(63 days)
The ZEUS System is intended for use by qualified medical professionals for the assessment ambulatory electrocardiogram (ECG) data from adult patients 18 years or older. The system is intended to be marketed as a service that downloads and analyzes up to 14 days of ECG data. Analysis metrics are provided in a report which is available for clinician review and printing. The reported ECG metrics include single-lead analysis on a beat by beat basis, heart rate measurement and rhythm analysis. The report does not contain diagnostic interpretation; the reported analysis is provided for review by the intended user to render a diagnosis based on clinical judgment and experience.
The ZIO ECG Utilization Service System ("ZEUS System") is an electrocardiogram (ECG) processing and analysis system, designed to handle continuously recorded, single-lead ECG data. Once the ECG data is downloaded, the data is processed through the algorithm and delivered to the QA Tool module where the results are reviewed and/or adjusted by iRhythm Certified Cardiographic Technicians (CCT's) for accuracy. iRhythm trained Patch in-take and CCT personnel follow internal procedures for processing and are of the algorithm performance anomalies. Any software anomalies are visible to and manually corrected by iRhythm Technologies CCT's during the QA review and/or Patch Report edits. The CCT generates a final report of the ECG findings contained within the data; thereby enabling the provision of a complete ECG processing and analysis service.
The provided text describes the ZEUS (ZIO ECG Utilization Services) System, a programmable diagnostic computer for ambulatory ECG data analysis. It details the device's indications for use, product description, and claims of substantial equivalence to a predicate device. However, it does not provide specific details about acceptance criteria, reported device performance metrics in a table, sample sizes for test/training sets, ground truth establishment methods in detail, or information regarding multi-reader multi-case studies.
Based on the provided document, here's what can be extracted and what information is missing:
1. Table of acceptance criteria and the reported device performance:
| Metric / Feature | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Automated ECG Analysis Performance | Satisfy requirements to minimize safety or efficacy concerns (specific thresholds not provided). | "The resulting statistics demonstrate sensitivity and positive predictivity levels which satisfy requirements and minimize safety or efficacy concerns." (No specific numbers or metrics are given). |
Missing Information: Specific numerical acceptance criteria (e.g., sensitivity > X%, specificity > Y%) and the actual reported numerical performance of the device against these metrics.
2. Sample size used for the test set and the data provenance:
Missing Information: The document states "Automated ECG analysis performance was quantified for any claimed analysis metrics," but does not provide the sample size of the test set used for this quantification. It also does not specify the data provenance (e.g., country of origin, retrospective or prospective).
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
Missing Information: The document states that the output of the algorithm is delivered to a "QA Tool module where the results are reviewed and/or adjusted by iRhythm Certified Cardiographic Technicians (CCT's) for accuracy." It also mentions "iRhythm trained Patch in-take and CCT personnel follow internal procedures for processing and are of the algorithm performance anomalies." However, it does not specify the number of experts (CCTs or other medical professionals) used to establish ground truth for a test set, nor their specific qualifications (e.g., years of experience beyond "Certified Cardiographic Technicians"). It also doesn't explicitly state that these CCTs established the ground truth for a separate test set, but rather that they review and adjust algorithm output in the operational workflow.
4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:
Missing Information: The document mentions that CCTs review and adjust results, implying a human-in-the-loop process. However, it does not describe a formal adjudication method (like 2+1 or 3+1 consensus) for establishing ground truth on a specific test set for performance evaluation.
5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:
Missing Information: The document does not mention a multi-reader multi-case (MRMC) comparative effectiveness study involving human readers with and without AI assistance. The system's output is intended for review by a clinician, but no study is described to quantify improvement with AI assistance.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
The document implies a system that includes human review: "Once the ECG data is downloaded, the data is processed through the algorithm and delivered to the QA Tool module where the results are reviewed and/or adjusted by iRhythm Certified Cardiographic Technicians (CCT's) for accuracy." And "Any software anomalies are visible to and manually corrected by iRhythm Technologies CCT's during the QA review and/or Patch Report edits." While the "Automated ECG analysis performance was quantified," the provided text does not explicitly state if a standalone algorithm-only performance evaluation was conducted entirely separate from this human-in-the-loop review process for regulatory submission. It seems the described system always includes human review as part of the service.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
The ground truth appears to be established by Certified Cardiographic Technicians (CCTs) who review and adjust the algorithm's output. The document states: "the results are reviewed and/or adjusted by iRhythm Certified Cardiographic Technicians (CCT's) for accuracy." This suggests expert review/adjudication as the basis for accuracy.
8. The sample size for the training set:
Missing Information: The document does not provide any information regarding the sample size used for training the algorithm.
9. How the ground truth for the training set was established:
Missing Information: The document does not describe how ground truth was established for the training set. It only mentions how "accuracy" for the analysis is handled, which seems to describe the operational workflow rather than a specific ground truth generation process for a dataset.
§ 870.1425 Programmable diagnostic computer.
(a)
Identification. A programmable diagnostic computer is a device that can be programmed to compute various physiologic or blood flow parameters based on the output from one or more electrodes, transducers, or measuring devices; this device includes any associated commercially supplied programs.(b)
Classification. Class II (performance standards).