(93 days)
The GEM Premier 3000, which is an upgraded version of the existing GEM Premier Plus (K961335), is a portable system for use by health care professionals to rapidly analyze whole blood samples at the point of health care delivery in a clinical setting. The instrument provides quantitative measurements of whole blood pH, pCO2, pO2, Na+, K+, Ca++, and Hct. These parameters along with derived parameters Base Excess, HCO3, TCO2 and sO2 aid in the diagnosis of a patient's acid/base status, oxygen delivery capacity, and electrolyte and metabolite balance.
The GEM Premier 3000, which is an upgraded version of the existing GEM Premier Plus (K961335), is a portable system for use by health care professionals to rapidly analyze whole blood samples at the point of health care delivery in a clinical setting. The instrument provides quantitative measurements of whole blood pH. pCO2, pO2, Na+, K+, Ca++, and Hct. These parameters along with derived parameters Base Excess, HCO3, TCO2 and sO2 aid in the diagnosis of a patient's acid/base status, oxygen delivery capacity, and electrolyte and metabolite balance.
Here's an analysis of the acceptance criteria and study detailed in the provided text for the GEM Premier 3000 device:
Acceptance Criteria and Device Performance for GEM Premier 3000
1. Table of Acceptance Criteria and Reported Device Performance
The provided document does not explicitly state pre-defined "acceptance criteria" in terms of specific thresholds for precision and method comparison. Instead, it presents the results of the studies and implicitly suggests that these results were deemed acceptable by the FDA for substantial equivalence. The predicate device (IL Synthesis and manual spun hematocrit) serves as the de facto acceptance benchmark.
Below are the reported performance metrics from the studies:
Precision Data (Combined Within Run and Total %CV / SD)
| Parameter | Mean (Level 1) | Within Run %CV/SD (L1) | Total %CV/SD (L1) | Mean (Level 2) | Within Run %CV/SD (L2) | Total %CV/SD (L2) | Mean (Level 3) | Within Run %CV/SD (L3) | Total %CV/SD (L3) |
|---|---|---|---|---|---|---|---|---|---|
| pH | 7.0886 | 0.0066 (SD) | 0.0180 (SD) | 7.4276 | 0.0038 (SD) | 0.0049 (SD) | 7.6263 | 0.0040 (SD) | 0.0082 (SD) |
| pCO₂ (mmHg) | 66.02 | 1.95 | 2.82 | 36.37 | 2.37 | 2.45 | 16.79 | 2.89 | 3.76 |
| pO₂ (mmHg) | 64.42 | 2.14 | 2.77 | 103.69 | 0.86 | 1.63 | 155.65 | 0.82 | 1.20 |
| Na+ (mmol/L) | 118.55 | 0.63 | 0.98 | 135.51 | 0.42 | 0.72 | 152.61 | 0.63 | 1.18 |
| K+ (mmol/L) | 2.32 | 1.92 | 2.04 | 3.782 | 0.62 | 0.92 | 6.143 | 0.57 | 0.83 |
| Ca++ (mmol/L) | 0.753 | 2.16 | 2.68 | 1.071 | 0.94 | 1.50 | 1.387 | 0.88 | 1.30 |
| Hct (%) | 24.34 | 1.18 | 1.43 | 43.93 | 1.11 | 1.57 | NA | NA | NA |
Method Comparison Data (GEM Premier 3000 vs. Predicate Device)
| Parameter | n | Slope | Intercept | r | Sample Range |
|---|---|---|---|---|---|
| pH | 128 | 1.0660 | -0.4754 | 0.9931 | 7.10 - 7.60 |
| pCO₂ (mmHg) | 130 | 0.9721 | 2.1653 | 0.9935 | 24.6 - 99.6 |
| pO₂ (mmHg) | 128 | 0.9977 | 1.3552 | 0.9990 | 32 - 538 |
| Na+ (mmol/L) | 85 | 1.0181 | -3.9552 | 0.9820 | 110 - 182 |
| K+ (mmol/L) | 84 | 0.9474 | 0.0736 | 0.9987 | 1.20 - 14.60 |
| Ca++ (mmol/L) | 80 | 0.9756 | 0.0202 | 0.9927 | 0.73 - 4.01 |
| Hct (%) | 117 | 1.0841 | -3.4184 | 0.9548 | 16.0 - 54.0 |
Based on the FDA's decision to clear the device (K992834), the reported precision and method comparison results for the GEM Premier 3000 were considered sufficient to demonstrate substantial equivalence to the predicate devices. The implicit acceptance criteria were therefore met.
2. Sample Size Used for the Test Set and Data Provenance
Precision Study:
- Sample Size: 60 replicates (4 per day for 14 days, twice on Day 1) on each of 7 different IL GEM 3000 instruments, totaling 420 replicates for each parameter at each control level.
- Data Provenance: The data was generated using "three levels of controls (GEM Check Plus) for pH, pCO2, pO3, Na+, K+, Ca++ and two levels of controls (GEM critCheck) for hematocrit." It is an in-vitro study using manufactured control materials. The country of origin is not explicitly stated but implies a controlled laboratory setting.
Method Comparison Study:
- Sample Size:
- pH, pCO₂, pO₂: n = 128-130
- Na+, K+, Ca++: n = 80-85
- Hct: n = 117
- Data Provenance: The samples were from "arterial, venous, heart bypass and liver transplant blood samples from hospital patients using heparinized syringes and from healthy volunteers using heparinized vacutainer tubes." This indicates a prospective collection of human whole blood samples in a clinical setting. The country of origin for the hospital patients and healthy volunteers is not specified, but it's likely within the US given the FDA submission.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
This type of device (blood gas analyzer) does not typically involve human "experts" to establish ground truth in the same way an imaging or diagnostic AI device does. The ground truth for the method comparison study was established by measurements from the predicate devices:
- For pH, pCO2, pO2, Na+, K+, and Ca++: the IL Synthesis (an existing, legally marketed blood gas analyzer).
- For Hct: manual spun hematocrit (a standard laboratory method).
No specific "experts" with defined qualifications were involved in establishing the ground truth beyond the inherent calibration and standard operating procedures of the predicate devices.
4. Adjudication Method for the Test Set
Not applicable. This is not a study requiring adjudication of interpretations (e.g., by multiple readers). The "ground truth" was obtained through direct measurements from the predicate devices.
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. The GEM Premier 3000 is an automated blood gas analyzer, not an AI-powered diagnostic tool that assists human readers/interpreters in a multi-reader, multi-case study context. It directly measures and reports values.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
Yes, in essence, the entire performance evaluation is a "standalone" assessment of the device's accuracy and precision. The GEM Premier 3000 operates as an automated system; its outputs are the direct measurements it provides. The method comparison study compares its standalone performance against established analytical methods (predicate devices), and the precision study evaluates its inherent measurement variability.
7. The Type of Ground Truth Used
The ground truth used for the method comparison study was comparative measurements from a legally marketed predicate device (IL Synthesis) and a standard laboratory method (manual spun hematocrit). This constitutes an established reference method or reference standard for the parameters being measured in the clinical laboratory context.
8. The Sample Size for the Training Set
Not applicable. The GEM Premier 3000 is a physical medical device (an analyzer) and not a machine learning or AI algorithm that requires a "training set" in the computational sense. Its underlying hardware and software are designed and validated through engineering processes, not by statistical training on a dataset.
9. How the Ground Truth for the Training Set Was Established
Not applicable, as there is no "training set" for this type of device. The device's operational parameters and calibration are established through engineering design, manufacturing standards, and internal validation, not through a data-driven training process.
§ 862.1120 Blood gases (P
CO2 , PO2 ) and blood pH test system.(a)
Identification. A blood gases (PCO2 , PO2 ) and blood pH test system is a device intended to measure certain gases in blood, serum, plasma or pH of blood, serum, and plasma. Measurements of blood gases (PCO2 , PO2 ) and blood pH are used in the diagnosis and treatment of life-threatening acid-base disturbances.(b)
Classification. Class II.