Junctional AV Ablation in CRT-D: JAVA-CRT: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Junctional AV Ablation in CRT-D: JAVA-CRT (NCT02946853), a randomized na study. The primary outcome is Number of Patients With Reduction ≥ 15% in Left Ventricular End-systolic Volume (LVESV).

2. Clinical Question

The clinical question is whether adding AVJ ablation to CRT-D is associated with a different chance of meeting the specified LVESV reduction threshold than CRT-D alone. The registry reports 5 participants meeting the threshold in the CRT-D group and 6 in the CRT-D and AVJ ablation group. The corresponding recorded group sizes for this outcome are 10 and 12. These counts are the basis for the primary comparison; they do not by themselves establish that the treatments differ or are equivalent.

The endpoint focuses on a change in cardiac structure over the baseline-to-6-month interval. It is not a direct measure of symptoms, survival, hospitalization, or quality of life. The registry separately lists change in left ventricular ejection fraction as a secondary outcome. Broader clinical implications therefore cannot be inferred from the primary endpoint alone.

3. Trial Design

JAVA-CRT (NCT02946853) is recorded as a randomized, single-masked, parallel-group trial. Its listed enrollment is 26. The registry does not report which people were masked, how masking was maintained, or further details about randomization procedures. The phase is recorded as NA. The sponsor is the University of Rochester.

Participants were assigned to CRT-D alone or CRT-D with AVJ ablation. The primary endpoint was assessed from baseline to 6 months. The registry reports that some patients did not undergo outcome measures because they could not attend follow-up due to COVID-19. For the primary outcome, the recorded group sizes are 10 for CRT-D and 12 for CRT-D with AVJ ablation; these are the denominators associated with the posted outcome, rather than a statement that every enrolled participant contributed an outcome.

The registry lists systolic heart failure and permanent atrial fibrillation as conditions. The brief summary notes that some people with heart failure cannot maintain sinus rhythm and that atrial fibrillation occurs among CRT patients. This provides the clinical context for examining AVJ ablation alongside CRT-D, but the registry record does not report a detailed mechanistic rationale or additional eligibility criteria.

The study started on 2016-10-01, reached primary completion on 2020-08-31, and is recorded as completed. The record lists 9 outcome measures and 9 analyses, while one secondary endpoint is posted in the results used for this comparison. The registry does not report a detailed protocol, a sample-size justification, or a prespecified analysis plan. The planning illustrations associated with the observed secondary result are not evidence of the trial's original design assumptions.

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: Some patients did not undergo outcome measures due to inability to come for follow-up due to COVID-19.

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 Number of Patients With Reduction ≥ 15% in Left Ventricular End-systolic Volume (LVESV), measured over Baseline to 6 months.

The primary endpoint counts participants whose LVESV decreased by at least 15% between baseline and 6 months. LVESV is a measure of the volume of blood remaining in the left ventricle at the end of contraction. The registry outcome is a threshold-based count: each participant either meets the specified reduction threshold or does not. The posted results therefore report the number meeting the threshold in each group, rather than each participant's continuous volume change.

A threshold can make a result easier to interpret as the proportion of participants reaching a defined response criterion, while retaining an outcome related to cardiac structure. However, categorizing a continuous change loses information about how far each person is above or below the cutoff. Two participants with different magnitudes of change can be counted alike if both meet the threshold. Conversely, small differences around the cutoff can place participants in different categories.

The endpoint is assessed over the baseline-to-6-month timeframe. The registry does not report the measurement procedures, the rationale for selecting the 15% cutoff, or how outcome assessors were masked. Nor does this structural measure directly establish effects on symptoms or other clinical outcomes. Some patients did not undergo outcome measures because they could not attend follow-up due to COVID-19, so the outcome counts describe those with posted measurements rather than necessarily every randomized participant.

7. Results Posted in the Registry

• CRT-D and AVJ Ablation: 50.0% (6/12) versus 50.0% (5/10) on control; risk difference 0.0 percentage points (95% CI -42.0 to 42.0), risk ratio 1.00 (0.43–2.31); unadjusted two-proportion z-test p = 1.000.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change in Left Ventricular Ejection Fraction (EF)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • CRT-D and AVJ Ablation: mean -8.60 versus -7.00 on control; difference -1.60 (95% CI -10.54 to 7.34), standardized effect d = -0.15; z-test p = 0.726.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between CRT-D and AVJ Ablation and the control. The interval gives the range of effects that remain compatible with the data.

The registry posts one secondary endpoint: change in left ventricular ejection fraction from baseline to 6 months. Its group summaries and the recalculated comparison do not show a clear separation, and the sponsor-reported Wilcoxon analysis also has a large p-value. The uncertainty around the recalculated difference is broad and crosses zero. These observations are directionally consistent with the primary comparison in that neither endpoint establishes a clear between-group difference in the posted results.

That consistency should not be overstated. EF change and the LVESV threshold are different outcomes, and one is continuous while the other counts participants meeting a cutoff. They capture different aspects of cardiac function and structure. Similar lack of clear separation across them does not establish that the interventions have the same effects, and it does not substitute for precise estimates.

Secondary results generally carry less inferential weight than a prespecified primary endpoint, particularly when several outcomes or analyses may have been considered. Multiple testing increases the chance that at least one result appears noteworthy by chance if each result is interpreted on its own. The registry lists 9 outcome measures and 9 analyses, but the results available for this reading include one posted secondary endpoint. The registry does not report a multiplicity adjustment or a hierarchy for secondary outcomes. Consequently, unadjusted secondary p-values should be treated as exploratory rather than as independent confirmatory evidence.

The registry does not provide enough information to know whether the posted secondary endpoint was selected in advance, whether other secondary results were measured but not posted, or how missing EF measurements were handled. The available secondary result can complement the primary outcome, but it cannot resolve the primary comparison's uncertainty or establish broader clinical effects.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change in Left Ventricular Ejection Fraction (EF)

9. Statistical Interpretation

10. Statistical Methods Explained

Two kinds of outcomes appear in the registry results. The primary outcome is binary at the participant level: each participant either had a reduction of at least 15% in LVESV or did not. The secondary outcome is continuous: change in EF is reported as a group mean with a standard deviation. Methods for comparing these outcome types answer related but distinct questions.

What does the primary comparison estimate?

A risk difference compares the proportion meeting the endpoint in one group with the proportion meeting it in the other. A value of zero means that the observed proportions are equal. A positive or negative value indicates the direction of the difference according to the stated comparison order. For the primary endpoint, the recalculated risk difference is 0.0. That point estimate is easy to read, but it does not show how precisely the underlying difference has been estimated.

A confidence interval gives a range of values compatible with the data under the statistical method and its assumptions. For this primary comparison, the interval crosses zero and extends substantially in each direction. Thus, the data do not pin down a narrow range of possible group differences. The interval does not assign a probability to the true effect, and it should not be read as proof that every value inside it is equally plausible.

The registry also reports the sponsor's logistic regression analysis as an odds ratio. Odds compare the chance of meeting the endpoint with the chance of not meeting it; an odds ratio compares those odds between groups. An odds ratio of 1.00 is the no-association point estimate. Odds ratios and risk differences use different scales, so their numerical values are not directly comparable. The registry does not give further details about the regression model, including whether it included covariates.

Why distinguish the reported analysis from a recalculation?

The posted group counts permit a simple, unadjusted comparison of the observed outcomes. That recalculation does not reproduce a regression model or account for any model terms that may have been used by the sponsor. The registry's analysis is the sponsor's reported inferential result and takes precedence. The recalculation is useful as a transparent check on the group counts, not as a replacement for the sponsor's method.

A p-value describes how surprising data at least as inconsistent with a null hypothesis would be under the assumptions of the specified test, if that null hypothesis were true. It is not the probability that the null is true, nor a measure of clinical importance. A large p-value does not demonstrate that groups are equivalent. Interpretation should consider the estimate, its interval, the outcome definition, and the available sample information together.

How should the continuous secondary endpoint be read?

For EF change, the registry reports a mean and standard deviation for each group. The mean describes the average observed change; the standard deviation describes variability among participants around that average. The recalculated mean difference is -1.6, with a confidence interval from -10.54 to 7.344. Since that interval includes zero and is broad, the observed group means do not precisely establish the direction or size of a difference.

The registry's own secondary analysis uses a Wilcoxon (Mann-Whitney) method and reports a net mean difference of -1.0 with p = 0.974. This method is based on the relative ordering of observations rather than simply comparing group means under a normal model. The registry labels the reported parameter as a mean difference, but the test method and the recalculated mean comparison are not identical procedures. Their outputs should not be treated as if they were the same calculation.

The standard deviation is relevant because it describes outcome variability, while the standard error describes uncertainty in an estimated group mean. A confidence interval for a difference incorporates uncertainty from both groups. When outcome variability is large relative to the separation between group means, estimates tend to be less precise. These are general statistical principles; the registry does not report additional distributional diagnostics for EF change.

What does the planning illustration mean?

The planning figures for EF are illustrations based on the observed effect, not the trial's original design or observed power. They assume a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The illustration gives participants per arm for 80% and 90% power when planning for 100%, 75%, and 50% of the observed effect. The displayed values are 664 and 889 for 100% of the observed effect; 1,180 and 1,580 for 75%; and 2,654 and 3,553 for 50%.

These figures show how planning requirements change when the target effect is made smaller, under the stated assumptions. They are not the number enrolled in this trial, do not show that the trial had a given probability of detecting an effect, and do not validate the assumptions for a future study. The registry does not report its own sample-size justification, target effect, or power calculation.

11. Confidence Intervals and Precision

Precision concerns how tightly the data locate an effect estimate. The primary risk-difference estimate is 0.0, but its confidence interval extends from -0.420 to 0.420. The interval includes no difference and values on either side of it, so the observed equality of proportions does not rule out a difference favoring either group. Its breadth indicates that the primary comparison is not estimated precisely.

The sponsor's reported odds ratio is 1.00, with a 95% confidence interval from 0.14 to 7.21. This interval also includes the no-difference value and permits a wide range of odds ratios. The risk difference and odds ratio have different scales and are derived through different analyses, so their interval limits should not be compared numerically as though they measured the same quantity. Both convey substantial uncertainty in the primary comparison.

For the continuous EF outcome, the recalculated mean difference is -1.6 and its interval runs from -10.54 to 7.344. This range includes zero and allows estimates in either direction. The interval should be read alongside the observed means and variation, not as a precise estimate of a likely effect.

Confidence intervals are compatible-value ranges under a statistical procedure and its assumptions; they are not guarantees about where the true value lies. The registry record does not provide enough detail to assess all assumptions underlying the sponsor's regression analysis. In general, limited outcome information tends to produce broad intervals, and that is an important constraint on conclusions from these results. A non-significant test cannot compensate for low precision, and a point estimate alone cannot convey the range of effects consistent with the data.

12. Analyses Reported by the Sponsor

The registry reports the sponsor's primary analysis as logistic regression, with an odds ratio of 1.00, a 95% confidence interval from 0.14 to 7.21, and p = 1.00. The reported point estimate indicates no observed difference on the odds-ratio scale, while the interval is wide and includes the no-difference value. The analysis is the sponsor's own reported result and takes precedence over a simple recalculation.

The unadjusted recalculation uses the posted participant counts: 5 meeting the threshold among 10 in the CRT-D group and 6 among 12 in the CRT-D and AVJ ablation group. It yields a risk difference of 0.0 and p = 1.0. This is a direct comparison of the observed counts, not a reproduction of the sponsor's logistic regression. The risk difference and odds ratio also express effects on different scales.

The two analyses have aligned point estimates and neither indicates an observed separation in these data. Their uncertainty, rather than the matching central estimates, is central to interpretation. The sponsor's interval allows a broad range of odds ratios, and the recalculated risk-difference interval likewise spans values favoring either group. The registry does not report enough model detail to determine whether or how adjustment affected the sponsor's estimate.

13. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when investigators evaluate several outcomes, time points, subgroups, or analyses. If each is tested separately without adjustment, the chance of at least one apparently positive finding can increase. The registry lists 9 outcome measures and 9 analyses, while the results include one posted secondary endpoint. The record does not report which outcomes were designated in advance as primary or secondary in a full analysis hierarchy, whether any multiplicity correction was used, or whether additional results were not posted. The secondary p-value should therefore not be read as confirmatory evidence independent of the primary result.

Missing outcome data are also relevant. The primary outcome description says that some patients did not undergo outcome measures because they were unable to attend follow-up due to COVID-19. The registry does not report the number missing by group, the reasons for any other missingness, or how missing outcomes were handled in the analyses. The outcome denominators are 10 for CRT-D and 12 for CRT-D with AVJ ablation, whereas the listed trial enrollment is 26. Without further accounting, it is not possible to determine the missing-data assumptions or whether the analysis preserved the original randomized groups.

Analysis populations matter because an analysis of all randomized participants can differ from one limited to those with measured outcomes or those who adhered to assigned treatment. The registry does not report a named analysis population, a participant flow, or details of treatment adherence for the primary endpoint. It also does not explain whether the logistic regression included adjustments or how incomplete records were treated.

These omissions limit assessment of possible bias and the robustness of the estimates. Missingness can affect a comparison if the chance of having an outcome recorded is related to treatment group or to the outcome itself. This is a general statistical concern, not a claim that such bias occurred in this trial. The posted counts remain interpretable as recorded outcomes, but conclusions about all randomized participants require information not reported in the registry record.

14. 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

Serious events with the highest arm-level rate: hospitalization with heart failure (7.7%).

The serious adverse event figures list 1 affected participant in each group, with 13 at risk in each group. Hospitalization with heart failure is the leading serious-event term, reported as 7.7%. The registry also records 0 participants with other adverse events in each group and deaths of 1 in the CRT-D group and 0 in the CRT-D and AVJ ablation group.

These figures are descriptive counts and should not be interpreted as proof that an event was caused by a treatment or that the groups have equivalent safety. The registry record does not provide further detail about event timing, severity beyond the serious-event classification, adjudication, exposure duration, or the relationship of events to the interventions. The listed at-risk counts also differ from the primary outcome denominators, so they should be understood within the safety reporting rather than substituted for the efficacy denominators.

With limited event information, apparent differences or similarities can be unstable. The reported counts do not support precise comparative conclusions about uncommon harms. They also do not address adverse outcomes that were not listed in the registry results.

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

The primary result describes the recorded number of participants who met a particular structural response threshold during baseline-to-6-month follow-up. The observed proportions were equal in the two groups, and the recalculated point estimate of the risk difference is zero. This is a statement about the posted observations, not a finding that AVJ ablation has no effect when added to CRT-D.

The confidence interval is broad and compatible with differences in either direction. Thus, the results do not establish equivalence, non-inferiority, or the absence of a clinically important effect. Those conclusions require an appropriate design and prespecified criteria; such criteria are not reported in the registry record. A p-value of 1.00 in the sponsor's analysis, or a large p-value in a recalculation, does not show that the treatments are identical.

The endpoint is a thresholded LVESV outcome, not a direct measure of how participants felt, functioned, or survived. The registry lists EF change as a secondary outcome, but neither outcome alone can establish the full clinical impact of the treatment strategies. The safety counts are also descriptive and too limited to establish comparative safety.

The trial record supports a cautious conclusion: among participants with posted primary outcome data, the same observed proportion in each group met the LVESV reduction threshold. The estimates remain imprecise, and follow-up outcome measurements were not available for some patients because of COVID-19-related inability to attend. Further interpretation depends on details not reported in the registry, including missing-data handling, analysis populations, and the full prespecified statistical plan.

16. Why This Trial Matters Statistically

This trial illustrates why a point estimate and a significance test are not enough to interpret a clinical comparison. The primary groups have matching observed proportions, producing a zero risk-difference estimate, but the confidence interval remains wide. Equal observed proportions in a small set of outcome observations do not establish that the underlying effects are equal. Precision and the range of effects compatible with the data are essential parts of the statistical reading.

The record also makes the distinction between analysis scales visible. The sponsor reports a logistic-regression odds ratio, while the recalculation presents a risk difference from the posted counts. These measures answer related questions but are not interchangeable. Their point estimates can be interpreted in context, and their intervals show how much uncertainty remains.

Missing outcome measurements matter as well as the nominal enrollment. The record notes COVID-19-related inability to attend follow-up, but does not report a detailed account of missingness or analysis populations. That gap prevents a full assessment of how the available outcome data relate to the randomized comparison.

Finally, planning calculations based on an observed secondary effect illustrate how uncertain pilot-scale estimates can inform future planning only under explicit assumptions. Those calculations are not observed power or the trial's own sample-size rationale. The registry does not report that rationale. Together, these features make JAVA-CRT useful for teaching the difference between an observed result, its precision, and the broader conclusions a trial can support.

17. 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 posted primary outcome is based on 10 participants in the CRT-D group and 12 in the CRT-D and AVJ ablation group, despite a listed enrollment of 26. The registry notes that some patients could not attend follow-up because of COVID-19, but it does not provide a complete account of missing outcomes by group or explain the analysis approach to missingness. This limits assessment of whether the observed counts represent the randomized comparison without bias.

The primary endpoint reduces a continuous change in LVESV to a threshold response. The registry does not report the rationale for the 15% cutoff, measurement procedures, or distributions of individual changes. The endpoint therefore cannot show how far participants' values changed or whether results are sensitive to the chosen threshold.

Further limitations include the absence of detailed eligibility criteria, a participant-flow account, masking details, model specifications, and a prespecified statistical plan in the registry record. The record lists 9 outcome measures and 9 analyses, but multiplicity handling is not reported. Only one secondary outcome is available in the results, limiting the ability to assess consistency across the broader set of outcomes.

The reported serious adverse event counts are sparse and lack detail about timing, attribution, or follow-up exposure. They cannot support precise safety comparisons. These limitations do not invalidate the posted counts, but they restrict the strength and scope of conclusions that can be drawn from them.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Junctional AV Ablation in CRT-D: JAVA-CRT — NCT02946853.
• Registry record last updated 2022-01-25; status last verified 2021-12.
• Results first posted 2022-01-25.
• 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.

21. Record Summary

JAVA-CRT, registered as NCT02946853, is a completed randomized, single-masked, parallel-group trial sponsored by the University of Rochester. It lists enrollment of 26 and compares CRT-D alone with CRT-D plus AVJ ablation in patients with systolic heart failure and permanent atrial fibrillation.

The primary endpoint is the number of participants with at least a 15% reduction in LVESV from baseline to 6 months. The posted counts are 5 of 10 for CRT-D and 6 of 12 for the combination group. The sponsor reports a logistic-regression odds ratio of 1.00, with a 95% confidence interval from 0.14 to 7.21 and p = 1.00. The unadjusted recalculation gives a risk difference of 0.0, with a broad interval that includes differences in either direction.

The record notes COVID-19-related inability to attend follow-up for some patients. One secondary endpoint, change in EF, is posted. The available estimates do not establish a between-group difference, but their imprecision means that they also do not establish equivalence or absence of effect.