(23 days)
M3290A: For central monitoring of multiple adult, pediatric, and neonatal patients; and where the clinician decides to monitor cardiac arrhythmia of adult, pediatric, and neonatal patients and/or ST segment of adult patients to gain information for treatment, to monitor adequacy of treatment, or to exclude causes of symptoms.
M4840A: For ambulatory and bedside monitoring of ECG and SpO2 parameters of adult and pediatric patients in healthcare facilities to gain information for treatment, to monitor adequacy of treatment, or to exclude causes of symptoms.
The Philips M3290A IntelliVue Information Center Software Release F.0 and M4840A Philips Telemetry System II with M4841A patient device.
This Philips submission for the M3290A IntelliVue Information Center Software Release F.0 and M4840A Philips Telemetry System II with M4841A patient device describes modifications (changes in ECG chest lead support, NBP limit alarms, and network functionality) and asserts substantial equivalence to a previously cleared predicate device (K040357). The documentation does not contain a detailed study with specific acceptance criteria and device performance metrics, as would typically be presented for de novo device approval or significant design changes requiring new clinical validation.
Instead, the submission states that:
"Verification, validation, and testing activities have successfully established the performance, functionality, and reliability characteristics of the new devices with respect to the predicates. Testing involved system level tests, integration tests, environmental tests, and safety testing from hazard analysis. Pass/Fail criteria were based on the specifications cleared for the predicate devices and test results showed substantial equivalence. The results successfully demonstrate that patient monitoring system functionality meets all reliability requirements and performance claims and is substantially equivalent to the predicate devices."
This indicates that the modifications were evaluated against the established performance specifications of the predicate device, and the testing confirmed that the changes did not degrade performance below those existing benchmarks.
Therefore, not all questions can be directly answered as the provided text does not contain a discrete study with defined acceptance criteria and performance data for this specific submission's modifications. The information below reflects what can be extracted or inferred from the provided text regarding the evaluation approach.
1. Table of Acceptance Criteria and Reported Device Performance
Based on the provided text, specific quantitative acceptance criteria and detailed device performance metrics (true positive rate, false positive rate, sensitivity, specificity, accuracy, AUC, F1-score) are not explicitly stated for the modified device. The document states that "Pass/Fail criteria were based on the specifications cleared for the predicate devices and test results showed substantial equivalence." This implies that the device maintained the performance characteristics of its predicate.
| Acceptance Criteria | Reported Device Performance (Implied) |
|---|---|
| Maintain performance specifications of predicate device (K040357) | Performance shown to be substantially equivalent to predicate device for all functions (including new ECG chest lead and NBP limit alarms) |
| Meet all reliability requirements | Successfully demonstrated |
| Meet all performance claims | Successfully demonstrated |
| Pass system level tests | Passed |
| Pass integration tests | Passed |
| Pass environmental tests | Passed |
| Pass safety testing from hazard analysis | Passed |
2. Sample size used for the test set and the data provenance
- Test set sample size: Not specified. The testing described includes "system level tests, integration tests, environmental tests, and safety testing from hazard analysis," which are typically internal engineering and validation tests rather than clinical studies with patient data in the context of AI/ML evaluation.
- Data provenance: Not specified. Given the nature of the modifications (adding support for an ECG chest lead, NBP alarms, network functionality), the testing likely involved controlled testing environments and simulated data, potentially with some real physiological data from internal archives if applicable, but no geographically or retrospectively/prospectively defined clinical dataset is mentioned.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
- Not applicable/Not specified. The evaluation focused on technical performance and substantial equivalence to a predicate device, not on diagnostic accuracy requiring expert ground truth labels for a dataset.
4. Adjudication method for the test set
- Not applicable/Not specified. There is no mention of a human adjudication process for establishing ground truth for a test set.
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
- No MRMC comparative effectiveness study was performed or mentioned. This submission does not describe an AI/ML algorithm intended to assist human readers.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done
- The device itself is a patient monitoring system, which includes automated capabilities like arrhythmia detection and ST segment monitoring. The testing described implicitly evaluates these "standalone" functionalities against their specifications, as part of assessing substantial equivalence to the predicate. However, specific performance metrics for individual algorithms (e.g., arrhythmia detection algorithm sensitivity) are not provided in this summary.
7. The type of ground truth used
- For the technical and safety testing conducted, the ground truth would be based on device specifications, engineering requirements, and regulatory standards. For example, in testing NBP alarms, the ground truth would be the defined NBP limits and whether the system correctly triggered an alarm when those limits were exceeded based on simulated or measured blood pressure values. For ECG lead support, the ground truth would be the accurate acquisition and display of the ECG signal via the new lead.
8. The sample size for the training set
- Not applicable. This device, as described in this 2004 submission, is a traditional medical device (patient monitor) with pre-specified algorithms, not a machine learning or AI device that undergoes a training phase with a distinct training set.
9. How the ground truth for the training set was established
- Not applicable. As noted above, this is not an AI/ML device that requires a training set and associated ground truth establishment.
§ 870.1025 Arrhythmia detector and alarm (including ST-segment measurement and alarm).
(a)
Identification. The arrhythmia detector and alarm device monitors an electrocardiogram and is designed to produce a visible or audible signal or alarm when atrial or ventricular arrhythmia, such as premature contraction or ventricular fibrillation, occurs.(b)
Classification. Class II (special controls). The guidance document entitled “Class II Special Controls Guidance Document: Arrhythmia Detector and Alarm” will serve as the special control. See § 870.1 for the availability of this guidance document.