Evaluation of a New Cardiac Biomarker Assay: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of a New Cardiac Biomarker Assay (NCT01608100), a na phase 3 study. The primary outcome is Clinical Performance - Area Under the Curve.

2. Clinical Question

The clinical question is about diagnostic performance, not whether a treatment changes a health outcome. The stated objective is to evaluate the performance and intended use of a new cardiac biomarker test, with assay results compared against whether an acute myocardial infarction occurred. Prognosis is listed among the secondary outcomes, but further details of that outcome are not reported in the registry record.

The target population is described only in broad terms as an intended-use population. The record does not provide details such as eligibility criteria, recruitment setting beyond presentation to the emergency department, or the characteristics of participants. Those omissions limit how specifically readers can judge whether the enrolled participants represent the people for whom the assay is intended.

3. Trial Design

This is a completed Phase 3, single-group device study with enrollment of 1101. The registry record lists the sponsor as Abbott Diagnostics Division, a start date of 2012-06, and a primary completion date of 2014-08. Allocation is NA and masking is none. The intervention is testing with the ARCHITECT STAT High Sensitive Troponin I Assay; all subjects in the described group had blood tested with the investigational assay.

This is not a randomized comparison of treatment groups. The registry record itself notes that the arms are not a like-for-like control comparison. Instead, the study evaluates diagnostic performance by comparing assay results with the diagnosis of whether acute myocardial infarction occurred. The diagnosis serves as the reference for assessing assay performance, rather than as a second study arm receiving a different intervention.

Specimens are described at three collection time points measured from presentation to the emergency department: 0-2 hours, >2 and up to 4 hours, and >4 hours and up to 9 hours. Assay results were generated from specimens in K2 EDTA, Lithium Heparin Separator, and Serum Separator tubes. The registry notes that not all time points or tube types could be collected for all subjects. For analysis, the number of subjects depended on having at least one collection time point in the relevant tube type.

The registry lists six outcome measures and 63 analyses. Those counts indicate that the record contains multiple outcome or analysis entries, but do not, on their own, identify which entries are primary, how they relate to one another, or which results were prespecified. Detailed protocol information, including the statistical analysis plan and any sample-size justification, is not reported in the registry record.

General design context: In a single-group diagnostic evaluation, the main evidence concerns how test measurements relate to a reference classification. It does not provide the treatment-effect interpretation that would come from comparing randomized intervention groups. The study design and the endpoint definition therefore need to be kept in view when interpreting the reported AUC.

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 number of subjects for analysis was determined by the number of subjects with at least one collection time point in the respective tube type (K2 EDTA, Lithium Heparin Separator, Serum Separator). Not all time points or collection tube types were able to be collected for all subjects.

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 Clinical Performance - Area Under the Curve, measured over Troponin value from three collection time points (0-2 hours, >2 and up to 4 hours, >4 hours and up to 9 hours) from subject presentation to the emergency department.

The primary endpoint is titled “Clinical Performance - Area Under the Curve.” The registry describes the AUC as a measure used to evaluate the assay’s clinical performance and to determine its optimum clinical sensitivity and specificity. It was assessed using Troponin-I assay results from specimens collected at the stated time points and in the three stated tube types. The unit is listed as probability, and the dispersion measure is a 95% confidence interval.

General statistical context: An ROC curve describes the trade-off between sensitivity and specificity across possible test thresholds. Its AUC condenses that curve into a summary of discrimination: a higher AUC generally indicates better separation of the reference-positive and reference-negative groups across thresholds. AUC does not, by itself, specify the sensitivity and specificity at a particular threshold, nor does it establish that a threshold is clinically appropriate. Those questions require threshold-specific results and the relevant clinical context.

The registry record does not report the primary AUC estimate or its 95% confidence interval. It also does not provide threshold-specific sensitivity or specificity values in the registry record for this analysis. Consequently, the endpoint’s definition can be explained, but its observed performance and precision cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary analysis is described as calculating an AUC for the ARCHITECT STAT High Sensitive Troponin-I assay. The registry gives the endpoint’s unit as probability and its dispersion as a 95% confidence interval. It says the objective of the AUC assessment is to evaluate clinical performance and inform optimum clinical sensitivity and specificity. The statistical methods used to calculate the AUC and interval are not reported in the registry record.

What does the AUC measure?

General statistical context: A receiver operating characteristic curve traces sensitivity against the false-positive rate as a test threshold varies. The AUC summarizes this curve across thresholds. Conceptually, it describes how well test values rank reference-positive cases above reference-negative cases. It is a measure of discrimination, not a direct measure of calibration, and it does not identify a single operating threshold. An AUC therefore cannot replace threshold-specific sensitivity and specificity when a particular decision threshold is under consideration.

The reference classification in this study is whether or not an acute myocardial infarction occurred. The registry says assay results were compared with that diagnosis, but does not describe the diagnostic criteria, timing, adjudication process, or blinding arrangements. The AUC is meaningful only in relation to the classification used as the reference. The registry record does not provide enough detail to assess the properties of that reference process.

What does the confidence interval contribute?

General statistical context: A confidence interval gives a range of values compatible with the data and the method’s assumptions, expressing sampling uncertainty around an estimate. It is not a guarantee that the true value lies in the interval, and it does not address all sources of bias. For this study, the stated interval level is 95%, but the interval itself and its calculation method are not reported in the registry record. Without the estimate and interval, neither the observed AUC nor its precision can be described.

Why are collection time and tube type relevant?

The primary endpoint description concerns results from three collection windows after emergency-department presentation: 0-2 hours, >2 and up to 4 hours, and >4 hours and up to 9 hours. It also identifies K2 EDTA, Lithium Heparin Separator, and Serum Separator specimens. These are dimensions of the assay evaluation, not separate treatment arms. The record says not all collection times and tube types were available for all participants, and that the analysis population for a tube type depended on having at least one time point in that tube type.

General statistical context: If repeated measurements from the same participant contribute to analyses, those observations may be statistically related. An analysis may need to account for that dependence when estimating uncertainty. The registry entries do not say whether the AUC analyses combined observations across collection times, estimated separate AUCs, or accounted for within-participant dependence. The precise analytic approach is not reported in the registry record, so it should not be inferred from the endpoint description alone.

What analysis details are unavailable?

The registry record does not describe the AUC estimator, confidence-interval method, handling of ties or missing measurements, or the criteria used to classify acute myocardial infarction. It does not report cutoffs, threshold-specific operating characteristics, covariate adjustment, or any subgroup analysis method. It also does not state whether the reported analyses were prespecified, whether there were interim analyses, or how multiple analyses were handled. These omissions prevent a complete assessment of the analysis plan.

The number of listed analyses is 63, and the number of outcome measures is six. Those counts do not establish how many AUC estimates were calculated, whether analyses were independent, or which analyses were treated as confirmatory. The registry’s primary endpoint description identifies AUC as the measure of interest, but the results for that endpoint are not reported in the registry record.

General distinction between analysis and design: An analysis method addresses how observed data are converted into an estimate and uncertainty measure. It cannot, by itself, correct every design limitation, such as an unclear reference standard or incomplete specimen availability. Interpretation should therefore consider the endpoint definition, participants contributing data, specimen collection, reference classification, and analysis method together. For this study, several of those details are not reported in the registry record.

No recalculated AUC, interval, or other effect estimate is available in the registry entries for this account. There is consequently no recalculation to compare with a sponsor analysis, and no numerical basis for describing the magnitude of performance. Any sample-size illustration associated with observed effects would be a planning illustration under its stated assumptions, not observed power; no such numerical illustration is reported in the registry entries used.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists six outcome measures and 63 analyses. It identifies the primary endpoint as clinical performance measured by AUC, but does not report a complete statistical analysis plan, the hierarchy among analyses, or a multiplicity adjustment strategy. It is therefore not possible to determine from the record whether the listed analyses were all prespecified, which were confirmatory, or how conclusions were selected across them. The counts should not be interpreted as evidence that every analysis tested a distinct hypothesis.

General statistical context: When many hypotheses are examined, the chance of obtaining apparently notable results by chance can increase. A prespecified hierarchy or adjustment procedure can help control that issue, but the appropriate approach depends on the objectives and analysis plan. Since no such plan is reported in the registry record, no claim about multiplicity control is justified. The missing primary AUC estimate also means that there is no reported primary result to compare with any other analyses.

Missing specimen availability is explicitly relevant to the analysis population. The registry says a subject was counted for an analysis according to having at least one collection time point in the relevant tube type, and that not all time points or tube types were collected for all subjects. It does not report how many participants contributed to each analysis, why specimens were unavailable, or whether any imputation or other missing-data method was used. Those details are not reported in the registry record.

General statistical context: Missing measurements can affect both precision and the population represented by an estimate. If specimen availability is associated with participant characteristics or clinical status, the analyzed subset may differ from the enrolled group. That is a general concern, not a claim that such an association occurred in this study. The registry entries do not allow the pattern or consequences of missingness to be assessed.

The record also does not describe the analysis population for the secondary prognosis outcome, or the methods used for that outcome. Accordingly, population definitions, missing-data handling, and multiplicity procedures remain unclear beyond the specimen-availability description for the primary evaluation.

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 0% reporting threshold.

The registry record lists 1101 participants at risk for serious adverse events and 0 affected. It also lists 1101 participants at risk for other adverse events and 0 affected. The number of deaths is not reported in the registry record. No top serious adverse-event terms are listed.

These entries are counts in the adverse-event reporting fields. They should be read as the recorded figures, not as proof that no event occurred outside the scope of collection or reporting. The record does not provide event definitions, surveillance procedures, reporting windows, severity criteria, or detail about how events were attributed. It also does not describe whether the absence of listed serious-event terms reflects no reported terms or a more limited reporting process.

General statistical context: A count of zero is not the same as evidence that risk is impossible. Safety interpretation depends on how events were sought and defined, the duration of observation, and completeness of reporting. Those details are not reported in the registry record, so the figures do not support a broader estimate of safety risk.

The study evaluates an assay in a single group rather than comparing randomized intervention groups. The adverse-event counts therefore do not provide a controlled comparison of event rates between different interventions. The separate missing death count should remain distinct from the zero values listed for the two adverse-event categories.

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

The registry record describes a diagnostic-performance study of the ARCHITECT STAT High Sensitive Troponin I Assay. It specifies an AUC endpoint, a 95% confidence interval as the dispersion measure, specimen collection windows, and three tube types. It also reports enrollment of 1101 and says that not all time points and tube types were available for every participant. These registry entries describe the intended evaluation and its data-collection context.

The record does not report the primary AUC estimate or its 95% confidence interval. It therefore does not provide the numerical evidence needed to characterize the assay’s discrimination or the precision of that estimate. No conclusion about the magnitude of diagnostic performance, threshold-specific sensitivity, or threshold-specific specificity can be drawn in the registry record. Missing numerical results should not be interpreted as a null result.

The design is single-group, with assay results compared against whether acute myocardial infarction occurred. This is not a randomized trial comparing treatments, and it cannot establish a treatment effect. The registry’s account also does not explain how the reference diagnosis was determined, so the strengths and limitations of that classification cannot be fully assessed.

General statistical context: A diagnostic-performance measure is one part of the evidence needed to judge a test. The population, reference classification, threshold-specific performance, uncertainty, and specimen handling all affect interpretation. AUC summarizes discrimination across thresholds but does not say what performance is achieved at a specific threshold or how a result should be used in clinical care.

The record reports zero affected participants in the serious and other adverse-event fields, with 1101 at risk in each field. The death count is not reported in the registry record. These entries are limited to the registry’s reported counts and do not supply the event definitions or monitoring details needed for a comprehensive safety assessment.

In short, the record is informative about the study question and broad design, but insufficient for judging the primary numerical finding. Its limitations are especially important when comparing analyses across different time points or tubes, because participant availability may vary and analysis-specific counts are not reported.

13. Why This Trial Matters Statistically

This study raises a useful statistical distinction: evaluating a diagnostic test is not the same as comparing treatments. The single-group design evaluates how assay measurements relate to a reference diagnosis, with AUC identified as the primary performance measure. Recognizing that structure helps prevent treatment-effect language from being applied to a diagnostic-accuracy question.

The study also illustrates why the analysis population matters. Specimen availability could differ across collection time points and tube types, and the record says not all of these were collected for all participants. If analyses use different subsets, estimates may not be directly comparable without knowing who contributed to each one. The registry does not report those analysis-specific counts or explain the missing-specimen process.

General statistical context: An AUC is useful as an overall discrimination summary, but it is not a complete account of test performance. Threshold-specific measures and their uncertainty address different questions. A confidence interval is also important because the point estimate alone does not show how precisely performance was estimated. Here, the endpoint and intended interval are described, but the AUC result and interval are not reported in the registry record.

The record lists 63 analyses and six outcome measures without giving the analysis hierarchy or multiplicity approach. That makes the distinction between primary and additional analyses important: counts of analyses do not establish which results were planned to support a primary conclusion. The record’s identification of an AUC primary endpoint is helpful, while its lack of numerical results and analysis-plan details limits interpretation.

Finally, the registry entry makes clear that adverse-event counts and diagnostic results answer different questions. The reported safety fields show zero affected for the listed event categories, while the primary diagnostic result is absent. Keeping those details separate avoids treating safety counts as evidence about diagnostic accuracy or treating a missing efficacy estimate as a negative result.

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 consequential limitation for interpreting the registry entry is that the primary AUC estimate and its 95% confidence interval are not reported in the registry record. The endpoint is defined, but its numerical performance and precision cannot be assessed. Threshold-specific sensitivity and specificity results are also not reported in the registry record for this analysis.

The reference diagnosis is described as whether acute myocardial infarction occurred, but its criteria and determination process are not described. The record does not say whether diagnosis assessors were blinded to assay results, whether the diagnosis was adjudicated, or how uncertain classifications were handled. These details are relevant to interpreting diagnostic accuracy and remain unknown.

The intended-use population is described broadly, without detailed eligibility criteria or participant characteristics. Although enrollment is reported as 1101, analysis-specific participant counts are not. The population could vary by tube type because inclusion depended on having at least one collection time point in that type. The record also states that not all specimens could be collected for all participants but does not describe the reasons or handling of missing measurements.

There is no randomized control group, allocation is listed as NA, and masking is listed as none. The study therefore is not a like-for-like comparison between intervention arms. The record does not provide further detail on conduct that would clarify the potential for measurement or classification bias.

Although six outcome measures and 63 analyses are listed, the analysis hierarchy, prespecification, multiplicity handling, AUC estimation method, and confidence-interval method are not reported in the registry record. Sample-size justification and detailed statistical planning are also not reported. These limitations leave important questions about precision, analysis selection, and generalizability unresolved.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Evaluation of a New Cardiac Biomarker Assay — NCT01608100.
• Registry record last updated 2015-03-17; status last verified 2015-01.
• Results first posted 2015-03-17.
• 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

NCT01608100 is a completed Phase 3, single-group evaluation of the ARCHITECT STAT High Sensitive Troponin I Assay, sponsored by Abbott Diagnostics Division. The record lists enrollment of 1101, a start date of 2012-06, and primary completion in 2014-08. Allocation is NA and masking is none.

The primary endpoint is clinical performance measured by AUC, with a 95% confidence interval. The assay was evaluated using specimens from three collection windows and three tube types, with analysis populations depending on specimen availability. The primary AUC estimate and its confidence interval are not reported in the registry record. The record lists six outcome measures and 63 analyses, without reporting the statistical analysis plan or multiplicity approach.

The serious and other adverse-event fields each list 0 affected among 1101 at risk; the death count is not reported in the registry record. The entry supports a description of the study design and endpoint, but not a quantitative conclusion about the assay’s observed diagnostic performance.