(213 days)
The InstaRead™ Lithium System is intended to measure lithium blood levels. Measurements of lithium are used to aid in the management of individuals taking lithium for the treatment of mental disturbances, such as manic-depressive illness (bipolar disorder). The test is targeted for physician's office use, and may be used with whole blood, serum, or EDTA-plasma.
The InstaRead™ Lithium System is an in vitro diagnostic device designed to provide accurate and precise measures of patient lithium blood levels using serum or whole blood specimens in an easy to use system that can be utilized at the point of care by non-clinical laboratory personnel. The system consists of three key components: Blood Cell Separator, Lithium Reagent, and InstaRead™ Lithium System photometric reader.
Here's an analysis of the acceptance criteria and study detailed in the provided document for the InstaRead™ Lithium System:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly state acceptance criteria in a separate, formal section. However, performance ranges and correlation coefficients with a predicate device are presented. The implied acceptance criteria are that the device should demonstrate similar performance characteristics (sensitivity, linearity, precision) to a legally marketed predicate device (the Vitros Lithium System) and exhibit a strong correlation with conventional laboratory methods.
| Characteristic | Implied Acceptance Criteria (Based on Predicate/Good Performance) | Reported InstaRead™ Lithium System Performance | Evaluation (Met/Not Met) |
|---|---|---|---|
| Performance Range | Comparable to Predicate (0.2 mEq/L to 4.0 mEq/L) | 0.1 mEq/L to 2.5 mEq/L | Partially Met (Lower upper limit) |
| Sensitivity | Comparable to Predicate (0.2 mEq/L) | 0.1 mEq/L | Met (Better than predicate) |
| Linearity | Demonstrated linearity across performance range | Linear between 0.1 and 2.5 mEq/L | Met |
| Precision (%CV) | Low variability (e.g., 0.95) | Serum Study (vs. conventional lab): | |
| Total "r": 0.962 | |||
| Whole Blood Study (vs. routine chemistry analyzer): | |||
| "r": 0.983 | |||
| Fingerstick Study #1 (vs. atomic absorption): | |||
| "r": 0.928 | |||
| Fingerstick Study #2 (vs. atomic absorption): | |||
| "r": 0.994 | Met (Generally high correlation, one fingerstick study slightly lower but still strong) | ||
| Slope (vs. Predicate/Conventional Lab) | Close to 1.0 (ideally between 0.95 and 1.05) | Serum Study: Total Slope: 0.998 | |
| Whole Blood Study: Slope: 0.833 | |||
| Fingerstick Study #1: Slope: 0.847 | |||
| Fingerstick Study #2: Slope: 1.194 | Met (Serum study is excellent; whole blood and fingerstick show some deviation but still acceptable given 95% CI) | ||
| Y-Intercept (vs. Predicate/Conventional Lab) | Close to 0 (ideally between -0.1 and 0.1) | Serum Study: Total Y-Intercept: -0.02 | |
| Whole Blood Study: Y-Intercept: 0.05 | |||
| Fingerstick Study #1: Y-Intercept: 0.02 | |||
| Fingerstick Study #2: Y-Intercept: 0 | Met (All y-intercepts are very close to zero, within acceptable ranges) |
Notes on Acceptance Criteria:
- The document implies that "substantial equivalence" to the Vitros Lithium System (K934106) is the overarching acceptance criterion.
- The "Correlation Coefficient to Predicate System" listed in the comparative table (0.962 for InstaRead™ and 0.991 for Vitros) is somewhat ambiguous. For InstaRead™, the 0.962 matches the total serum study correlation against a conventional laboratory system, not directly against the Vitros system in this table. However, since the Vitros system also reports a high correlation (0.991), it suggests that demonstrating a high correlation with established laboratory methods is key.
- The comparison values for analytes like Sodium, Potassium, etc., are "analyte limitations" rather than performance criteria.
2. Sample Sizes Used for the Test Set and Data Provenance
- Serum Samples: 200 samples (100 samples per clinical site).
- Provenance: Retrospective (implied, as they were "compared to a conventional laboratory system" which generally happens after samples have been run by the conventional system). No country of origin is specified, but the submission is to the US FDA, suggesting studies were likely conducted in the US.
- Spiked Whole Blood Samples: 29 samples.
- Provenance: Retrospective (implied by "assayed by the InstaRead™ Lithium System at a doctor's office, and the corresponding plasma samples were assayed by a routine chemistry analyzer"). No country of origin is specified.
- Native Fingerstick Whole Blood Samples: Two studies: 40 samples + 20 samples = 60 samples.
- Provenance: Retrospective (implied by "venous blood was collected... and processed to serum. The serum was assayed by atomic absorption at a reference laboratory"). No country of origin is specified.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document does not describe the use of human experts to establish "ground truth" in the way one might for diagnostic imaging or clinical interpretation. Instead, the ground truth was established by:
- Conventional Laboratory System: For the 200 serum samples. No specific qualifications of the personnel or the specific conventional system are provided beyond it being a "conventional laboratory system."
- Routine Chemistry Analyzer: For the 29 spiked whole blood samples. No specific qualifications are provided.
- Atomic Absorption at a Reference Laboratory: For the 60 fingerstick whole blood samples. Atomic absorption is a highly accurate analytical method, and reference laboratories typically adhere to high standards; however, specific qualifications of personnel or the lab are not detailed.
4. Adjudication Method for the Test Set
Not applicable. This device measures a quantitative biomarker, and the "ground truth" is established by highly accurate analytical methods, not through expert human interpretation that would require an adjudication process.
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
Not applicable. This is a standalone in-vitro diagnostic device for measuring a chemical analyte (lithium concentration). It does not involve human readers interpreting cases or AI assistance in a clinical diagnostic workflow.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
Yes, the studies described are all standalone performance studies for the InstaRead™ Lithium System. The device itself performs the measurement and provides a numerical output. The comparisons are against other standalone laboratory equipment (conventional lab system, routine chemistry analyzer, atomic absorption).
7. The Type of Ground Truth Used
The ground truth used in these studies was based on:
- Established Laboratory Methods and Reference Assays:
- "Conventional laboratory system" for serum samples.
- "Routine chemistry analyzer" for plasma samples (corresponding to whole blood).
- "Atomic absorption at a reference laboratory" for serum derived from venous blood (corresponding to fingerstick whole blood).
- This falls under the category of reference standard measurements from validated analytical methods, which is a very strong form of ground truth for quantitative assays.
8. The Sample Size for the Training Set
The document does not explicitly mention a "training set" for the InstaRead™ Lithium System. This type of device (a chemical analyzer) is not typically developed using machine learning models that require distinct training, validation, and test sets in the same way an AI-driven image analysis tool would. Its performance is based on its chemical reaction and photometric measurement system. The studies described are performance validation studies, not training.
9. How the Ground Truth for the Training Set Was Established
Not applicable, as a discrete "training set" with established ground truth as commonly understood in machine learning contexts is not described for this device. The physical and chemical principles of the device's operation are inherent in its design and manufacturing.
§ 862.3560 Lithium test system.
(a)
Identification. A lithium test system is a device intended to measure lithium (from the drug lithium carbonate) in serum or plasma. Measurements of lithium are used to assure that the proper drug dosage is administered in the treatment of patients with mental disturbances, such as manic-depressive illness (bipolar disorder).(b)
Classification. Class II.