TXA vs. Amicar in Total Knee and Hip Arthroplasty: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for TXA vs. Amicar in Total Knee and Hip Arthroplasty (NCT02030821), a randomized phase 4 study. The primary outcome is Total Blood Loss Over Course of Stay (Intraoperative and Postoperatively Until Discharge).

2. Clinical Question

3. Trial Design

The registry identifies NCT02030821 as a completed phase 4, randomized, parallel-group study sponsored by Duke University. The record gives enrollment as 246. It lists a start date of 2015-01 and a primary completion date of 2017-11-12. The study title describes a comparison of TXA and Amicar in total knee and hip arthroplasty, with effectiveness, safety, and cost analysis among its aims.

The allocation is described as randomized, and masking is listed as none. The record does not report the randomization sequence method, whether allocation was concealed, or any steps taken to limit bias in outcome measurement. The arm results separate the operations into total hip arthroplasty and total knee arthroplasty, giving four groups: TXA and Amicar for each procedure type. That separation matters when reading the analysis: the reported test covers all four groups rather than a single pooled two-arm comparison.

The registry says blood loss, transfusion need, preoperative and lowest postoperative hematocrit and hemoglobin, and complications would be collected during hospitalization. The listed primary endpoint is total blood loss over the hospital stay. Length of hospitalization and cost of hospitalization are listed as secondary outcomes. Although the study summary describes a cost-analysis objective, numerical results for cost and length of stay are not reported in the registry record.

The registry does not report the planned sample-size justification, details of the protocol, or a full statistical analysis plan. It also does not describe how participants were assigned across the joint-procedure groups or whether any planned analyses pooled the procedures. Those omissions limit what can be concluded about the intended design beyond the randomized, unmasked, parallel structure recorded.

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: Patients were excluded from the trial after enrollment if their surgery was cancelled.

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.

Population: All subjects who completed the study

6. Primary Endpoint

The registry lists the primary outcome as Total Blood Loss Over Course of Stay (Intraoperative and Postoperatively Until Discharge), measured over Participants will be followed for the duration of hospital stay, an expected average of 5 days.

The primary endpoint is total blood loss over the course of the hospital stay, combining intraoperative loss with postoperative loss until discharge. It is measured in milliliters. The record says participants were followed for the duration of hospitalization, described as an expected average of 5 days. The endpoint therefore covers more than the operation itself, but its follow-up ends at discharge rather than at a fixed time after surgery.

The registry reports the endpoint as a median with an inter-quartile range. A median describes the center of the observed distribution: half of the observations are at or below it and half at or above it. The inter-quartile range describes the span from the lower to the upper quartile and gives a sense of variation among participants without assuming that the data are symmetric. This is a useful descriptive approach when a distribution may be skewed or affected by unusually high values. That explanation is general; the registry does not report why this summary was selected.

Blood loss is a direct measure of one of the study's stated concerns in major joint surgery. The study summary also describes transfusion need and blood measurements as collected information, but those are not the primary endpoint reported in the posted results. The record does not provide a clinical threshold for interpreting a difference in milliliters or explain how the primary endpoint relates to transfusion decisions.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is summarized with medians and inter-quartile ranges. In general, the median is a measure of location that is less affected by extreme observations than the arithmetic mean. The inter-quartile range gives the lower and upper quartiles, which contain the middle half of the observations. Reading the median together with this range helps distinguish a group’s central value from the spread of participant-level results. These summaries are descriptive: they do not, on their own, provide an estimate of the uncertainty in a difference between treatments.

What does the Kruskal-Wallis test assess?

The registry names a Kruskal-Wallis test for the four hip-and-knee, TXA-and-Amicar groups and reports p < 0.05. In general, the Kruskal-Wallis test is a rank-based method for comparing distributions across independent groups. It uses the ordering of observations rather than relying on normally distributed measurements. The test evaluates an overall null hypothesis that the groups have the same distribution, subject to its assumptions. A small p-value indicates that the observed rank pattern would be unusual under that null hypothesis; it does not give the probability that the null hypothesis is true.

For this trial, the test is reported across four groups, not as a treatment-only comparison. The result therefore does not isolate TXA versus Amicar within hip procedures or within knee procedures. It also does not identify which group or groups account for the overall signal. Pairwise follow-up tests, estimates of contrasts, and adjustment for multiple comparisons are not reported in the registry record. The test result should be read at the level of the analysis that was actually reported.

Why does distributional spread matter?

In general, over-dispersion means that the observed variability is greater than a model expects. Ignoring extra variation can make standard errors too small and confidence intervals too narrow, making evidence appear more precise than it is. Rank-based analyses can avoid some distributional assumptions about the raw measurements, but they do not remove the need for suitable study design, independent observations, or clear reporting of the analysis population. The registry does not report a specific assessment of over-dispersion or a model-based analysis for this endpoint.

The inter-quartile ranges reported for each group are measures of participant-level spread, not standard errors. They should not be treated as uncertainty intervals around the medians or around treatment effects. A confidence interval, in general, describes a range of values compatible with the data and a specified statistical procedure. No confidence interval for a TXA-versus-Amicar treatment contrast is reported for this endpoint.

What do the planning illustrations represent?

Sample-size illustrations based on observed effects are planning calculations, not evidence about the trial's actual power after completion. The registry record's stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The target effect is described as a stated fraction of the observed effect. Such calculations explore how many participants per arm might be needed to detect an effect under those assumptions; they do not show that the trial had a particular power or that the effect is reproducible. The record does not report the trial's own sample-size justification, so the illustrations should not be mistaken for the design rationale used when enrollment was planned.

For a planning calculation, the assumed effect and analysis must match the intended clinical contrast. Here the registry reports four outcome groups and an overall Kruskal-Wallis test, while a simple two-arm illustration assumes a two-arm comparison. Those are different analysis structures. The illustrations cannot replace a prespecified analysis plan, account for every feature of the study design, or determine whether the posted test was adequately powered.

What information would support a fuller assessment?

Interpretation would be strengthened by a prespecified contrast, an effect estimate with a confidence interval, and details of any adjustment for operation type or other baseline factors. Information about missing observations, protocol deviations, analysis populations, and the handling of transfusions or discharge timing would also help assess how the reported endpoint was constructed. These details are not reported in the registry record. Their absence does not change the recorded medians or test result, but it limits independent evaluation of the analysis and its precision.

10. Analyses Reported by the Sponsor

The registry reports the sponsor's analysis of the primary endpoint as a Kruskal-Wallis test across the four procedure-and-treatment groups, with p < 0.05. This is the inferential result to distinguish from the group medians and inter-quartile ranges, which are descriptive summaries. The reported test indicates an overall difference in distributions among the groups but does not identify a particular pairwise contrast.

The posted results do not include a recalculated two-arm treatment estimate or confidence interval that can be set alongside the registry analysis. The four group summaries alone do not supply the participant-level ranks or other information needed to reproduce the reported Kruskal-Wallis calculation. They also do not define a single pooled TXA-versus-Amicar comparison. Accordingly, there is no independent recalculation in the registry record that contradicts or replaces the sponsor's test.

The record does not report a treatment-effect estimate adjusted for operation type, pairwise p-values, or a confidence interval for the treatment contrast. The sponsor's reported p-value should therefore be interpreted as an overall four-group test, not as proof of a treatment difference between TXA and Amicar within either operation type. The record lists no discrepancy between a sponsor estimate and a recalculated estimate.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 5 outcome measures and 4 analyses. It identifies total blood loss as the primary endpoint and lists length of hospitalization stay and cost of hospitalization as secondary outcomes. The registry record does not describe a multiplicity plan, such as which outcomes or group comparisons were designated for confirmatory testing or whether any adjustment to account for multiple testing was planned. The reported p-value is for a four-group Kruskal-Wallis analysis; pairwise follow-up tests are not reported.

Multiplicity matters in general because conducting many statistical tests increases the chance of obtaining at least one apparently unusual result even when there is no true difference. This does not make the reported p-value invalid, but it means its interpretation depends on the broader analysis plan and on how many outcomes and contrasts were considered. The record does not provide enough detail to determine how the reported test fits into a prespecified testing hierarchy.

Missing-data handling is not reported in the registry record. The group-specific outcome counts are reported, but the record does not explain whether every randomized participant contributed primary-endpoint data, why any data might be absent, or how missing observations were treated. It also does not report a flow of participants from randomization through analysis. The enrollment total and group result counts should not be treated as a substitute for that information.

The analysis population is not reported in the registry record. It is therefore unclear whether the reported summaries use all randomized participants, only participants who received assigned treatment, or another defined subset. The record does not state how protocol deviations or crossovers were handled. These details affect the interpretation of a randomized comparison because analyses based on different populations can answer different questions.

The registry summary says complications and transfusion need were collected, but the posted primary-endpoint results do not include a transfusion analysis. Results for the listed secondary outcomes are also not reported in the registry record. A complete accounting of outcome reporting and missingness cannot be reconstructed from the information available in the record.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: deep infection (6.5%), acute pulmonary embolism (2.7%), nstemi (1.4%).

The registry reports serious adverse events and other adverse events separately for each of the four procedure-and-treatment groups. For total hip arthroplasty, serious adverse events were reported for 3 of 46 participants in the TXA group and 0 of 44 in the Amicar group. Other adverse events were reported for 3 of 46 and 1 of 44, respectively. For total knee arthroplasty, serious adverse events were reported for 3 of 73 participants in the TXA group and 0 of 72 in the Amicar group. Other adverse events were reported for 1 of 73 and 1 of 72, respectively. No deaths are reported in any group.

The registry also lists Deep Infection (6.5%), Acute Pulmonary Embolism (2.7%), and NSTEMI (1.4%) among the top serious adverse event terms. It does not give the group assignment or event-level denominator alongside those terms in the record, so they should not be attributed to a particular treatment group on that basis.

These figures are descriptive event counts. The record does not report a formal comparative safety analysis, confidence intervals, event timing, or enough detail to assess how events were defined and collected. A count of zero in a group does not establish that the underlying risk is zero, and differences in counts alone do not demonstrate a treatment-related effect.

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

The registry results establish that blood-loss summaries were posted for four groups defined by treatment and operation type, and that the sponsor reported an overall Kruskal-Wallis result of p < 0.05. The medians and inter-quartile ranges describe the observed distributions in those groups. The reported test supports the conclusion that the four distributions were not all alike under the test's null hypothesis, but it does not show which groups differ or quantify a direct TXA-versus-Amicar effect.

The results do not establish that one drug reduces blood loss more than the other. To answer that question directly, readers would need a clearly specified treatment contrast, preferably analyzed separately by procedure or with a justified model that accounts for procedure type, together with an effect estimate and its uncertainty. Those items are not reported in the registry record. The group summaries are not a substitute for such an estimate.

The p-value does not measure the size or clinical importance of a difference. In general, a p-value summarizes how compatible observed data are with a specified null model; it is not the probability that a treatment works, nor does it indicate how likely a result is to be reproduced. The record does not specify a clinically important blood-loss difference or provide a confidence interval for the treatment contrast, so neither clinical importance nor precision can be determined from the reported test.

The study summary states that cost and hospitalization length were secondary goals, and the safety results include adverse-event counts. However, cost and length-of-stay results are not reported in the registry record, and the posted safety counts are not accompanied by a comparative analysis. The stated objectives should not be mistaken for findings.

In general, a randomized design can help balance known and unknown factors between assigned groups, but the interpretation also depends on how randomization was implemented, whether participants remained in their assigned groups for analysis, and how missing data were handled. These details are not reported in the registry record. The safest reading is therefore limited to the posted group distributions, the overall test, and the descriptive adverse-event counts.

14. Why This Trial Matters Statistically

This trial is statistically notable because its reported primary-endpoint analysis combines two dimensions: treatment and type of joint procedure. Reporting four groups preserves that distinction in the descriptive results, but the inferential analysis is an overall comparison among the four groups. That structure makes it especially important to align the statistical question with the clinical question. A broad test can detect an overall pattern without answering whether the two drugs differ within either procedure type.

The use of medians and inter-quartile ranges communicates distributional location and spread without reducing the outcome to a mean and standard deviation. This can be informative for blood-loss data, where values may not follow a symmetric distribution. The Kruskal-Wallis test is also rank based, which is a different inferential approach from a standard analysis that assumes normally distributed outcomes. These are general methodological considerations; the registry does not report a rationale for choosing these methods or diagnostic checks.

The record also illustrates why a p-value alone is not an adequate account of comparative evidence. The overall p-value is reported, but there is no effect estimate or confidence interval for a treatment contrast and no pairwise testing description. Readers can see the group-level medians and ranges, yet cannot assess the precision of a direct TXA-versus-Amicar comparison from those summaries alone.

Finally, the study's stated interests extend beyond blood loss to transfusion, safety, length of stay, and hospitalization cost. The registry posts some adverse-event counts but does not report results for the listed secondary outcomes. This distinction between planned objectives and posted results is important: a trial's broad goals do not guarantee that each outcome is available for interpretation. Statistical reporting is most useful when it connects each objective to a clearly defined estimate, uncertainty measure, analysis population, and accounting of missing observations.

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 most important trial-specific limitation in the registry results is that the reported inferential test is across four groups, while the central treatment question concerns TXA versus Amicar. The test does not provide a within-procedure treatment contrast or identify which groups differ. The group medians are descriptive and should not be treated as adjusted estimates of treatment effect.

The record does not report confidence intervals for treatment contrasts, a model adjusting for joint procedure, or pairwise comparisons. It gives no explanation of whether hip and knee procedures were intended to be analyzed together or separately for the primary question. Because blood-loss distributions are reported separately by procedure, this missing analytical detail limits interpretation of a pooled treatment comparison.

The registry does not report the sample-size justification, a full statistical analysis plan, or the randomization and allocation-concealment procedures. It also does not describe the analysis population, missing-data methods, participant flow, or how protocol deviations were handled. These omissions make it difficult to assess precision, attrition, and the extent to which the analysis follows the randomized allocation.

The primary endpoint is observed until discharge, so follow-up duration depends on the hospital stay. The record describes an expected average of 5 days but does not report each participant's follow-up duration or how discharge timing was handled in endpoint measurement. It also does not report results for length of stay, cost, transfusion need, or several collected blood measurements. The posted adverse-event counts lack a comparative analysis and additional event-level detail. These limitations concern the completeness and interpretability of the registry record; they do not alter the values that are reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: TXA vs. Amicar in Total Knee and Hip Arthroplasty — NCT02030821.
• Registry record last updated 2018-12-05; status last verified 2018-11.
• Results first posted 2018-12-05.
• 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

NCT02030821 is a completed, randomized, unmasked, parallel-group phase 4 study sponsored by Duke University, with enrollment of 246. It compares TXA and Amicar in total hip and total knee arthroplasty. The primary endpoint is total blood loss during hospitalization, reported as medians and inter-quartile ranges for four procedure-and-treatment groups.

The registry reports a Kruskal-Wallis analysis across those groups with p < 0.05. This is an overall four-group result, not a direct estimate of the TXA-versus-Amicar effect within either operation type. The record also reports serious and other adverse-event counts, with no deaths listed in any group. Results for the listed secondary outcomes of hospitalization length and cost are not reported in the registry record.

The main interpretive limits are the absence of a treatment-contrast estimate and confidence interval, and the lack of reported details on multiplicity, missing data, analysis populations, and the full statistical plan. The posted results support descriptive and overall-test interpretation, but not a precise, procedure-specific treatment comparison.