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 Preferences for Open Vs. Endovascular Repair for Abdominal Aortic Aneurysm (NCT03115346), a randomized na study. The primary outcome is Veteran Preference for Repair Type.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03115346 |
| Phase | NA |
| Condition | Abdominal Aortic Aneurysm |
| Allocation | Randomized |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 241 |
| Arms in the primary outcome | n/a |
| Lead sponsor | VA Office of Research and Development |
| Results posted | Yes |
2. Clinical Question
Population
Veterans with an abdominal aortic aneurysm who were candidates for endovascular or open surgical repair were eligible for the study, according to the registry summary. The primary outcome is described for Veterans, with the registry noting that its reported outcome numbers differ from participant flow because some survey answers were omitted. The record does not provide further eligibility criteria or a detailed description of the analyzed population.
Intervention
At intervention sites, Veterans received a decision aid containing information about open and endovascular repair and completed a survey about their preferences. The study was designed to examine whether this decision aid was associated with alignment between treatment preference and repair type. The record does not describe the decision aid’s full content or how it was delivered.
Comparator
Veterans at control sites completed the same survey but did not receive the decision aid. This comparison therefore concerns the addition of the decision aid to the survey process, rather than a comparison between the two surgical procedures themselves. The record does not report whether other care or decision-support practices were standardized between groups.
Primary question
The central question is whether offering the decision aid changes Veterans’ preferences for repair type and the agreement between preference and the repair ultimately received. The listed primary outcome, “Veteran Preference for Repair Type,” is described as measuring preference and domains associated with preference for endovascular or open repair, with a 30-day timeframe. Although the study summary describes a comparison of preference–repair agreement, the registry record does not report the outcome results or the statistical analysis of that comparison.
3. Trial Design
The registry identifies the study as randomized, unmasked, and single-group in its design classifications. Its summary describes assignment at the medical-center level: participating VA Medical Centers were assigned to decision-aid or control conditions. The summary states that 12 sites were assigned to each group and describes 22-24 VA Medical Centers as participating. Since assignment is described by site, observations from Veterans at the same site may be related. That clustering is important when interpreting any group comparison; it is not equivalent to assigning each Veteran independently.
The intervention was a decision aid about open and endovascular repair for abdominal aortic aneurysm. Veterans at intervention sites received it and completed a preference survey. Veterans at control sites completed the survey without receiving the aid. The summary also describes a later survey about satisfaction after surgery. Decision Regret Scale is listed as a secondary outcome. The registry reports three outcome measures overall, but the detailed specification of every measure is not reported in the registry record.
The study was sponsored by the VA Office of Research and Development. Its status is completed. The recorded start date is 2017-04-15, and the recorded primary completion date is 2020-12-10. Enrollment is listed as 241. These registry entries describe the study’s administrative record; they do not establish how many participants contributed usable responses to each outcome. The primary outcome description specifically notes that its numbers differ from participant flow because some survey answers were omitted.
Although the registry indicates randomization, further details about the randomization procedure, site stratification, allocation concealment, or any changes to the design are not reported in the registry record. Masking is listed as none. The record does not describe how the study addressed the possibility that awareness of receiving a decision aid could influence survey responses.
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 | Veteran's Hospital Vascular Surgery Clinics Using Decision Aid | Veterans Hospital Vascular Surgery Clinics Providing Usual Care |
|---|---|---|
| STARTED | 126 | 109 |
| COMPLETED | 95 | 87 |
| NOT COMPLETED | 31 | 22 |
Reasons for not completing this period:
| Reason | Veteran's Hospital Vascular Surgery Clinics Using Decision Aid | Veterans Hospital Vascular Surgery Clinics Providing Usual Care |
|---|---|---|
| Did not have AAA surgery | 28 | 15 |
| Death | 3 | 1 |
| Withdrawal by Subject | 0 | 3 |
| Lost to Follow-up | 0 | 3 |
Analysis population for the primary outcome, as defined in the registry: These numbers differ from the participant flow module due to omitted survey answers by participants.
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.
Population: The Overall Number of Baseline Participations reflects the Number of Participations who started the study as described in the Participant Flow, Overall Study Table.
| Characteristic | Unit | Decision Aid | Control | Total |
|---|---|---|---|---|
| Participants analysed | 126 | 109 | 235 | |
| Age, Continuous | years | 73 (66 to 80) | 74 (66 to 82) | 74 (66 to 82) |
| Sex: Female, Male: Female | Participants | 0 | 1 | 1 |
| Sex: Female, Male: Male | Participants | 126 | 108 | 234 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 1 | 2 | 3 |
| 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 | 4 | 7 | 11 |
| Race (NIH/OMB): White | Participants | 111 | 97 | 208 |
| Race (NIH/OMB): More than one race | Participants | 7 | 3 | 10 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 3 | 0 | 3 |
6. Primary Endpoint
The registry lists the primary outcome as Veteran Preference for Repair Type, measured over 30 days.
| Element | Registry entry |
|---|---|
| Result type | Count Of Participants |
| Measure of spread | not applicable |
| Unit | Participants |
| Endpoint classification | Binary Endpoint |
The primary outcome is titled “Veteran Preference for Repair Type.” It is recorded as a count of participants, and its description says the investigators’ survey measured the effect of the decision aid on Veterans’ preferences for abdominal aortic aneurysm repair and on domains associated with preference for endovascular or open repair. Its stated timeframe is 30 days. The summary places that outcome in the broader context of assessing agreement between a Veteran’s preference and the repair type received.
Preference is relevant to the study question because the intervention is intended to inform a choice between two repair approaches. A measure of preference can describe what participants favor, while an agreement measure can address whether that preference corresponds to the repair type ultimately received. These are related but not identical concepts. The registry’s primary outcome description focuses on preference and associated domains; the summary also refers to agreement. The exact operational definition used to classify agreement is not reported in the registry record.
The registry notes that outcome numbers differ from participant flow because some survey answers were omitted. It does not report the numbers of participants with each preference, the distribution of preference domains, or the agreement results. It also does not describe how missing survey answers were handled in the primary endpoint analysis. Consequently, the endpoint definition identifies the intended subject of measurement, but does not by itself establish the direction or size of any observed group difference.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record establishes a clinically focused research question: whether a decision aid is associated with Veterans’ repair preferences and the agreement between preference and repair type. It does not report the primary outcome findings. No group-specific preference counts, agreement estimates, effect estimate, confidence interval, or p-value are reported in the registry record. A treatment effect, statistical significance, or degree of precision therefore cannot be inferred from the recorded design description alone.
That absence matters particularly because the study’s assignment was described at the medical-center level. Veterans recruited at the same site may have more similar experiences or responses than Veterans recruited at different sites. A valid comparison needs to account for the site-based assignment rather than treating every survey response as if it arose from an independently randomized participant. The registry itself cautions that a simple individual-level re-analysis would be misleading and indicates that the sponsor’s own analysis is the relevant analysis. However, the sponsor’s analysis results are not reported in the registry record.
The listed enrollment is 241, but that figure should not be read as the denominator for the primary outcome. The outcome description explicitly says the outcome numbers differ from participant flow because some survey answers were omitted. The registry does not report the number with a usable response in each group, how omissions differed across groups, or how the omissions were handled analytically. Without those details, the extent to which the recorded outcome population reflects all enrolled participants cannot be assessed.
The decision aid and control conditions also should not be confused with the open and endovascular procedures. The study compares a decision-support approach with a survey-only approach; it is not described as randomizing Veterans to receive one repair rather than the other. Its question concerns preference and preference–repair alignment, not a direct comparison of the clinical effects of the surgical procedures.
The record identifies Decision Regret Scale as a secondary outcome, but does not report its results. It also lists three outcome measures overall, while the remaining outcome details and analysis results are not reported in the registry record. The completed status means the study is recorded as completed, not that a particular result is established. Interpretation of effectiveness, uncertainty, and applicability requires the outcome estimates and analysis details that are absent from this record.
9. Statistical Methods Explained
Randomization is intended to make comparison groups comparable, on average, with respect to factors that could influence outcomes. In this study, the registry summary describes assignment of medical centers rather than individual Veterans. This is a cluster-randomized structure: the site is the unit assigned to a condition, while Veterans provide outcome measurements within sites. When people within a site share clinicians, processes, or local context, their outcomes may be correlated. General statistical principle: accounting for that correlation is important because treating clustered observations as independent can make uncertainty appear smaller than it is.
The registry explicitly warns that a simple individual-level re-analysis would be misleading and says the sponsor’s own analysis takes precedence. This design-specific warning is important when reading any estimate or test. An analysis that ignores site assignment may use the wrong uncertainty calculation, even if its group percentages or counts look plausible. The registry record does not report the sponsor’s model, its assumptions, the estimated effect, or its uncertainty. Details such as whether site was modeled as a random effect or handled through another cluster-adjusted method are not reported in the registry record.
For a binary endpoint, outcomes are often represented as counts of participants with an event or outcome category, together with an appropriate denominator. The registry describes the primary measure as a count of participants, but does not provide the preference-category counts or a binary definition of preference–repair agreement. It therefore is not possible to identify from this record the exact effect measure used for the primary question. In general, an effect estimate describes the size and direction of a group contrast, while a confidence interval describes the range of values compatible with the data and the analysis assumptions. Neither is reported for the primary result.
Cluster designs also require attention to the amount of information contributed by sites, not only the number of individuals. General statistical principle: if responses within the same site are correlated, additional participants from one site may add less independent information than participants recruited across separate sites. The degree of within-site correlation and the method used to account for it are not reported in the registry record. The listed enrollment alone cannot establish the precision of a cluster-adjusted comparison.
The registry describes a planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It says the effect to detect is a stated fraction of the observed effect. These assumptions describe a planning exercise, not the study’s original design justification, and any resulting sample-size values are planning illustrations rather than observed power. The observed effect used for that illustration, the displayed planning values, and the trial’s own sample-size justification are not reported in the registry record. The record does not report whether the original design used different assumptions or how clustering informed its planning.
General explanation: a two-sided test assesses evidence for a difference in either direction, whereas a one-sided test specifies a direction in advance. A significance level is a decision threshold used in a testing framework; it does not measure the probability that a conclusion is true. A normal approximation uses a mathematical approximation to obtain inference, and its suitability depends on the data and design. These concepts explain the stated planning assumptions, but do not establish which inferential procedure was used for the study’s primary analysis.
Missing survey answers are a specific concern because the primary outcome description says outcome numbers differ from participant flow due to omitted answers. General statistical principle: missingness can affect interpretation if the likelihood of a response is related to treatment group, preference, or other factors associated with the outcome. Analyses may use complete responses, imputation, weighting, or other methods, but the registry record does not report which approach was used, whether reasons for missingness were examined, or whether sensitivity analyses were performed.
The study is listed as unmasked. In general, participants who know whether they received a decision aid may report preferences differently because of awareness, expectations, or the information they encountered. That is not proof that responses were biased; it is a potential interpretive consideration for a preference-based survey outcome. The registry record does not report procedures intended to assess or reduce such effects.
The primary outcome timeframe is 30 days. A timeframe helps define when a measurement is intended to be collected, but the record does not provide the detailed survey schedule, timing windows, or rules for handling responses outside the planned window. It also does not report an analysis protocol or statistical plan specifying the primary comparison. These omissions limit assessment of whether the reported methods, if available elsewhere, would align with the registered outcome description.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists three outcome measures overall and identifies Decision Regret Scale as a secondary outcome. The primary outcome is preference for repair type. The detailed descriptions of all outcome measures, their hierarchy, and any prespecified secondary comparisons are not reported in the registry record. It also records no analysis entries. As a result, the analysis methods and results for the primary and secondary outcomes are not available in this record.
When a study evaluates multiple outcomes or comparisons, the chance of finding an apparently unusual result can depend on how many analyses are conducted and how they are interpreted. General statistical principle: investigators may prespecify a primary outcome, adjust for multiple testing, or treat secondary findings as exploratory. The registry record does not report whether a multiplicity strategy was specified, whether any adjustment was made, or how secondary outcomes were intended to support the primary conclusion. No inference about those choices should be made from the outcome list alone.
Missing data are directly relevant to the primary outcome because the registry says its outcome numbers differ from participant flow due to omitted survey answers. The number of missing answers by group, reasons for omission, and analysis population are not reported in the registry record. The record also does not say whether all randomized sites or participants were included in each analysis, whether participants were analyzed according to their site’s assigned group, or whether any alternative populations were used.
These gaps mean that the completeness and robustness of the outcome comparisons cannot be evaluated from the registry record. In particular, the record does not report a missing-data strategy, sensitivity analyses, interim analyses, or a statistical analysis plan. Those details are necessary to judge how outcome selection, missing responses, and multiple comparisons may have influenced interpretation.
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 |
|---|---|---|---|
| Decision Aid | 0 | 126 | 0.0% |
| Control | 0 | 109 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Decision Aid | 8 of 126 |
| Control | 5 of 109 |
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 |
|---|---|
| Decision Aid | 0 of 126 |
| Control | 0 of 109 |
The adverse-event record lists the Decision Aid group with 126 participants at risk, zero serious events, zero other events, and 8 deaths. The Control group is listed with 109 participants at risk, zero serious events, zero other events, and 5 deaths. No serious-event terms are listed. These are registry-reported counts; they are not the primary preference outcome and should not be used as its denominators.
The record does not provide event descriptions, causes of death, timing, relatedness to the intervention, or an explanation of how deaths were collected. It also does not explain the relationship between these safety-group counts and the survey population, which the primary outcome description says differs from participant flow because some answers were omitted. The counts alone therefore do not establish whether events were related to the decision aid, the repair procedure, or other circumstances. They also do not permit a comparison of risk without additional information about follow-up and ascertainment. Zero recorded serious or other events should be read as the counts entered in this registry record, not as evidence that no such events occurred outside the recorded reporting process.
12. What the Result Does, and Does Not, Mean
This study addresses a decision-making question for Veterans who are candidates for open or endovascular repair of an abdominal aortic aneurysm. Its intervention is a decision aid about the repair options, and the control condition involves completing the survey without the aid. The intended focus is on stated preference and the relationship between that preference and repair type. It does not directly compare the clinical effects of open and endovascular surgery.
The record establishes that the study is completed and describes its design, population, intervention, and outcome framework. It does not report the primary results. In particular, no group-level preference findings, preference–repair agreement estimate, confidence interval, or p-value are reported in the registry record. Therefore, the record cannot establish whether the decision aid changed preferences, whether agreement differed between groups, or how precise any such difference was.
The distinction between a study’s question and its answer is important. A registered outcome definition explains what investigators intended to measure; it is not itself evidence of an effect. Similarly, enrollment and completion status do not tell readers whether an intervention changed an outcome. Conclusions about the observed findings require the analysis results and sufficient details about how those results were obtained.
The site-level assignment described in the summary also affects what would count as an appropriate analysis. Any comparison should account for the clustering of Veterans within medical centers, as the registry cautions against a simple individual-level re-analysis. The sponsor’s own analysis results are not reported in the registry record, so neither its estimate nor its uncertainty can be evaluated.
Finally, missing survey answers are acknowledged, but their amount, distribution, and handling are not reported. This limits assessment of how representative the analyzed responses may be of enrolled participants. The registry information supports a description of the study’s aims and design, but not a conclusion about effectiveness, lack of effect, clinical benefit, or the direction of the primary result.
13. Why This Trial Matters Statistically
Preference-sensitive decisions are not captured fully by asking only which procedure was received. A person’s stated preference and the treatment ultimately received are related aspects of decision-making, and the study’s summary identifies agreement between them as a central concern. Measuring these outcomes can help distinguish the effect of decision support on preference from the separate question of whether preferences align with subsequent care. The registry describes that question without reporting its answer.
The study is also statistically instructive because assignment is described at the medical-center level. When sites, rather than individual participants, are randomized, the analysis must respect the unit of assignment. Participants treated within one center may share features that make their responses correlated. The registry’s warning against a simple individual-level re-analysis highlights why design and analysis cannot be separated when interpreting clinical research.
Missing survey answers add another important layer. The record explicitly notes that outcome figures differ from participant flow because some answers were omitted. That makes the outcome-specific analysis population distinct from the enrollment figure, while the exact group denominators and missing-data approach remain unreported. The distinction matters for judging how much information supports any estimate and whether nonresponse could affect interpretation.
The study therefore illustrates both the value and the limits of a registry record. Its description identifies a decision aid, a survey-only comparator, a cluster-based assignment structure, and a preference-related endpoint. But without the sponsor’s analysis and outcome results, the statistical question remains unresolved in the registry record. Readers can understand what was intended to be studied and which design features matter, but cannot quantify or assess the reported effect.
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 main limitation for interpreting this trial from the registry record is the absence of primary outcome results and sponsor analysis details. The record does not report preference counts, a measure of preference–repair agreement, an effect estimate, a confidence interval, or a p-value. It also cautions that individual-level re-analysis would be misleading because the study is cluster-randomized, yet the cluster-adjusted analysis itself is not reported.
The primary outcome description states that some survey answers were omitted and that outcome numbers consequently differ from participant flow. The record does not give outcome-specific denominators, reasons for missingness, group-specific missingness, or the approach used to handle incomplete responses. This prevents assessment of whether the analyzed responses could differ systematically from those without recorded answers.
Further design and analysis details are absent. The record does not report the randomization procedure, the statistical model, cluster-correlation assumptions, analysis population, multiplicity strategy, or whether sensitivity analyses were performed. It does not give a detailed definition for classifying agreement between preference and repair type, despite describing that agreement as a study aim. The record also does not explain how unmasked assignment may have affected self-reported preferences.
The registry’s safety entries provide event counts but do not describe event timing, causes, relatedness, or collection procedures. These figures therefore have limited interpretability and should not be treated as a complete account of safety outcomes. Overall, the registry record supports a careful description of the study’s intent and design, but not a complete appraisal of its findings or their precision.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Preferences for Open Vs. Endovascular Repair for Abdominal Aortic Aneurysm — NCT03115346.
- Registry record last updated 2024-09-26; status last verified 2024-09.
- Results first posted 2024-09-26.
- 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
NCT03115346 is a completed, randomized, unmasked study sponsored by the VA Office of Research and Development. The registry lists enrollment of 241 and records a start date of 2017-04-15 and primary completion date of 2020-12-10. Its summary describes site-level assignment to a decision-aid intervention or a survey-only control condition for Veterans who were candidates for open or endovascular repair of abdominal aortic aneurysm.
The primary outcome is Veteran Preference for Repair Type, with a 30-day timeframe. Decision Regret Scale is listed as a secondary outcome. The registry notes omitted survey answers and cautions that simple individual-level re-analysis would be misleading because of the cluster-randomized design. Primary outcome results, the sponsor’s analysis, and detailed handling of missing data are not reported in the registry record.