mean. The registry also posts an unadjusted mean-difference comparison for each of these outcomes. In general, a confidence interval expresses the range of estimates reasonably compatible with the data under the method used; a wide interval indicates less precision than a narrow one. The posted intervals for secondary comparisons are unadjusted and should not be interpreted as accounting for multiple outcomes or covariates.

For the report-of-falls outcome, the registry presents percentages by arm and an unadjusted risk-difference comparison. A risk difference is an absolute contrast between the proportions in two groups. Its confidence interval describes uncertainty in that contrast under the calculation method. These are general explanations of the measures, not evidence that any particular secondary finding is reliable or clinically important. The registry record does not report an adjusted analysis for these comparisons.

What does the planning illustration mean?

The sample-size illustration is based on the observed effect and is explicitly not the trial’s own design calculation. It assumes a two-sided significance level of 0.05, equal allocation, and a normal approximation; its alternative effect sizes are stated fractions of the observed effect. The entries describe participants per arm for specified power levels under those assumptions. Such planning values are hypothetical design illustrations, not observed power and not a retrospective measure of whether the trial was adequately powered. The registry record does not report the original sample-size justification or the assumptions used when the trial was planned.

In general, power calculations depend on the effect size to be detected, outcome variability, significance threshold, allocation, and analysis method. Changing those assumptions changes the resulting planning requirement. A calculation based on the observed effect does not validate that effect estimate, and a planning illustration cannot repair missing uncertainty estimates or an unreported primary analysis.

Why do unadjusted results need caution?

An unadjusted comparison evaluates a group contrast without accounting for other covariates or multiplicity. In general, examining several outcomes creates more opportunities for apparently notable findings by chance. A prespecified testing strategy or multiplicity adjustment can address that issue, but neither is described for the posted secondary results. The registry lists no reported analyses. Thus, the posted secondary comparisons are useful descriptive evidence, but they should not be treated as a complete, confirmatory analysis of all study outcomes.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes, including cost, knee-related measures, range of motion, gait speed, pain, falls, and re-hospitalization. It does not report a statistical analysis plan, a hierarchy of primary and secondary tests beyond the named primary endpoint, or a procedure for controlling multiplicity. The registry lists zero reported analyses. Consequently, the posted unadjusted secondary comparisons should be read as separate exploratory evidence rather than as a family of results with a documented overall error-control strategy.

Missing-data methods are not reported in the registry record. In particular, the record does not describe how missing cost, functional, or patient-reported outcome measurements were handled; whether any imputation was used; or whether sensitivity analyses assessed the impact of missing observations. The primary endpoint population is defined as participants randomized and receiving eligible surgery who completed the study through 12 weeks post surgery. The record does not explain how that completion requirement was applied or how the results relate to all randomized participants.

The analysis population is also not described using standard labels such as intention-to-treat or per-protocol. The arm-level endpoint counts are reported, but the registry does not explain exclusions, post-randomization eligibility decisions, or the availability of each measure by group. The protocol’s sample-size justification, interim analyses, and detailed statistical plan are not reported in the registry record. The planning illustration is expressly based on the observed effect rather than the original trial design and should not be used to fill in those missing design details.

These omissions restrict conclusions about robustness and confirmatory status. They do not demonstrate that a particular method was or was not used; they mean that the method is not documented in the registry record.

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

The posted primary data describe health service use costs at 12 weeks for the virtual and traditional physical therapy groups. The arm-level medians and inter-quartile ranges allow readers to see the central cost summaries and the spread reported for each group. They do not provide a formal estimate of the between-group contrast or its precision. The primary result should therefore be described as a comparison of reported summaries, not as a definitive estimate of a causal cost difference.

The trial’s randomized design is relevant to comparing the assigned rehabilitation approaches, but interpretation also depends on how outcomes were collected and analyzed. The registry does not report the primary comparison method, missing-data handling, or a detailed analysis population beyond the stated completion criterion. Those gaps make it impossible to determine from the registry record alone how sensitive the cost summaries are to incomplete follow-up or analytic choices.

The results do not establish that one approach is clinically superior, nor do they establish that the approaches are equivalent. A lack of a posted primary inferential analysis is not evidence that the true contrast is zero. Likewise, a difference between arm medians is not on its own a measure of statistical certainty, clinical importance, or cost-effectiveness. The registry record does not provide a cost-effectiveness analysis, a threshold for a meaningful cost difference, or the detailed components included in health service use costs.

The secondary outcomes broaden the study’s focus but do not remove these limitations. Their measures differ in type and timing, and the registry does not report a multiplicity plan or a complete inferential analysis. The most defensible reading is therefore bounded: the posted data describe selected outcomes in the trial groups, while the strength and precision of the overall comparative conclusions remain incompletely documented.

12. Why This Trial Matters Statistically

This trial raises a statistical question that is important in rehabilitation research: how should a study compare approaches that may differ in both service delivery and patient outcomes? A cost endpoint can capture a dimension that functional and patient-reported measures do not, while those other measures address outcomes that cost alone cannot represent. Keeping the outcome domains distinct helps prevent a descriptive cost result from being treated as a complete account of recovery or value.

The design also illustrates why the analysis population and endpoint definition matter. The primary population is described as participants who were randomized, received eligible surgery, and completed follow-up through 12 weeks. That definition gives context for the posted summaries, but the registry does not explain missing-data handling or how that population relates to all randomized participants. For a randomized comparison, clear documentation of inclusion, outcome availability, and analysis methods is necessary to judge how well the estimate reflects the planned groups.

Another useful statistical lesson is that multiple secondary outcomes require a clear interpretive framework. When many domains are assessed, isolated unadjusted results can be difficult to distinguish from chance variation. A prespecified hierarchy or multiplicity approach can help readers understand which findings are confirmatory and which are exploratory. The registry record does not report such a framework, so the secondary findings should be considered supportive context rather than a single unified verdict.

Finally, the record demonstrates the value of distinguishing descriptive summaries from comparative inference. Medians and inter-quartile ranges convey observed cost distributions, but a comparative estimate with uncertainty is needed to quantify how precisely the groups differ. Transparent reporting of that analysis, along with missing-data and sample-size methods, would make the statistical evidence easier to evaluate.

13. 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 main limitation specific to this registry record is that the primary endpoint is reported through arm-level medians and inter-quartile ranges, without a posted between-group estimate, confidence interval, or test. The record also does not define the components of total health service use costs or explain how those costs were measured and valued. As a result, readers cannot assess from the registry entry how the reported cost measure was constructed or how precise the comparative cost result is.

The primary endpoint population is restricted to participants who received eligible surgery and completed follow-up through 12 weeks. The record does not report missing-data procedures or explain whether the population differs from the randomized groups in ways that could affect interpretation. The arm counts are posted, but detailed participant flow and outcome availability are not described in the endpoint information.

The study was unmasked, and the record does not describe blinding of outcome collection or cost classification. This may be relevant to outcomes involving reporting or decisions about service use, although the registry entry does not establish that any specific bias occurred. The record also does not provide detailed allocation procedures, the original sample-size rationale, interim-analysis information, or the statistical analysis plan.

For secondary outcomes, multiple unadjusted comparisons are posted without a reported multiplicity strategy. Their findings are therefore less definitive, particularly because the registry does not identify a clinical threshold for meaningful change or report analyses for every listed secondary outcome. These limitations concern the completeness of the registry reporting; they do not demonstrate that the trial investigators did not perform additional analyses outside the registry record.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Virtual vs. Traditional Physical Therapy Following Total Knee Replacement — NCT02914210.
• Registry record last updated 2021-11-02; status last verified 2021-10.
• Results first posted 2020-01-18.
• 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.

17. Record Summary

NCT02914210 is a completed randomized, unmasked, parallel-group study of virtual versus traditional physical therapy after total knee replacement. The registry lists enrollment as 306 and identifies Duke University as the sponsor. The primary endpoint is total health service use cost at 12 weeks after surgery, summarized by arm using medians and inter-quartile ranges. The endpoint population is described as randomized patients who received eligible surgery and completed the study through 12 weeks.

The virtual arm used an in-home rehabilitation program with remote physical therapist support and monitoring. The comparator was standard home health and/or outpatient clinic physical therapy as prescribed. Selected secondary outcomes have posted summaries, but the registry lists no reported analyses. The primary comparative estimate, missing-data methods, multiplicity strategy, and detailed statistical plan are not reported in the registry record.