Ultrasound Elastography of Breast Lesions: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ultrasound Elastography of Breast Lesions (NCT00716482), a registered phase not stated study. The primary outcome is Estimates of Effect of Selectively Upgrading BIRADS Category 3 and Downgrading BIRADS 4a Masses Based on SWE Features. Overall Specificity and Sensitivity of BI-RADS Score Using Conventional B-mode Ultrasound vs. B-mode + Certain SWE Characteristics.

2. Clinical Question

3. Trial Design

NCT00716482 is a completed study titled “Ultrasound Elastography of Breast Lesions.” Its official title describes an assessment of the clinical value of SuperSonic Shear Wave Elastography in ultrasound evaluation of breast lesions. The sponsor is SuperSonic Imagine. The recorded study start was 2008-07, and primary completion was 2010-06. Enrollment is listed as 1681.

The study concerns diagnostic imaging rather than a randomized comparison of treatments. Its primary outcome describes estimates associated with selectively upgrading BI-RADS category 3 masses and downgrading BI-RADS 4a masses based on SWE features. It also names overall sensitivity and specificity for conventional B-mode ultrasound compared with B-mode plus selected SWE characteristics. The registry record says the analysis was per protocol and based on lesions, with one lesion analyzed per participant. It does not report how participants or imaging assessments were allocated to approaches, and the record flags that the arms are not a like-for-like control comparison.

The conservative strategy used E-homogeneity, E-max and E-color to upgrade BI-RADS 3 lesions to BI-RADS 4a' or downgrade BI-RADS 4a lesions to BI-RADS 3'. The aggressive strategy used the same features but changed the classification of more lesions. The analysis description identifies 939 lesions, comprising 289 malignant and 650 benign masses. These analysis counts describe the primary outcome population; the record does not explain how that population relates to overall enrollment.

The registry lists secondary outcomes concerning qualitative intraobserver reproducibility of SWE homogeneity, intraobserver reliability of quantitative SWE measurements, and interobserver agreement for B-mode ultrasound and SWE features. It lists 4 outcome measures and 21 analyses. The detailed protocol, recruitment flow, imaging workflow and statistical analysis plan are not reported in the registry record.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Per protocol. Analysis includes 289 malignant and 650 benign masses. One lesion was analyzed per participant.

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: There were 1681 "started" participants and 1562 "completed" participants. Reasons for rejected participants are listed in "participant flow" section

6. Primary Endpoint

The registry lists the primary outcome as Estimates of Effect of Selectively Upgrading BIRADS Category 3 and Downgrading BIRADS 4a Masses Based on SWE Features. Overall Specificity and Sensitivity of BI-RADS Score Using Conventional B-mode Ultrasound vs. B-mode + Certain SWE Characteristics, measured over 2 years.

The primary endpoint is framed around estimates of how selected SWE features affect the classification of BI-RADS 3 and BI-RADS 4a masses, alongside overall sensitivity and specificity for conventional B-mode ultrasound versus B-mode combined with selected SWE characteristics. Its unit is participants, and its measure is recorded as a number. The analysis description, however, is expressed in terms of 939 lesions, with one lesion analyzed per participant.

The reference standard defines a positive result as malignant cytologic or histopathologic findings. A negative result is defined as a BI-RADS 2 lesion, a BI-RADS 3 lesion with benign histopathology, or a BI-RADS 3 lesion with a 1 year follow-up ultrasound showing resolved or decreased lesion size. The endpoint timeframe is listed as 2 years.

In general, sensitivity describes how often an assessment identifies cases classified as positive by the reference standard; specificity describes how often it classifies reference-standard negative cases as negative. These measures help characterize diagnostic classification, but they do not by themselves describe the consequences of a classification for an individual. The registry record gives the endpoint definitions and analysis population, but does not report the resulting sensitivity or specificity estimates, their confidence intervals, or the classification counts needed to assess them.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry presents a binary diagnostic endpoint, with the primary measure recorded as a number of participants and the clinical outcome described in terms of malignant or benign masses. In general, a binary outcome assigns observations to categories such as positive and negative. For diagnostic evaluation, the reference standard supplies the classification against which an imaging assessment is judged. Here, malignant cytologic or histopathologic results define positive cases. The negative category includes BI-RADS 2 lesions, BI-RADS 3 lesions with benign histopathology, and BI-RADS 3 lesions with a 1 year follow-up ultrasound showing resolved or decreased lesion size.

What do sensitivity and specificity represent?

In general, sensitivity is the proportion of reference-standard positive cases classified as positive by the test or assessment. Specificity is the proportion of reference-standard negative cases classified as negative. They describe different aspects of classification: a change in the threshold or rule used to label lesions can affect the balance between identifying positive cases and correctly classifying negative cases. A strategy that classifies more lesions into a higher BI-RADS category may have a different sensitivity-specificity profile from a more conservative strategy, but the direction and magnitude cannot be determined without the observed results.

The primary outcome focuses on selectively upgrading BI-RADS 3 masses and downgrading BI-RADS 4a masses using SWE features. E-homogeneity, E-max and E-color are named as the features used in both strategies. The aggressive strategy is described as making more upgrades and downgrades than the conservative strategy. These descriptions identify the decision rules at a broad level, but the registry record does not report the exact threshold rules or the number of lesions reclassified under either strategy.

Why does the reference standard matter?

Diagnostic measures depend on how positive and negative status is defined. A reference standard based on cytology or histopathology is not the same pathway as a negative classification based on imaging category or follow-up. Because the registry includes multiple routes to a negative reference classification, the resulting specificity describes performance relative to that composite definition. The record does not report counts by reference-standard route, missing or indeterminate classifications, or the procedures used to resolve disagreement between sources.

In general, a confidence interval expresses the uncertainty around an estimated quantity, given the data and statistical method used. Narrower intervals generally indicate greater precision under the method's assumptions, while wider intervals indicate less precision. The posted registry information does not include confidence intervals for sensitivity, specificity, or the effect of reclassification. Consequently, it is not possible to assess the precision of the endpoint estimates from this record.

What does the analysis population tell readers?

The endpoint description identifies a per-protocol analysis of 939 lesions, including 289 malignant and 650 benign masses, and states that one lesion was analyzed per participant. Per-protocol indicates that the reported analysis was defined around a protocol-adherent population, but the registry record does not describe the criteria used to determine adherence, exclusions, or the handling of incomplete assessments. Enrollment is listed as 1681; the record does not explain why the primary analysis includes the stated lesion population or how participants outside that analysis were handled. No additional analysis population or intention-to-diagnose approach is described in the available details.

The registry names secondary assessments of qualitative intraobserver reproducibility for SWE homogeneity, reliability of quantitative SWE measurements within an observer, and agreement between observers for B-mode ultrasound and SWE features. In general, reproducibility concerns whether repeated assessments by the same observer are consistent, while interobserver agreement concerns consistency across observers. The exact statistics, reading procedures and numerical results for these outcomes are not reported in the registry record.

The registry lists 4 outcome measures and 21 analyses. Multiple outcomes or analyses can raise multiplicity concerns because examining more hypotheses can increase the chance of a nominally notable result when no adjustment is made. Whether that concern was addressed depends on the prespecified analysis plan, the relationships among analyses, and any multiplicity procedure. None of those details is reported, so no claim can be made about adjustment or control of error rates.

The stated planning illustration is based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. These are planning assumptions, not a report of the trial's own design justification. Any sample-size figures based on them are planning illustrations, not observed power. The registry record does not report the original sample-size justification or establish that the study was designed around these assumptions.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 4 outcome measures and 21 analyses. That indicates several outcomes or analyses were recorded, but does not establish which were prespecified as primary, how they were ordered, or whether a multiplicity adjustment was used. In general, multiple statistical tests can increase the chance of obtaining a nominally noteworthy finding by chance. The appropriate interpretation depends on the analysis plan and the relationships among the tests. Those details are not reported in the registry record, so the presence or absence of multiplicity control cannot be determined.

The primary analysis is described as per protocol, with 939 lesions, including 289 malignant and 650 benign masses, and one lesion analyzed per participant. The registry does not define protocol adherence, provide a participant or lesion flow, or explain how the analysis population relates to enrollment of 1681. It also does not state how incomplete imaging, missing reference-standard information, indeterminate assessments, or exclusions were handled. These items are not reported in the registry record.

The outcome description defines positive and negative reference standards, including a negative classification based on a 1 year follow-up ultrasound showing resolved or decreased lesion size. It does not give counts for each reference-standard pathway or describe how unavailable follow-up was treated. Without those details, readers cannot assess whether missingness could affect the diagnostic estimates.

The registry does not report an intention-to-diagnose analysis, sensitivity analyses, prespecified subgroup analyses, interim analyses, or a detailed statistical analysis plan. Their use should not be assumed. The posted details therefore support a description of the endpoint and analysis population, but not a full appraisal of missing-data handling or multiplicity procedures.

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

The record establishes that the study asked whether selected SWE characteristics could inform BI-RADS reclassification and the overall sensitivity and specificity of breast ultrasound assessment. It identifies the features used, the conservative and aggressive strategies, the endpoint timeframe, and the reference-standard definitions. It also describes a per-protocol analysis involving 939 lesions, including malignant and benign masses.

Those design details do not establish the direction or size of diagnostic-performance results. The registry record does not report numerical sensitivity or specificity estimates, confidence intervals, classification counts, or statistical comparisons. As a result, the posted information does not show how often either approach correctly classified lesions, how precise any estimate was, or whether performance differed between strategies. The hypothesis in the study summary is a rationale for evaluation, not a finding.

Even with numerical diagnostic measures, sensitivity and specificity would describe performance relative to the chosen reference standard and study analysis. They would not, by themselves, establish how an imaging approach affects patient outcomes, the consequences of individual classifications, or the suitability of a classification for a particular patient. The registry record does not report those outcomes, and these statistics should not be treated as medical advice.

The record describes a nonrandomized comparison and states that allocation was not stated and the arms are not a like-for-like control comparison. Therefore, any contrast between imaging assessments should not be presented as a randomized causal effect. The record also does not report reader blinding or detailed assessment procedures, which limits evaluation of possible sources of bias.

The defensible conclusion is narrow: the registry documents a completed diagnostic-imaging study and its planned classification question, but the numerical endpoint findings and their uncertainty are not reported in the registry record. No claim that adding SWE improves, worsens or leaves diagnostic performance unchanged can be supported from the posted information alone.

12. Why This Trial Matters Statistically

This study is statistically notable for asking how imaging features may change diagnostic classification rather than evaluating a treatment effect. It sets out a comparison between conventional B-mode ultrasound and B-mode ultrasound combined with selected SWE characteristics, with sensitivity and specificity as central performance concepts. That framing highlights an important feature of diagnostic research: changes in classification rules can alter how positive and negative cases are identified, so both measures are relevant to interpretation.

The registry also makes the classification strategies explicit at a broad level. The conservative and aggressive strategies use E-homogeneity, E-max and E-color, while the aggressive approach is described as changing the classification of more lesions. This distinction matters statistically because a more expansive reclassification rule may shift the balance between false-negative and false-positive classifications. The actual balance cannot be determined without the outcome counts and estimates, which are not reported in the registry record.

The use of a defined reference standard is another important aspect. Positive status is based on malignant cytologic or histopathologic findings, while negative status includes specified imaging, histopathology and follow-up criteria. Diagnostic estimates are meaningful only in relation to the standard used to label cases. The record's composite negative definition therefore belongs in any interpretation of the intended endpoint.

The study also lists reproducibility and observer-agreement outcomes, recognizing that imaging assessment may involve judgments and measurements that could vary within or between observers. However, their results and methods are not reported in the registry record. Overall, the entry is useful for identifying the diagnostic question and its statistical structure, while the absence of numerical results prevents an assessment of observed performance or precision.

13. 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 the primary outcome is that its numerical results are not reported in the registry record. Sensitivity, specificity, confidence intervals, p-values and classification counts are unavailable, so the magnitude, direction and precision of diagnostic performance cannot be assessed.

The analysis population is described as per protocol and includes 939 lesions, comprising 289 malignant and 650 benign masses, with one lesion analyzed per participant. Enrollment is listed as 1681, but no participant or lesion flow explains the relationship between enrollment and this analysis population. Eligibility criteria, protocol-adherence definitions and reasons for exclusion are not reported in the registry record.

The reference standard includes more than one route to a negative classification, including imaging categories, benign histopathology and follow-up ultrasound. Counts by route and handling of missing or indeterminate reference assessments are not reported. The endpoint timeframe is listed as 2 years, while one negative reference-standard pathway uses a 1 year follow-up ultrasound; the registry does not explain further how these time descriptions relate.

The registry does not report the detailed classification thresholds, observer training, reading workflow, blinding, or independence of the reference-standard assessment from the imaging interpretation. Allocation is not stated, and the record flags that the arms are not a like-for-like control comparison. These gaps limit assessment of potential bias and comparability. The statistical analysis plan, multiplicity procedures, missing-data methods and original sample-size justification are also not reported in the registry record.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Ultrasound Elastography of Breast Lesions — NCT00716482.
• Registry record last updated 2013-09-09; status last verified 2013-09.
• Results first posted 2013-09-09.
• 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.

17. Record Summary

NCT00716482, “Ultrasound Elastography of Breast Lesions,” is a completed study sponsored by SuperSonic Imagine. The recorded start was 2008-07, primary completion was 2010-06, and enrollment was 1681. The study asks whether adding ShearWave Elastography to conventional ultrasound can inform assessment of breast lesions.

The primary outcome concerns BI-RADS reclassification using E-homogeneity, E-max and E-color, as well as overall sensitivity and specificity for conventional B-mode ultrasound compared with B-mode plus selected SWE characteristics. The endpoint description identifies a per-protocol analysis of 939 lesions, including 289 malignant and 650 benign masses, with one lesion analyzed per participant. Positive and negative reference-standard definitions are specified.

The registry lists secondary outcomes on SWE reproducibility and observer agreement, 4 outcome measures and 21 analyses. Numerical primary results and detailed methods for the analysis are not reported in the registry record, limiting conclusions about diagnostic performance and its precision.