The ClinicalTrials.gov record identifies a randomized comparison with one primary endpoint: the incidence of catheter-related venous thrombosis as confirmed by diagnostic ultrasound. The record states that the study was prematurely terminated and that any data related to the study has been lost.
1. Trial at a Glance
SOLO2 was a randomized, parallel-group, single-masked study comparing two PICC devices in patients indicated for a PICC for any medical condition. The registry lists 57 enrolled participants, 2 arms, and a primary endpoint focused on catheter-related venous thrombosis.
| Feature | SOLO2 |
|---|---|
| Study acronym | SOLO2 |
| ClinicalTrials.gov identifier | NCT01921114 |
| Official study title | The PROOF Study: The PICC Related Obstruction Of Flow Study |
| Phase | NA |
| Status | Terminated |
| Enrollment | 57 |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Single |
| Primary purpose | Treatment |
| Number of arms | 2 |
| Lead sponsor | Angiodynamics, Inc. |
| Sponsor type | Industry |
| Start date | 2013-10 |
| Primary completion date | 2014-07 |
2. Clinical Question
The registry describes a study of patients indicated for a PICC for any medical condition. The statistical question can be expressed as whether the incidence of catheter-related venous thrombosis, confirmed by diagnostic ultrasound, differs between the two randomized PICC-device groups.
Population
Patients indicated for a peripherally inserted central catheter for any medical condition.
Intervention
BioFlo™ Peripherally Inserted Central Catheter (PICC).
Comparator
Bard® Dual-Lumen PowerPICC SOLO2®.
Primary question
How does the incidence of catheter-related venous thrombosis differ between the two randomized device groups?
3. Trial Design
BioFlo™ PICC
- BioFlo™ Peripherally Inserted Central Catheter (PICC)
- Randomized device assignment
- Evaluated for catheter-related venous thrombosis
Bard® Dual-Lumen PowerPICC SOLO2®
- Bard® Dual-Lumen PowerPICC SOLO2® device
- Randomized device assignment
- Evaluated for catheter-related venous thrombosis
The parallel-group structure is statistically important because each participant is assigned to one device group and the endpoint is compared between those groups. Randomization is intended to make the groups comparable, on average, with respect to factors that could otherwise confound a device comparison.
4. Endpoints
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Primary endpoint | Incidence of Catheter-related Venous Thrombosis as Confirmed by Diagnostic Ultrasound | 10 (+/-3) days and any time there is clinically suspected venous thrombosis (as per standard of care) |
What the endpoint measures
The primary endpoint is an incidence measure rather than a continuous measurement. Each participant can be classified according to whether catheter-related venous thrombosis is confirmed by diagnostic ultrasound during the specified assessment framework.
The time frame has two components. The first is a scheduled assessment at 10 (+/-3) days. The second is an assessment at any time when venous thrombosis is clinically suspected, according to standard of care. This means that endpoint ascertainment is tied both to a scheduled assessment window and to clinically triggered diagnostic evaluation.
For a simple binary endpoint, this produces a proportion within each device group. A between-group comparison can then quantify the relative or absolute difference in observed incidence.
5. Statistical Methodology
Binary endpoint analysis
Because the registered primary endpoint is an incidence of confirmed venous thrombosis, the fundamental statistical data structure is binary: an endpoint event is either confirmed or it is not within the defined assessment framework. A primary comparison would ordinarily begin by summarizing the number and proportion of participants with confirmed thrombosis in each randomized arm.
For a randomized two-arm comparison, an analysis could report an absolute risk difference, a risk ratio, or an odds ratio, together with an appropriate confidence interval. The choice should follow the prespecified statistical analysis plan and the amount and structure of available endpoint data.
Risk difference
The difference in event proportions between the two device groups. It answers how many more or fewer endpoint events were observed per unit of participants evaluated.
Risk ratio
The ratio of the incidence proportion in one device group to the incidence proportion in the other. A value of 1 represents equal observed incidence proportions.
Odds ratio
The ratio of the odds of confirmed thrombosis between groups. It is mathematically distinct from the risk ratio, particularly when the endpoint is not rare.
Exact methods
With a relatively small randomized study, exact methods can be considered when event counts are sparse. The appropriate method depends on the observed contingency table and prespecified analysis.
Confidence intervals
A confidence interval is essential for an incidence comparison because a point estimate alone does not show how precisely the treatment contrast has been estimated. For example, an absolute risk difference can be accompanied by a confidence interval describing uncertainty around that difference.
The width of the interval is particularly relevant in a study with 57 enrolled participants. Small studies can produce imprecise estimates, especially when the endpoint event is uncommon. Precision depends not only on total enrollment but also on the allocation of participants between groups and the number of observed events.
The two measures answer different questions. The risk difference describes the absolute separation in incidence, while the risk ratio describes the multiplicative relationship between the two incidence proportions.
Randomization and analysis population
Randomization is the central design feature supporting a between-device causal comparison. An analysis that retains participants according to their randomized assignment preserves the principal comparison established by the randomization process, subject to the limitations created by early termination and missing information.
The registry does not report analysis populations beyond the enrollment total. It therefore does not establish how many randomized participants contributed evaluable endpoint observations or how endpoint observations were handled after early termination.
Diagnostic confirmation
The endpoint requires confirmation by diagnostic ultrasound. This is important statistically because the outcome definition is not simply a clinical suspicion of thrombosis. A suspected event enters the endpoint framework through diagnostic assessment, while the registered endpoint specifically concerns thrombosis confirmed by ultrasound.
6. Statistical Methods Explained
Why is this primarily a binary-outcome analysis?
The endpoint asks whether catheter-related venous thrombosis was confirmed by diagnostic ultrasound within the registered assessment framework. That naturally produces an event/no-event classification for each participant. The primary summary therefore consists of incidence proportions rather than means or medians.
Why use a risk difference?
A risk difference directly describes the absolute separation between the two device groups. If one group had an observed incidence proportion of p1 and the other had p0, the difference p1 − p0 would express the absolute contrast. This can be easier to interpret clinically than a relative measure when event rates are available.
Why also consider a risk ratio?
A risk ratio expresses how the incidence proportion in one group compares multiplicatively with the other. A value below 1 would indicate a lower observed incidence in the numerator group, while a value above 1 would indicate a higher observed incidence. The ratio does not describe the absolute number of events prevented or increased.
Why does the assessment window matter?
An incidence estimate is inseparable from its observation period. Here, the registry specifies an assessment at 10 (+/-3) days and additional assessment whenever venous thrombosis is clinically suspected according to standard of care. The timing therefore affects which events can enter the endpoint definition and how participants contribute information.
Why is diagnostic confirmation important?
Diagnostic confirmation creates a distinction between suspicion and an endpoint event. A clinical suspicion triggers evaluation, but the registered primary endpoint is catheter-related venous thrombosis confirmed by diagnostic ultrasound. This distinction reduces ambiguity in the outcome definition.
Why are confidence intervals important in a small study?
With 57 enrolled participants, the statistical uncertainty around an observed difference can be substantial. A confidence interval communicates that uncertainty rather than presenting the point estimate as if it were known without error. A wide interval can remain compatible with several materially different underlying effects.
Why does premature termination affect statistical interpretation?
Early termination can reduce the amount of information available for the planned comparison. If fewer endpoint observations are collected than anticipated, estimates can become less precise. It can also prevent the prespecified statistical analysis from being completed as originally intended.
7. Planned Analysis
For the registered primary endpoint, the natural starting point would be a comparison of the incidence of catheter-related venous thrombosis confirmed by diagnostic ultrasound between the two randomized device groups. The analysis would ordinarily present the number of endpoint events, the number of participants contributing evaluable observations, and the corresponding incidence proportion for each group.
A between-group effect could then be expressed as an absolute risk difference and/or a relative measure such as a risk ratio, accompanied by a confidence interval. If event counts were sparse, an exact test or exact confidence interval could be appropriate. The final choice would depend on the prespecified statistical analysis plan and the actual endpoint counts.
| Statistical component | Typical role for this endpoint |
|---|---|
| Event count | Number of participants with catheter-related venous thrombosis confirmed by diagnostic ultrasound. |
| Incidence proportion | Describes the proportion of evaluable participants experiencing the primary endpoint. |
| Risk difference | Measures the absolute difference in incidence between randomized device groups. |
| Risk ratio | Measures the relative incidence between device groups. |
| Confidence interval | Quantifies uncertainty around the estimated treatment contrast. |
| Hypothesis test | Could assess whether the observed between-group difference is compatible with a specified null hypothesis, subject to the prespecified analysis plan. |
For a binary endpoint, the p-value is not a measure of effect magnitude. It describes the compatibility of the observed data with a specified null hypothesis under the chosen statistical model or test. The clinical and statistical interpretation should therefore consider the estimated incidence in each arm, the effect measure, and its confidence interval together.
8. Interpreting Incidence Comparisons
An incidence proportion describes the fraction of evaluable participants who experienced the defined endpoint during the relevant observation framework. It is an absolute frequency measure, not a measure of instantaneous hazard.
A risk difference compares the two incidence proportions directly. A negative value would indicate a lower observed incidence in the first group, while a positive value would indicate a higher observed incidence. The magnitude is expressed on an absolute probability scale.
A risk ratio compares the incidence proportion multiplicatively. A ratio of 1 indicates equal incidence proportions. A ratio below 1 indicates lower incidence in the numerator group, but it does not by itself indicate the absolute number of events involved.
The confidence interval describes statistical uncertainty around the estimated effect. It does not describe the range of outcomes that individual patients will experience. In a small study, limited event information can lead to substantial uncertainty even when the point estimate appears large.
A p-value depends on both the observed effect and the amount of information available. A small p-value does not necessarily represent a large clinical effect, while a non-small p-value does not prove that two devices have identical event rates. Effect estimates and confidence intervals provide the information needed to understand magnitude and precision.
9. The Importance of the 10-Day Assessment
The registry specifies assessment at 10 (+/-3) days. This creates a defined early observation window around the scheduled endpoint assessment. The additional provision for assessment whenever venous thrombosis is clinically suspected extends ascertainment beyond that scheduled assessment when clinical circumstances trigger diagnostic evaluation.
From a statistical perspective, the assessment schedule determines how participants enter the denominator and how endpoint events are classified. A scheduled assessment does not mean that all participants necessarily have identical follow-up information after that point, particularly in a prematurely terminated study.
The phrase "any time there is clinically suspected venous thrombosis (as per standard of care)" also means that endpoint ascertainment has a clinically triggered component. The statistical analysis therefore needs to remain faithful to the registry definition rather than treating the endpoint as though it were based exclusively on a single fixed-time ultrasound for every participant.
10. Randomization and Causal Interpretation
Randomization is important because it creates the design framework for comparing the two devices without intentionally assigning participants according to prognostic characteristics. If randomization is successfully implemented and followed by appropriate analysis, differences between groups can be interpreted within a stronger causal framework than differences from a nonrandomized observational comparison.
That advantage depends on having sufficient follow-up and endpoint information. The registry's statement that the study was prematurely terminated and that any data related to the study has been lost substantially limits what can be learned from the randomized comparison.
Supports comparison of the two assigned device groups.
Each participant belongs to one randomized device group.
The registry identifies the study as single-masked.
The primary endpoint requires confirmation by ultrasound.
11. What Can and Cannot Be Concluded From the Registry
The registry establishes that the study was randomized, parallel, single-masked, and conducted in patients indicated for a PICC. It also identifies the two devices, the enrollment of 57 participants, and the primary endpoint and its assessment framework.
However, the record states that the study was prematurely terminated and that any data related to the study has been lost. Consequently, the registered information does not provide an observed incidence comparison from which an effect estimate can be interpreted.
| Question | What the registry establishes |
|---|---|
| How many participants enrolled? | 57 |
| How many arms were planned? | 2 |
| Was allocation randomized? | Yes |
| Was the design parallel? | Yes |
| Was the study single-masked? | Yes |
| What was the primary endpoint? | Incidence of catheter-related venous thrombosis as confirmed by diagnostic ultrasound. |
| When was the endpoint assessed? | 10 (+/-3) days and any time there is clinically suspected venous thrombosis, as per standard of care. |
| Were formal statistical analyses posted? | No formal statistical analyses were posted to ClinicalTrials.gov. |
| Why was the study limited? | The record states that the study was prematurely terminated and that any data related to the study has been lost. |
12. Limitations
- Premature termination: the ClinicalTrials.gov record states that the study was prematurely terminated.
- Loss of study data: the record states that any data related to the study has been lost.
- No formal statistical analyses: the registry does not provide posted statistical-analysis estimates for the primary endpoint.
- Unknown evaluable sample: enrollment was 57, but the available record does not establish how many participants contributed evaluable primary-endpoint observations.
- Unknown event count: the registry-defined endpoint is specified, but an observed count of confirmed catheter-related venous thromboses is not available in the information represented here.
- Unreported arm sizes: the record identifies 2 arms and total enrollment of 57 but does not report the number assigned to each device group.
- Assessment complexity: the endpoint combines a scheduled 10 (+/-3)-day assessment with clinically triggered assessment whenever thrombosis is suspected.
- Limited inferential precision: even a completed study with 57 participants could have limited precision for an uncommon binary endpoint; premature termination further restricts the information available.
13. Why This Trial Matters Statistically
SOLO2 is a useful statistical teaching case because it illustrates an important distinction between trial design and available evidence. A study can have a clear randomized structure and a well-defined primary endpoint while still yielding insufficient information for a quantitative treatment-effect interpretation.
| Statistical concept | How it appears in SOLO2 |
|---|---|
| Randomization | The allocation is identified as randomized. |
| Parallel-group comparison | Two device groups are compared in a parallel design. |
| Binary endpoint | Catheter-related venous thrombosis confirmed by diagnostic ultrasound can be represented as an event/no-event outcome. |
| Incidence proportion | The primary endpoint is an incidence measure. |
| Risk difference | A natural absolute comparison for the two incidence proportions. |
| Risk ratio | A natural relative comparison for the two incidence proportions. |
| Confidence interval | Needed to communicate uncertainty around an estimated incidence contrast. |
| Small-sample inference | 57 total enrolled participants makes the choice of inferential method particularly relevant if endpoint counts are sparse. |
| Endpoint ascertainment | Confirmation requires diagnostic ultrasound, with scheduled and clinically triggered assessments. |
| Missing information | Premature termination and reported loss of study data prevent quantitative interpretation of the endpoint. |
14. Statistical Interpretation vs Clinical Interpretation
Statistical interpretation
The appropriate statistical target is a comparison of the incidence of ultrasound-confirmed catheter-related venous thrombosis between the randomized device groups. A complete analysis would require the underlying endpoint observations and evaluable denominators.
Clinical interpretation
The clinical question concerns whether the incidence of the defined catheter-related thrombosis endpoint differed between the two devices. The available registry information does not provide an observed between-device effect estimate with which to characterize that difference.
This distinction is important because statistical design does not substitute for observed data. Randomization determines how the comparison should be interpreted if adequate data are available; it does not itself establish the direction or magnitude of an endpoint difference.
15. Understanding the Missing Quantities
A complete statistical analysis of the primary endpoint would ordinarily require several pieces of information beyond total enrollment. Each one answers a different part of the inferential question.
How many participants were assigned to each device?
The total enrollment of 57 does not by itself establish the denominator for either randomized arm.
How many thromboses were confirmed?
The numerator is required to calculate an incidence proportion.
How many participants contributed endpoint observations?
This determines the denominator used in the incidence calculation and affects precision.
How was endpoint ascertainment completed?
The registered assessment framework includes both the 10 (+/-3)-day assessment and clinically suspected thrombosis assessments.
What was the between-device contrast?
An incidence comparison could be summarized with a risk difference, risk ratio, odds ratio, or another prespecified measure.
16. Statistical Lessons From an Early-Terminated Study
Premature termination is not merely an administrative feature of a clinical trial. Statistically, it changes the information available for estimation. If a study stops before enough endpoint observations accumulate, the resulting analysis may have substantially less precision than the original design anticipated.
There is also an important distinction between lack of evidence and evidence of no difference. When endpoint information is unavailable, the absence of a reported effect estimate cannot be interpreted as evidence that the two devices had identical thrombosis incidence. Conversely, the randomized design alone cannot establish that one device had a lower incidence.
Precision is determined by the amount of information contributing to the estimate. For a binary endpoint, both the number of participants and the number of observed events influence how precisely a between-group effect can be estimated.
The loss of study data described in the registry is therefore especially consequential. Without the endpoint observations and corresponding denominators, the statistical comparison cannot be characterized numerically.
17. Design Features in Context
| Feature | Registry description | Statistical relevance |
|---|---|---|
| Allocation | Randomized | Creates the planned basis for a comparative treatment-effect analysis. |
| Design model | Parallel | Participants remain in one of two device groups for the comparison. |
| Masking | Single | The registry identifies masking as single; the available record does not provide further masking details here. |
| Primary purpose | Treatment | Frames the study as an intervention comparison. |
| Arms | 2 | Defines the two-group statistical comparison. |
| Primary endpoint | Ultrasound-confirmed catheter-related venous thrombosis incidence | Determines the appropriate binary-outcome statistical framework. |
| Assessment | 10 (+/-3) days and clinically suspected thrombosis | Defines the endpoint ascertainment framework. |
| Status | Terminated | Limits the completeness of the planned evidence. |
18. Sources
- ClinicalTrials.gov: NCT01921114, the official registry record for SOLO2.
Continue through the Clinical Biostats clinical-trial library
Explore randomized-trial design, endpoint definitions, and statistical interpretation across clinical research.
19. Record Summary
SOLO2 was a randomized, parallel, single-masked study enrolling 57 participants in two PICC-device groups: BioFlo™ Peripherally Inserted Central Catheter (PICC) and Bard® Dual-Lumen PowerPICC SOLO2®. The registry-defined primary endpoint was the incidence of catheter-related venous thrombosis as confirmed by diagnostic ultrasound, assessed at 10 (+/-3) days and whenever venous thrombosis was clinically suspected according to standard of care.
Statistically, this endpoint is naturally analyzed as a binary incidence outcome, with the between-device comparison potentially expressed through an absolute risk difference or a relative measure such as a risk ratio, accompanied by a confidence interval. The interpretation would depend on the actual number of confirmed events and the number of evaluable participants in each randomized group.
The central limitation is the status of the study itself. The ClinicalTrials.gov record states that the study was prematurely terminated and that any data related to the study has been lost. The record therefore defines the statistical question and its assessment framework but does not provide the quantitative endpoint information needed to characterize the observed difference between the two devices.