(392 days)
The Multifunction Keito devices are intended to be used by general public without prescription, operated with coins, to measure personal health parameters such as weight, estimation of body fat index, blood pressure, and pulse rate. Multifunction Keito can be installed in pharmacies or in other public sites. It is not a diagnostic device, and only offers data that users can consult with their personal physicians. The results are either printed on a ticket or saved on a card.
The Multifunction Keito K6, K7 & K8 are intended for public use to measure height, weight, systolic and diastolic blood pressures, to calculate pulse rate and body mass, and to estimate the percentage of body fat by using a non-invasive bio impedance analyzer. The devices have 2 coin entries allowing the selection of two operation modes: partial or total cycle of measurements. Partial cycle measures weight, height, and estimation of body fat. Total cycle measures weight, height, blood pressure, pulse rate, and estimation of body fat.
Here's a breakdown of the acceptance criteria and the studies that illustrate how the Keito K6, K7, and K8 multifunction devices meet these criteria, based on the provided text:
Acceptance Criteria and Device Performance
The provided document details the accuracy specifications of the Keito devices and presents comparative studies against predicate devices and observed/manual measurements to demonstrate substantial equivalence. The "acceptance criteria" here are implied by the comparison to established predicate devices and the accuracy claims.
Table 1: Acceptance Criteria and Reported Device Performance
| Parameter | Acceptance Criteria (from predicate/accuracy claim) | Keito K7 Reported Performance (Mean Difference vs. Observers) | Keito K5 Reported Performance (Mean Difference vs. Observers) | Keito K8 Reported Performance (Error Medium vs. Tanita) |
|---|---|---|---|---|
| Blood Pressure | ||||
| Systolic Accuracy | 1 mmHg (claimed) | -0.6 mmHg (K7-Observers) | 1.0 mmHg (K5-Observers) | N/A |
| Diastolic Accuracy | 1 mmHg (claimed) | 0.4 mmHg (K7-Observers) | -0.7 mmHg (K5-Observers) | N/A |
| Pulse Accuracy | Not explicitly stated | 0.9 bpm (K7-Observer) | -1.4 bpm (K5-Observer) | N/A |
| Weight | 100 g (claimed) | N/A | N/A | N/A |
| Height | 1 cm (claimed) | N/A | N/A | N/A |
| Body Fat | 0.1% (claimed) | N/A | N/A | -0.4% (K8-Tanita TBF-300) |
Note: The "Acceptance Criteria" column is derived from the stated accuracy of the devices and their predicates. The "Reported Device Performance" comes directly from the study results presented in the tables for K7 Blood Pressure and K8 Body Fat. The overall conclusion for blood pressure is that K7 performance is "very similar" to K5, implying K7 meets K5's established performance. For Body Fat, the results are considered "safe and accurate as TANITA TBF-300."
Study Details
2. Sample Sizes and Data Provenance
- Blood Pressure Measurement (Multifunction K7 Test Results):
- Test Set Sample Size: 267 observations for Systolic, Diastolic, and Pulse measurements. 89 patients for Age, Wrist/Arm circumference, and Sex data.
- Data Provenance: Not explicitly stated, but the submission is from a Spanish company (
Aguiflail Ibérica, S.L., Barcelona, Spain), suggesting the study was likely conducted in Spain. The data is presented as a "clinical test," implying prospective collection for the purpose of this submission.
- Body Fat Measurement (Keito K8 vs. TANITA TBF-300):
- Test Set Sample Size: 177 observations.
- Data Provenance: Not explicitly stated, but similar to the blood pressure study, it's likely from Spain and prospectively collected for this comparative study.
3. Number of Experts and Qualifications (for Ground Truth)
- Blood Pressure Measurement: The study compares the Keito K7 (and K5) devices against "Observers."
- Number of Experts: "Observer1" and "Observer2" are mentioned for inter-observer variability, and the device is compared against these "Observers." This implies at least two human observers.
- Qualifications: The qualifications of these observers are not specified in the provided document. It is not mentioned if they are medical professionals like cardiologists or nurses, nor is their experience level given.
- Body Fat Measurement: The Keito K8 is compared against the
TANITA TBF-300device. The ground truth for body fat is not established by human experts but by an established predicate device.
4. Adjudication Method for the Test Set
- Blood Pressure Measurement: For "Observer1 - Observer2" comparison, a simple comparison of their readings is used to assess inter-rater agreement. For "Keito K7 - Observers" (and K5), the device readings are compared directly against the readings of the human observers. No specific adjudication method (like 2+1 or 3+1 consensus) is described to resolve discrepancies between human observers or between the device and human observers. The "Mean Difference" and "SD of Differences" statistical metrics are used for comparison.
- Body Fat Measurement: The Keito K8 device measurements are compared directly to the TANITA TBF-300 predicate device measurements. No human adjudication is involved as the predicate device serves as the reference.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No MRMC study evaluating human reader improvement with AI assistance was explicitly performed or detailed. The studies here are focused on demonstrating the standalone performance of the device (Keito K7/K8) in comparison to either human observers (for BP) or a predicate device (for Body Fat), not on how AI (or the device's algorithms) assists human readers. The Keito devices are intended for public self-use and are not diagnostic tools requiring human interpretation of their output.
6. Standalone (Algorithm Only) Performance
- Yes, standalone performance was assessed. The studies presented (7.1.1 and 7.1.2 for Blood Pressure, 7.2 for Body Fat) represent the standalone performance of the Keito K7, K5, and K8 devices, respectively. The data shows the device's measurements compared to a reference (human observers for BP, predicate device for Body Fat) without human intervention in the device's measurement process itself. The devices operate "without prescription" and provide results directly to the user.
7. Type of Ground Truth Used
- Blood Pressure Measurement: The ground truth for blood pressure and pulse rate is established by human observers (referred to as "Observer1" and "Observer2") directly measuring the same parameters. This effectively serves as a "clinical reference standard" in this context.
- Body Fat Measurement: The ground truth for body fat percentage is established by a predicate device, the
TANITA TBF-300. This is a comparative study where the predicate device acts as the reference standard.
8. Sample Size for the Training Set
- The document does not provide details on the sample size used for the training set for the algorithms within the Keito devices. The studies described are validation/testing studies for device performance. It states that the K6, K7, and K8 are an "evolution" of the K5 unit, implying that the underlying technology and software have a history of development, but the specifics of training data are not disclosed.
9. How the Ground Truth for the Training Set was Established
- As the document does not specify a training set or its sample size, it also does not detail how the ground truth for any training set was established. We can infer that the original Keito K5, and presumably its predecessors or early development, would have been built and refined using data, likely compared against established measurement techniques or other reference devices, similar to the validation studies presented here. However, this is an inference and not explicitly stated in the provided text.
§ 870.1130 Noninvasive blood pressure measurement system.
(a)
Identification. A noninvasive blood pressure measurement system is a device that provides a signal from which systolic, diastolic, mean, or any combination of the three pressures can be derived through the use of tranducers placed on the surface of the body.(b)
Classification. Class II (performance standards).