Determination of Fractional Flow Reserve by Anatomic Computed Tomographic Angiography: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Determination of Fractional Flow Reserve by Anatomic Computed Tomographic Angiography (NCT01233518), a na na study. The primary outcome is Percentage of Participants With Diagnostic Accuracy of CCTA Plus FFRCT and CCTA Alone.

2. Clinical Question

This is a diagnostic-accuracy question, rather than a question about the effect of assigning participants to different treatments. The registry describes a single-group study in which participants were to receive CCTA, invasive coronary angiography, FFR, and FFRCT according to the protocol. It does not describe a randomized control group. The relevant comparison is between diagnostic classifications and the reference standard, not between outcomes in separate treatment arms.

The endpoint description specifies that the participant-level analysis used the vessel with the most adverse clinical status to represent each participant. FFR and FFRCT were measured on a continuous scale and then classified at the stated 0.80 threshold; CCTA stenosis classification used the stated 50% threshold. The registry record does not report the resulting diagnostic accuracy estimates, confidence intervals, or the numbers classified correctly and incorrectly. It therefore does not establish from the posted results how closely either diagnostic approach agreed with invasive FFR.

3. Trial Design

The registry describes a prospective, multicenter diagnostic study of CT-FLOW, an investigational technology for noninvasive assessment of coronary lesion significance using CCTA. The listed condition is coronary artery disease. Enrollment was 285, and the planned setting included up to 20 centers in the United States, Canada, Europe, and Asia. The study is marked completed; the recorded start was 2010-10 and primary completion was 2011-11.

The design is listed as single group, with allocation recorded as not applicable and no masking. The arm description says participants would receive CCTA, invasive coronary angiography, FFR, and FFRCT per protocol. These features fit a diagnostic evaluation in which test results are assessed against a reference method. They do not constitute a randomized treatment comparison, and the single arm should not be read as a control-versus-treatment contrast.

The primary outcome was diagnostic accuracy for CCTA plus FFRCT and CCTA alone at the participant level, with invasive FFR as the reference standard. The stated timeframe was 1 day. A secondary outcome is described as subject-level diagnostic performance of CCTA plus FFRCT. The registry lists two outcome measures, but does not provide analysis records.

The trial is listed as not applicable for phase. HeartFlow, Inc. is named as sponsor. The registry description characterizes the study as prospective and multicenter, but details of recruitment, the number of centers that actually enrolled participants, and the sequence or timing of all assessments beyond the stated endpoint timeframe are not reported in the registry record. The protocol's full statistical analysis plan and sample-size justification are also not reported in the registry record. The design description alone cannot establish how complete the analyzed dataset was or whether every enrolled participant contributed to every diagnostic comparison.

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

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 Percentage of Participants With Diagnostic Accuracy of CCTA Plus FFRCT and CCTA Alone, measured over 1 day.

The primary endpoint concerns the percentage of participants whose diagnostic classification for CCTA plus FFRCT and CCTA alone was accurate when compared with invasive FFR. The target condition was the presence or absence of at least one hemodynamically significant coronary artery stenosis. This participant-level framing asks whether the method correctly classifies a participant, rather than whether it correctly classifies each coronary vessel separately.

For the per-participant analysis, the vessel with the most adverse clinical status was selected to represent that participant. The description says FFR and FFRCT measurements were recorded continuously and dichotomized at 0.80: values ≤0.80 were considered diseased, while values >0.80 were described as nondiagnostic for ischemia. CCTA stenosis was classified at 50%, with stenoses ≥50% considered obstructive and stenoses <50% considered nonsignificant. These thresholds define how the diagnostic categories were formed; they are not estimates of diagnostic performance.

Diagnostic accuracy is used to summarize how often classifications correspond to the reference standard in the analyzed participants. A confidence interval conveys uncertainty around that estimate: a narrower interval indicates greater precision, while a wider interval indicates less precision. The registry specified a 95% confidence interval and stated that the prespecified primary endpoint was expected to be greater than 70% of its lower bound. The observed accuracy, the interval, and whether that criterion was met are not reported in the registry record.

Accuracy can be useful as an overall summary, but by itself it does not describe every type of classification error. The registry record does not report the underlying counts needed to distinguish positive from negative classifications or to assess sensitivity and specificity.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The statistical ideas involved in this trial are those of diagnostic accuracy, classification against a reference standard, and uncertainty around an estimated proportion. The registry gives the endpoint definition and the intended thresholds, but does not report the numerical primary outcome or a full analysis plan. The explanations below are general unless explicitly tied to the registry description.

What does accuracy measure?

In general, diagnostic accuracy is the proportion of assessed participants whose test classification agrees with the reference classification. It combines correct positive and correct negative classifications into a single summary. That compact summary can be useful for an overall view, but it does not show whether errors were false positives or false negatives. Sensitivity and specificity separate those error patterns by considering positive and negative reference-standard groups, respectively. The registry record does not report these measures or the underlying classification counts.

In this study, invasive FFR was identified as the reference standard. The primary outcome concerns whether CCTA plus FFRCT and CCTA alone correctly determined the presence or absence of at least one hemodynamically significant stenosis at the participant level. The endpoint description says that the vessel with the most adverse clinical status represented a participant in the per-patient analysis. This participant-level rule matters because diagnostic performance can differ depending on whether the unit is a vessel or a participant. It also means the result should not be interpreted as an estimate of accuracy for every individual coronary vessel.

Why were measurements dichotomized?

General statistical explanation: a continuous measurement can be converted into a categorical decision by choosing a threshold. That makes it possible to classify results as positive or negative for a target condition and compare them with a reference standard. However, categorization discards information about how far a measurement lies from the threshold. Classification near a cutoff can be sensitive to measurement variability, and performance depends in part on the cutoff selected.

The registry describes FFR and FFRCT measurements as continuous before dichotomization at 0.80. Values ≤0.80 were considered diseased, while values >0.80 were called nondiagnostic for ischemia. It also specifies CCTA stenosis categories at 50%: stenoses ≥50% were considered obstructive and stenoses <50% nonsignificant. These are the study's classification conventions as described in the endpoint. The record does not provide additional detail about how borderline, unreadable, or unavailable measurements were handled.

How should the confidence interval be read?

General statistical explanation: a confidence interval is a range of values compatible with the observed estimate under the assumptions of the method used. It communicates precision, not certainty that a particular value is true. A narrower interval generally indicates a more precise estimate, while a wider interval indicates less precision. The width depends on factors including the number of observations and the variability of the data. A confidence interval does not correct for bias in how participants were selected, how measurements were made, or how missing results were handled.

The primary endpoint description specifies a 95% confidence interval and states that the prespecified primary endpoint was expected to be greater than 70% of the lower bound of that interval. The endpoint estimate, the interval limits, and the evaluation of that criterion are not reported in the registry record. Without those values, the stated criterion cannot be assessed. The record also does not describe the confidence-interval calculation method, such as the particular interval procedure used for a proportion.

What does the reference-standard comparison establish?

General statistical explanation: evaluating an index test against a reference standard assesses agreement with the chosen standard in the study sample. The reference standard is not automatically error-free; its role is to define the comparator for the analysis. Diagnostic-accuracy results are specific to the population studied, the way participants were selected, the test procedures, the threshold rules, and the prevalence and spectrum of disease in the sample.

Here, invasive FFR is the stated reference standard, and the diagnostic methods of interest are CCTA plus FFRCT and CCTA alone. The registry describes the comparison but does not give the results or details sufficient to assess the accuracy of the reference process, blinding of test readers, or handling of discordant and missing measurements. The study is listed as unmasked. The record does not clarify whether any masking was used in particular measurement or interpretation steps.

What does over-dispersion mean in general?

General explanation: over-dispersion occurs when observed variation is greater than a statistical model assumes. If a model treats observations as more independent or less variable than they actually are, standard errors may be too small and confidence intervals too narrow. Correlation among observations, such as multiple measurements from the same participant or clustering by study center, can create this concern. Appropriate methods may account for clustering or extra variability when the design and data require it.

The registry describes a multicenter trial and a participant-level primary analysis, but it does not report the analytic model or whether center-level clustering was considered. It also does not report whether multiple vessel-level observations entered any analysis. Therefore, no conclusion about over-dispersion or its treatment can be drawn from the registry record.

How do planning illustrations differ from trial results?

General explanation: a sample-size calculation estimates how many observations might be needed under specified assumptions to detect a target effect with a chosen statistical approach. Such an illustration is not an observed result and is not a measure of the probability that a completed study has found a true effect. The registry's planning assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. The posted assumptions characterize these as planning illustrations rather than the trial's own design.

Those assumptions do not change the trial's listed single-group diagnostic design, and they should not be treated as evidence that the study used two randomized arms. No specific planning sample-size values or original sample-size justification are reported in the registry record. A planning calculation cannot replace the observed diagnostic estimates, their confidence intervals, or a description of the actual analysis population.

10. Multiplicity, Missing Data and Analysis Populations

Multiplicity concerns arise when a study evaluates multiple outcomes, comparisons, time points, or analysis populations. In general, conducting many statistical tests can increase the chance of a nominally notable result occurring by chance. A prespecified hierarchy, adjustment method, or other strategy may address that issue, but the appropriate approach depends on the study's objectives and analysis plan.

The registry lists two outcome measures: the primary diagnostic-accuracy outcome and a secondary outcome described as subject-level diagnostic performance of CCTA plus FFRCT. It does not report a multiplicity strategy, an outcome hierarchy beyond identifying the primary endpoint, or numerical results for either outcome. The number of analysis records is listed as zero. Whether additional outcomes or comparisons were assessed outside the registry entries is not reported in the registry record.

Missing-data handling is also not described. The record does not state how participants with incomplete CCTA, FFRCT, angiography, or invasive FFR results were treated, whether all tests were successfully interpreted, or how unavailable measurements affected denominators. The enrolled total is listed as 285, while the safety listing gives 252 participants at risk; the reason for that safety denominator is not reported in the registry record. The primary endpoint's analyzed denominator is not reported.

The analysis populations are not defined in enough detail to determine whether the primary result was based on all enrolled participants, participants with complete paired test results, or another group. Details about exclusions, protocol deviations, and sensitivity analyses are not reported in the registry record. Without that information, the impact of missing data and the comparability of analysis populations cannot be assessed.

11. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: retroperitoneal bleed (0.4%), hypotension (0.4%), septic shock (0.4%).

The registry lists a single arm for safety, with 252 participants at risk. It reports 2 serious adverse events and 0 other adverse events. The top serious-event terms are retroperitoneal bleed, hypotension, and septic shock, each listed as 0.4%. The death count is not reported in the registry record.

These figures describe the adverse-event entries as recorded; they do not establish whether an event was caused by a study procedure or by the investigational technology. The registry information does not provide event narratives, severity details beyond the serious-event category, timing, attribution, or the number of affected individuals for each term. It also does not clarify whether a participant could contribute more than one event or appear under more than one term. The listed percentages should therefore be read as registry entries, not as a complete account of individual-level risk.

The safety denominator differs from the enrollment count, but the reason is not reported in the registry record. The adverse-event information is distinct from the diagnostic-accuracy endpoint and cannot establish whether CCTA plus FFRCT or CCTA alone agreed with invasive FFR. It should be interpreted as a limited registry safety summary rather than a comparative safety analysis.

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

The registry describes a prospective, multicenter investigation of diagnostic accuracy in coronary artery disease. Its primary question is whether CCTA plus FFRCT and CCTA alone correctly classify the presence or absence of at least one hemodynamically significant coronary stenosis, using invasive FFR as the reference standard. It also provides the classification thresholds, the participant-level analysis rule, and the intended confidence-interval framework.

What the posted information does not provide is the primary accuracy estimate, the confidence interval, the analyzed participant count, or the classification counts. The secondary outcome's numerical findings are also not reported in the registry record. Thus, the registry does not show how accurate either approach was in the analyzed sample, how precise any estimate was, or whether the stated expectation for the lower confidence bound was met. Enrollment and endpoint definitions cannot answer those questions.

Even a reported accuracy value would need to be interpreted within the study's diagnostic context. General statistical explanation: performance measured against a reference standard applies to the participants and procedures studied; it does not automatically carry over to other populations or settings. An overall accuracy measure also does not reveal the balance between false-positive and false-negative classifications. That distinction requires additional results, such as sensitivity and specificity or the underlying classification counts, which are not reported in the registry record.

The design is not a randomized comparison of treatment groups. The study is listed as a single group, with no masking, and the primary objective is diagnostic assessment rather than estimation of a treatment effect. The record therefore does not support claims about comparative clinical outcomes or treatment benefit. The safety counts are limited registry entries and do not fill the gap in diagnostic results.

The most defensible reading is that the registry records the intended diagnostic comparison and its methods, but does not post the primary numerical findings needed to judge performance. Any interpretation beyond that would require results and analysis details not reported in the registry record.

13. Why This Trial Matters Statistically

Diagnostic studies require more than a plausible test and a stated threshold: they need a clearly defined target condition, an explicit reference standard, and a transparent account of how test results are classified and analyzed. This registry supplies several of those design elements. It identifies invasive FFR as the reference standard, distinguishes CCTA alone from CCTA plus FFRCT, states the thresholds used, and specifies a participant-level endpoint based on the vessel with the most adverse clinical status.

Those choices make the intended statistical question understandable. The use of a participant-level outcome also clarifies that the aim was not simply to count correct classifications across all vessels. The stated 95% confidence interval and the prespecified expectation concerning its lower bound indicate that precision was part of the endpoint framework, although the registry does not report the estimate or interval needed to evaluate it.

The record also illustrates why design information and results must be considered separately. A completed study with enrollment of 285 may have a clearly described objective, but the primary diagnostic estimates, confidence limits, and analysis denominator remain necessary for judging the evidence. Here, those results are not reported in the registry record. The absence of analysis records and limited detail on missing data further restricts statistical interpretation.

Finally, the design distinction matters. This is listed as a single-group diagnostic study, not a randomized comparison of treatment strategies. Its statistical value lies in the intended assessment of classification against a reference standard, not in estimating effects between intervention arms. The record makes that aim visible while leaving the numerical performance evidence unavailable.

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 most important limitation for statistical interpretation is that the registry does not report the primary diagnostic-accuracy estimate or its 95% confidence interval. It also does not report whether the prespecified expectation based on the lower confidence bound was met. Without these values, the primary result's magnitude and precision cannot be evaluated.

The primary analyzed denominator and the counts of correct and incorrect classifications are not reported. Consequently, the posted information does not permit examination of the balance of false-positive and false-negative results, or separate calculation of sensitivity and specificity. Details of missing measurements, exclusions, and the analysis population are also not reported in the registry record.

The study is listed as single group and unmasked. The registry does not describe masking procedures for test interpretation, the sequence and timing of all assessments, or the handling of discordant measurements. It lists a multicenter design and up to 20 planned centers, but the number of centers that actually enrolled participants and any approach to accounting for center differences are not reported. These omissions prevent assessment of possible sources of measurement or center-related variability.

The registry lists two outcome measures but provides zero analysis records. It does not report a multiplicity plan, a full statistical analysis plan, or an original sample-size justification. Although planning assumptions are described, no specific planning estimates are reported in the registry information. The safety denominator is 252 compared with enrollment of 285, and the reason for this difference is not reported. Together, these gaps limit evaluation of both the completeness of the analyses and the generalizability of any findings.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Determination of Fractional Flow Reserve by Anatomic Computed Tomographic Angiography — NCT01233518.
• Registry record last updated 2021-02-09; status last verified 2020-12.
• Results first posted 2021-01-22.
• 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

NCT01233518 is a completed, prospective, multicenter, single-group diagnostic study sponsored by HeartFlow, Inc. It enrolled 285 participants and assessed CCTA plus FFRCT and CCTA alone against invasive FFR for identifying hemodynamically significant coronary stenosis. The primary endpoint was participant-level diagnostic accuracy, with a stated 1 day timeframe and a 95% confidence interval.

The registry record describes the diagnostic thresholds and analysis concept but does not report the primary accuracy estimate, confidence interval, analyzed denominator, or secondary outcome results. It lists two outcome measures and zero analyses. The safety listing reports 2 serious adverse events among 252 participants at risk and 0 other adverse events; the death count is not reported in the registry record.