This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Safe and Timely Antithrombotic Removal - Direct Oral Anticoagulants Apixaban & Rivaroxaban (STAR-D) (NCT05093504), a randomized na study. The primary outcome is Incidence of Perioperative Bleeding.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT05093504 |
| Phase | NA |
| Condition | Hemorrhage, Surgical, Hemorrhage Postoperative, Blood Loss, Surgical |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Enrollment | 9 |
| Arms in the primary outcome | n/a |
| Lead sponsor | CytoSorbents, Inc |
| Results posted | Yes |
2. Clinical Question
Population
Patients taking apixaban or rivaroxaban who were undergoing urgent cardiothoracic surgery. The registry describes the study as prospective and multicenter. Further eligibility criteria and the characteristics of enrolled participants are not reported in the registry record.
Intervention
Standard of care plus the DrugSorb-ATR system. The study was designed to evaluate removal of apixaban and rivaroxaban by this device in the urgent-surgery setting.
Comparator
Standard of care with a sham set-up. The registry describes this as the control arm. The exact details of standard of care and the sham procedure are not reported in the registry record.
Primary question
Whether use of DrugSorb-ATR was associated with a difference in clinically significant perioperative bleeding, assessed with a ranked composite endpoint through the first 48 hours post-operation. The registry contains no adjudicated primary outcome data, so it does not answer this comparison.
3. Trial Design
STAR-D was registered as a prospective, multicenter, randomized, quadruple-masked, parallel study. Its official title describes a double-blind study, while the registry’s masking field specifies quadruple masking. The record identifies the interventions as devices: DrugSorb-ATR and a sham comparator. The control arm was standard of care with a sham set-up; the intervention arm was standard of care plus DrugSorb-ATR. The allocation ratio and operational details of masking are not reported in the registry record.
The registry lists enrollment as 9 and phase as NA. The recorded start date is 2021-12-27, and the primary completion date is 2024-01-29. The status is terminated. The sponsor is CytoSorbents, Inc. The brief summary describes the purpose as evaluating removal of apixaban and rivaroxaban to reduce the likelihood of serious bleeding in patients undergoing urgent cardiothoracic surgery.
The primary outcome was incidence of perioperative bleeding, defined in the record as clinically significant perioperative bleeding events evaluated through a ranked composite. Secondary outcomes covered direct oral anticoagulant drug removal, chest tube drainage, platelet and packed red blood cell transfusions, perioperative bleeding-event severity, and surgical re-exploration for bleeding. The registry lists 10 outcome measures and 1 analysis, but that does not establish which analyses were completed or what statistical procedures were used. The primary endpoint data were not cleaned and adjudicated before termination, and the planned Clinical Events Committee was not convened. Consequently, the record has no primary bleeding outcome results.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Overall Study
| Stage | Control | DrugSorb-ATR Intervention |
|---|---|---|
| STARTED | 5 | 4 |
| COMPLETED | 5 | 3 |
| NOT COMPLETED | 0 | 1 |
Reasons for not completing this period:
| Reason | Control | DrugSorb-ATR Intervention |
|---|---|---|
| Death | 0 | 1 |
Analysis population for the primary outcome, as defined in the registry: The trial was terminated before the outcome measure data were cleaned and adjudicated. All bleeding outcomes data were to be adjudicated by a Clinical Events Committee, which was not convened because the trial was terminated. Therefore, no bleeding outcomes were available.
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.
| Characteristic | Unit | Control | DrugSorb-ATR Intervention | Total |
|---|---|---|---|---|
| Participants analysed | 5 | 4 | 9 | |
| Age, Continuous | years | 62.8 (SD 14.7) | 68.5 (SD 12.2) | 65.3 (SD 13.1) |
| Sex: Female, Male: Female | Participants | 2 | 3 | 5 |
| Sex: Female, Male: Male | Participants | 3 | 1 | 4 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 1 | 1 | 2 |
| Race (NIH/OMB): White | Participants | 3 | 3 | 6 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 1 | 0 | 1 |
| Region of Enrollment: United States | Participants | 5 | 4 | 9 |
| Hypertension | Participants | 5 | 2 | 7 |
| Heart failure | Participants | 1 | 1 | 2 |
| Coronary artery disease | Participants | 1 | 1 | 2 |
| Subjects with apixaban | Participants | 4 | 2 | 6 |
| Subjects with rivaroxaban | Participants | 1 | 2 | 3 |
6. Primary Endpoint
The registry lists the primary outcome as Incidence of Perioperative Bleeding, measured over Through the first 48 hours post-operation.
| Element | Registry entry |
|---|---|
| Result type | N/A |
| Measure of spread | not applicable |
| Unit | not stated |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint was the incidence of clinically significant perioperative bleeding events, evaluated as a ranked composite through the first 48 hours post-operation. A composite endpoint groups multiple outcomes into a single assessment. Ranking can represent differences in clinical importance among event categories rather than treating every event as equivalent. That is a general explanation of the method; the exact components, ranking rules, and criteria for clinically significant bleeding are not reported in the registry record.
In general, an endpoint focused on perioperative bleeding is relevant to a study intended to evaluate a device used during urgent surgery, because it measures bleeding events in the period specified by the study. A composite may capture more than one type or severity of event, but interpretation depends on the prespecified definitions and the way the ranking is applied. Those details cannot be assessed from the registry description alone.
The record states that the trial ended before bleeding outcome data were cleaned and adjudicated. All bleeding outcomes were to be adjudicated by a Clinical Events Committee, which was not convened because the trial was terminated. Therefore, no bleeding outcomes were available. There is no posted primary endpoint estimate, comparison between arms, or precision measure to interpret.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The central statistical limitation is that STAR-D has no available primary bleeding outcome results. The registry explicitly attributes this to termination before the outcome data were cleaned and adjudicated and notes that the planned Clinical Events Committee was not convened. Without adjudicated endpoint data, the intended comparison between standard care plus DrugSorb-ATR and standard care with a sham set-up cannot be evaluated using the primary outcome.
This absence of results is not evidence that bleeding did or did not occur, and it does not establish that the arms were alike or different. It also does not show whether the device removed either anticoagulant or whether removal was associated with bleeding. Such conclusions would require outcome measurements and an analysis that are not available in the registry record. The study’s stated enrollment is 9, but enrollment alone cannot substitute for an endpoint estimate or demonstrate what the trial could detect.
The adverse-event entries are separate from the primary endpoint. The record lists serious adverse events and other adverse events by arm, along with deaths and serious-event terms. These entries provide safety information as recorded, but they are not a substitute for the adjudicated ranked composite of perioperative bleeding. The registry does not report that the listed events were caused by the intervention, nor does a comparison of the counts establish a causal effect. With the small recorded group sizes and no reported analysis plan or uncertainty estimates for these entries, descriptive counts should not be treated as a definitive comparison.
Precision cannot be judged for the primary outcome because there is no primary effect estimate or confidence interval. In general, a confidence interval expresses the range of effects reasonably compatible with the observed data under the analysis assumptions; it is not available for the primary endpoint in this record. The registry also does not report an inferential result for that endpoint. The appropriate reading is therefore that the planned primary comparison is unresolved in the posted record, rather than that it produced a null result.
The study design and endpoint description identify a clinically focused question, but design intent does not establish a result. Randomization can support a fair comparison when outcome ascertainment and analysis are completed, while masking can help limit some forms of bias. In STAR-D, the lack of cleaned, adjudicated primary data prevents readers from assessing how the intended comparison turned out. The secondary outcome names likewise do not establish that usable results were collected or analyzed.
9. Statistical Methods Explained
The registry describes STAR-D as randomized, parallel, multicenter, prospective, and quadruple-masked. These are design features, not evidence that a particular statistical analysis was completed. Randomization assigns participants to study arms by chance and is generally intended to reduce systematic differences between groups at the start of a trial. A parallel design means participants are assigned to one study arm rather than crossing between interventions. The registry does not report the allocation ratio or the sequence-generation and allocation-concealment procedures, so those features cannot be assessed.
What does masking contribute?
In general, masking means that specified people involved in a study do not know which intervention a participant received. It can reduce the influence of expectations on care, outcome assessment, or reporting. The registry uses the term quadruple masking and the official title uses double-blind, but it does not identify which roles were masked or describe how masking was maintained. It is therefore possible to describe the recorded design label, but not to judge its implementation from the registry record.
How should a ranked composite be understood?
A composite outcome combines multiple clinical events into one endpoint. A ranked composite additionally orders outcomes according to a hierarchy or severity rule, so the analysis can distinguish events by their assigned rank. In general, this approach can preserve distinctions that would be lost if all events were counted as equivalent. It also requires transparent definitions: readers need to know which events qualify, how ties are handled, and how the ranking is translated into a statistical comparison. The registry says the primary endpoint was a ranked composite of clinically significant perioperative bleeding, but does not report its component definitions, ranking algorithm, or analysis method. In this trial, the endpoint was not adjudicated, so no ranked-composite result is available.
What can be inferred from the analysis information?
The registry lists 1 analysis. That entry alone does not specify the analysis population, statistical model, effect measure, adjustment variables, handling of repeated or tied outcomes, or uncertainty calculation. These details are not reported in the registry record. A general statistical interpretation should distinguish between a design description and a completed analysis: randomized allocation describes how treatment assignment was intended, while an analysis result requires outcome data and a defined method. The absent primary outcome data mean that no estimate of the intended arm comparison can be assessed.
How should planning illustrations be read?
In general, sample-size planning asks how many participants would be needed to detect a chosen effect under specified assumptions. The registry’s planning description identifies a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It says that the effect to detect is a stated fraction of the observed effect. For time-to-event illustrations, the description refers to the total number of events; for other illustrations, it refers to participants per arm. Such planning quantities are illustrations, not observed trial results and not observed power. They do not describe the actual allocation ratio or establish the trial’s statistical plan.
The primary outcome is not reduced to a single comparison in the registry’s analysis information because it has several categories or time points. That matters for interpreting any planning exercise: a single effect size cannot be assumed to represent the full ranked composite without a defined summary and analysis. No numerical sample-size illustration or numerical effect estimate is reported in the registry record for this endpoint. Therefore, there is no basis to infer a target sample size, a detectable effect, or the precision that would have been achieved.
Why does adjudication matter?
Adjudication is a structured review of whether reported events meet defined endpoint criteria. In general, an independent or otherwise specified committee can apply consistent definitions across treatment arms, limiting the influence of treatment knowledge or local interpretation. The registry says all bleeding outcomes were to be adjudicated by a Clinical Events Committee, but that committee was not convened because the trial was terminated. The endpoint data were not cleaned and adjudicated before termination. This leaves no validated primary outcome dataset for the planned comparison, regardless of the original randomization design.
Statistical methods for missing outcomes, multiplicity, interim monitoring, and analysis populations are not reported in the registry record. General methods such as imputation, sensitivity analyses, or multiplicity adjustment cannot be attributed to STAR-D without documentation. The key methodological distinction is between information describing what the trial planned to study and information showing that an endpoint was measured, adjudicated, and analyzed. The registry supports the former description but reports that the primary bleeding outcome was unavailable.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 10 outcome measures, including the primary bleeding endpoint and secondary measures of drug removal, drainage, transfusions, bleeding-event severity, and re-exploration. Testing several outcomes can raise the chance of at least one apparently notable result arising by chance. In general, multiplicity procedures may control that risk, but the registry record does not report a prespecified hierarchy, adjustment method, or other multiplicity strategy. The listed count of 1 analysis does not reveal how multiple outcomes were handled.
For the primary endpoint, the issue is not a reported estimate with incomplete follow-up: the record says the outcome data were not cleaned and adjudicated before the trial was terminated. The Clinical Events Committee was not convened, and no bleeding outcomes were available. The reasons for termination beyond this statement, the extent of raw outcome collection, and any missing-data mechanism are not reported in the registry record. It would be inappropriate to assume that absent endpoint results represent no events or that the missingness was random.
The registry does not report the analysis populations, such as whether analyses were to follow randomized assignment or include only participants with particular data. It also does not report imputation, sensitivity analyses, interim analyses, or a detailed statistical analysis plan. These omissions limit assessment of how missing observations and multiple outcomes would have been addressed. They also mean that no analysis population or missing-data method should be inferred from the design labels or from the presence of adverse-event entries.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Control | 1 | 5 | 20.0% |
| DrugSorb-ATR Intervention | 2 | 4 | 50.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Control | 0 of 5 |
| DrugSorb-ATR Intervention | 1 of 4 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Control | 2 of 5 |
| DrugSorb-ATR Intervention | 2 of 4 |
Most frequent serious adverse event terms
| Event term | Organ system | Control | DrugSorb-ATR Intervention |
|---|---|---|---|
| Acute Respiratory Failure | Respiratory, thoracic and mediastinal disorders | 1 of 5 | 1 of 4 |
| Hypotension | Vascular disorders | 1 of 5 | 0 of 4 |
| Hypoxia | Respiratory, thoracic and mediastinal disorders | 0 of 5 | 1 of 4 |
| Carotid artery dissection | Nervous system disorders | 0 of 5 | 1 of 4 |
| Cerebral ischaemi | Nervous system disorders | 0 of 5 | 1 of 4 |
| Post procedural hemorrhage | Injury, poisoning and procedural complications | 1 of 5 | 0 of 4 |
| Pericarditis | Cardiac disorders | 0 of 5 | 1 of 4 |
| Bradycardia | Cardiac disorders | 1 of 5 | 0 of 4 |
| Thrombocytopenia | Blood and lymphatic system disorders | 0 of 5 | 1 of 4 |
| Anaemia | Blood and lymphatic system disorders | 0 of 5 | 1 of 4 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Control | DrugSorb-ATR Intervention |
|---|---|---|---|
| Hyperglycemia | Metabolism and nutrition disorders | 1 of 5 | 1 of 4 |
| Hypoxia | Respiratory, thoracic and mediastinal disorders | 0 of 5 | 2 of 4 |
| Pleural effusion | Respiratory, thoracic and mediastinal disorders | 1 of 5 | 1 of 4 |
| Hypotension | Vascular disorders | 0 of 5 | 2 of 4 |
| Coagulopathy | Blood and lymphatic system disorders | 0 of 5 | 1 of 4 |
| Atrial fibrillation | Cardiac disorders | 0 of 5 | 1 of 4 |
| Oedema peripheral | General disorders | 1 of 5 | 0 of 4 |
| Pyrexia | General disorders | 0 of 5 | 1 of 4 |
| Clostridium difficile infection | Infections and infestations | 1 of 5 | 0 of 4 |
| Urinary tract infection | Infections and infestations | 0 of 5 | 1 of 4 |
Serious events with the highest arm-level rate: acute respiratory failure (25.0%), hypoxia (25.0%), carotid artery dissection (25.0%), cerebral ischaemi (25.0%), pericarditis (25.0%).
The registry lists serious adverse events for Control with 5 at risk and 1 affected, and for DrugSorb-ATR Intervention with 4 at risk and 2 affected. It lists other adverse events for Control with 5 at risk and 2 affected, and for DrugSorb-ATR Intervention with 4 at risk and 2 affected. The death entries are 0 for Control and 1 for DrugSorb-ATR Intervention. These are descriptive registry entries; they do not by themselves establish that an event was caused by the device or indicate a statistically reliable difference between arms.
The record’s serious-event terms include Acute Respiratory Failure, Hypoxia, Carotid artery dissection, Cerebral ischaemi, Pericarditis, and Thrombocytopenia, each reported as 25.0%. These term-level entries should be read as recorded, not as an adjudicated comparison of the primary bleeding endpoint. The registry does not provide enough detail in these entries to determine event timing, relationship to the intervention, or how multiple events within a participant were handled. The absence of a primary bleeding result remains separate from these safety listings.
12. What the Result Does, and Does Not, Mean
STAR-D addresses a defined statistical question: in urgent cardiothoracic surgery patients taking apixaban or rivaroxaban, compare standard care plus DrugSorb-ATR with standard care and a sham set-up on a ranked composite of clinically significant perioperative bleeding through the first 48 hours post-operation. The registry record establishes that this was the intended question and describes the randomized, parallel, masked design. It does not provide the primary outcome data needed to estimate the comparison.
Accordingly, the record does not establish whether the intervention arm had more, fewer, or similar bleeding events than the control arm. It does not provide an effect size, confidence interval, or inferential result for the primary endpoint. The absence of those statistics is not equivalent to a finding of no effect. Nor can the planned study objective, device description, or enrollment figure be used to fill the evidentiary gap.
The adverse-event entries provide descriptive counts and term labels, but they do not resolve the primary question. They are not a substitute for the trial’s ranked bleeding endpoint, and they do not establish causality. The registry does not report enough detail to support a broader treatment-effect conclusion from those entries.
In general, a randomized trial can provide an informative comparison when participant outcomes are measured consistently, endpoint definitions are applied, and the planned analysis is reported. Here, the registry states that bleeding outcomes were not cleaned and adjudicated and that the Clinical Events Committee was not convened. Readers should therefore distinguish the trial’s intended design from the results that are actually available. The record describes a trial that was undertaken and terminated, but leaves its primary efficacy question unanswered.
13. Why This Trial Matters Statistically
This trial is statistically notable because its central endpoint depended on both outcome collection and adjudication, and the registry says neither produced an available primary result before termination. A randomized design can reduce bias in treatment assignment, but randomization cannot compensate for the absence of an analyzed endpoint. When the primary endpoint is unavailable, the intended comparison remains unresolved even if the design question is clearly stated.
The ranked-composite structure also matters. Composite endpoints can combine several event types, while ranking attempts to preserve distinctions in severity or importance. That approach requires clearly specified components and rules. The registry does not report those details, and the adjudication committee did not convene. Readers therefore cannot verify how events would have been classified or how the planned ranking would have affected the comparison.
The record also illustrates why descriptive safety listings and primary outcome analyses answer different questions. The adverse-event counts are part of the registry record, but they do not provide the adjudicated bleeding comparison that the study set out to evaluate. Treating them as a replacement would conflate different outcomes and overstate what the data can support.
Finally, a planning illustration is not evidence that a trial had a particular level of power or that it observed a treatment effect. The registry’s planning description lists assumptions, but it does not supply a completed primary analysis. In general, statistical interpretation depends on the observed endpoint, the analysis method, and uncertainty around the estimate. For STAR-D, that chain is interrupted at outcome availability, making the limits of inference as important as the stated design.
14. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
The principal trial-specific limitation is that the primary bleeding outcome data were not cleaned and adjudicated before termination. The planned Clinical Events Committee was not convened, and the registry states that no bleeding outcomes were available. This prevents estimation or interpretation of the primary arm comparison. The record does not report whether any preliminary or unadjudicated bleeding data existed, or why the trial was terminated.
The registry describes a ranked composite but does not give its detailed event components, ranking rules, or primary analysis method. It also does not report the allocation ratio, masking roles, analysis populations, handling of missing outcomes, multiplicity strategy, or a detailed statistical analysis plan. These gaps limit assessment of the intended methods independently of the missing outcome results.
Enrollment is listed as 9, and the available arm-level safety entries are descriptive. The registry does not provide primary endpoint estimates or uncertainty measures, and the adverse-event listings do not establish causality. The status is terminated, but the record does not state whether termination was related to safety, recruitment, operational issues, or another reason. These limitations mean that conclusions should remain confined to the design description and the entries actually reported, rather than extending to the intervention’s effect on perioperative bleeding.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Safe and Timely Antithrombotic Removal - Direct Oral Anticoagulants Apixaban & Rivaroxaban (STAR-D) — NCT05093504.
- Registry record last updated 2025-05-04; status last verified 2025-04.
- Results first posted 2025-05-04.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
STAR-D (NCT05093504) was a prospective, multicenter, randomized, quadruple-masked, parallel device study sponsored by CytoSorbents, Inc. It compared standard of care plus DrugSorb-ATR with standard of care and a sham set-up in patients undergoing urgent cardiothoracic surgery while taking apixaban or rivaroxaban. The registry lists enrollment as 9 and status as terminated.
The primary endpoint was clinically significant perioperative bleeding assessed with a ranked composite through the first 48 hours post-operation. The record states that bleeding outcomes were not cleaned and adjudicated before termination, the Clinical Events Committee was not convened, and no bleeding outcomes were available. Descriptive adverse-event entries are recorded, but they do not answer the primary comparison. The registry does not report a primary effect estimate or detailed statistical analysis plan.