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
SNOCAT was a randomized, parallel-group, quadruple-masked phase 2/3 trial evaluating sodium nitrite in participants with out-of-hospital cardiac arrest. The registered primary endpoint was survival to hospital admission, defined as having a sustained pulse at arrival to the emergency department.
| Feature | SNOCAT |
|---|---|
| Trial name | SNOCAT |
| ClinicalTrials.gov identifier | NCT03452917 |
| Phase | Phase 2/3 |
| Status | Completed |
| Therapeutic area | Cardiovascular |
| Condition | Out-of-hospital cardiac arrest |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Primary purpose | Treatment |
| Enrollment | 1,502 |
| Lead sponsor | University of Washington |
| Start | February 8, 2018 |
| Primary completion | December 31, 2019 |
2. Clinical Question
The central question was whether sodium nitrite administered at 45 mg or 60 mg could improve survival to hospital admission compared with placebo in people experiencing out-of-hospital cardiac arrest.
Population
Participants with out-of-hospital cardiac arrest.
Intervention
Sodium nitrite 45 mg or sodium nitrite 60 mg.
Comparator
Placebo.
Primary question
Does either sodium nitrite dose change the number of participants who survive to hospital admission within 24 hours after out-of-hospital cardiac arrest?
The registry classifies the primary endpoint as binary and defines survival to hospital admission as the patient having a sustained pulse at arrival to the emergency department. That definition makes the primary outcome clinically immediate: it focuses on whether circulation has been restored and sustained through arrival at the emergency department rather than on later hospital survival.
3. Trial Design
Placebo
- Placebo drug
- 496 participants were at risk for the reported serious-adverse-event summary
Sodium nitrite
- Sodium nitrite 45 mg
- Sodium nitrite 60 mg
- 499 participants were at risk for the 45 mg serious-adverse-event summary
- 497 participants were at risk for the 60 mg serious-adverse-event summary
4. Trial Timeline
Trial start
The registry lists February 8, 2018 as the study start date.
Primary completion
The registry lists December 31, 2019 as the primary completion date.
Registry status
The current registry status posted on ClinicalTrials.gov for the study is completed.
5. Endpoints
| Endpoint | Registry definition / time frame | Type |
|---|---|---|
| Number of Participants Who Survived to Hospital Admission | Within 24 hours after out-of-hospital cardiac arrest. Patient has sustained pulse at arrival to the emergency department. | Binary |
| Survival to Discharge | Within 3-6 months after cardiac arrest. | Time-to-event |
| Number of Days in ICU | Within first 3 months of hospital admission. | Continuous |
| Survival to 24 Hours | Within first 24 h after hospital admission. | Time-to-event |
| Re-arrest | Before hospital arrival. | Binary |
| Survival to 48 Hours After Admission to Hospital | Within first 48 hours after admission to hospital. | Time-to-event |
| Survival to 72 Hours After Admission to Hospital | Within first 72 hours after admission to hospital. | Time-to-event |
The primary endpoint is binary because each participant is classified according to whether the specified survival state was achieved. Several secondary outcomes are labeled time-to-event in the registry, although the posted statistical analyses for those outcomes use chi-squared tests and report mean differences in final values rather than hazard ratios.
6. Statistical Methodology
Primary endpoint analysis
The registry reports a chi-squared test for each primary comparison: placebo versus sodium nitrite 45 mg and placebo versus sodium nitrite 60 mg. The reported effect measure is a mean difference in final values.
Placebo vs sodium nitrite 60 mg
Both comparisons address the same primary endpoint, survival to hospital admission within 24 hours after out-of-hospital cardiac arrest, but evaluate the two sodium-nitrite doses separately.
Chi-squared testing
A chi-squared test is a standard method for comparing categorical outcomes between independent groups. For a binary endpoint such as survival to hospital admission, the underlying data can be represented in a two-by-two table: treatment group by survival status.
Treatment a b
Control c d
The chi-squared statistic evaluates whether the observed treatment-group distribution is compatible with the null hypothesis of no association between treatment assignment and the binary outcome.
Wilcoxon / Mann-Whitney testing
The registry reports the Wilcoxon (Mann-Whitney) test for the secondary endpoint Number of Days in ICU. This nonparametric comparison is useful when a continuous outcome may not be adequately represented by a normal-distribution assumption. The registry reports mean difference as the effect measure even though the inferential method is nonparametric.
Confidence intervals
Confidence intervals quantify statistical uncertainty around the reported effect estimates. The two primary analyses use 95% one-sided confidence intervals, whereas the posted secondary analyses use 95% two-sided confidence intervals.
Superiority hypothesis
The posted analyses identify the hypothesis type as superiority. A superiority analysis asks whether the observed treatment-group difference is inconsistent with the null hypothesis of no treatment difference in the specified direction or framework. A non-significant superiority test does not establish equivalence; it indicates that the analysis did not provide sufficient statistical evidence for superiority under the reported test.
7. Primary Results
The registry reports two formal statistical analyses for the primary endpoint, one for each sodium-nitrite dose versus placebo. Both use chi-squared testing and report a mean difference in final values with a 95% one-sided confidence interval.
Survival to Hospital Admission: Sodium Nitrite 45 mg vs Placebo
Reported mean difference
95% one-sided CI: -8.0 · P = 0.82
Endpoint: survival to hospital admission within 24 hours after out-of-hospital cardiac arrest.
| Feature | Placebo vs Sodium Nitrite 45 mg |
|---|---|
| Analysis | Chi-squared |
| Effect measure | Mean Difference (Final Values) |
| Estimate | -2.9 |
| Confidence level | 95% |
| Confidence interval | -8.0, one-sided |
| P-value | 0.82 |
| Hypothesis | Superiority |
The reported estimate of -2.9 is expressed on the registry's reported mean-difference scale for the final values. Its negative direction indicates that the estimated difference favored the placebo group on that reported scale rather than the 45 mg sodium-nitrite group.
The estimate does not mean that sodium nitrite caused a 2.9-unit reduction in an individual's probability of survival. The registry does not provide the underlying two-by-two cell counts needed to translate this posted estimate into an independently calculated absolute risk difference, risk ratio, or odds ratio.
The 95% one-sided confidence interval extends to -8.0. This indicates uncertainty around the estimated difference and does not describe the range of effects that individual patients might experience.
The P = 0.82 value is evidence about compatibility with the null hypothesis under the specified test; it is not a measure of effect size. A p-value of 0.82 should not be interpreted as an 82% probability that the treatment has no effect, nor does it quantify clinical importance.
Survival to Hospital Admission: Sodium Nitrite 60 mg vs Placebo
Reported mean difference
95% one-sided CI: -6.5 · P = 0.66
Endpoint: survival to hospital admission within 24 hours after out-of-hospital cardiac arrest.
| Feature | Placebo vs Sodium Nitrite 60 mg |
|---|---|
| Analysis | Chi-squared |
| Effect measure | Mean Difference (Final Values) |
| Estimate | -1.3 |
| Confidence level | 95% |
| Confidence interval | -6.5, one-sided |
| P-value | 0.66 |
| Hypothesis | Superiority |
The reported estimate of -1.3 is smaller in magnitude than the corresponding 45 mg estimate, but both estimates are reported on the same registry-defined mean-difference scale. The estimate alone should not be interpreted as a patient-level probability or as a relative treatment effect.
The one-sided 95% confidence interval reaches -6.5, showing that the estimated difference is subject to substantial uncertainty. A confidence interval is about uncertainty in the estimated treatment contrast, not about the probability that an individual patient will survive.
The P = 0.66 result does not measure the magnitude of the treatment effect. It describes the evidence against the null hypothesis under the reported chi-squared superiority analysis. It should not be converted into a probability that sodium nitrite is ineffective.
Because two dose-specific primary comparisons are reported, interpretation also requires attention to the trial's overall multiplicity strategy. The registry data presented here do not specify an alpha-allocation or multiplicity-adjustment procedure for these two primary comparisons.
8. Comparing the Two Primary Estimates
| Primary endpoint | Comparison | Estimate | 95% CI | P-value |
|---|---|---|---|---|
| Survival to hospital admission | Placebo vs sodium nitrite 45 mg | -2.9 | -8.0, one-sided | 0.82 |
| Survival to hospital admission | Placebo vs sodium nitrite 60 mg | -1.3 | -6.5, one-sided | 0.66 |
The two dose-specific estimates point in the same reported direction, but neither primary analysis produces a small p-value under the posted superiority test. The difference in estimates between the two doses should not be interpreted as evidence that one dose is superior to the other: the registry reports placebo-versus-dose comparisons, not a direct 45 mg-versus-60 mg statistical comparison.
9. Secondary Endpoint Results
Survival to Discharge
| Comparison | Estimate | 95% CI | P-value | Method |
|---|---|---|---|---|
| Placebo vs Sodium Nitrite 45 mg | -1.7 | -6.0 to 2.6 | 0.44 | Chi-squared |
| Placebo vs Sodium Nitrite 60 mg | -0.4 | -4.9 to 4.0 | 0.85 | Chi-squared |
Survival to discharge was assessed within 3-6 months after cardiac arrest. The 45 mg estimate was -1.7 with a 95% two-sided confidence interval from -6.0 to 2.6 and P = 0.44. The 60 mg estimate was -0.4 with a 95% two-sided confidence interval from -4.9 to 4.0 and P = 0.85.
Both confidence intervals cross zero, so the posted analyses do not provide statistical evidence of a difference from placebo under these comparisons. The intervals also show why a non-significant result should not be equated with proof of identical outcomes: a range of positive and negative differences remains compatible with the reported uncertainty.
Number of Days in ICU
| Comparison | Estimate | 95% CI | P-value | Method |
|---|---|---|---|---|
| Placebo vs Sodium Nitrite 45 mg | 2.2 days | -0.9 to 5.4 | 0.12 | Wilcoxon (Mann-Whitney) |
| Placebo vs Sodium Nitrite 60 mg | 0.5 days | -2.5 to 3.6 | 0.41 | Wilcoxon (Mann-Whitney) |
The 45 mg comparison has a reported mean difference of 2.2 days, with a 95% two-sided confidence interval from -0.9 to 5.4 and P = 0.12. The 60 mg comparison has a reported mean difference of 0.5 days, with a 95% two-sided confidence interval from -2.5 to 3.6 and P = 0.41.
The positive point estimates indicate more ICU days for the sodium-nitrite groups on the reported mean-difference scale, but both confidence intervals include zero. The estimates therefore do not establish a statistically detectable difference in ICU days under the reported analyses. The Wilcoxon/Mann-Whitney test is nonparametric, whereas the registry's effect measure is reported as a mean difference; those are distinct aspects of the analysis and should not be conflated.
Survival to 24 Hours
| Comparison | Estimate | 95% CI | P-value |
|---|---|---|---|
| Placebo vs Sodium Nitrite 45 mg | -2.4 | -8.3 to 3.5 | 0.42 |
| Placebo vs Sodium Nitrite 60 mg | -2.1 | -8.0 to 3.9 | 0.49 |
Both analyses used chi-squared testing. The confidence intervals cross zero in both dose comparisons, and the reported p-values are 0.42 and 0.49.
Re-arrest
| Comparison | Estimate | 95% CI | P-value |
|---|---|---|---|
| Placebo vs Sodium Nitrite 45 mg | 0.4 | -7.8 to 8.6 | 0.93 |
| Placebo vs Sodium Nitrite 60 mg | 5.3 | -2.8 to 13.3 | 0.20 |
Re-arrest was defined for the time frame before hospital arrival. The 45 mg estimate was 0.4, with a 95% two-sided confidence interval from -7.8 to 8.6 and P = 0.93. The 60 mg estimate was 5.3, with a 95% two-sided confidence interval from -2.8 to 13.3 and P = 0.20.
Survival to 48 Hours After Admission to Hospital
| Comparison | Estimate | 95% CI | P-value |
|---|---|---|---|
| Placebo vs Sodium Nitrite 45 mg | -2.0 | -7.7 to 3.8 | 0.50 |
| Placebo vs Sodium Nitrite 60 mg | -2.1 | -7.8 to 3.7 | 0.48 |
Both comparisons used chi-squared testing. The reported confidence intervals cross zero and the p-values are 0.50 and 0.48, respectively.
Survival to 72 Hours After Admission to Hospital
| Registry comparison | Estimate | 95% CI | P-value |
|---|---|---|---|
| Placebo | -1.7 | -6.0 to 2.6 | 0.49 |
| Placebo | -0.4 | -4.9 to 4.0 | 0.85 |
The registry lists two posted chi-squared analyses for survival to 72 hours after admission, with estimates of -1.7 and -0.4. Both use 95% two-sided confidence intervals. The registry comparison field identifies “Placebo” for both analyses rather than identifying a sodium-nitrite dose, so the treatment comparison cannot be more specifically labeled from the posted registry fields.
10. Secondary Results Summary
| Endpoint | Comparison | Estimate | 95% CI | P-value |
|---|---|---|---|---|
| Survival to discharge | Placebo vs 45 mg | -1.7 | -6.0 to 2.6 | 0.44 |
| Survival to discharge | Placebo vs 60 mg | -0.4 | -4.9 to 4.0 | 0.85 |
| Number of days in ICU | Placebo vs 45 mg | 2.2 | -0.9 to 5.4 | 0.12 |
| Number of days in ICU | Placebo vs 60 mg | 0.5 | -2.5 to 3.6 | 0.41 |
| Survival to 24 hours | Placebo vs 45 mg | -2.4 | -8.3 to 3.5 | 0.42 |
| Survival to 24 hours | Placebo vs 60 mg | -2.1 | -8.0 to 3.9 | 0.49 |
| Re-arrest | Placebo vs 45 mg | 0.4 | -7.8 to 8.6 | 0.93 |
| Re-arrest | Placebo vs 60 mg | 5.3 | -2.8 to 13.3 | 0.20 |
| Survival to 48 hours | Placebo vs 45 mg | -2.0 | -7.7 to 3.8 | 0.50 |
| Survival to 48 hours | Placebo vs 60 mg | -2.1 | -7.8 to 3.7 | 0.48 |
| Survival to 72 hours | Registry lists Placebo | -1.7 | -6.0 to 2.6 | 0.49 |
| Survival to 72 hours | Registry lists Placebo | -0.4 | -4.9 to 4.0 | 0.85 |
11. Serious Adverse Events
The registry reports serious adverse events by treatment arm as the number affected divided by the number at risk. These data provide a treatment-group safety summary distinct from the efficacy endpoints.
| Arm | Serious adverse events affected / at risk |
|---|---|
| Placebo | 280 / 496 |
| Sodium Nitrite 45 mg | 249 / 499 |
| Sodium Nitrite 60 mg | 268 / 497 |
The numerator and denominator are the registry's reported values. The corresponding percentages can be obtained directly from these fractions, but the registry's posted safety information is most transparently represented by preserving the affected/at-risk counts.
Serious adverse events are safety outcomes, so their interpretation differs from a superiority test for the primary efficacy endpoint. The affected/at-risk counts describe the observed frequency of serious adverse events in each arm; they do not by themselves establish causality, severity distribution, timing, or statistical significance.
The ClinicalTrials.gov record does not provide a formal comparative statistical analysis for these serious-adverse-event counts. Consequently, they should be interpreted descriptively rather than as evidence of a statistically tested treatment difference.
12. Statistical Methods Explained
Why was a chi-squared test used for the primary endpoint?
The primary endpoint is binary: each participant either survived to hospital admission according to the registered definition or did not. A chi-squared test is designed for comparing categorical outcome distributions between independent groups. Conceptually, it asks whether the observed treatment-group differences are larger than would be expected from random variation under the null hypothesis.
What does the primary estimate of -2.9 mean?
The registry identifies the effect measure as “Mean Difference (Final Values),” so -2.9 is the reported difference on that scale for the placebo versus 45 mg comparison. It should not be relabeled as a risk ratio, odds ratio, or hazard ratio.
Why does the confidence interval matter?
A point estimate gives one estimate of the treatment contrast, whereas the confidence interval shows the uncertainty surrounding that estimate under the specified statistical framework. For example, the primary 45 mg analysis has a 95% one-sided confidence interval extending to -8.0. That interval communicates considerably more information about uncertainty than the point estimate -2.9 alone.
Why does P = 0.82 not mean that the treatment has an 82% chance of being ineffective?
A p-value is calculated under a null hypothesis and measures how unusual the observed data, or data more extreme, would be under that hypothesis. It is not the probability that the null hypothesis is true and it is not a direct measure of treatment efficacy or harm. Effect estimates and confidence intervals provide the complementary information needed to understand magnitude and precision.
Why are the primary confidence intervals one-sided while the secondary intervals are two-sided?
The registry reports 95% one-sided confidence intervals for both primary analyses and 95% two-sided confidence intervals for the posted secondary analyses. One-sided and two-sided intervals answer different inferential questions, so they should not be interchanged when interpreting the reported results.
Why is a Wilcoxon/Mann-Whitney test used for ICU days?
Number of days in ICU is a continuous outcome that may not have a symmetric or approximately normal distribution. The Wilcoxon/Mann-Whitney test provides a rank-based nonparametric comparison between groups and therefore does not require the same distributional assumptions as a conventional two-sample t-test. The registry nevertheless reports the effect measure as a mean difference, which is a separate descriptive quantity from the nonparametric test itself.
13. Time-to-Event Endpoints: An Important Distinction
Several secondary outcomes are classified in the registry as time-to-event endpoints: survival to discharge, survival to 24 hours, survival to 48 hours, and survival to 72 hours. Time-to-event analysis is often associated with Kaplan-Meier estimation, log-rank testing, and hazard ratios because those methods explicitly account for event timing and censoring.
For SNOCAT, however, the posted analyses reported by the registry use chi-squared tests for these survival outcomes and report mean differences in final values. That is materially different from a conventional Kaplan-Meier/Cox analysis.
Time-to-event model → "When did the event occur, accounting for censoring?"
A fixed-time survival endpoint can be analyzed as a binary outcome if the analysis is defined around a particular time point. A full time-to-event analysis instead retains information about the timing of events across follow-up and handles right censoring explicitly.
This distinction is especially important when reading a registry record: an endpoint labeled “time-to-event” does not automatically mean that a Cox proportional-hazards model or Kaplan-Meier estimator was used. The actual posted method determines how the result should be interpreted.
14. Randomization and Masking
Randomization is a central design feature because treatment assignment is determined independently of participants' subsequent outcomes. In principle, this allows baseline prognostic factors to be distributed between groups without requiring every potential confounder to be measured and adjusted for.
The registry describes SNOCAT as quadruple-masked. Masking can reduce the possibility that knowledge of treatment assignment influences behavior, outcome assessment, or other trial processes. Its importance depends on the particular endpoint and how outcomes are ascertained.
Randomization
Creates the basis for a causal treatment comparison by assigning participants to treatment groups through the randomized design.
Quadruple masking
Reduces opportunities for knowledge of assignment to influence participants or trial personnel involved in the study.
The registry does not report a randomization ratio or stratification factors in the information summarized here. Those features therefore should not be inferred from the treatment-group results.
15. Multiplicity and the Two Dose Comparisons
The primary endpoint was evaluated through two posted superiority comparisons: placebo versus sodium nitrite 45 mg and placebo versus sodium nitrite 60 mg. When more than one hypothesis is tested, the probability of obtaining at least one apparently positive result by chance can increase unless the statistical design accounts for multiplicity.
| Question | Registry information |
|---|---|
| Primary endpoint | One registered primary endpoint |
| Primary comparisons posted | 2 |
| 45 mg comparison | Chi-squared; 95% one-sided CI; P = 0.82 |
| 60 mg comparison | Chi-squared; 95% one-sided CI; P = 0.66 |
| Multiplicity adjustment | Not reported in the ClinicalTrials.gov record |
| Alpha-spending / interim strategy | Not reported in the ClinicalTrials.gov record |
Because the registry information does not specify an alpha-allocation or multiplicity procedure, the two p-values should be read as the posted results of the individual analyses rather than assuming an unreported familywise error-control strategy.
16. Analysis Population and Missing Data
The registry record identifies the overall enrollment as 1,502 and provides arm-specific denominators for the serious-adverse-event summary. The statistical-analysis fields do not provide a complete description of the analysis population, missing-data handling, censoring rules, or imputation strategy for each endpoint.
These details matter because an analysis of a binary survival endpoint depends on how participants with incomplete endpoint information are classified. Similarly, continuous ICU-duration analyses can be affected by incomplete follow-up or by how zero days and missing observations are handled. The appropriate interpretation therefore remains tied to the analysis actually posted in the registry.
17. Limitations
- Effect-measure terminology: the primary binary endpoint is reported with “Mean Difference (Final Values)” as the effect measure. The underlying event counts and calculation details are not provided in the statistical-analysis fields summarized here.
- Different analysis conventions: several endpoints are classified as time-to-event, but their posted analyses use chi-squared testing rather than a Kaplan-Meier/Cox framework.
- One-sided primary intervals: the primary analyses report 95% one-sided confidence intervals, which should not be interpreted as though they were two-sided intervals.
- Two primary comparisons: the 45 mg and 60 mg doses were each compared with placebo. The ClinicalTrials.gov record does not specify how multiplicity across those comparisons was controlled.
- No direct dose comparison: the posted primary analyses do not provide a statistical comparison of 45 mg versus 60 mg.
- Incomplete methodological detail: the ClinicalTrials.gov record does not report detailed missing-data procedures, imputation, stratification factors, interim-analysis rules, or Bayesian methods.
- Survival-to-72-hours labeling: the registry identifies “Placebo” in the comparison field for both posted analyses, limiting how specifically those two results can be attributed to a treatment-versus-placebo contrast.
- Safety interpretation: serious-adverse-event counts are available by arm, but a formal comparative safety analysis is not provided in the registry-reported statistical-analysis fields.
18. Why This Trial Matters Statistically
SNOCAT provides a useful example of why endpoint labels and statistical methods should be read together rather than in isolation. The trial includes a binary primary endpoint, several secondary survival outcomes, a continuous ICU-duration outcome, two dose-specific treatment comparisons, and a combination of parametric-looking effect reporting with nonparametric or categorical hypothesis tests.
| Concept | How it appears in SNOCAT |
|---|---|
| Randomization | Participants were randomized in a parallel-group design. |
| Masking | The registry describes the trial as quadruple-masked. |
| Binary endpoint | Survival to hospital admission is the registered primary endpoint. |
| Chi-squared test | Used for both primary analyses and multiple secondary outcomes. |
| Confidence interval | Primary analyses use 95% one-sided intervals; secondary analyses use 95% two-sided intervals. |
| Wilcoxon/Mann-Whitney | Used for Number of Days in ICU. |
| Superiority testing | All posted statistical analyses identify superiority as the hypothesis type. |
| Multiple comparisons | Two sodium-nitrite doses are each compared with placebo for the primary endpoint. |
| Time-to-event terminology | Several secondary endpoints are classified as time-to-event despite posted chi-squared analyses. |
| Safety analysis | Serious adverse events are reported as affected participants divided by those at risk. |
19. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The two posted primary comparisons have estimates of -2.9 and -1.3, with P-values of 0.82 and 0.66, respectively. Under the reported chi-squared superiority analyses, neither comparison provides statistical evidence of superiority.
Clinical interpretation
The registry results do not provide evidence from the posted primary analyses that either sodium-nitrite dose improved the registered survival-to-hospital-admission endpoint. Secondary outcomes should be considered separately because they address different clinical time frames and outcomes.
A statistically non-significant result should not be interpreted as proof that the treatment and placebo are identical. The confidence intervals quantify uncertainty around the reported estimates, and several secondary intervals include both negative and positive values. The distinction between “no statistically detected difference” and “demonstrated equivalence” is fundamental.
20. Statistical Reading of the Confidence Intervals
The reported estimate is -2.9, with a 95% one-sided confidence interval extending to -8.0. The interval communicates uncertainty around the treatment contrast and should be interpreted on the registry's reported mean-difference scale.
The reported estimate is -1.3, with a 95% one-sided confidence interval extending to -6.5. The confidence interval again does not establish a patient-level probability of survival or a relative risk.
The secondary two-sided intervals for survival to discharge, ICU days, survival to 24 hours, re-arrest, and survival to 48 hours include zero. This means that the posted uncertainty intervals encompass no difference on the reported effect scale.
21. What the P-values Do — and Do Not — Mean
The primary p-values are 0.82 for the 45 mg comparison and 0.66 for the 60 mg comparison. These values are generated under the statistical tests specified in the registry.
A p-value does not measure the size of an effect. A small p-value can occur with a modest effect in a sufficiently large study, while a potentially meaningful estimate can fail to achieve a small p-value when uncertainty is substantial. Conversely, a large p-value is not evidence that two treatments are exactly equivalent.
The most informative reading therefore combines the p-value with the effect estimate, confidence interval, endpoint definition, and analysis method. For SNOCAT, this combination indicates that neither posted primary comparison provides statistical evidence of superiority on the registered primary endpoint.
22. What the Registry Does Not Report in the Available Record
The available registry information does not report several design and analysis details that can materially affect statistical interpretation. These include a non-inferiority margin, crossover procedures, a factorial structure, a formal interim-analysis plan, a missing-data or imputation strategy, a Bayesian analysis, and randomization stratification factors.
These omissions are important precisely because they prevent assumptions from being substituted for documented methodology. For example, the absence of a reported non-inferiority margin means the trial should not be described using non-inferiority logic; the posted analyses are identified as superiority analyses.
| Design topic | Registry information |
|---|---|
| Non-inferiority margin | Not reported |
| Crossover | Not reported |
| Factorial design | Not reported; the design model is parallel |
| Interim analysis | Not reported |
| Missing-data / imputation method | Not reported |
| Bayesian methods | Not reported |
| Randomization stratification | Not reported |
23. Related Tutorials
Learn more about the methods used in this trial:
24. Related Calculators
25. Sources
- ClinicalTrials.gov: NCT03452917 — SNOCAT.
- PubMed: PMID 33433575.
Continue through the Clinical Biostats trial library
Connect clinical-trial endpoints with the statistical concepts used to design, analyze, and interpret randomized studies.
26. Record Summary
SNOCAT is a randomized, parallel, quadruple-masked phase 2/3 cardiovascular trial with 1,502 enrolled participants evaluating sodium nitrite in out-of-hospital cardiac arrest. Its registered primary endpoint was survival to hospital admission within 24 hours after out-of-hospital cardiac arrest, defined as a sustained pulse at arrival to the emergency department.
The two posted primary analyses separately compared placebo with sodium nitrite 45 mg and placebo with sodium nitrite 60 mg. The reported estimates were -2.9 and -1.3, respectively, with 95% one-sided confidence intervals extending to -8.0 and -6.5 and p-values of 0.82 and 0.66. Neither analysis provides statistical evidence of superiority under the reported chi-squared tests.
The secondary analyses show the same broader statistical lesson. Survival outcomes were analyzed using chi-squared tests, while ICU days were analyzed using a Wilcoxon/Mann-Whitney test. The reported confidence intervals for the secondary outcomes generally encompass zero, and the p-values do not provide statistical evidence of a treatment difference under the posted analyses.