FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

The Calibrators for the OEM Instruments are aqueous reagents with salts (chemical constituents) added to obtain desired analyte levels to provide the desired calibration.

AI/ML Overview

The provided text describes performance characteristics for various calibrating materials. The acceptance criteria are implicitly defined by establishing "substantial equivalence" of the Mission Diagnostics calibrators to OEM calibrating materials. The studies aim to demonstrate that the Mission Diagnostics devices perform "equally" to the predicate OEM devices.

Here's an analysis of the provided information for the devices where detailed performance data is given:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criterion for all studies is that the Mission Diagnostics calibrator performs "equally" to the OEM calibrator. This is demonstrated through comparative studies, where the performance of the Mission device, when used to calibrate an instrument, results in QC and sample recoveries that correlate strongly with, and are comparable to, the recoveries obtained when the OEM device is used for calibration. This "equally" is quantified by high correlation coefficients ($r^2$) and low standard errors of the estimate (S(y.x)) in linear regression analysis, alongside comparable %CV values for precision.

Let's focus on three specific devices for which detailed performance data is provided: CD-105670D (Hct Slope), CD-471818D (634 Ca/pH Slope Standard), and CD-473692D (pH Blood Gas Slope).

Device/Analyte	Acceptance Criteria (Implicit: Performance "equally" to OEM, high correlation, low S(y.x))	Reported Device Performance (Mission vs. OEM)
CD-105670D (Hct Slope)	High $r^2$ and low S(y.x) between Mission and OEM calibrated results. Similar %CV for QC samples.	Regression: Mission = 0.999 * Corning - 0.818, Range = 26 & 48 %Hct, $r^2$ = 0.997, df = 95, n = 96, S(y.x) = 0.58 %Hct.
QC Precision (Mission): QC-A: 2.2% CV; QC-B: 1.6% CV.
QC Precision (OEM): QC-A: 1.0% CV; QC-B: 1.2% CV.	Met. Very high correlation ($r^2$=0.997) and minimal difference in %CVs.
CD-471818D (634 Ca/pH Slope Standard)	High $r^2$ and low S(y.x) between Mission and OEM calibrated results for both serum samples and QC samples. Similar %CV for QC samples.	Serum Sample Regression: Mission = 1.018 * Corning - 0.019, Range = 0.37 to 4.57 mmol/L, $r^2$ = 0.998, df = 68, n = 69, S(y.x) = 0.05 mmol/L.
QC Sample Regression (Ca): Mission = 1.028 * Corning - 0.031, Range = 0.55 - 2.18, $r^2$ = 0.998, df = 32, n = 33, S(y.x) = 0.02 mmol/L Ca.
QC Sample Regression (pH): Mission = 1.045 * Corning - 0.36, Range = 7.05 - 7.62, $r^2$ = 0.937, df = 32, n = 33, S(y.x) = 0.05.
QC Precision (Mission Hct): pH: 0.18-0.62% CV; Ca: 0.62-1.50% CV.
QC Precision (OEM Hct): pH: 0.21-0.67% CV; Ca: 0.84-1.90% CV.	Met. Very high correlation for Ca ($r^2$=0.998) and strong correlation for pH ($r^2$=0.937). %CVs are comparable.
CD-473692D (pH Blood Gas Slope)	High $r^2$ and low S(y.x) between Mission and OEM calibrated results for QCs. Similar %CV for QC samples.	QC Regression (pH): Mission = 1.053 * Corning - 0.392, Range = 7.02 to 7.60, $r^2$ = 0.987, df = 112, n = 113, S(y.x) = 0.02.
QC Regression (pCO2): (No data provided for pCO2 regression statistics in the excerpt.)
QC Regression (pO2): Mission = 1.001 * Corning + 0.661, Range = 71 to 142 mm Hg pO2, $r^2$ = 0.972, df = 112, n = 113, S(y.x) = 4.19 mm Hg.
QC Precision (Mission): pH: 0.02-0.30% CV; pCO2: 2.1-2.6% CV; pO2: 1.6-3.7% CV.
QC Precision (OEM): pH: 0.23-0.26% CV; pCO2: 2.4-3.0% CV; pO2: 1.8-3.8% CV.	Met. High correlation for pH ($r^2$=0.987) and pO2 ($r^2$=0.972). %CVs are very comparable.

2. Sample Sizes Used for the Test Set and the Data Provenance

The studies are designed as comparative performance evaluations between the Mission Diagnostics calibrators and the OEM calibrators.

CD-105670D (Hct Slope):
- Test Set Sample Size: n=96 individual QC observations for regression analysis (derived from 7 days, 2 runs each for Mission and OEM, with 2 QC samples measured in duplicate for each run, so 7 days * 2 runs/day * 2 QC samples * 2 duplicates = 56 for Mission and same for OEM, which might imply n=112 total measurements, or 96 for the regression points between Mission and OEM).
- Data Provenance: Not explicitly stated, but the context of an in vitro diagnostic device for clinical analyzers implies laboratory testing, likely prospective. Country of origin not specified.
CD-471818D (634 Ca/pH Slope Standard):
- Test Set Sample Size:
  - Serum samples: n=69 for regression analysis.
  - QC samples (Ca): n=33 for regression analysis.
  - QC samples (pH): n=33 for regression analysis.
  - QC precision studies: 12 replicates for each of 3 QC levels for both Mission and OEM for both Ca and pH (total 1232*2 = 144 measurements for QCs alone).
- Data Provenance: Not explicitly stated, but implies laboratory testing using spiked/diluted serum samples and QC materials. Likely prospective. Country of origin not specified.
CD-473606D (654 Lithium Slope Solution 2.50 mmol/L):
- Test Set Sample Size:
  - Serum samples: n=28 for regression analysis.
  - QC precision studies: 20 replicates for each of 3 QC levels for both Mission and OEM for Na, K, and Li (total 2032*3 = 360 measurements for QCs alone).
- Data Provenance: Not explicitly stated, but implies laboratory testing using spiked/diluted serum samples and QC materials. Likely prospective. Country of origin not specified.
CD-473692D (pH Blood Gas Slope):
- Test Set Sample Size:
  - QC samples: n=113 for pH regression analysis, n=113 for pO2 regression analysis.
  - Precision studies: N varies by analyte and level (e.g., pH Level 1 has 36 observations for Mission, 37 for OEM). Total observations for QC precision for pH, pCO2, pO2 across 3 levels over 5 days, 2 runs/day with 4 replicates per run (implied n=8 per QC sample per day for each calibrator) = (36+38+40) + (37+40+39) = 230 for pH. Similar counts for pCO2 and pO2.
- Data Provenance: Not explicitly stated, but implies laboratory testing using QC materials. Likely prospective. Country of origin not specified.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and the Qualifications of Those Experts

Not applicable. These studies do not involve human expert interpretation of results to establish ground truth. The "ground truth" (or reference values) for quality control materials and serum samples in these types of in vitro diagnostic studies are typically established by certified reference methods, inter-laboratory comparisons, or assigned values by the manufacturer of the QC materials, under a strict quality management system in a laboratory setting. The predicate device (OEM calibrator) acts as the comparative standard.

4. Adjudication Method for the Test Set

Not applicable. There is no human adjudication process described, as the evaluation is based on quantitative analytical results comparing the performance of two calibrators.

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. These are studies for in vitro diagnostic calibrators, not AI-assisted human reader studies.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done

Yes, these are standalone performance studies of the calibrator materials in conjunction with the specified automated analyzers. The calibrators themselves are not an "algorithm" in the typical sense, but rather a chemical reagent. The performance is determined by how well the analytical system (instrument + calibrator) performs compared to the system calibrated with the OEM device. The results (e.g., QC recoveries, analyte concentrations) are measured directly by the instrument.

7. The Type of Ground Truth Used (expert consensus, pathology, outcomes data, etc.)

The ground truth or reference standard is the performance achieved when the respective analyzer is calibrated using the OEM calibrator. The study's objective is to show that the Mission Diagnostics calibrator produces equivalent results to the OEM calibrator, implying that the OEM calibrator's performance represents the acceptable "truth" or standard. Additionally, the QC samples used have established target values (though not explicitly stated as "ground truth" in the text, it is implied as a reference for assessing accuracy and precision). For serum samples, the "ground truth" is typically the reference values established through the OEM calibration.

8. The Sample Size for the Training Set

Not applicable. These are not machine learning (AI) models that require a training set. The "training" in this context refers to the calibration of the analyzer, which uses the calibrator material itself, not a separate data set.

9. How the Ground Truth for the Training Set was Established

Not applicable, as there is no training set for an AI model. The calibrators "train" (calibrate) the instrument by providing known reference points for the instrument to adjust its measurement scale. The values of these calibrators are established by the manufacturer, Mission Diagnostics, and are designed to be equivalent to the OEM's calibrators.

Summary

Regulation Number and Section

§ 862.1150 Calibrator.

(a)
Identification. A calibrator is a device intended for medical purposes for use in a test system to establish points of reference that are used in the determination of values in the measurement of substances in human specimens. (See also § 862.2 in this part.)(b)
Classification. Class II (special controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 862.9.