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
SHOCK is a phase 3 randomized, multicenter clinical trial evaluating extracorporal cytokine removal in patients with septic shock. The registry describes a three-arm, parallel-group design comparing standard medical therapy with two Cytosorb treatment schedules.
| Feature | SHOCK |
|---|---|
| Trial acronym | SHOCK |
| ClinicalTrials.gov identifier | NCT04742764 |
| Title | Extracorporal Cytokin Removal in Septic Shock: a Prospective, Randomized, Multicenter Clinical Trial |
| Phase | Phase 3 |
| Status | Suspended |
| Condition | Septic Shock |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Primary purpose | Treatment |
| Enrollment | 135 |
| Number of arms | 3 |
| Lead sponsor | University of Pecs |
| Start date | 2024-01-01 |
| Primary completion date | 2026-10-31 |
2. Clinical Question
The clinical question is whether adding extracorporal cytokine removal treatment using Cytosorb to standard medical therapy is associated with shock reversal in patients with septic shock, and whether the timing of shock reversal differs according to the cytokine-removal schedule.
Population
Patients with septic shock enrolled in the prospective, randomized, multicenter SHOCK trial.
Intervention
Standard medical therapy plus cytokine removal treatment using Cytosorb, with the adsorber changed every 12 hours.
Comparator
Standard medical therapy alone. A second randomized comparison used standard medical therapy plus Cytosorb with the adsorber changed every 24 hours.
Primary question
How does cytokine-removal treatment relate to the proportion achieving shock reversal and the time from treatment initiation to shock reversal?
3. Trial Design
Standard medical therapy
- Standard medical therapy
- Registry intervention type: combination product
Cytosorb every 12 hours
- Standard medical therapy
- Cytokine removal treatment using Cytosorb
- Adsorber changed every 12 hours
Cytosorb every 24 hours
- Standard medical therapy
- Cytokine removal treatment using Cytosorb
- Adsorber changed every 24 hours
The registry characterizes the allocation as randomized and the design model as parallel. This is important statistically because each participant is assigned to one of the three treatment strategies rather than being deliberately exposed to each strategy in sequence.
4. Trial Population, Intervention, and Comparator Structure
The three-arm structure creates two distinct cytokine-removal schedules alongside the standard-medical-therapy arm. Statistically, this allows the study to address both the broader question of cytokine removal versus standard therapy and the more specific question of whether the 12-hour and 24-hour adsorber-change schedules differ in their association with shock reversal.
| Arm | Standard medical therapy | Cytosorb | Adsorber schedule |
|---|---|---|---|
| Standard therapy | Yes | No | Not applicable |
| Cytokine removal, 12-hour schedule | Yes | Yes | Changed every 12 hours |
| Cytokine removal, 24-hour schedule | Yes | Yes | Changed every 24 hours |
Because the registry specifies three arms but does not provide arm-specific enrollment counts here, the statistical interpretation should not assume equal allocation. A three-arm randomized design can be balanced or unbalanced, and the precision of each pairwise comparison depends on the number of participants contributing to that comparison.
5. Primary Endpoints
| Endpoint | Registry time frame | Registry definition |
|---|---|---|
| Shock reversal | At the time of shock reversal assessed up to 5 days | Proportion of patients achieving shock reversal, defined as follows: no need (or minimal need, meaning max. the 10% of the maximum dose) of vasopressore for 3 hours, with haemodynamic measurements, and arterial, central venous blood gas analysis, arterial lactate level measurement, venous and arterial pCO2-gap and O2 saturation measurements to confirm cardiorespiratory stability. |
| Time to shock reversal | From the start of the treatment until shock reversal assessed up to 5 days | The time from the start of the treatment (T0) until shock reversal. |
Why the two endpoints complement each other
The first endpoint is a binary outcome: whether a patient achieves the registry-defined state of shock reversal during the specified assessment period. The second endpoint uses the timing of that same clinical event.
This distinction is statistically important. A proportion summarizes the occurrence of an event within a defined window, whereas a time-to-event endpoint preserves information about when the event occurred. Two treatment groups could have similar proportions achieving shock reversal but different timing distributions, or similar timing among responders but different proportions ultimately reaching the endpoint.
For a binary endpoint, the treatment comparison concerns the probability of achieving the specified clinical state during the defined assessment period. The appropriate effect measure depends on the prespecified analysis plan and can include an absolute difference, risk ratio, odds ratio, or another model-based contrast.
For time to shock reversal, the event time is anchored to treatment initiation. Patients who do not experience shock reversal during the observation window create a right-censoring or competing-event issue that must be handled according to the prespecified statistical analysis plan.
6. Clinical Definition of Shock Reversal
The registry's definition is more specific than a simple improvement in blood pressure. Shock reversal requires no need, or minimal need, for vasopressors for 3 hours, with the maximum minimal requirement defined in the registry as 10% of the maximum dose. The registry also specifies a set of physiologic measurements used to confirm cardiorespiratory stability.
Vasopressor requirement
No need, or minimal need defined as a maximum of 10% of the maximum dose, for 3 hours.
Hemodynamic assessment
Hemodynamic measurements are part of the confirmation of cardiorespiratory stability.
Blood-gas assessment
Arterial and central venous blood gas analysis are specified.
Additional physiologic measures
Arterial lactate, venous and arterial pCO2-gap, and O2 saturation measurements are specified.
From a statistical perspective, a carefully defined composite clinical state can improve consistency of endpoint ascertainment because the event is tied to explicit criteria. At the same time, the more components required for an event, the more important it becomes that each component is measured consistently across treatment groups and trial centers.
7. Planned Analysis
The registry currently contains the two registered primary endpoints but no posted statistical analyses. The endpoints therefore define what is intended to be measured, while the numerical treatment effects and formal hypothesis-test results are not reported on ClinicalTrials.gov.
Analysis of shock reversal
Shock reversal is a binary endpoint assessed within a maximum of 5 days. A conventional analysis for this type of endpoint would compare the proportion achieving reversal between randomized groups and report an effect measure with an appropriate confidence interval. With three randomized arms, the analysis could involve prespecified pairwise comparisons or a model containing all three treatment groups, depending on the statistical analysis plan.
The key statistical quantity is the between-group contrast in the probability of shock reversal. For example, an absolute risk difference can be written as:
An absolute difference is expressed in probability units. It answers how much the observed probability of shock reversal differs between two randomized groups during the specified assessment period.
Analysis of time to shock reversal
Time to shock reversal is a time-to-event endpoint beginning at T0, the start of treatment, and assessed through 5 days. A conventional analysis for such an endpoint would use a survival-analysis framework, such as Kaplan-Meier estimation and a time-to-event comparison, provided that the event and censoring rules are prespecified appropriately.
A Cox proportional-hazards model is another common approach for estimating a hazard ratio for a time-to-event endpoint. If used, its proportional-hazards assumption would need to be considered rather than treated as automatically valid.
For a time-to-shock-reversal analysis, a hazard ratio above 1 would indicate a higher estimated instantaneous event rate in one group relative to another, under the fitted model. It would not directly equal a difference in the proportion achieving reversal by 5 days.
Three-arm comparisons
A three-arm randomized trial creates more than one possible treatment contrast. The standard-therapy arm provides the natural reference for evaluating each Cytosorb schedule, while the two Cytosorb arms permit a direct comparison of the 12-hour and 24-hour schedules.
| Potential contrast | Statistical question |
|---|---|
| 12-hour Cytosorb vs standard therapy | Does the 12-hour adsorber-change strategy differ from standard medical therapy for shock reversal? |
| 24-hour Cytosorb vs standard therapy | Does the 24-hour adsorber-change strategy differ from standard medical therapy for shock reversal? |
| 12-hour Cytosorb vs 24-hour Cytosorb | Does the adsorber-change schedule correspond to a difference in shock reversal? |
These contrasts are not automatically equivalent to three independent confirmatory hypotheses. If multiple comparisons are formally tested, the statistical analysis plan should specify how multiplicity is handled and which comparisons constitute the primary confirmatory questions.
8. Statistical Methodology
Binary outcome analysis
The shock-reversal endpoint is defined as a proportion. The basic data structure is therefore a count of patients achieving reversal and a count not achieving reversal in each randomized group.
A simple comparison can be expressed as a two-by-two table for any selected pair of treatment groups. For three arms, the full analysis can instead be represented as a treatment-group-by-outcome table and analyzed with a prespecified regression or hypothesis-testing framework.
Here, Y=1 represents achievement of shock reversal. A treatment coefficient can be translated into an odds ratio, while other effect measures such as risk differences or risk ratios may be more directly interpretable clinically.
Time-to-event analysis
Time to shock reversal differs from simply recording whether reversal occurred by day 5 because it uses the timing of the event. Kaplan-Meier methods are commonly used to describe a time-to-event distribution when right censoring is present.
The Kaplan-Meier estimator updates the estimated probability of remaining event-free whenever an event occurs. For shock reversal, the interpretation depends on whether the analysis is framed around time to reversal or time without reversal.
Because the registry defines the event as shock reversal rather than death or disease progression, the statistical treatment of patients who die or otherwise become unable to experience the endpoint is particularly important. Such observations may not be equivalent to ordinary noninformative censoring. The exact handling should therefore be determined by the prespecified analysis plan rather than assumed.
Cox proportional-hazards model
If a Cox model is selected for time to shock reversal, the hazard ratio summarizes the relative instantaneous rate of achieving reversal between treatment groups. The model does not directly estimate the difference in the probability of reversal at 5 days.
A hazard ratio is a relative time-to-event measure. It should be interpreted together with an absolute measure, such as the probability of shock reversal by 5 days, when both are available. The hazard ratio also depends on model assumptions, including the proportional-hazards assumption when that model is used.
Confidence intervals
A confidence interval quantifies statistical uncertainty around an estimated treatment effect under the specified statistical model and sampling framework. A narrow interval indicates greater statistical precision than a wide interval, all else being equal.
A confidence interval should not be interpreted as the range of effects that individual patients will experience. It describes uncertainty about the estimated population-level treatment contrast.
P-values
For a prespecified hypothesis test, a p-value measures the compatibility of the observed data with the null hypothesis under the assumptions of the test. It is not a measure of the magnitude of the treatment effect, the clinical importance of an effect, or the probability that the null hypothesis is true.
9. Statistical Methods Explained
Why are there two primary endpoints?
The two endpoints capture different dimensions of shock reversal. The first asks whether the registry-defined state is achieved during the 5-day assessment period. The second asks how long it takes to reach that state after treatment begins. A proportion can summarize event occurrence, while a time-to-event analysis can retain information about the timing of that occurrence.
Why is time to shock reversal a time-to-event endpoint?
Because the event has a measurable occurrence time beginning at T0. Patients can reach shock reversal at different times, so collapsing all observations into a simple yes/no outcome can discard information about when reversal occurred.
Why does the 5-day window matter?
The registered shock-reversal endpoint is assessed up to 5 days. This creates a common observation horizon for the primary endpoint and prevents an event occurring much later from being treated as equivalent to an event within the prespecified assessment period.
Why is the 3-hour criterion important?
The registry does not define reversal merely as a transient reduction in vasopressor requirement. The patient must have no need, or only minimal need, for vasopressors for 3 hours, together with specified measurements to confirm cardiorespiratory stability. Statistically, this makes endpoint ascertainment depend on a defined clinical state rather than a single measurement.
Why does the three-arm design require attention to multiplicity?
With three treatment groups, several pairwise comparisons are possible. If all are tested formally, the chance of at least one false-positive conclusion can increase unless the testing strategy accounts for multiplicity. A prespecified hierarchy or multiplicity-adjustment procedure can control the relevant type I error rate.
Why might Kaplan-Meier analysis be useful?
Kaplan-Meier methods can describe the distribution of time until an event while accounting for right-censored observations. For this trial, such an approach is relevant to the time-to-shock-reversal endpoint, although the appropriate treatment of patients who die or otherwise cannot experience reversal requires explicit prespecification.
Why does the definition of shock reversal matter statistically?
Endpoint definitions determine which observations are classified as events. Requiring vasopressor criteria, a 3-hour duration, and confirmation of cardiorespiratory stability can reduce ambiguity, but it also means that consistent measurement of every component is necessary for comparable endpoint ascertainment across treatment groups.
10. Missing Data and Censoring Considerations
The registry defines shock reversal as a clinical event assessed over a maximum of 5 days, but it does not report a missing-data or censoring strategy in the information summarized here.
For the binary endpoint, missing outcome information can affect the denominator and the estimated treatment proportions. The statistical analysis plan would normally specify how participants without an evaluable endpoint are handled and whether sensitivity analyses are performed.
For time to shock reversal, censoring requires particular attention. A patient who has not achieved reversal by the end of the assessment window is not automatically equivalent, from a time-to-event perspective, to a patient who is lost to follow-up earlier. Similarly, an event such as death can prevent shock reversal from occurring and may require competing-risk methods rather than ordinary censoring, depending on the estimand.
11. Randomization and Causal Interpretation
Randomization is central to the causal interpretation of a treatment comparison because treatment assignment is determined by the trial design rather than by clinical characteristics chosen by investigators after enrollment.
In a parallel randomized trial, the treatment-group comparison can therefore be interpreted as a comparison generated by the randomized assignment, subject to adherence, missing data, post-randomization events, and the precise estimand used for analysis.
The presence of three arms is especially useful analytically because it separates the question of whether cytokine removal differs from standard therapy from the question of whether two Cytosorb schedules differ from one another.
12. Multiplicity and Three-Arm Testing
The registry identifies two primary endpoints and three treatment arms. That combination creates a statistical design in which the definition of the confirmatory testing family is important.
| Dimension | What must be defined statistically |
|---|---|
| Endpoints | How shock reversal and time to shock reversal contribute to the confirmatory question. |
| Treatment comparisons | Which of the three-arm contrasts are formally tested. |
| Type I error | How the overall false-positive rate is controlled when more than one hypothesis is tested. |
| Analysis sequence | Whether a hierarchical or other prespecified testing strategy is used. |
| Effect measures | Which treatment contrast and confidence interval are used for each endpoint. |
The registry information does not specify an alpha-allocation or multiplicity-adjustment procedure. Consequently, statistical interpretation should distinguish the existence of multiple potential comparisons from any particular multiplicity strategy.
13. Interim Analysis, Bayesian Methods, and Crossover
The registry information summarized here does not report an interim efficacy analysis, an alpha-spending procedure, Bayesian methods, or a crossover design. The trial is registered as randomized and parallel rather than as a crossover study.
Interim analysis
No interim statistical analysis is reported in the registry information summarized here.
Bayesian methods
No Bayesian statistical method is reported in the registry information summarized here.
Crossover
The registry identifies a parallel design; no crossover structure is reported.
Factorial design
The registry identifies a three-arm parallel design rather than a factorial design.
14. Trial Timeline
Trial start
The registry lists January 1, 2024 as the study start date.
Randomized multicenter study
The study is registered as a phase 3, randomized, parallel clinical trial with three treatment arms and planned enrollment of 135.
Primary completion date
The registry lists October 31, 2026 as the primary completion date.
Suspended
The ClinicalTrials.gov record identifies the study status as suspended.
15. What a Treatment Effect Would Mean
A treatment effect on shock reversal would describe a difference in the probability of achieving the registry-defined clinical state within the specified 5-day assessment period. It would not by itself describe how quickly patients who achieved reversal reached that state.
A treatment effect on time to shock reversal would describe a difference in the timing of the event from treatment initiation. If expressed as a hazard ratio, the estimate would be relative and model-based; it would not directly equal an absolute increase in the proportion achieving reversal.
A binary endpoint and a time-to-event endpoint can provide complementary information. The first focuses on whether the clinical state is reached during the specified window; the second focuses on when it is reached. Interpretation is strongest when the two are consistent and their estimands are clearly distinguished.
16. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The trial has two registered primary endpoints: a binary shock-reversal endpoint assessed up to 5 days and a time-to-shock-reversal endpoint measured from treatment initiation through 5 days. Formal numerical analyses are not posted in the registry.
Clinical interpretation
The trial is designed to examine whether extracorporal cytokine removal using Cytosorb, with either a 12-hour or 24-hour adsorber-change schedule, is associated with the specified clinical state of shock reversal in septic shock.
17. Important Limitations and Interpretation Issues
- No posted numerical analyses: the ClinicalTrials.gov record does not report estimates, confidence intervals, or p-values for the two registered primary endpoints.
- Three-arm structure: multiple clinically relevant treatment contrasts are possible, so the prespecified comparison hierarchy matters for confirmatory interpretation.
- Two primary endpoints: shock reversal and time to shock reversal address related but distinct questions and require clear definition of the confirmatory testing framework.
- Endpoint complexity: shock reversal depends on vasopressor requirements, a 3-hour criterion, hemodynamic measurements, blood-gas measurements, arterial lactate, pCO2-gap measurements, and O2 saturation measurements.
- Time-to-event assumptions: if a Cox model is used, proportional hazards would need to be assessed as an assumption of the model rather than presumed automatically.
- Censoring and competing events: patients who do not reach shock reversal by 5 days and patients who become unable to experience the event may require different statistical handling.
- Arm-specific enrollment: the registry information summarized here gives total enrollment of 135 but does not provide arm-specific enrollment counts.
- Masking details: the registry identifies the study as single-masked but does not specify the masked party in the information summarized here.
- Generalizability: applicability outside the enrolled septic-shock population depends on the eligibility criteria and clinical setting represented by the multicenter trial.
18. Why This Trial Matters Statistically
SHOCK is a useful statistical teaching case because it combines a three-arm randomized design with both binary and time-to-event primary endpoints. The same underlying clinical concept—shock reversal—is represented in two different statistical forms: whether reversal occurs during a defined period and the time required to reach reversal.
| Concept | How it appears in SHOCK |
|---|---|
| Randomization | Patients are allocated randomly among three treatment arms. |
| Parallel-group design | The registry identifies the design model as parallel. |
| Three-arm comparison | Standard medical therapy is compared with two Cytosorb schedules. |
| Binary endpoint | Shock reversal is defined as a proportion achieving the specified clinical state. |
| Time-to-event endpoint | Time to shock reversal is measured from treatment start until reversal. |
| Clinical endpoint definition | Reversal requires specified vasopressor, duration, hemodynamic, blood-gas, lactate, pCO2-gap, and O2 saturation criteria. |
| Censoring | The time-to-event endpoint is assessed through a maximum of 5 days, making event and censoring rules important. |
| Multiplicity | Two primary endpoints and three arms create multiple possible statistical comparisons. |
| Hazard ratio | A conventional potential measure for a time-to-event comparison, if specified by the analysis plan. |
| Confidence interval | Would quantify uncertainty around a treatment-effect estimate when results are available. |
| p-value | Would address compatibility with a prespecified null hypothesis rather than effect magnitude. |
| Estimand | Clarifies whether the question concerns probability of reversal, timing of reversal, or another treatment-effect definition. |
19. The Difference Between Risk and Time-to-Event Measures
One of the central statistical lessons in SHOCK is that a proportion and a time-to-event measure are not interchangeable.
| Measure | Primary question | Interpretive focus |
|---|---|---|
| Shock reversal proportion | Did the patient achieve reversal during the specified 5-day period? | Absolute or relative difference in event probability |
| Time to shock reversal | When did reversal occur after treatment started? | Distribution of event times and relative event rate |
Suppose two randomized groups ultimately had the same proportion of patients achieving reversal, but one group reached reversal substantially earlier. A binary endpoint alone could fail to capture that distinction. Conversely, a time-to-event model can summarize timing but may be less immediately interpretable than an absolute 5-day probability.
This is why reporting both endpoints can be statistically informative, provided that the analysis population, event definition, censoring rules, and treatment contrasts are consistent across the two analyses.
20. Why Endpoint Definition and Measurement Are Statistically Important
The SHOCK shock-reversal definition illustrates an important principle in clinical-trial statistics: statistical analysis begins with endpoint construction. The outcome is not simply "shock improved." It is a prespecified clinical state requiring several criteria to be satisfied.
Threshold
The registry specifies a maximum of 10% of the maximum dose for the minimal vasopressor requirement.
Duration
The vasopressor criterion must be maintained for 3 hours.
Physiologic confirmation
Hemodynamic and multiple blood-gas and oxygenation measurements are part of confirmation.
Observation window
The primary shock-reversal endpoint is assessed up to 5 days.
Every component can affect classification. If a patient meets the vasopressor criterion but does not satisfy the specified physiologic confirmation requirements, the statistical classification of the endpoint may differ from a simpler definition of clinical improvement. This makes endpoint ascertainment procedures an important component of trial validity.
21. Understanding Three Treatment Arms
A three-arm trial can answer a richer set of questions than a simple two-arm comparison, but that additional structure comes with a larger set of possible contrasts.
| Question | Reference comparison | Statistical meaning |
|---|---|---|
| Does Cytosorb with a 12-hour schedule differ from standard therapy? | 12-hour arm vs standard therapy | Tests one Cytosorb schedule against the non-Cytosorb arm. |
| Does Cytosorb with a 24-hour schedule differ from standard therapy? | 24-hour arm vs standard therapy | Tests the second Cytosorb schedule against standard therapy. |
| Do the two Cytosorb schedules differ? | 12-hour arm vs 24-hour arm | Tests the schedule itself rather than cytokine removal versus no cytokine removal. |
These questions have different clinical interpretations. A comparison of either Cytosorb arm with standard therapy concerns the presence of cytokine removal in addition to standard medical therapy. The direct 12-hour versus 24-hour comparison instead concerns the treatment schedule.
22. Statistical Interpretation of the Registry Status
The ClinicalTrials.gov record identifies SHOCK as suspended. The registry also lists a primary completion date of 2026-10-31.
Trial status and statistical results are separate concepts. A status designation describes the operational state of the study in the registry; it does not itself establish whether an intervention is effective or ineffective. Likewise, the absence of posted numerical analyses does not provide evidence for either direction of a treatment effect.
23. Sources
- ClinicalTrials.gov: SHOCK — NCT04742764.
- PubMed: PubMed record associated with the SHOCK trial literature.
Continue through Clinical Biostats
Explore statistical methods, calculators, and additional clinical-trial analyses covering the methods used across randomized clinical research.
24. Record Summary
SHOCK is a phase 3, randomized, multicenter, parallel clinical trial in septic shock with planned enrollment of 135 and three treatment arms. The study compares standard medical therapy with two Cytosorb cytokine-removal schedules, one using an adsorber changed every 12 hours and the other every 24 hours. Its two registered primary endpoints focus on the same clinical concept from complementary statistical perspectives: the proportion achieving shock reversal within 5 days and the time from treatment initiation to shock reversal.
The statistical structure makes several principles especially important. The binary endpoint requires a clear definition of the event and an appropriate comparison of proportions. The time-to-event endpoint requires explicit rules for event timing, censoring, and patients who may become unable to experience shock reversal. The three-arm design creates multiple possible treatment contrasts, while the two primary endpoints create additional multiplicity considerations. A formal analysis must therefore specify the estimand, analysis population, treatment contrasts, missing-data and censoring rules, effect measures, confidence intervals, and type I error strategy.
The ClinicalTrials.gov record currently reports no formal statistical analyses for the two primary endpoints. Accordingly, the registry establishes the clinical and statistical questions the trial is designed to address, while numerical treatment effects remain unreported in the record.