Search Results

The i-STAT Alinity instrument with i-STAT tests is intended for use in point of care or clinical laboratory settings. The i-STAT Alinity system is intended for the quantitative measurement of various analytes in arterial and venous whole blood.

The i-STAT Hematocrit test is intended for use in the in vitro quantification of packed red blood cell volume fraction in arterial or venous heparinized whole blood, or in arterial or venous non-anticoagulated whole blood.

Hematocrit measurements can aid in the determination of normal or abnormal total red cell volume status that can be associated with conditions including anemia and erythrocytosis.

The i-STAT Hematocrit test with the i-STAT Alinity System has not been evaluated in neonates.

The i-STAT Hematocrit test with the i-STAT Alinity System is not for use with capillary samples.

For in vitro diagnostic use.

Device Description

The i-STAT Alinity System is a handheld, in vitro diagnostic analytical device designed to run i-STAT test cartridges. The system is designed for use at or near point of patient care, by trained medical professionals and is for prescription use only.

The i-STAT Alinity System is comprised of the instrument, rechargeable battery, base station, electronic simulator, control material, printer and i-STAT test cartridges. The i-STAT Alinity Instrument features a barcode scanner, user interface with touch screen display and wireless capability. The instrument reports quantitative results within approximately 2 minutes.

The i-STAT test cartridge contains sensors which are located on the biosensors chips. The instrument interacts with the cartridge to move fluid across the sensors and generate a quantitative result. Cartridges require two to three drops of whole blood which are typically applied to the cartridge using a syringe.

AI/ML Overview

The provided document describes the FDA 510(k) premarket notification for the i-STAT Hematocrit test with the i-STAT Alinity System. This is a point-of-care device for quantitative measurement of packed red blood cell volume fraction (hematocrit) in whole blood.

Here's an analysis of the acceptance criteria and the study proving the device meets them, based on the provided text:

Important Note: This document primarily focuses on demonstrating substantial equivalence to a predicate device (i-STAT Hematocrit test with i-STAT 1 Wireless Analyzer). Therefore, the "acceptance criteria" are largely implied by the performance characteristics demonstrated to be comparable or better than the predicate, or to fall within acceptable laboratory limits for precision, linearity, etc. There isn't an explicit table of pre-defined acceptance criteria with pass/fail thresholds in the same way one might find for, say, an AI diagnostic algorithm's sensitivity/specificity. The criteria are implicitly the successful demonstration of analytical performance, rather than clinical efficacy against a true disease state, as it's a quantitative measurement device.

1. Table of Acceptance Criteria and Reported Device Performance

As mentioned, explicit, pre-defined acceptance criteria with numerical thresholds (e.g., "sensitivity > 90%") are not stated for this type of device and submission. Instead, the acceptance is based on demonstrating sound analytical performance characteristics comparable to or better than a legally marketed predicate device. The performance data presented below indicate that the device met its internal performance specifications, which are implicitly the "acceptance criteria" for this submission type.

Performance Characteristic	Implicit Acceptance Criteria (based on common lab standards/comparison to predicate)	Reported Device Performance
Precision (Aqueous Materials)	Acceptable within-run, between-run, and total precision (%CV and SD) for a hematocrit measuring device.	CV L2 (16.9%PCV): ST 0.46, CVt 2.72%; Sr 0.44, CVr 2.60%; Srr 0.09, CVrr 0.53%; Sdd 0.09, CVdd 0.53% CV L3 (33.9%PCV): ST 0.51, CVt 1.50%; Sr 0.48, CVr 1.42%; Srr 0.13, CVrr 0.38%; Sdd 0.11, CVdd 0.32% CV L4 (55.2%PCV): ST 0.49, CVt 0.89%; Sr 0.47, CVr 0.85%; Srr 0.12, CVrr 0.22%; Sdd 0.09, CVdd 0.16% CV L5 (65.0%PCV): ST 0.39, CVt 0.60%; Sr 0.37, CVr 0.57%; Srr 0.10, CVrr 0.15%; Sdd 0.09, CVdd 0.14%
Precision (Whole Blood)	Acceptable within-instrument and total precision (%CV and SD) across different hematocrit levels at multiple sites.	Abnormal Low (< 38%PCV): Total SD 0.51, %CV 1.49 (Site 1); Total SD 0.44, %CV 1.24 (Site 2); Total SD 0.50, %CV 1.45 (Site 4) Normal (38-51%PCV): Total SD 0.51, %CV 1.13 (Site 1); Total SD 0.52, %CV 1.22 (Site 2); Total SD 0.44, %CV 1.03 (Site 4) Abnormal High (> 51%PCV): Total SD 0.48, %CV 0.88 (Site 1); Total SD 0.30, %CV 0.57 (Site 2); Total SD 0.22, %CV 0.41 (Site 4)
Linearity	Demonstration of linearity across the claimed reportable range (15 - 75 %PCV) with an appropriate regression model.	Demonstrated over the reportable range (15 - 75 %PCV). Best fitting regression was a third-order model, with non-linearity ranging from 0.19 to 0.81 %PCV. (This is generally considered acceptable if the non-linearity is small enough.)
Recovery	Acceptable percentage recovery across the reportable range.	% recovery ranged from 100.1% to 102.8%. (This is a very good recovery range, indicating accurate measurement across the range.)
Limit of Quantitation (LoQ)	LoQ should be at or below the claimed low end of the reportable range.	LoQ determined to be 14.0 %PCV, which is less than the low end of the reportable range (15 %PCV). (Meets criteria.)
Interference	Identification of compounds that do/do not interfere within specified concentrations (difference from reference > 10.8% of mean reference value).	Non-interfering: Bromide (< 37.5 mmol/L), Bilirubin (≤ 0.342 mmol/L), Sodium Thiosulfate (≤ 6.7 mmol/L), Triglyceride (≤ 37 mmol/L), White Blood Cells (21,700 WBC/μL). Interfering: Total Protein (> 12 g/dL) at 26.5-31.5 %PCV Hct level; White Blood Cells (> 50,000 WBC/µL) at 26.5-31.5 %PCV Hct level. (Crucial information for labeling and use.)
Anticoagulant Study	Comparability between heparinized and non-anticoagulated whole blood samples.	Deming regression results: slope of 1.00 and correlation coefficient of 1.00. (Indicates excellent agreement.)
Microhematocrit Reference Study	Strong correlation and negligible bias when compared to the microhematocrit reference method (K2EDTA).	Deming regression: slope of 1.02, intercept of -0.53, R² of 1.00 (vs K2EDTA reference); and slope of 1.02, intercept of -0.41, R² of 1.00 (vs K2EDTA microhematocrit). (Indicates excellent agreement.)
Method Comparison with Predicate Device	Strong correlation and agreement between the new device and the predicate device across the reportable range.	Weighted Deming regression for all 3 sites combined: slope of 1.016 and correlation coefficient (r) of 0.995. (Indicates very strong agreement, supporting substantial equivalence.)

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

For a quantitative measurement device like this, there isn't a single "test set" in the context of an AI model. Instead, various test samples are used for different analytical performance studies:

Precision (Aqueous Materials): 4 levels of aqueous materials tested over 20 days. N=80 for each level (total of 320 measurements).
Precision (Whole Blood): Venous whole blood samples (native or altered) at low abnormal, normal, and high abnormal hematocrit levels. N=21 for each level at each of 3 sites (total of 9 samples, each tested 21 times, so 9 * 21 = 189 measurements described in the table, across an unspecified total number of individual blood samples).
Linearity: A series of whole blood samples used to span the reportable range. (Specific N not provided, but typically would involve multiple dilutions).
Recovery: A series of whole blood samples spanning the reportable range. (Specific N not provided).
Limit of Quantitation (LoQ): Whole blood samples altered to low hematocrit levels (< 15 %PCV). (Specific N not provided).
Interference: Whole blood test samples. (Specific N not provided, but "Testing was conducted at two hematocrit levels" implies multiple samples for each compound).
Anticoagulant Study: 40 blood samples spanning the reportable range (15 to 75 %PCV).
Microhematocrit Reference Study: 40 lithium heparinized whole blood samples spanning the reportable range (15 to 75 %PCV).
Method Comparison with Predicate Device: 240 whole blood (venous or arterial) samples covering the reportable range (15 to 75 %PCV).

Data Provenance:

Country of Origin: Not explicitly stated, but the submission is to the U.S. FDA, and the company (Abbott Point of Care, Inc.) is based in Princeton, NJ, implying a U.S. context for the studies.
Retrospective/Prospective: These are analytical performance studies using prepared samples or patient samples collected specifically for the study. Therefore, they are prospective in nature, as the data collection and measurements were conducted specifically for the purpose of validating the device performance.
Sites:
- Precision (Aqueous Materials): One site.
- Precision (Whole Blood): 3 point-of-care sites.
- Method Comparison with Predicate Device: 3 point-of-care sites.
- Other studies: Sites not specified, but implied to be laboratory or clinical settings.

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

For a quantitative measurement device like the i-STAT Hematocrit test, "ground truth" is typically established by:

Reference Methods: Such as the microhematocrit method, which is a gold standard for hematocrit measurement.
Predicate Device Measurements: Used for method comparison studies to show equivalence.
Preparation of Spiked/Altered Samples: Where the known concentration/value is set by the preparation method.

Experts Required: The studies are primarily laboratory-based analytical evaluations. Therefore, the "ground truth" is not established by human readers/experts interpreting medical images or clinical data, but by established laboratory instrumentation and validated methods. The qualifications would be typical laboratory personnel, scientists, and statisticians who conduct and analyze such studies according to CLSI guidelines. The document does not specify a number of "experts" as it would for, e.g., radiology image interpretation.

4. Adjudication Method for the Test Set

Adjudication methods (like 2+1, 3+1) are common in studies where multiple human readers interpret data, and a consensus or majority rule is needed to establish ground truth for ambiguous cases. This is not applicable here because the ground truth for hematocrit measurements is established by quantitative reference methods and chemical/biological preparations, not by human interpretation or consensus.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done

No, an MRMC study was not done. MRMC studies are specific to evaluating devices that assist human readers in tasks like image interpretation (e.g., radiologists reading scans with or without AI). The i-STAT Hematocrit test is a standalone quantitative measurement device; it does not assist a human reader in making a diagnosis or interpretation in the same way an AI for image analysis would. Its effectiveness is assessed by its measurement accuracy, precision, and agreement with reference methods and predicate devices.

Effect Size: Not applicable as it's not a human-in-the-loop diagnostic AI.

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

Yes, the studies reported are 'standalone' performance studies for the device. The i-STAT Alinity System with the Hematocrit test cartridge takes a blood sample and provides a quantitative result. The performance characteristics described (precision, linearity, LoQ, interference, recovery, method comparisons) all evaluate the device's ability to accurately and reliably measure hematocrit on its own, without human real-time intervention or interpretation of its output in a diagnostic workflow (beyond reading the numerical result).

7. The Type of Ground Truth Used

The ground truth for the analytical performance studies was established through:

Reference Methods: Primarily the microhematocrit method (K2EDTA reference), which is a widely accepted laboratory standard for determining packed red blood cell volume.
Predicate Device Output: For method comparison studies, the results from the legally marketed i-STAT Hematocrit test with the i-STAT 1 Wireless Analyzer served as a comparative "ground truth" or reference.
Prepared Samples: For studies like linearity, recovery, and LoQ, the "ground truth" concentrations or values are precisely prepared in the laboratory.
Spiked Samples: For interference studies, samples were spiked with known concentrations of potentially interfering substances.

8. The Sample Size for the Training Set

This document describes the regulatory submission (510(k)) and performance studies for a medical device (a hematocrit measuring instrument and its test cartridge). It is not an AI/ML device in the typical sense that would require a "training set" for model development. The measurement principle is based on a conductivity method, not a learned algorithm trained on data. Therefore, the concept of a "training set" is not applicable to this device.

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

As explained above, there is no "training set" for this type of device. The ground truth for the test samples (used to evaluate performance) was established by reference methods, predicate device results, and precise laboratory preparation as detailed in point 7.

Ask a Question

Ask a specific question about this device

Page 1 of 1