Dual Energy CT as a Noninvasive Method to Screen for Gastroesophageal Varices: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dual Energy CT as a Noninvasive Method to Screen for Gastroesophageal Varices (NCT03910413), a na na study. The primary outcome is To Measure the Accuracy of Dual Energy CT for Detecting Any Varices and High-risk Varices in Patients With Cirrhosis Presenting for Upper Gastrointestinal Endoscopy..

2. Clinical Question

The clinical question concerns diagnostic classification, rather than whether a treatment changes a health outcome. The planned CT grades were no varices, small varices, and large or high-risk varices. Endoscopy used corresponding grades as the reference. This makes agreement between the two assessments central to the intended question: CT classifications would be judged against endoscopic classifications, with particular interest in identifying any varices and the large or high-risk category.

The registry describes the motivation for evaluating a noninvasive imaging approach in the setting of variceal screening. Its summary notes limitations of endoscopy and describes potential information available from CT. Those statements explain the rationale for the study; they are not findings from this trial. The key distinction is between a reason to investigate a test and evidence about how well the test performed in the enrolled study.

Although the question was specified, the registry says the primary outcome was not collected after premature termination due to covid. No diagnostic accuracy estimate or comparison of CT with endoscopy is reported in the registry record. The intended clinical question therefore remains unanswered by the posted study results.

3. Trial Design

The registry describes NCT03910413 as a single-group diagnostic-test study of Dual Energy CT for gastroesophageal varices. The intervention is listed as “Duel Energy CT” in the registry. The study was sponsored by the University of Alabama at Birmingham. Allocation is recorded as NA, masking as none, and the study status as terminated. The phase is listed as NA.

The registry lists enrollment of 11. The recorded start date is 2019-06-05, and the primary completion date is 2019-09-25. These dates describe the record; they do not establish that the planned diagnostic outcome was completed or analyzed. The registry explicitly states that the study ended prematurely due to covid and that the primary outcome data were not collected.

Endoscopy was the reference standard for grading varices. The intended CT assessment was to occur no more than 2 weeks from endoscopy. The outcome description defines three grades for both assessments: 0 for no varices, 1 for small varices, and 2 for large or high-risk varices. The posted record does not report results for these categories.

This design does not include a randomized control arm or a treatment comparison. Instead, it was intended to evaluate a diagnostic test against a reference assessment. The registry lists one group for the intervention and does not describe a separate comparator arm. Thus, the important comparison was between the CT and endoscopic classifications, rather than between two assigned treatment groups.

The record does not report recruitment details beyond the listed enrollment, participant characteristics, reasons for individual nonparticipation, or a statistical analysis plan. It also does not report a planned enrollment target or why enrollment ended at the recorded level. These omissions limit interpretation of the conduct and intended analysis, independently of the fact that the outcome data were not collected.

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

Analysis population for the primary outcome, as defined in the registry: The study was ended prematurely due to covid. This data was not collected.

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.

6. Primary Endpoint

The registry lists the primary outcome as To Measure the Accuracy of Dual Energy CT for Detecting Any Varices and High-risk Varices in Patients With Cirrhosis Presenting for Upper Gastrointestinal Endoscopy., measured over DECT will be no more than 2 weeks from the time of endoscopy.

The primary endpoint was intended to assess the accuracy of Dual Energy CT for detecting any varices and large or high-risk varices in patients with cirrhosis presenting for upper gastrointestinal endoscopy. The registry specifies endoscopic grading as the reference standard and planned the CT assessment no more than 2 weeks from endoscopy.

Both assessments were to classify findings as 0 for no varices, 1 for small varices, and 2 for large or high-risk varices. In the outcome description, small varices are described as less than 5 mm and large varices as at least 5 mm. The endpoint therefore depended on comparing CT grades with endoscopic grades. Diagnostic measures such as sensitivity describe how often a test identifies people who have the condition according to the reference standard; specificity describes how often it identifies people who do not. These are general explanations of diagnostic statistics, not results reported for this study.

Such an endpoint is useful for evaluating whether an imaging test classifies findings similarly to the chosen reference assessment. However, the registry states that the outcome data were not collected because the study ended prematurely due to covid. It reports no endpoint values, diagnostic accuracy estimates, or grade-by-grade comparison. The planned endpoint therefore cannot be interpreted as an observed measure of test performance.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Diagnostic accuracy studies ask how well an index test classifies a condition when judged against a reference standard. In this study, Dual Energy CT was the index test and endoscopic grading was the reference standard. This is a general explanation of the design. The registry reports the intended comparison but does not report the outcome data needed to calculate accuracy.

What does a reference standard do?

A reference standard provides the basis for deciding which test classifications count as correct for the study question. Here, the registry identifies endoscopy as that standard. CT and endoscopy were each to use the grades 0, 1, and 2, representing no varices, small varices, and large or high-risk varices. The planned interval was no more than 2 weeks. Keeping assessments close in time can matter in general because diagnostic comparisons are most interpretable when the condition being assessed is not separated by a long interval. That general point does not establish what happened to participants in this study, and the record does not report completed paired assessments.

How is accuracy usually expressed?

For a binary diagnostic question, sensitivity is the proportion of reference-positive cases that the test identifies as positive. Specificity is the proportion of reference-negative cases that the test identifies as negative. Positive and negative predictive values describe the probability of the condition among test-positive and test-negative results, respectively. These concepts are general; the registry does not report values for any of these measures.

The outcome description includes graded categories and asks about both any varices and large or high-risk varices. A graded test can be evaluated by examining agreement across categories, or by defining a particular category as positive for a focused question. The exact analysis choices matter: they determine how intermediate grades are treated and which classification errors count for each question. The registry does not report the planned statistical definitions for these analyses. It would therefore be inappropriate to assume a specific thresholding or calculation method beyond the categories and questions stated in the endpoint description.

Why is uncertainty important?

An accuracy estimate is a summary of observed classifications, while a confidence interval describes the statistical uncertainty around that estimate under the analysis assumptions. A narrow interval generally indicates greater precision, and a wide interval indicates less precision. This is a general explanation. The registry reports no accuracy estimates or confidence intervals, so there is no trial-specific precision to interpret. Enrollment of 11 does not by itself reveal the value or uncertainty of an accuracy estimate, especially when the outcome data were not collected.

In diagnostic studies, the mix of participants and the frequency of findings can affect how well an estimate applies to other settings. Predictive values, in particular, depend on how common the condition is in the population being assessed. This is general statistical context, not a finding about this study. The registry does not report the observed distribution of endoscopic grades, so the prevalence of the relevant findings among assessed participants cannot be described.

What does the termination imply for analysis?

The registry states that the study ended prematurely due to covid and that the primary outcome data were not collected. Without those measurements, the planned diagnostic comparison cannot be performed from the posted results. The record lists no statistical analyses. It does not say whether an analysis plan was written, whether data were reviewed before termination, or whether any unreported analyses were conducted. Those details are not reported in the registry record.

Missingness can be handled in different ways depending on why data are missing and what was measured. General approaches may include describing the missing observations, defining analysis populations, or using statistical methods suited to the missing-data mechanism. Such methods cannot recreate a diagnostic comparison when the required outcome classifications were not collected. The registry does not report an imputation method, a missing-data strategy, or an analysis population for the primary endpoint.

What can safety counts tell us?

Safety summaries count recorded events among people considered at risk; they are not measures of diagnostic accuracy. The posted record lists 0 serious adverse events, 0 other adverse events, and 0 deaths among 11 at risk in the Dual Energy CT group. These figures describe the registry entries, but do not indicate the precision of an estimate of event risk or establish that no events could occur in other use. The record does not list serious adverse event terms.

The registry’s sample-size assumptions describe a planning illustration based on an observed effect, with equal allocation between two arms, a two-sided significance level of 0.05, and a normal approximation. Those planning details do not match the posted single-group diagnostic-test record in a way that establishes a trial-specific power analysis. The record does not report a design rationale, a sample-size justification, or the effect assumptions for the primary diagnostic question. Sample-size illustrations should be understood as planning scenarios, not as observed power or evidence about the uncollected endpoint.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure: accuracy of Dual Energy CT for detecting any varices and large or high-risk varices against endoscopic grading. The outcome description contains more than one diagnostic question, but the record does not report how these questions were prioritized for statistical testing, whether separate tests were planned, or how multiplicity would be handled. The statistical analysis plan is not reported in the registry record.

The primary outcome data were not collected, according to the registry, after the study ended prematurely due to covid. This is not a case where a reported endpoint result can be assessed alongside a specified missing-data adjustment. The record does not describe which individual assessments, if any, were missing, nor does it report an analysis population for the primary endpoint. No imputation method, complete-case rule, or sensitivity analysis for missing outcome data is reported.

The record lists no statistical analyses. It does not report interim analyses, stopping rules, a prespecified analysis plan, or adjustments for multiple comparisons. These details are not reported in the registry record; their absence should not be interpreted as evidence that a particular method was or was not used outside the posted record.

Because the endpoint was not collected, there are no reported estimates, p-values, or confidence intervals to evaluate for multiplicity or missing-data effects. The posted record therefore supports a description of what was planned and what was not collected, but not an assessment of whether statistical adjustments would have changed an outcome conclusion.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The registry’s safety entries for the Dual Energy CT group list 11 at risk, 0 affected by serious adverse events, 0 affected by other adverse events, and 0 deaths. The record lists no serious adverse event terms. These are descriptive entries for the reported safety fields; they should not be read as results about the primary diagnostic endpoint.

A zero in a safety summary means that the registry reports no affected cases in that field. It does not establish that the risk is absent, and it does not provide a precise estimate of how frequently an event might occur outside the recorded group. The registry record does not provide additional detail about safety surveillance, event definitions, follow-up duration, or the collection process. Those details are not reported in the registry record.

The listed safety counts also do not resolve the absence of the primary outcome. The diagnostic question required CT and endoscopy grading data, while the safety fields concern adverse events and deaths. The registry states that the primary outcome was not collected. Consequently, the safety entries should be interpreted narrowly as the reported safety counts, not as evidence for or against diagnostic accuracy or as a complete account of clinical risk.

12. What the Result Does, and Does Not, Mean

The posted record establishes that a diagnostic study was planned: Dual Energy CT was to be compared with endoscopic grading for detection of varices in patients with cirrhosis. It also establishes that the study was terminated and that the registry says the primary outcome data were not collected. The record does not establish how accurately CT classified varices in this study.

There is no reported agreement measure, sensitivity, specificity, predictive value, or other diagnostic estimate. No confidence interval is available to show the precision of an estimate, because no primary diagnostic estimate is reported. It is therefore not possible to state whether CT classifications matched endoscopic classifications, how often they differed, or whether performance varied across the specified grades.

The outcome’s absence should not be interpreted as evidence that the test performed poorly. Nor should the background rationale or descriptions of CT’s potential advantages be treated as proof that it performed well in this trial. Both a positive and a negative conclusion about diagnostic accuracy would go beyond what the posted results show.

The safety entries are limited to the registry’s recorded counts: 0 serious adverse events, 0 other adverse events, and 0 deaths among 11 at risk. Those counts do not answer the diagnostic question and do not demonstrate that an event is impossible. The registry does not report enough detail to characterize the safety monitoring or the precision of safety estimates.

The most defensible conclusion is about the evidence available in the record: the study had a stated diagnostic objective and a named reference standard, but the primary outcome was not collected. The registry record therefore does not provide an evaluable estimate of diagnostic accuracy. It also does not report the statistical plan or further analysis details that could clarify how the intended results would have been summarized.

13. Why This Trial Matters Statistically

This study is statistically relevant because diagnostic-test evaluation depends on having paired information from the test under evaluation and a reference standard. The registry identifies Dual Energy CT as the test and endoscopic grading as the reference. It specifies corresponding categories and a planned interval of no more than 2 weeks. These features define the intended comparison, but they do not substitute for the observed classifications required to estimate accuracy.

The study also illustrates an important distinction between a protocol question and a result. A precise endpoint description can identify what investigators intended to measure, including detection of any varices and large or high-risk varices. Yet without collected endpoint data, neither accuracy nor uncertainty can be quantified from the posted record. The enrollment entry of 11 cannot stand in for the missing diagnostic comparison.

For readers evaluating evidence, this distinction guards against treating background statements, trial objectives, or safety counts as outcome findings. The registry’s rationale describes why a noninvasive method was of interest. Its safety fields describe recorded adverse-event counts. Neither provides the CT-versus-endoscopy classifications needed to answer the primary diagnostic question.

The record also highlights why transparency about missing outcomes and analysis plans matters. Here, the registry explicitly states that the primary outcome was not collected, while the statistical analysis plan and handling of missing data are not reported in the registry record. This leaves no basis for judging the intended precision, analysis populations, or any planned treatment of multiple diagnostic questions.

Accordingly, the statistical importance of this record is not a measured level of diagnostic performance. It is the clear separation between a planned assessment and an analyzable result: the study documents the former, while the latter is absent from the posted results.

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 is that the registry states the primary outcome was not collected. Without CT and endoscopic grading results, the planned diagnostic accuracy question cannot be answered from this record. The study status is terminated, and the registry attributes premature ending to covid.

The registry lists enrollment of 11, but does not report a planned enrollment target, a rationale for the enrollment level, or a sample-size justification for the diagnostic endpoint. It does not report participant characteristics or the distribution of endoscopic grades. Consequently, the record does not allow assessment of the population represented or the frequency of the target findings among those assessed.

The statistical analysis plan is not reported in the registry record. The record does not specify how the questions of detecting any varices and large or high-risk varices would be analyzed, which diagnostic measures would be emphasized, or how uncertainty would be presented. It also does not report analysis populations, missing-data methods, or whether there were interim analyses.

The planned interval between CT and endoscopy was no more than 2 weeks, but the record does not report whether completed paired assessments were available before termination. It provides no diagnostic results that could be checked against this planned timing. The safety summary is also limited: event counts are listed, but definitions, follow-up details, and surveillance procedures are not reported in the registry record.

These limitations are specific to the posted record and should not be extended into claims about unreported study conduct. They mean that diagnostic performance, its precision, and the generalizability of any result cannot be evaluated from the registry results.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Dual Energy CT as a Noninvasive Method to Screen for Gastroesophageal Varices — NCT03910413.
• Registry record last updated 2023-08-14; status last verified 2023-08.
• Results first posted 2023-08-14.
• 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.

18. Record Summary

NCT03910413, titled “Dual Energy CT as a Noninvasive Method to Screen for Gastroesophageal Varices,” is registered as a single-group diagnostic-test study sponsored by the University of Alabama at Birmingham. The intervention is Dual Energy CT, and endoscopic grading was the planned reference standard. The registry lists enrollment of 11, a start date of 2019-06-05, a primary completion date of 2019-09-25, and status as terminated.

The primary outcome was intended to assess accuracy for detecting any varices and large or high-risk varices, using matching grades for CT and endoscopy. The registry states that the study ended prematurely due to covid and that the primary outcome data were not collected. No statistical analyses or diagnostic accuracy results are reported.

The safety entries list 0 serious adverse events, 0 other adverse events, and 0 deaths among 11 at risk in the Dual Energy CT group. The statistical analysis plan, sample-size justification, and missing-data methods are not reported in the registry record.