This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
STOP AF First was a randomized, parallel-group clinical trial evaluating cryoablation versus antiarrhythmic drug therapy in participants with antiarrhythmic-drug-naive paroxysmal atrial fibrillation. The trial enrolled 225 participants and registered two primary binary endpoints assessed from randomization to 12 months.
| Feature | STOP AF First |
|---|---|
| Trial name | STOP AF First |
| Brief title | STOP AF First: Cryoballoon Catheter Ablation in an Antiarrhythmic Drug Naive Paroxysmal Atrial Fibrillation |
| Condition | Atrial Fibrillation; Atrial Fibrillation New Onset |
| Phase | NA |
| Status | COMPLETED |
| Allocation | RANDOMIZED |
| Design model | PARALLEL |
| Masking | NONE |
| Primary purpose | TREATMENT |
| Enrollment | 225.0 |
| Interventions | Cryoablation (device); Antiarrhythmic drug (drug) |
| Lead sponsor | Medtronic Cardiac Ablation Solutions |
| Sponsor type | INDUSTRY |
| Start | 2017-06-23 |
| Primary completion | 2020-06-25 |
| Results posted | Yes |
| Outcome measures posted | 4 |
2. Clinical Question
The central statistical question is whether treatment with cryoablation produces a different rate of treatment success over the period from randomization to 12 months compared with antiarrhythmic drug therapy in the trial population.
Population
Participants with paroxysmal atrial fibrillation represented by the trial's condition fields and described in the brief title as antiarrhythmic-drug naive.
Intervention
Cryoablation using the Arctic Front Advance™ Cardiac CryoAblation Catheter, as specified in the registered treatment-success definition.
Comparator
Antiarrhythmic drug therapy.
Primary question
How does treatment success from randomization to 12 months compare between the randomized treatment strategies, while also evaluating the prespecified primary safety endpoint?
3. Trial Design
Cryoablation
- Intervention type: device
- Pulmonary vein isolation ablation procedure utilizing the Arctic Front Advance™ Cardiac CryoAblation Catheter
- Included in the primary treatment-success definition
Antiarrhythmic drug
- Intervention type: drug
- Antiarrhythmic drug treatment
- Included in the primary treatment-success definition
The design is parallel rather than crossover: participants were randomized to one of two treatment strategies, and the registry description does not identify a crossover component. The trial was also unmasked, so masking is not part of the design's protection against differential knowledge of treatment assignment.
4. Trial Timeline
Trial start
The registry record identifies 2017-06-23 as the trial start date.
Primary completion
The registry record identifies 2020-06-25 as the primary completion date.
Registry status
The ClinicalTrials.gov record lists the study status as COMPLETED and indicates that results were posted.
5. Primary Endpoints
STOP AF First registered two primary endpoints. Both are binary outcomes, and both use a time frame extending from randomization to 12 months.
| Endpoint | Time frame | Type | Registry description available in the ClinicalTrials.gov record |
|---|---|---|---|
| Percentage of Participants With Treatment Success at 12 Months After Antiarrhythmic Drug (AAD) Initiation or Ablation Utilizing the Arctic Front Advance™ Cardiac CryoAblation Catheter. | Randomization to 12 months | Binary | Treatment success after AAD initiation in the control arm or pulmonary vein isolation ablation procedure utilizing the Arctic Front Advance™ Cardiac CryoAblation Catheter in the treatment arm. Treatment success is the opposite of treatment failure. |
| Primary Safety Endpoint - Rate of Composite List of Serious Adverse Events. | Randomization to 12 months | Binary | Measured by rate of a composite list of serious adverse events in the cryoablation-treated as-randomized arm. |
Primary Endpoint 1: Treatment Success
The registry definition posted on ClinicalTrials.gov for this endpoint states that treatment success is the opposite of treatment failure. The registry-reported definition identifies treatment failure as including acute procedural failure in the treatment arm and documented AF/AT/AFL on ambulatory monitoring/12-lead ECG after the 90-day post-ablation blanking period (treatment)/AAD optimization period (control arm). Minimum of 30 seconds on ambulatory monitoring or 10 seconds on 12-lead ECG. * Any subsequent AF surgery or ablation in the left atrium. * Any subsequent cardioversion after the 90-day post-ablation blanking period (treatment)/AAD optimization period (control arm). * Class I or III antiarrhythmic drug (or sotalol) use after the 90-day blanking period (treatment arm only). The registry excerpt ends at that point, so no additional definition is introduced here.
Primary Endpoint 2: Composite Serious Adverse Events
The primary safety endpoint is a rate of a composite list of serious adverse events in the cryoablation-treated as-randomized arm. The registry-reported definition identifies, among its components, TIA within 7 days, cerebrovascular accident within 7 days, major bleeding requiring transfusion or producing a 20% or greater fall in hematocrit within 7 days, and development of a significant pericardial effusion within 30 days. The registry definition continues beyond the excerpt provided in the trial data.
6. Statistical Methodology
The ClinicalTrials.gov record classifies both primary endpoints as binary. That classification determines the natural statistical framework: each participant contributes an outcome category, such as treatment success versus treatment failure or serious adverse event versus no serious adverse event, within the relevant analysis period.
Binary endpoint analysis
For a randomized two-arm trial with a binary primary endpoint, the treatment groups can be summarized using event proportions and compared using an appropriate two-group test. Depending on the prespecified statistical analysis plan and expected cell counts, a Pearson chi-square test or Fisher's exact test may be used for an unadjusted comparison. Effect measures can include an absolute risk difference, risk ratio, or odds ratio, each answering a somewhat different question.
For treatment success, the event is success. For a safety endpoint, the event is occurrence of the defined serious-adverse-event outcome. The denominator must match the analysis population specified for that endpoint.
Why the registry endpoint type matters
The registry explicitly classifies the two primary endpoints as binary rather than time-to-event endpoints. That means a conventional binary-outcome analysis is the natural starting point. It would be inappropriate to substitute a Kaplan-Meier hazard ratio merely because the study has a 12-month follow-up period; the time frame alone does not turn a binary endpoint into a time-to-event endpoint.
Randomization as the basis for comparison
Because the allocation is randomized, the principal comparison can be framed around the difference between the two assigned treatment groups. Randomization is intended to balance measured and unmeasured baseline characteristics in expectation, allowing an observed between-group difference in the prespecified outcome to be interpreted within the randomized-trial framework.
Unmasked design
The trial is registered as having no masking. For a binary clinical outcome, lack of masking can matter particularly when outcome ascertainment or treatment decisions can be influenced by knowledge of assignment. The extent of that concern depends on how the endpoint was measured and adjudicated. The ClinicalTrials.gov record does not provide additional adjudication or masking details, so no stronger conclusion is warranted.
7. Planned Analysis
Results are posted for the trial, but the ClinicalTrials.gov record contains no formal statistical analyses. Therefore, the appropriate statistical interpretation is to distinguish the registry's posted endpoint and safety information from a formal treatment-effect analysis that is not reported here.
What can be established from the ClinicalTrials.gov record
Two primary endpoints were registered, both with a time frame of randomization to 12 months.
The ClinicalTrials.gov record does not provide a formal statistical-analysis record containing an effect estimate, confidence interval, or P-value for either primary endpoint.
Treatment-success endpoint
The registry states that treatment success was evaluated after AAD initiation in the control arm or after pulmonary vein isolation ablation using the Arctic Front Advance™ Cardiac CryoAblation Catheter in the treatment arm. Because the endpoint is binary, a typical comparative analysis would estimate the treatment-success proportion in each randomized group and then quantify the between-group difference using an appropriate binary effect measure.
A risk difference would answer the question, "How many additional treatment successes per unit of population were observed in one group compared with the other?" A risk ratio would answer the relative question, "How large was the treatment-success probability in one group relative to the other?" An odds ratio would instead compare the odds of success.
Primary safety endpoint
The primary safety endpoint is a composite serious-adverse-event rate. For a binary composite, the primary comparison would ordinarily summarize the proportion of participants meeting the composite definition in each relevant group and compare those proportions using a prespecified binary-outcome method.
The ClinicalTrials.gov record specifically identifies the cryoablation-treated as-randomized arm in the endpoint definition. That wording matters: the formal analysis population and denominator should be taken from the prespecified statistical analysis plan or registry results rather than reconstructed from unrelated counts.
8. Reported Safety Information
The ClinicalTrials.gov record reports serious adverse events by arm using affected participants over the stated at-risk population.
| Arm | Affected / at risk | Interpretation |
|---|---|---|
| Randomized and Treated With Cryoablation | 16/104 | 16 participants were reported as affected among 104 at risk. |
| Randomized and Treated With Antiarrhythm | 14/99 | 14 participants were reported as affected among 99 at risk. |
These counts are descriptive safety information. They should not be converted into a comparative safety claim without specifying the analysis population, denominator definition, statistical method, and uncertainty measure. The ClinicalTrials.gov record does not provide a formal confidence interval or P-value for this comparison.
Cryoablation arm
The registry-reported safety count is 16 affected participants among 104 at risk.
Antiarrhythmic arm
The registry-reported safety count is 14 affected participants among 99 at risk.
The primary safety endpoint is also a composite. Composite endpoints require careful interpretation because the individual components can differ in clinical frequency, severity, and relevance. The registry's registry-reported definition therefore matters more than a generic statement about "serious adverse events."
9. Statistical Methods Explained
Why is randomization important in this trial?
Randomization creates the comparison structure for the trial. Participants are assigned to the treatment strategies by the randomized design rather than by an investigator's choice. In expectation, this balances both measured and unmeasured prognostic factors across groups. The statistical analysis can therefore focus on the contrast between randomized groups rather than attempting to recreate comparability entirely through covariate adjustment.
Why is treatment success analysed as a binary endpoint?
The registry classifies the treatment-success endpoint as binary. Each participant ultimately contributes to one of two outcome categories for the defined assessment: treatment success or treatment failure. The appropriate summary is consequently a proportion or another effect measure based on those two categories rather than a continuous mean or a survival-time hazard ratio.
What would a risk difference tell us?
A risk difference compares the event probability directly between two groups:
For treatment success, a positive risk difference would indicate a higher observed success proportion in the cryoablation group under that coding. For a safety event, the same mathematical quantity would represent an absolute difference in event probability.
The advantage of a risk difference is that it remains on an absolute probability scale. Its interpretation depends on which outcome is coded as the event.
What would a risk ratio tell us?
A risk ratio compares two probabilities multiplicatively:
A risk ratio of 1 would indicate equal observed event probabilities. Values above or below 1 describe relative differences in the probability of the specified event.
For a treatment-success endpoint, the event definition is success. For a safety endpoint, the event definition is the specified serious-adverse-event composite. The direction therefore depends on the endpoint being analysed.
Why is an odds ratio different from a risk ratio?
An odds ratio compares odds rather than probabilities. If the event is relatively uncommon, the odds ratio and risk ratio may be numerically similar, but they are not the same estimand. When the event is more common, the distinction becomes more important. A statistical analysis should therefore name the effect measure rather than simply describing an "effect size."
Why does the 12-month time frame not automatically require a survival analysis?
The registry defines the primary endpoints as binary and gives the assessment period as randomization to 12 months. A fixed-time binary endpoint asks whether the participant met the specified endpoint during the defined period. A time-to-event endpoint would instead use the timing of the event and censoring information. Without a registered time-to-event estimand, a hazard ratio should not be assumed merely because follow-up extends over time.
Why should safety counts not be treated as an efficacy result?
The treatment-success endpoint and the serious-adverse-event endpoint answer different questions. One concerns the registered definition of treatment success; the other concerns a composite of serious adverse events. Even when both are binary, their clinical interpretation, direction of benefit or harm, and appropriate summary should remain separate.
10. Interpreting the Safety Denominators
The reported safety figures are expressed as affected participants over participants at risk: 16/104 for the randomized-and-treated cryoablation group and 14/99 for the randomized-and-treated antiarrhythmic group.
The denominator is part of the statistical result. Reporting only the affected count would conceal how many participants were considered at risk. Conversely, using the total enrollment of 225 as the denominator would not reproduce the registry-reported safety definition.
16/104 and 14/99 are descriptive counts. An inferential comparison would additionally require a defined estimand, analysis population, statistical test or model, and uncertainty measure. None of those formal comparison results is reported in the trial's the posted statistical analyses.
A composite serious-adverse-event endpoint combines specified event types. A single composite rate can therefore conceal differences in the frequency or nature of its individual components. The component definitions should be retained when interpreting the overall safety endpoint.
11. What the Registry Data Supports — and Does Not Support
| Question | What the ClinicalTrials.gov record supports |
|---|---|
| Was the trial randomized? | Yes. Allocation is listed as RANDOMIZED. |
| Was it parallel-group? | Yes. The design model is PARALLEL. |
| Was it masked? | No. Masking is listed as NONE. |
| How many participants were enrolled? | 225.0. |
| How many primary endpoints were registered? | 2. |
| What type were the primary endpoints? | Binary. |
| What was the primary assessment period? | Randomization to 12 months. |
| Were results posted? | Yes. |
| Are formal statistical-analysis estimates reported in the ClinicalTrials.gov record? | No. No formal statistical analyses were posted. |
| Are serious-adverse-event counts reported? | Yes: 16/104 for randomized and treated with cryoablation and 14/99 for randomized and treated with antiarrhythm. |
12. Design Features That Matter Statistically
Randomization
Random allocation provides the foundation for the between-group comparison and reduces systematic treatment-assignment bias in expectation.
Parallel design
Each randomized participant belongs to one treatment pathway rather than a crossover sequence, so the principal comparison is between parallel randomized groups.
No masking
The registry lists masking as NONE. Knowledge of treatment assignment can be relevant when outcomes involve assessment or treatment decisions.
Binary endpoints
Both primary endpoints are classified as binary, directing the analysis toward proportions and two-group binary-outcome comparisons.
The ClinicalTrials.gov record does not report a factorial design, crossover, non-inferiority margin, interim analysis, Bayesian method, multiplicity procedure, missing-data imputation method, or stratification scheme. Those design elements are therefore not incorporated into the statistical interpretation of this page.
13. Independent Statistical Interpretation
The most important statistical feature of STOP AF First in the ClinicalTrials.gov record is the alignment between the randomized parallel design and the binary primary endpoints. The study asks a fixed-period treatment-success question and a fixed-period composite safety question. That structure is substantially different from a trial whose primary estimand is a hazard ratio over continuous follow-up.
For treatment success, the key quantity is the probability that a participant satisfies the registry's definition of success during the specified 12-month period. The comparison can be expressed on an absolute scale, such as a risk difference, or a relative scale, such as a risk ratio or odds ratio. Those estimands should not be conflated: an apparently large relative difference can coexist with a small absolute difference when the underlying event is uncommon, and the reverse can also occur.
The safety endpoint requires an additional layer of caution because it is composite. A composite can improve statistical efficiency by combining events, but the resulting overall rate is meaningful only in the context of the exact components included in the definition. Here, the registry definition includes several serious events with different clinical characteristics and different time windows.
The absence of formal statistical-analysis entries in the ClinicalTrials.gov record is also important. A page presenting an estimated treatment effect, confidence interval, or P-value would necessarily be adding information that is not contained in the ClinicalTrials.gov record.
14. Limitations
- No formal analysis estimates reported: no formal statistical analyses were posted, so no treatment-effect estimate, confidence interval, or P-value is reported on this page.
- Posted outcomes are not the same as a the ClinicalTrials.gov record: the trial is marked as having results posted, but in the registry record does not contain the corresponding formal statistical comparisons.
- Binary endpoint structure: a fixed-period binary analysis does not describe when an event occurred within the period unless timing is separately incorporated into the estimand.
- Composite safety endpoint: the overall serious-adverse-event rate can combine clinically different event types, so the composite should not automatically be interpreted as a uniform clinical outcome.
- Unmasked design: the registry lists masking as NONE. The potential influence of treatment knowledge depends on how outcomes were measured and assessed.
- Denominator specificity: safety counts use the registry-reported randomized-and-treated populations of 104 and 99, rather than the overall enrollment of 225.0.
- Incomplete registry-reported endpoint definition: the provided registry definition for treatment failure ends within the registry-reported excerpt, so this page does not add criteria beyond the text reported in the ClinicalTrials.gov record.
- No unsupported design assumptions: the ClinicalTrials.gov record does not identify crossover, factorial allocation, non-inferiority margins, interim analyses, Bayesian methods, or missing-data procedures, so these are not used to interpret the trial.
15. Why This Trial Matters Statistically
STOP AF First provides a useful teaching example of how statistical interpretation begins with the estimand and endpoint structure rather than with a favorite statistical test. The trial is randomized and parallel, but its two primary endpoints are binary. That combination makes it a practical case for learning how treatment effects are defined, how denominators are chosen, and how descriptive event counts differ from formal inference.
| Concept | How it appears in STOP AF First |
|---|---|
| Randomization | The allocation is registered as RANDOMIZED. |
| Parallel-group design | The design model is PARALLEL. |
| Binary outcomes | Both registered primary endpoints are classified as binary. |
| Fixed follow-up period | Both primary endpoints use randomization to 12 months. |
| Risk estimation | Binary endpoints can be summarized using event proportions and absolute or relative effect measures. |
| Composite endpoint | The primary safety endpoint combines specified serious adverse events. |
| Analysis population | The ClinicalTrials.gov record explicitly identifies randomized-and-treated populations and their at-risk denominators. |
| Unmasked trial | Masking is listed as NONE, making knowledge of assignment part of the design context. |
16. Related Tutorials
Learn more about the statistical concepts most relevant to this trial:
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: STOP AF First, NCT03118518.
- PubMed: PMID 38007219.
- PubMed: PMID 33197158.
Continue through Clinical Biostats
Explore tutorials and statistical calculators that build the concepts used to interpret randomized clinical trials and binary endpoints.
19. Record Summary
STOP AF First is a randomized, parallel-group clinical trial with 225.0 enrolled participants, two treatment arms, no masking, and two registered binary primary endpoints assessed from randomization to 12 months. The intervention was cryoablation and the comparator was antiarrhythmic drug therapy. The primary efficacy endpoint concerns treatment success, while the primary safety endpoint concerns a composite list of serious adverse events.
The ClinicalTrials.gov record includes serious-adverse-event counts of 16/104 among participants randomized and treated with cryoablation and 14/99 among participants randomized and treated with antiarrhythmic therapy. However, no formal statistical analyses were posted. The appropriate statistical interpretation is therefore to explain how the binary outcomes would normally be compared without creating treatment-effect estimates or inferential statistics that are not present in the ClinicalTrials.gov record.