Assessment of Primary Prevention Patients Receiving An ICD - Systematic Evaluation of ATP: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Assessment of Primary Prevention Patients Receiving An ICD - Systematic Evaluation of ATP (NCT02923726), a randomized na study. The primary outcome is All-Cause Shock.

2. Clinical Question

In practical terms, the comparison concerns whether an ATP-first programming strategy changes the occurrence of a first shock from any cause relative to shock-only programming. The endpoint is not limited to shocks considered appropriate: an external committee adjudicated delivered shocks, and both appropriate and inappropriate shocks counted. That makes the primary result a broad measure of shock occurrence rather than a result about one specific reason for a shock.

3. Trial Design

APPRAISE ATP (NCT02923726) is recorded as a completed, randomized, single-masked, parallel-group device trial sponsored by Boston Scientific Corporation. Its registry enrollment is 2692. The phase field is listed as NA. The recorded start date is 2016-09-30, and the primary completion date is 2023-07-06.

Participants were assigned to one of two device-programming strategies. In the ATP and Shock arm, ATP was delivered before shock therapy once tachycardia had been detected and the required duration was met. In the Shock Only arm, shock therapy was delivered without preceding ATP after those same detection and duration conditions. The registry describes the study population as primary-prevention patients indicated for ICD therapy and programmed according to current guidance of higher rate cut-offs and therapy delays.

The primary endpoint is all-cause shock, specified as a count of participants over a 5-year timeframe. The registry says the endpoint population was all randomized subjects. The listed arm counts for this endpoint are 1302.0 in ATP and Shock and 1293.0 in Shock Only. The record also describes a time-to-first-all-cause-shock objective and reports a log-rank analysis, so results include both participant-count summaries and a time-to-event analysis.

The record lists four secondary outcomes: all-cause shock or death from any cause, appropriate shock, inappropriate shock, and all-cause mortality. Results are posted for the first three; results for all-cause mortality are not reported in the registry record. Details of the randomization procedure, who was masked, and the original sample-size justification are not reported in the registry record.

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: Of the 2692 participants enrolled,19 were consent ineligible and 12 were intent and therefore not randomized. 2595 participants were randomized. Consented and Ineligible: Consent signed but eligibility criteria was not met at time of enrollment • Intent: All eligibility criteria met and consent signed, but participant was not randomized due to: * no longer meeting eligibility criteria at the time of randomization * no longer willing to participate in the study

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized subjects.

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 All-Cause Shock, measured over 5 years.

The primary endpoint is the incidence of all-cause shocks, expressed as the count of participants with an event over the 5 years specified by the registry. The description says the primary comparison concerns shocks in participants programmed to shock only versus participants receiving standard therapy, defined as ATP and shock. It also frames the objective in terms of time to first all-cause shock.

All delivered shocks were adjudicated by an external committee as appropriate or inappropriate. Both types counted toward the primary endpoint. This broad definition captures whether a participant received a shock without restricting the outcome to a particular clinical classification of that shock. The endpoint therefore addresses the frequency or timing of shock occurrence, not simply whether ATP was delivered or whether an individual shock was judged appropriate.

General statistical context: time-to-event outcomes can use information about when an event occurs, rather than treating every participant as if their follow-up ended at the same point. They can also account for participants whose event time is not observed through the full follow-up period, using censoring methods. The registry reports a log-rank analysis for the time-to-first-shock objective. The handling of censoring and the precise endpoint rules for participants without a recorded event are not reported in the registry record.

The registry also posts participant counts for the endpoint. Those counts describe participants with an event, whereas the time-to-event result compares event timing over follow-up. These are related but not identical summaries, so their estimates and p-values need not match.

7. Results Posted in the Registry

• Shock Only: 13.8% (178/1293) versus 9.9% (129/1302) on control; risk difference 3.9 percentage points (95% CI 1.4 to 6.3), risk ratio 1.39 (1.12–1.72); unadjusted two-proportion z-test p = 0.002.

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.

All-Cause Shock or Death From Any Cause

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Shock Only: 24.9% (322/1293) versus 22.8% (297/1302) on control; risk difference 2.1 percentage points (95% CI -1.2 to 5.4), risk ratio 1.09 (0.95–1.25); unadjusted two-proportion z-test p = 0.211.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Shock Only and the control. The interval gives the range of effects that remain compatible with the data.

Appropriate Shock

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Shock Only: 10.2% (132/1293) versus 7.4% (96/1302) on control; risk difference 2.8 percentage points (95% CI 0.7 to 5.0), risk ratio 1.38 (1.08–1.78); unadjusted two-proportion z-test p = 0.011.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Shock Only than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Inappropriate Shock

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Shock Only: 4.8% (62/1293) versus 3.1% (40/1302) on control; risk difference 1.7 percentage points (95% CI 0.2 to 3.2), risk ratio 1.56 (1.06–2.31); unadjusted two-proportion z-test p = 0.024.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Shock Only than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes extend the primary question in two directions: a combined outcome of all-cause shock or death, and separate summaries of appropriate and inappropriate shocks. Across the posted results, the shock-specific outcomes show a pattern in the same broad direction as the primary participant-count comparison. The combined shock-or-death outcome is less conclusive in its unadjusted comparison, and its sponsor-reported time-to-event analysis is not statistically significant by the reported p-value. This broad pattern suggests that conclusions differ by endpoint rather than applying uniformly across every measure.

The registry lists all-cause mortality as a secondary outcome, but its results are not reported in the registry record. Results are posted for all-cause shock or death from any cause, appropriate shock, and inappropriate shock. That absence means the listed all-cause mortality outcome cannot be incorporated into an overall interpretation of the secondary findings.

The separate shock classifications help describe what contributed to the all-cause shock outcome, while the combined endpoint includes death alongside shock. These endpoints overlap conceptually and may also involve participants counted in more than one outcome; the registry does not explain the extent of overlap. They should not be treated as independent confirmations of the primary finding.

General statistical context: secondary analyses often involve multiple opportunities to observe a statistically notable result. If many hypotheses are tested, the chance of at least one apparently positive result can rise unless the analysis plan accounts for multiplicity. The registry does not report a multiplicity adjustment or a hierarchy controlling interpretation across endpoints. Consequently, unadjusted secondary results carry less confirmatory weight than a prespecified primary analysis, particularly when the full testing plan is unavailable. The secondary results can add context, but they do not independently establish the primary conclusion or resolve the equivalency question.

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

• All-Cause Shock or Death From Any Cause
• Appropriate Shock
• Inappropriate Shock
• All Cause Mortality

9. Statistical Interpretation

10. Statistical Methods Explained

The registry includes two types of statistical summaries for the primary endpoint: participant-count comparisons and a sponsor-reported time-to-event analysis. Each answers a related but distinct question. The count comparison focuses on whether a participant had an all-cause shock during the endpoint period. The time-to-event analysis concerns the timing of the first shock and uses the log-rank method.

What does the risk difference describe?

The reported risk difference compares the proportion with an event in one arm with the proportion in the other. In the posted recalculation, the contrast is Shock Only minus ATP and Shock. The estimate is 0.039, with a 95% confidence interval from 0.014 to 0.063. A positive value means the recorded participant risk is higher in the first-named arm of that contrast. This is an unadjusted summary based on the posted counts and denominators; it does not account for event timing or covariates.

General statistical context: an unadjusted comparison is useful for describing the observed arm-level data, but it is not automatically the same estimand or method as the analysis chosen by a trial sponsor. An estimand is the precise treatment contrast a statistical analysis is intended to describe, including the outcome and how events over time are handled.

What does the log-rank analysis add?

The sponsor's analysis is reported as a log-rank test with a hazard ratio. A log-rank test compares event-time patterns between groups over follow-up. A hazard ratio summarizes a relative contrast in event rates over time under the analysis framework. The record gives the hazard ratio as 0.72, a 95.9% confidence interval from 0.57 to 0.92, and p = 0.005. It does not state the reference coding or provide enough detail to interpret the direction of the hazard ratio independently of that coding. The result is marked as a superiority analysis.

General statistical context: time-to-event methods can incorporate the timing of events and observations that are censored, such as follow-up without an observed event through the analysis period. Such methods depend on assumptions and analysis choices. The registry does not report the censoring rules, model details beyond the log-rank method, or whether other adjustments were used.

Why can the analyses differ?

The participant-count estimate treats the endpoint as the presence or absence of an event in the posted count summary. The log-rank analysis evaluates when events occur. A count comparison can therefore differ from a time-to-event estimate even when both use the same trial arms and endpoint concept. A difference in their numerical results is not, by itself, evidence of an error; the methods answer different statistical questions.

The registry's primary endpoint description also uses the phrase time to first all-cause shock, while the posted outcome is described as incidence and recorded as a count of participants. Reading both descriptions together is important: the count summary gives the number of participants with an event, and the sponsor's log-rank result gives the reported time-to-event comparison. Further details of how the endpoint was operationalized and how the two analyses relate are not reported in the registry record.

How should the p-values be read?

The unadjusted participant-count comparison reports p = 0.002. The sponsor's log-rank analysis reports p = 0.005. General statistical context: a p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of the test. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. Interpretation also depends on whether the hypothesis and analysis were prespecified and whether multiple outcomes or analyses were considered. The registry's full statistical analysis plan and multiplicity procedure are not reported.

For a superiority test, a small p-value is commonly interpreted as evidence against the null hypothesis of no difference under the chosen test. That interpretation should not be transferred to an equivalence claim. General statistical context: equivalence requires prespecified margins and an analysis designed to determine whether the uncertainty interval is sufficiently contained within those margins. The registry record does not report equivalence margins, so the reported superiority classification does not establish equivalence.

What is the role of the confidence interval?

The interval around an estimate communicates uncertainty and a range of effect values compatible with the data under the analysis method. For the risk difference, the reported interval is entirely above zero. The log-rank interval is reported on the hazard-ratio scale and is from 0.57 to 0.92. These intervals are method-specific; they should not be treated as interchangeable or as if they measured precisely the same effect.

General statistical context: a confidence interval is not a guarantee that the true effect lies in the displayed range, and it does not account for sources of uncertainty omitted from the analysis. It does help distinguish a relatively precise estimate from a wide one and shows which values are compatible with the data under the model. The registry does not report all model assumptions or analysis details, so the posted intervals should be interpreted as summaries of the stated analyses.

11. Confidence Intervals and Precision

The unadjusted risk difference for all-cause shock is 0.039, with a 95% confidence interval from 0.014 to 0.063. The entire interval is above zero, indicating that the posted unadjusted comparison is compatible with a positive difference for Shock Only minus ATP and Shock under this calculation. The interval communicates uncertainty around the estimate; it does not show that every value in the interval is equally likely or prove a particular clinical interpretation.

The sponsor's log-rank analysis reports a hazard ratio of 0.72 with a 95.9% confidence interval from 0.57 to 0.92. The interval does not include 1, the value representing no relative contrast on the hazard-ratio scale. However, the registry does not state the hazard-ratio reference coding, so the direction should not be assigned without that information. The interval and p-value belong to the sponsor's time-to-event analysis, not to the unadjusted count comparison.

General statistical context: narrower intervals indicate greater statistical precision for the particular estimate and method; wider intervals indicate more uncertainty. Precision is not the same as clinical importance, and an interval cannot resolve limitations such as unreported missing-data handling, analysis choices, or multiplicity. The registry does not report the complete statistical analysis plan or enough detail to assess all assumptions underlying the interval estimates.

The primary count comparison and the time-to-event analysis use different scales and summaries. Their intervals should therefore be read separately rather than compared as if they were two versions of the same quantity. Neither interval alone addresses the registry's equivalency objective, because equivalence margins are not reported in the registry record.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures illustrate how many participants per arm would be needed under stated assumptions to detect selected fractions of the observed effect. The displayed figures are participants per arm, not total events. They use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each planning scenario uses the stated fraction of the observed effect: 100%, 75%, or 50%. These are planning assumptions for the illustration; the trial's own original sample-size assumptions and justification are not reported in the registry record.

The 80% power column describes the number of participants per arm associated with an 80% chance of detecting the chosen effect under the planning assumptions, if that effect and the assumptions held. The 90% power column gives the corresponding planning number for a 90% chance. These are prospective design calculations, not evidence about how much power the completed trial had.

The pattern is that smaller target effects require more participants per arm. For the primary endpoint, the figures range from 1,100 participants per arm for planning around 100% of the observed effect at 80% power to 5,469 participants per arm for planning around 50% of the observed effect at 90% power. This change reflects the general statistical principle that subtle differences are harder to distinguish from random variation than larger differences.

The planning numbers are illustrations based on the observed effect, not observed power. They do not establish that the trial was adequately or inadequately designed, and they do not show the probability that the trial's result is correct. The original design inputs, including any assumptions used by the sponsor when planning the trial, are not reported in the registry record.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.099; treatment proportion p₁ = 0.138; average p̄ = 0.118.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4569 = 0.8954.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4560 = 0.3838.
4. n = (0.8954 + 0.3838)² / (0.0386)² = 1,099.1, rounded up to 1,100 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.0990783410138249                          # control proportion
p1 <- 0.137664346481052                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2923726)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 1,100 (80% power), 1,471 (90% power)
 75% of observed effect: n per arm = 1,886 (80% power), 2,524 (90% power)
 50% of observed effect: n per arm = 4,086 (80% power), 5,469 (90% power)
power.prop.test: 1099.1 
Simulated power at n = 1,100 per arm: 0.800 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor's primary result is a log-rank analysis of time to first all-cause shock. It reports a hazard ratio of 0.72, a 95.9% confidence interval from 0.57 to 0.92, and p = 0.005, and is classified in the registry as a superiority analysis. The reported groups are ATP and Shock and Shock Only. The registry does not specify the reference coding for the hazard ratio, so its direction should be read cautiously.

The recalculation is a simple, unadjusted comparison of the posted participant counts. It reports a risk difference of 0.039 for Shock Only minus ATP and Shock, with a 95% confidence interval from 0.014 to 0.063 and p = 0.002. The recalculation is not a substitute for the sponsor's time-to-event analysis: it summarizes event occurrence from counts, while the sponsor analysis uses event timing and the log-rank method.

The registry marks the discrepancy indicator as false. The estimates are not identical because they describe results using different statistical summaries. The sponsor's reported analysis should be read as the trial's own posted time-to-event result, and the recalculation as a separate descriptive comparison. The full sponsor analysis plan and any further model details are not reported in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary endpoint and lists four secondary outcomes. It reports results for three of the secondary outcomes, but results for all-cause mortality are not reported in the registry record. It does not report a multiplicity-adjustment procedure, a hierarchy for testing outcomes, or how the family of hypotheses was defined. Without those details, the secondary p-values should be treated as unadjusted evidence rather than as results controlled for the number of analyses considered.

The registry specifies the primary endpoint population as all randomized subjects. It also posts arm-level participant counts for the endpoint. However, the record does not describe how missing outcome information was handled, whether any participants were excluded from particular analyses, or how incomplete follow-up was addressed. For the time-to-event analyses, censoring rules and the treatment of withdrawals are not reported in the registry record.

Details needed to assess the analysis population are also limited. The record does not report the exact analysis-set definition beyond the primary population statement, whether the analysis followed assigned treatment regardless of later treatment or programming changes, or how protocol deviations were handled. These details matter because analyses can differ depending on whether they include all randomized participants, only those with available outcomes, or another defined group.

The record describes the trial as randomized and single-masked but does not identify who was masked or provide the randomization procedure. The original statistical analysis plan, prespecified equivalence margins, and sample-size justification are not reported. These omissions do not invalidate the posted results, but they limit independent assessment of whether the analyses followed the intended plan and how missing information or multiple comparisons affected interpretation.

16. 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: multiple heart failure symptoms (13.1%), pulmonary (7.3%), gastrointestinal (6.3%), atrial fibrillation (af) (5.3%), heart failure symptoms - unspecified (4.3%).

The registry reports serious adverse events affecting 633 participants in the ATP and Shock arm and 618 in the Shock Only arm. Other adverse events are reported for 420 participants and 448 participants, respectively. Deaths are listed as 196 and 175, respectively. The registry also lists common serious adverse-event terms, including multiple heart failure symptoms, pulmonary, gastrointestinal, atrial fibrillation, unspecified heart failure symptoms, and cancer, with the associated percentages posted in the record.

These figures are descriptive safety summaries. They do not by themselves show whether an event was caused by the assigned programming strategy, and the registry does not provide an adjusted comparison or a statistical analysis for these safety counts. The record also does not specify in this summary whether participants could appear under more than one event term, the detailed event-collection rules, or the exposure time used for comparisons. The listed groups should therefore be read as reported counts, not as proof of a causal safety difference between strategies.

Deaths are shown as a separate figure from the serious and other adverse-event counts. The registry does not explain how those categories relate to one another, so the figures should not be added together or interpreted as mutually exclusive categories. The posted safety information provides context about events recorded during the trial but cannot establish comparative safety on its own.

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

The posted results address whether all-cause shocks occurred, and when the first shock occurred, under two device-programming strategies in a randomized primary-prevention ICD trial. The participant-count comparison shows a higher recorded risk in the Shock Only arm. The sponsor's time-to-event analysis is reported as a superiority result. These are findings about the endpoint and analyses recorded in the registry, not a universal statement about every outcome or every person receiving an ICD.

The trial's stated objective mentions assessing equivalency. Equivalence is a specific statistical claim: it requires prespecified margins defining how small a difference must be to count as equivalent, along with an analysis designed to test that claim. Those margins are not reported in the registry record. A superiority analysis, even when it reports evidence of a difference, does not establish equivalence. Nor does a non-significant result for a secondary endpoint demonstrate that the arms are equivalent.

The primary endpoint includes both appropriate and inappropriate shocks, so it is a broad measure of shock occurrence. Secondary outcomes separate shock classifications and combine shock with death, but they do not provide a complete account of every possible benefit, harm, or clinical consideration. Results for the listed all-cause mortality outcome are not reported in the registry record.

General statistical context: statistical evidence and clinical importance are related but distinct. A p-value is not a measure of the size or importance of an effect, and a confidence interval does not resolve missing information about the analysis plan or outcome handling. The registry record does not report enough detail to assess all such considerations, including missing-data procedures, censoring rules, multiplicity handling, or equivalence margins.

These results should therefore be interpreted narrowly: they describe the posted primary and secondary endpoint analyses for the randomized programming comparison. They do not supply individualized medical advice, and they do not establish that either programming strategy is appropriate for a particular patient.

18. Why This Trial Matters Statistically

This trial is statistically notable because its stated goal concerns equivalency while its posted primary sponsor analysis is classified as a superiority analysis. Those frameworks answer different questions. A difference may be detected by a superiority test, but equivalence requires margins specified in advance and evidence that the effect is contained within those margins. The margins are not reported in the registry record, so the equivalency objective cannot be evaluated from the posted analysis classification alone.

The record also makes clear why estimands and analysis methods matter. The participant-count comparison describes whether a participant experienced a shock, while the log-rank analysis uses time to first shock. The methods can produce different estimates and p-values because they summarize event information differently. Reporting both requires readers to keep their interpretations distinct rather than treating one as a verification of the other.

The secondary outcomes add context but raise a multiplicity issue. Several outcomes are listed, and results are not reported for one of them. Without a reported testing hierarchy or adjustment, secondary p-values are less definitive than a result from a prespecified primary analysis. The absence of an all-cause mortality result also limits the ability to interpret the full set of outcomes named in the registry.

Finally, the sample-size planning illustrations show how strongly required sample sizes depend on the effect targeted for detection. They are not the trial's original planning rationale and are not observed power. Together, these features illustrate why statistical interpretation depends on the endpoint definition, analysis scale, uncertainty, and reporting completeness—not only on whether a p-value is small.

19. 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.

Several trial-specific reporting gaps limit independent interpretation. The registry describes an equivalency objective but does not report equivalence margins or the full criteria used to assess equivalence. The sponsor result is classified as superiority, and the record does not provide enough information to determine how the stated equivalency objective was operationalized in the analysis.

The sponsor's hazard ratio is reported with the groups named, but the reference coding is not stated. This limits interpretation of the estimate's direction from the registry summary alone. The registry does not report the detailed model specification, censoring rules, or handling of participants whose event status or follow-up was incomplete.

The registry does not report the randomization procedure, who was masked, the original sample-size justification, or the full statistical analysis plan. It also does not describe missing-data handling, analysis-set details beyond the primary population statement, or procedures for protocol deviations. These omissions constrain assessment of the analysis choices and their potential effect on the estimates.

For secondary outcomes, results for all-cause mortality are not reported in the registry record. The record also does not report a multiplicity procedure or testing hierarchy, making the secondary findings less confirmatory. Safety counts are posted, but detailed event-collection rules and adjusted between-arm comparisons are not reported. These limitations concern the detail available in the registry record; they do not independently establish that the trial analyses were incorrect.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Assessment of Primary Prevention Patients Receiving An ICD - Systematic Evaluation of ATP — NCT02923726.
• Registry record last updated 2024-10-22; status last verified 2024-09.
• Results first posted 2024-10-22.
• 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.

23. Record Summary

NCT02923726, APPRAISE ATP, is recorded as a completed randomized, single-masked, parallel-group device trial sponsored by Boston Scientific Corporation. The study compared ATP before shock therapy with shock-only programming in primary-prevention ICD patients. The registry lists enrollment of 2692 and a primary endpoint of all-cause shock over 5 years among all randomized subjects.

The primary participant-count comparison reports a higher risk in the Shock Only arm, while the sponsor's log-rank analysis reports a superiority result for time to first all-cause shock. These are different statistical summaries and should be interpreted separately. The registry also lists four secondary outcomes, with results not reported for all-cause mortality.

Equivalence margins, the full analysis plan, missing-data procedures, and the original sample-size justification are not reported in the registry record. The posted results therefore support interpretation of the analyses described, while leaving important design and analysis details unavailable.