cobas b 101 system: The cobas b 101 instrument is a multi-assay system designed to quantitatively analyze cobas reagent discs. The system is intended for professional, in vitro diagnostic use in a clinical laboratory setting or point-of-care (PoC) locations.

HbA1c test: The cobas HbA1c Test is an in vitro diagnostic test designed to quantitatively determine glycated hemoglobin (HbA1c) in human capillary and venous whole blood on the cobas b 101 instrument. The system is intended for professional use in a clinical laboratory setting or point-of-care (PoC) locations. Measurement of hemoglobin A1c is used to assess the level of control of a patient's diabetes and to monitor long term blood glucose control. Elevated levels of hemoglobin A1c indicate uncontrolled diabetes in a patient.

Device Description

The cobas b 101 system is a bench top analyzer which measures HbA1c. The system is fully automated, self-contained and utilizes a single use reagent disc. The system has the ability to measure capillary or venous whole blood samples. Sample is applied directly from the fingerstick or via a pipette when testing venous whole blood. The operator simply applies sample to the disc and places the disc in the instrument. There are no pre-analytics needed as the disc is self-filling by capillary forces. There is no intervention by the operator during measurement. At completion of the test, the instrument displays a quantitative result. No calculations or interpretation are required by the operator.

Calibration of the instrument is completed as part of the manufacturing process. Calibration information is contained on each disc and is read when the disc is loaded on the instrument. No calibration intervention is required by the operator.

An optional barcode scanner can be provided to read barcode information for patient identification. The barcode scanner uses LED as the light source. Results can be printed out by using an optional external printer.

A connection to a Data Management System is possible either via a USB interface to a local PC or via an Ethernet converted to a Laboratory Information Management System (LIMS). The communication protocol is defined according to the CLSI approved POCT1-A2 standard.

HbA1c (glycated hemoglobin) can be determined by using samples from capillary blood directly from the fingertip or from venous whole blood with heparin or EDTA (K2 or K3) anticoagulant. The blood sample is diluted and mixed with TRIS buffer to release hemoglobin from the erythrocytes. A fraction of the sample is conveyed into a reaction chamber where it is mixed with sodium lauryl sulfate (SLS). SLS is used to form the SLS-hemoglobin complex. The concentration of total hemoglobin is calculated by measuring SLS-hemoglobin complex with a wavelength of 525 nm. Hemoglobin A1c (HbA1c) in another fraction of the sample is first denaturated by potassium ferricyanide and sucrose laurate. The denatured HbA1c bonds with HbA1c antibody on the latex particle. Latex agglutination inhibition reaction then occurs by reacting the agglutinator that has synthetic antigen which can bond with HbA1c antibody. The concentration of HbA1c is calculated by measuring the latex agglutination inhibition reaction with a wavelength of 625 nm. The % hemoglobin A1c value is measured using a ratio of concentrations of HbA1c to total hemoglobin.

AI/ML Overview

The provided document describes the cobas b 101 system and cobas HbA1c Test, an in vitro diagnostic device for quantitative determination of glycated hemoglobin (HbA1c).

Here's an analysis of the acceptance criteria and the studies performed to meet them:

1. Table of Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state a single table of "acceptance criteria" against which all performance metrics are measured. Instead, it discusses various performance evaluations with implicit acceptance criteria (e.g., "no significant interference," "within acceptance criteria," "met the predefined acceptance criteria," "non-significant interference was defined as ≤ 10% difference").

Based on the provided information, I can construct a summary table and highlight what can be inferred as acceptance criteria for those tests:

Evaluation Category	Acceptance Criteria (Inferred from study statements)	Reported Device Performance and Remarks
Precision	Not explicitly stated, but generally looking for low CV% (coefficient of variation) and SD (standard deviation) values.	Internal Precision (n=84 per sample): Total %CV ranged from 1.3% to 2.9% for samples and 1.8% to 1.9% for controls. External Precision (n=504 combined per sample): Total %CV ranged from 1.3% to 2.7% for samples and 3.1% to 3.7% for controls. Overall, results demonstrate good precision.
Linearity/Reportable Range	Not explicitly stated, but implies correlation close to 1, slope near 1, and intercept near 0 within the claimed range. Deviation from linearity must not exceed acceptance criteria for a first-order regression.	Slope: 0.996, Intercept: -0.014, Pearson r: 0.9961. Claimed Measuring Range: 4 – 12% HbA1c. The data was considered linear after higher-order models were tested as not significant or deviations were within acceptable limits.
Hemoglobin Linearity	All results within acceptance criteria.	All results were within the acceptance criteria, supporting linear HbA1c measurements within total Hb concentrations of 6 - 20 g/dL. (Range tested: ~4.5 - 14% HbA1c, high total Hb up to ~22 g/dL).
Endogenous Interferences	No significant interference found up to tested concentrations.	No significant interference found up to tested concentrations (e.g., Albumin 77.5 g/L, Bilirubin 85 mg/dL, Lipemia 750 mg/dL, Glucose 2800 mg/dL, RF 1200 IU/mL, Total Protein 126 g/L). Labeling claims reflect no significant interference up to slightly lower ranges for most substances.
Exogenous Interferences (Drugs)	No interference found at therapeutic concentrations using common drug panels.	No interference was found at therapeutic concentrations for a panel of 15 common drugs (concentrations listed in Table 6).
Cross-Reactivity (Carbamylated Hb, Acetylated Hb, Labile HbA1c)	No relevant cross-reactivity found up to the listed concentrations.	No relevant cross-reactivity found up to 3000 mg/dL for all three, which are significantly higher than physiologically occurring concentrations.
Cross-Reactivity (HbA0)	No cross-reactions with HbA0 at physiologically occurring concentrations.	A correlation analysis between cobas b 101 and cobas c 501 showed that at physiologically occurring concentrations, no cross-reactions with HbA0 were found.
Cross-Reactivity (HbA1a, HbA1b)	Results met the predefined acceptance criteria. No cross-reactions at physiologically occurring concentrations.	Results met predefined acceptance criteria. This supports the claim that at physiologically occurring concentrations, no cross-reactions with HbA1a and HbA1b were found.
Hemoglobin Variants	Non-significant interference was defined as ≤ 10% difference between candidate and reference methods.	Results met defined acceptance criteria for HbAS, HbAC, HbAE, and HbA2. HbF interference was excluded up to 10% concentration. (Labeling includes limitations for HbF > 10%).
Method Comparison (vs. Reference)	Not explicitly stated, but a good correlation (Pearson's r close to 1) and linear regression (slope close to 1, intercept close to 0) are implicit.	Capillary, EDTA K2, Lithium Heparin: For all three sample types across three sites, Pearson's r was 0.99. Regression lines were very close to y=x (slopes of 0.97-1.00, intercepts of -0.20 to 0.04). The reference method was the Tosoh G8 HPLC Analyzer.
Matrix Comparison	Not explicitly stated, but a good correlation (Pearson's r close to 1) and linear regression (slope close to 1, intercept close to 0) are implicit.	EDTA (K2) vs EDTA (K3) whole blood: N=91, Regression Line y = 1.03x - 0.00, Pearson's r = 0.99.
Stability	Acceptable and supports the claimed stability.	Studies were found to be acceptable and support the claimed stability of 16 months at 2-30 °C for the reagent disks.

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

Precision (Internal): 8 K2EDTA or K3EDTA venous whole blood samples and 2 control samples. Each sample measured 84 times (duplicate in 2 runs per day for 21 days).
Precision (External): 4 human sample pools and 3 lots of controls. Each assessed 168 times (duplicate 2x/day for 21 days per site), for a total of 504 measurements per sample/control pool combined across 3 sites.
Linearity/Assay Reportable Range: 11 dilutions of a low (3.6% HbA1c) and high (12.9% HbA1c) K2EDTA whole blood sample. (Number of replicates per dilution not specified but implied to be sufficient for CLSI EP06-A).
Hemoglobin Linearity: 4 K2EDTA and Li-Heparin venous whole blood samples. For each, a 10-level dilution series (using NaCl as diluent) was prepared and tested in triplicate.
Endogenous Interferences: Two HbA1c levels (normal and pathological range) for each interferent. Sample pools were K2EDTA venous whole blood. Varying dilutions mixed to create dilution series. (Number of replicates per dilution not specified but implied to be sufficient for CLSI EP07-A2).
Exogenous Interferences (Drugs): K2EDTA venous whole blood pools. Each sample tested in 5 replicates.
Cross-Reactivity (Carbamylated Hb, Acetylated Hb, Labile HbA1c): K2EDTA venous whole blood samples at normal and pathological HbA1c levels. Each dilution level tested in singlicate on 3 instruments.
Cross-Reactivity (HbA0): Approximately 60 K2EDTA whole blood samples.
Cross-Reactivity (HbA1a, HbA1b): 11 K2EDTA samples of varying HbA1a+b concentrations. Each measured in triplicate on the cobas b 101 instrument.
Hemoglobin Variants: 130 samples total, covering HbAS (20), HbAC (20), HbAD (20), HbAE (20), Elevated F (20), Elevated A2 (10). Each sample tested once.
Method Comparison: Capillary whole blood (N=125, 133, 121 per site), EDTA (K2) whole blood (N=125, 133, 121 per site), Lithium Heparin whole blood (N=125, 130, 117 per site). All from prospective blood sampling. Tested in singlicate.
Matrix Comparison: 91 samples of EDTA (K2) and EDTA (K3) whole blood. Tested in singlicate.

Data Provenance: Most studies use human blood samples, either K2EDTA/K3EDTA venous whole blood or capillary whole blood. The External Precision and Method Comparison studies were conducted at "3 Point-of-Care sites," implying real-world or near-real-world clinical settings, likely within the country where the studies were conducted (not specified, but often the manufacturer's location or major market). The data appears to be prospective for the clinical performance (Method Comparison) as it mentions "prospective blood sampling" and "No contrived samples were tested." For other non-clinical performance studies (e.g., interferences, linearity), samples were prepared according to CLSI guidelines, which may involve spiking or diluting samples to achieve specific concentrations.

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

The document does not mention the use of "experts" to establish ground truth in the traditional sense of consensus reading for image-based diagnostics. For HbA1c measurements, the ground truth is established by reference methods or predicate devices, which are themselves highly standardized and validated laboratory instruments.

For Method Comparison vs. Reference, the reference method used was the Tosoh G8 HPLC Analyzer. This is a well-established and validated laboratory instrument for HbA1c measurement.
For Hemoglobin Variants, reference methods were "selected based on no interference for a particular variant." The specific reference methods for each variant are not listed, but they are implied to be established, accurate methods.
For Cross-Reactivity (HbA0), the reference system was the cobas c 501.

Therefore, the "ground truth" is analytical, derived from established laboratory methods, not expert consensus on qualitative interpretation.

4. Adjudication Method (for the test set)

Not applicable. This device provides quantitative measurements of HbA1c. "Adjudication method" typically refers to how disagreements among multiple human readers are resolved in qualitative or semi-quantitative assessments (e.g., radiology reads). Here, the comparison is against an objective reference measurement.

5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was done

No, an MRMC comparative effectiveness study was not done. This type of study is primarily relevant for diagnostic imaging or similar scenarios where human readers interpret cases, and the AI's impact on reader performance (e.g., accuracy, efficiency) is evaluated. The cobas b 101 system and cobas HbA1c Test is an in vitro diagnostic device for quantitative chemical analysis, not a device interpreted by human readers in the same manner.

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

Yes, the entire document describes the standalone performance of the cobas b 101 system and cobas HbA1c Test. All the precision, linearity, interference, cross-reactivity, and method comparison studies were performed on the device itself comparing its quantitative output to established reference methods or accepted criteria. There is no human-in-the-loop component in the direct measurement and reporting of HbA1c values by the device. The "user" interaction involves sample application, but the analytical process is fully automated within the instrument.

7. The Type of Ground Truth Used

The ground truth used is primarily analytical measurements from established reference methods.

Tosoh G8 HPLC Analyzer: Used as the reference method for the main method comparison studies for capillary, EDTA (K2), and Lithium Heparin whole blood.
cobas c 501: Used as a reference system for HbA0 cross-reactivity.
Selected reference methods: Used for hemoglobin variant testing.
Calculated values: For linearity studies and some cross-reactivity studies, target concentrations were prepared by dilution from known high/low samples.

8. The Sample Size for the Training Set

The document does not explicitly mention a "training set" for the cobas b 101 system or the cobas HbA1c Test. This is typical for an IVD device like this, which likely uses a rule-based algorithm and/or pre-calibrated sensor systems rather than a machine learning model that requires a distinct training phase. The "calibration information is contained on each disc and is read when the disc is loaded on the instrument," suggesting an internal calibration process defined by the manufacturer rather than a data-driven training set used by the end-user.

If there were any internal development or calibration data used by the manufacturer during the design of the reagent or instrument, it is not disclosed as a "training set" in this 510(k) summary.

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

As no specific "training set" is identified for a machine learning model, this question is not directly applicable. For the development and calibration of the device (if that's considered analogous to "training"), the ground truth would have been established through a rigorous process of using known standards, calibrators, and reference materials, similar to how the "ground truth" for the test set was established (i.e., against established analytical reference methods and validated materials). The document states the "Calibration of the instrument is completed as part of the manufacturing process," implying this process.

Summary

{0}------------------------------------------------

Image /page/0/Picture/1 description: The image shows the logo for the U.S. Department of Health & Human Services. The logo consists of a stylized caduceus symbol, which is a staff with two snakes coiled around it. The caduceus is positioned to the right of the text "DEPARTMENT OF HEALTH & HUMAN SERVICES - USA", which is arranged in a circular fashion around the symbol.

Food and Drug Administration 10903 New Hampshire Avenue Document Control Center - WO66-G609 Silver Spring, MD 20993-0002

July 28, 2017

ROCHE DIAGNOSTICS OPERATIONS PATTY BATES REGULATORY AFFAIRS PRINCIPAL 9115 HAGUE ROAD INDIANAPOLIS, IN 46250

Re: K163633

Trade/Device Name: cobas HbA1c Test, cobas b 101 System Regulation Number: 21 CFR 864.7470 Regulation Name: Glycosylated hemoglobin assay Regulatory Class: II Product Code: LCP. JJE Dated: June 15, 2017 Received: June 16, 2017

Dear Patty Bates:

We have reviewed your Section 510(k) premarket notification of intent to market the device referenced above and have determined the device is substantially equivalent (for the indications for use stated in the enclosure) to legally marketed predicate devices marketed in interstate commerce prior to May 28, 1976, the enactment date of the Medical Device Amendments, or to devices that have been reclassified in accordance with the provisions of the Federal Food. Drug. and Cosmetic Act (Act) that do not require approval of a premarket approval application (PMA). You may, therefore, market the device, subject to the general controls provisions of the Act. The general controls provisions of the Act include requirements for annual registration, listing of devices, good manufacturing practice, labeling, and prohibitions against misbranding and adulteration. Please note: CDRH does not evaluate information related to contract liability warranties. We remind you, however, that device labeling must be truthful and not misleading.

If your device is classified (see above) into either class II (Special Controls) or class III (PMA), it may be subject to additional controls. Existing major regulations affecting your device can be found in the Code of Federal Regulations, Title 21, Parts 800 to 898. In addition, FDA may publish further announcements concerning your device in the Federal Register.

Please be advised that FDA's issuance of a substantial equivalence determination does not mean that FDA has made a determination that your device complies with other requirements of the Act or any Federal statutes and regulations administered by other Federal agencies. You must comply with all the Act's requirements, including, but not limited to: registration and listing (21 CFR Part 807); labeling (21 CFR Parts 801 and 809); medical device reporting (reporting of medical device-related adverse events) (21 CFR 803); good manufacturing practice requirements as set forth in the quality systems (QS) regulation (21 CFR Part 820); and if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR 1000-1050.

{1}------------------------------------------------

Page 2-Ms. Bates

If you desire specific advice for your device on our labeling regulations (21 CFR Parts 801 and 809), please contact the Division of Industry and Consumer Education at its toll-free number (800) 638 2041 or (301) 796-7100 or at its Internet address

http://www.fda.gov/MedicalDevices/Resourcesfor You/Industry/default.htm. Also, please note the regulation entitled. "Misbranding by reference to premarket notification" (21 CFR Part 807.97). For questions regarding the reporting of adverse events under the MDR regulation (21 CFR Part 803), please go to

http://www.fda.gov/MedicalDevices/Safety/ReportaProblem/default.htm for the CDRH's Office of Surveillance and Biometrics/Division of Postmarket Surveillance.

You may obtain other general information on your responsibilities under the Act from the Division of Industry and Consumer Education at its toll-free number (800) 638-2041 or (301) 796-7100 or at its Internet address

http://www.fda.gov/MedicalDevices/ResourcesforYou/Industry/default.htm.

Sincerely yours,

Kellie B. Kelm -S

for Courtney H. Lias, Ph.D. Director Division of Chemistry and Toxicology Devices Office of In Vitro Diagnostics and Radiological Health Center for Devices and Radiological Health

Enclosure

{2}------------------------------------------------

Indications for Use

510(k) Number (if known)

K163633

Device Name cobas b 101 system cobas HbA1c Test

Indications for Use (Describe)

HbA1c test: The cobas HbA1c Test is an in vitro diagnostic test designed to quantitatively determine glycated hemoglobin (HbA ic) in human capillary and venous whole blood on the cobas b 101 instrument. The system is intended for professional use in a clinical laboratory setting or point-of-care (PoC) locations. Measurement of hemoglobin A 1c is used to assess the level of control of a patient's diabetes and to monitor long term blood glucose control. Elevated levels of hemoglobin Alc indicate uncontrolled diabetes in a patient.

Type of Use (Select one or both, as applicable)

X Prescription Use (Part 21 CFR 801 Subpart D)

| Over-The-Counter Use (21 CFR 801 Subpart C)

CONTINUE ON A SEPARATE PAGE IF NEEDED.

This section applies only to requirements of the Paperwork Reduction Act of 1995.

DO NOT SEND YOUR COMPLETED FORM TO THE PRA STAFF EMAIL ADDRESS BELOW.

The burden time for this collection of information is estimated to average 79 hours per response, including the time to review instructions, search existing data sources, gather and maintain the data needed and complete and review the collection of information. Send comments regarding this burden estimate or any other aspect of this information collection, including suggestions for reducing this burden, to:

Department of Health and Human Services Food and Drug Administration Office of Chief Information Officer Paperwork Reduction Act (PRA) Staff PRAStaff(@fda.hhs.gov

"An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB number."

{3}------------------------------------------------

cobas b 101 system 510(k) Summary K163633

This summary of 510(k) safety and effectiveness information is being submitted in accordance with the requirements of 21 CFR 807.92.

In accordance with 21 CFR 807.87, Roche Diagnostics hereby submits official notification as required by Section 510(k) of the Federal Food, Drug and Cosmetics Act of our intention to market the device described in this Premarket Notification 510(k).

The purpose of this Traditional 510(k) Premarket Notification is to obtain FDA review and clearance for the cobas b 101 system and cobas HbA1c Test.

{4}------------------------------------------------

Submitter Name	Roche Diagnostics
Address	9115 Hague RoadP.O. Box 50416Indianapolis, IN 46250-0457
Contact	Patty BatesPhone: (317) 521-4572FAX: (317) 521-2324Email: patty.bates@roche.comTracy BushPhone: (317) 521-3723FAX: (317) 521-2324Email: tracy.bush@roche.com
Date Prepared	November 30, 2016
Proprietary Name	cobas HbA1c Testcobas b 101 system
Common Name	Test system for HbA1cDiscrete photometric chemistry analyzer for clinical use
Classification Name	Glycosylated hemoglobin assay – Class IIAnalyzer, chemistry (Photometric, discrete), for clinical use – Class I
Product Codes,Regulation Numbers	LCP, 21 CFR 864.7470JJE, 21 CFR 862.2160
Predicate Devices	DCA Vantage

{5}------------------------------------------------

1. DEVICE DESCRIPTION

1.1. System: cobas b 101 system

Reagent: cobas HbA1c Test 1.2.

{6}------------------------------------------------

denaturated by potassium ferricyanide and sucrose laurate. The denatured HbA1c bonds with HbA1c antibody on the latex particle. Latex agglutination inhibition reaction then occurs by reacting the agglutinator that has synthetic antigen which can bond with HbA1c antibody. The concentration of HbA1c is calculated by measuring the latex agglutination inhibition reaction with a wavelength of 625 nm. The % hemoglobin A1c value is measured using a ratio of concentrations of HbA1c to total hemoglobin.

2. INDICATIONS FOR USE

2.1. System: cobas b 101 system

The cobas b 101 instrument is a multi-assay system designed to quantitatively analyze cobas reagent discs. The system is intended for professional, in vitro diagnostic use in a clinical laboratory setting or point-of-care (PoC') locations.

Reagent: cobas HbA1c Test 2.2.

The cobas HbA1c Test is an in vitro diagnostic test designed to quantitatively determine glycated hemoglobin (HbA1c) in capillary finger-stick or venous whole blood, collected in EDTA (K2 or K3) or lithium heparin tubes, on the cobas b 101 instrument. This test is intended for professional use in a clinical laboratory setting or point-of-care (PoC) locations. This test is not for screening or diagnosis of diabetes or neonatal use. Measurement of hemoglobin A1c is used to monitor long term blood glucose control in patients previously diagnosed with diabetes.

TECHNOLOGICAL CHARACTERISTICS 3.

The following tables compare the cobas b 101 system with its predicate devices.

Candidate Device Name	Predicate Device Name	K-Number
cobas b 101 system and cobasHbA1c Test	DCA Vantage	K071466

{7}------------------------------------------------

Feature	Predicate Device: DCA Vantage (K071466)	Candidate Device: cobas b 101system/cobas HbA1c Test
Instrument Intended Use	The DCA VantageTM is a semi-automated,benchtop system. It is designed toquantitatively measure the percentHemoglobin A1c in blood and lowconcentrations of albumin in urine(microalbuinuria), creatinine in urine, andalbumin/creatinine ratio in urine.	The cobas b 101 instrument is amulti-assay system designed toquantitatively analyze cobasreagent discs. The system isintended for professional, in vitrodiagnostic use in a clinicallaboratory setting or point-of-care(PoC) locations.
Assay Intended Use	This assay provides a convenient,quantitative method for measuring thepercent concentration of hemoglobin A1c inblood.	The cobas HbA1c Test is an invitro diagnostic test designed toquantitatively determine glycatedhemoglobin (HbA1c) in capillaryfinger-stick or venous wholeblood, collected in EDTA (K2 orK3) or lithium heparin tubes, onthe cobas b 101 instrument. Thistest is intended for professionaluse in a clinical laboratory settingor point-of-care (PoC) locations.This test is not for screening ordiagnosis of diabetes or neonataluse.
Assay Indications for Use	The measurement of hemoglobin A1cconcentration is recommended formonitoring the long-term care of personswith diabetes.	Measurement of hemoglobin A1cis used to monitor long termblood glucose control in patientspreviously diagnosed withdiabetes.
Assay Method	Latex agglutination - inhibition immunoassay	Same
Detection Method	photometry	Same
Applications/Test Time	6.5 minutes	<6 minutes
Test Platform	Single use	Same
Sample Type/Matrix	Whole blood: capillary and venous	Same
Sample Anticoagulants	EDTA, heparin, fluoride/oxalate, citrate	EDTA, lithium heparin
Sample Total Hemoglobin Range	7-24 g/dL	6-20 g/dL
Sample Application	Sample collected in capillary holder. Holderinserted into cartridge.	Sample collected directly to thedisc or by transfer of sampleusing pipette or dropper.
Sample Volume	1 µL	2 µL
Reagent Test Principle	Monoclonal antibody agglutination reaction;intensity of light transmitted is measured byspectrophotometry.	Latex agglutination - inhibitionimmunoassay reaction todetermine HbA1c by change oftransmission of light; extent oflight scatter is measured byspectrophotometry.
Feature	Predicate Device: DCA Vantage (K071466)	Candidate Device: cobas b 101system/cobas HbA1c Test
Calibration Method	Calibration card for each reagent lot; nocalibration by operator	Calibration information read fromeach reagent disc; no calibrationby operator
Traceability/Standardization	IFCC, NGSP	Same
Reagent Stability	3 months at room temp	Store at 2-30 °C (36-86 °F) for 16months
Operating Temperature Range	15-32 °C61-88 °F	15-32 °C59-90 °F
Measuring Range	2.5-14% HbA1c	4-12% HbA1c

Table 1: Instrument and Assay Comparison

{8}------------------------------------------------

NON-CLINICAL PERFORMANCE EVALUATION 4.

The following internal performance data were provided in support of the substantial equivalence determination:

Precision •
Linearity •
Hemoglobin Linearity •
Endogenous Interferences Bilirubin/Lipemia/Glucose/Rheumatoid Factor •
Exogenous Interferences Drugs •
Cross-reactivity Carbamylated Hemoglobin, Acetylated Hemoglobin, Labile HbA1c, • HbA0, HbA1a, HbA1b
Hemoglobin Variants •
Stability

Precision 4.1.

Internal precision of the cobas HbA1c Test was evaluated according to CLSI EP05-A3 on one cobas b 101 instrument with 2 cobas HbA1c Test disc lots and 1 cobas HbA1c Control lot. Two control samples and 8 K2EDTA or K3EDTA venous whole blood samples were measured over 21 days. The protocol consisted of measuring the sample material in duplicate in 2 runs per day

{9}------------------------------------------------

for 21 days producing n=84 results per sample. Samples and runs were randomized each day. Table 3 below summarizes the precision results.

			Repeatability		Between run		Between day		Total
Sample	N	Mean (%)	SD	% CV	SD	% CV	SD	% CV	SD	% CV
Sample 1	84	5.2	0.14	2.8	0.00	0.0	0.04	0.8	0.15	2.9
Sample 2	84	5.5	0.10	1.8	0.03	0.6	0.05	1.0	0.12	2.1
Sample 3	84	7.8	0.13	1.7	0.00	0.0	0.04	0.5	0.14	1.7
Sample 4	84	10.3	0.11	1.1	0.00	0.0	0.06	0.5	0.13	1.2
Sample 5	84	5.0	0.08	1.7	0.00	0.0	0.05	1.0	0.10	2.0
Sample 6	84	5.6	0.07	1.2	0.02	0.3	0.04	0.7	0.08	1.5
Sample 7	84	6.9	0.11	1.6	0.00	0.0	0.05	0.8	0.12	1.8
Sample 8	84	8.2	0.10	1.2	0.02	0.3	0.03	0.4	0.11	1.3
HbA1c Control Level 1	84	5.1	0.08	1.7	0.02	0.4	0.05	0.9	0.10	1.9
HbA1c Control Level 2	84	9.3	0.10	1.0	0.04	0.5	0.13	1.3	0.16	1.8

Table 2: Internal Precision Data

Note: SD of zero due to variance contributed by particular component was below stated significant figure.

External precision of the cobas b 101 HbA1c assay was performed according to CLSI EP05-A3 at 3 Point-of-Care sites. At each site, the samples were measured in duplicate 2x/day for 21 days for a total of n = 84 measurements. Six cobas b 101 instruments with 3 reagent disc lots were used in the study. Each PoC site assessed 2 reagent disc lots on 2 cobas b 101 instruments. Three lots of controls and 4 human sample pools were tested. The results are summarized in the table below.

Table 3: External Precision Data
-----------------------------------

				Repeatability		Between run		Between day		Total
Sample	Site	N	Mean (%)	SD	% CV	SD	% CV	SD	% CV	SD	% CV
Sample 1	Site 1	168	5.3	0.06	1.2	0.04	0.9	0.04	0.7	0.09	1.6
	Site 2	168	5.2	0.07	1.4	0.00	0.0	0.05	0.9	0.09	1.7
	Site 3	168	5.1	0.08	1.5	0.03	0.6	0.02	0.3	0.09	1.7
	Combined	504	5.2	0.07	1.4	0.03	0.6	0.06	1.1	0.10	1.9
Sample 2	Site 1	168	5.4	0.07	1.3	0.03	0.5	0.06	1.2	0.10	1.8

{10}------------------------------------------------

Sample	Site	N	Mean (%)	Repeatability		Between run		Between day		Total
				SD	% CV	SD	% CV	SD	% CV	SD	% CV
	Site 2	168	5.3	0.07	1.3	0.04	0.8	0.03	0.6	0.09	1.6
	Site 3	168	5.3	0.08	1.5	0.03	0.7	0.02	0.4	0.09	1.7
	Combined	504	5.3	0.07	1.4	0.04	0.7	0.06	1.1	0.10	1.9
Sample 3	Site 1	168	8.0	0.13	1.6	0.13	1.6	0.00	0.0	0.18	2.3
	Site 2	168	7.9	0.10	1.3	0.04	0.5	0.06	0.7	0.12	1.5
	Site 3	168	8.0	0.24	3.0	0.08	1.0	0.16	2.0	0.30	3.8
	Combined	504	8.0	0.17	2.1	0.09	1.1	0.11	1.3	0.22	2.7
Sample 4	Site 1	168	10.9	0.11	1.0	0.01	0.1	0.04	0.3	0.11	1.0
	Site 2	168	10.8	0.13	1.2	0.08	0.8	0.00	0.0	0.16	1.4
	Site 3	168	10.8	0.12	1.1	0.02	0.2	0.06	0.5	0.13	1.2
	Combined	504	10.8	0.12	1.1	0.05	0.5	0.06	0.6	0.14	1.3
Control Level 1	Site 1	168	5.2	0.10	2.0	0.09	1.8	0.17	3.2	0.22	4.2
	Site 2	168	5.1	0.09	1.8	0.05	1.0	0.17	3.4	0.20	4.0
	Site 3	168	5.2	0.13	2.5	0.00	0.0	0.07	1.3	0.14	2.8
	Combined	504	5.2	0.11	2.1	0.06	1.1	0.14	2.8	0.19	3.7
Control Level 2	Site 1	168	10.2	0.17	1.7	0.13	1.2	0.26	2.6	0.34	3.3
	Site 2	168	10.1	0.14	1.4	0.19	1.9	0.14	1.4	0.27	2.7
	Site 3	168	9.9	0.15	1.6	0.08	0.8	0.18	1.8	0.25	2.5
	Combined	504	10.1	0.16	1.5	0.14	1.4	0.23	2.3	0.31	3.1

Linearity/Assay Reportable Range 4.2.

The linearity study was conducted to demonstrate that measurements across the claimed measuring range for the cobas HbA1c Test are linear. The study was performed according to CLSI guideline EP06-A.

The linearity study was performed on the cobas b 101 system by measuring 11 dilutions of a low (3.6% HbA1c) and high (12.9% HbA1c) K2EDTA whole blood sample. The linearity data were analyzed with regards to linear, quadratic and cubic polynomials according to CLSI EP06-A. In a first step, a linearity check was performed with a first order (linear) regression and then with higher order models (quadratic and cubic). The coefficients obtained for the higher order models were then tested for significance. In the event that the coefficients obtained for the higher order

{11}------------------------------------------------

models were not significant, the data was considered linear. Otherwise, non-linearity was detected.

In the latter case, the higher order model must not deviate more than the acceptance criteria for the first order regression at the corresponding concentration values. Calculation of deviation from linearity was based on the third order model. The linearity results are in the table below:

Table 4: Linearity Results

Slope	Intercept	Pearson r	Claimed Measuring Range
0.996	-0.014	0.9961	4 – 12% HbA1c

Hemoglobin Linearity 4.3.

The hemoglobin linearity study was conducted to assess the linearity of HbA 1c measurements throughout the specified total hemoglobin concentration range of the cobas HbA1c Test.

One lot of reagent was utilized in this study. Four K2EDTA and Li-Heparin venous whole blood samples across the HbA1c concentration range of approx. 4.5 - 14% HbA1c and a high total hemoglobin concentration of approx. 22g/dL were used. For each sample, a 10-level dilution series was prepared using NaCl as diluent.

Each sample and dilution level was tested in triplicate. The median of these 3 replicates was used for calculations. In addition to the calculated Hb-concentration, each dilution level sample was measured on the Sysmex KX-21 (FDA clearance K981761).

All results are within the acceptance criteria. The cobas b 101 measures HbA1c in presence of total hemoglobin concentrations between 6 - 20 g/dL.

Analytical Sensitivity 4.4.

Please see linearity study.

{12}------------------------------------------------

Endogenous Interferences 4.5.

Albumin/Bilirubin/Lipemia/Glucose/Rheumatoid Factor/ /Total Protein 4.5.1.

The effect on quantitation of HbA1c in the presence of endogenous interfering substances was determined on the cobas b 101 system using the cobas HbA1c Test. The interference study was performed according to CLSI guideline EP07-A2. Sample pool material was KJEDTA venous whole blood with a concentration of approximately 5.5 and > 9.0 % HbA1c.

For each interferent, a separate stock solution containing the potentially interfering material was prepared. All pools were prepared using KsEDTA venous whole blood. Two HbA1c levels, one in the normal and another in the pathological range, were evaluated for each of the endogenous substances.

To achieve the target high concentration interferent pool (high sample pool), pooled whole blood was spiked with the interferent. Another pool (low sample pool), without interferent, was created by dilution of the pooled whole blood with the same volume of diluent as the high sample pool. This pool contains no interferent and serves as the reference pool for the testing. The high and low sample pools were mixed in different ratios to vield a dilution series with varying concentrations of the interferent.

There was no significant interference found up to the tested concentrations. The table below summarizes the results and claims for the endogenous substances.

EndogenousSubstance	Highest Level Tested with NoSignificant Interference	Labeling Claim:No significant interference up to
Albumin	77.5 g/L	60 g/L
Conjugated Bilirubin	85 mg/dL	60 mg/dL
Unconjugated Bilirubin	85 mg/dL	60 mg/dL
Lipemia / Intralipid	750 mg/dL	500 mg/dL
Lipemia / Native Triglycerides	694 mg/dL	500 mg/dL
Glucose	2800 mg/dL	2000 mg/dL
Rheumatoid Factor (RF)	1200 IU/mL	750 IU/mL
Total Protein	126 g/L	120 g/L

Table 5: Potentially Interfering Endogenous Substances Results Summary

{13}------------------------------------------------

Exogenous Interferences – Drugs 4.6.

The effect on quantitation of HbA1c in the presence of exogenous interfering substances was determined on the cobas b 101 system with the cobas HbA1c Test.

The table below lists the potential exogenous interferents and the concentrations tested:

Drug	Highest Level Tested with NoSignificant Interference (mg/L)
Acetyl Cysteine	1660
Ampicillin-Na	1000
Ascorbic acid	300
Cyclosporine	5
Cefoxitin	2500
Heparin	5000 U/L
Levodopa	20
Methyldopa +1,5	20
Metronidazole	200
Phenylbutazone	400
Doxycycline	50
Acetylsalycilic acid	1000
Rifampicin	60
Acetaminophen	200
Ibuprofen	500
Theophylline	100

Table 6: Potentially Interfering Drugs and Test Concentrations

The interference study was performed according to CLSI guideline EP07-A2. For each substance, a separate stock solution containing the potentially interfering drug was prepared. All pools were prepared using K2EDTA venous whole blood. To achieve the target concentration drug pool, pooled whole blood was spiked with the drug. Another pool, without drug, was created by dilution of the pooled whole blood with the same volume of diluent as the high concentration drug pool. This pool contains no drug and serves as the reference pool for the testing. Each sample was tested in 5 replicates and the mean value used for the assessment. No interference was found at therapeutic concentrations using common drug panels.

{14}------------------------------------------------

Cross-Reactivity 4.7.

Carbamylated Hemoglobin, Acetylated Hemoglobin and Labile HbA1c 4.7.1.

The effect on quantitation of HbA1c in the presence of potential cross-reactants was determined on the cobas b 101 system with the cobas HbA1c Test. The study was performed according to CLSI guideline EP07-A2. Sample pool material was K₂EDTA venous whole blood with a concentration of approximately 5.5 and > 9.0 % HbA1c.

To prepare the carbamylated Hb high sample pool, sodium cyanate (final concentration is 3000 mg/dL) in isotonic saline is added to the whole blood sample. For the low sample pool, the same volume of isotonic saline (no sodium cyanate) is added to the whole blood sample.

To prepare the acetylated Hb high sample pool, acetylsalicylic acid in 75% Ethyl Alcohol (final concentration is 3000 mg/dL) is added to the whole blood sample. For the low sample pool, the same volume of 75% Ethyl Alcohol (no acetylsalicylic acid) is added to whole blood sample.

To prepare the labile HbA1c high sample pool, glucose in isotonic saline (final concentration is 3000 mg/dL) is added to the whole blood sample and incubated for one hour at 37°C. For the low sample pool, the same volume of isotonic saline (no glucose) is added to the whole blood sample.

For each interferent, a separate stock solution containing the potentially interfering cross-reactant was prepared. All pools were prepared using K2EDTA venous whole blood. Two HbA1c levels, one in the normal and another in the pathological range, were evaluated for each of the crossreactants.

To achieve the target high concentration cross-reactant pool (high sample pool), pooled whole blood was spiked with the cross-reactant. Another pool (low sample pool), without crossreactant, was created by dilution of the pooled whole blood with the same volume of diluent as the high sample pool. This pool contains no cross-reactant and serves as the reference pool for the testing. The high and low sample pools were mixed in different ratios to yield a dilution series with varying concentrations of the cross-reactant. Each dilution level was tested in singlicate on 3 instruments and the median was used for calculations.

{15}------------------------------------------------

No relevant cross-reactivity was found up to the listed concentrations. The table below summarizes the results and claims for the hemoglobin derivative cross-reactants.

Cross-reactant	Highest Level Tested with NoSignificant Interference	Physiologically occurringconcentration
Carbamylated Hb	3000 mg/dL	750 -1100 mg/dL
Acetylated Hb	3000 mg/dL	320 - 500 mg/dL
Labile HbA1c	3000 mg/dL	430 - 650 mg/dL

Table 7: Hemoglobin Derivative Cross-reactant Results and Claims

4.7.2. HbA0

A correlation analysis was performed. Correlation was determined between increasing amounts of hemoglobin (and hence HbA0) and the % relative bias of measured HbA1c using the cobas b 101 and the reference system cobas c 501. The HbA1c values were measured from approximately 60 K2EDTA whole blood samples with cobas b 101 and cobas c 501 as the reference system. Total hemoglobin was measured with the reference system cobas c 501.

The relative bias between HbA1c results from cobas b 101 and cobas c 501 was calculated and plotted against the measured amounts of hemoglobin. A linear regression and the corresponding Pearson's correlation coefficient was calculated.

The data supports the claims that at physiologically occurring concentrations, no cross reactions with HbA0 were found.

HbA1a and HbA1b 4.7.3.

A separate stock solution containing the potential cross-reactant HbA1a+b was prepared. This stock solution consisted of ~10.35 mg/ml HbA1a+b diluted in buffer. Ten dilutions of the stock were prepared at varying HbA1a+b concentrations. These samples were mixed with a normal and pathological K2EDTA whole blood sample to create 11 K2EDTA samples of varying HbA1a+b concentrations.

{16}------------------------------------------------

All K2EDTA samples were measured in triplicate on the cobas b 101 instrument resulting in n=33 results per sample type (normal and pathological). The median value for each sample was used for calculation.

Because the instrument only reports the HbA 1c concentrations and not the individual concentrations of A1c and total Hemoglobin, the expected HbA1c concentrations were calculated based on the c501 results for A1c and Hb total.

The results obtained met the predefined acceptance criteria. These results support the crossreactivity claims in the labeling that at physiologically occurring concentrations, no cross reactions with HbA1a and HbA1b were found.

4.8. Hemoglobin Variants

Hemoglobin variant testing was conducted to determine if there is any significant interference with the major hemoglobin variants and the cobas HbA1c Test. Hemoglobin variants are structurally altered hemoglobin molecules with at least one amino acid exchange compared to the normal beta chain of hemoglobin. These changes are caused by mutations in the coding region of the globin genes which encode the protein part of hemoglobin. The most common hemoglobin variants are HbS, HbC, HbD and HbE. Moreover, in some conditions, the fetal hemoglobin HbF is elevated. Also, the erythrocytes of some patients (e.g. beta thalassemia minor) contain elevated levels of HbA2. It is crucial to ensure accurate HbA1c results from patients who are carriers of these variants.

Variant Type	Number of Samples	% Variant
AS	20	29 - 41%
AC	20	28 - 36%
AD	20	36 - 42%
AE	20	20 - 27%
Elevated F	20	3 - 29%
Elevated A2	10	4.3 - 6.2%
Total	130

Table 8: Hemoglobin Variant Samples

Each sample was tested once in at least one run on 1 cobas b 101 instrument. A total of 130 samples were measured with a sample range of 4.35-11.58% HbA1c. Reference methods were

{17}------------------------------------------------

selected based on no interference for a particular variant. Results obtained were compared to those obtained with the reference methods. Mean values of triplicate measurements on the reference systems were used for comparison. Non-significant interference was defined as ≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤≤ 10% difference between the candidate and reference methods.

The results obtained met the defined acceptance criteria for HbAS, HbAC, HbAE and HbA2. HbF interference can be excluded up to an HbF concentration of 10%.

The sponsor has the following limitations in the labeling:

Heterozygous presence of the most common hemoglobin variants (HbAS, HbAD, HbAE, HbA2) does not interfere.

Specimens containing high amounts of HbF (> 10 %) may result in lower than expected % HbA1c values (DCCT/NGSP).

Traceability 4.9.

The cobas HbA1c Test has been standardized against the approved IFCC reference method for the measurement of hemoglobin A1c in human blood. The cobas b 101 instrument reports values in % hemoglobin A1c traceable to DCCT/NGSP by calculation. The cobas b 101 is certified with the National Glycohemoglobin Standardization Program (NGSP). The certification expires in one year. See NGSP website for current certification at http://www.ngsp.org.

4.10. Stability

The studies were performed according to CLSI guideline EP025-A. The cobas reagent disks can be stored at 2-30 °C for 16 months. The stability studies were found to be acceptable and support the claimed stability.

4.11. Expected Values

In 2016, the American Diabetes Association (ADA) recommended a reasonable A1c goal for many nonpregnant adults is < 7 % (53 mmol/mol). Providers might reasonably suggest more stringent A1C goals (such as 6.5 % (48 mmol/mol)) for selected individual patients if this can be achieved without significant hypoglycemia or other adverse effects of treatment. Appropriate patients might include those with short duration of diabetes, type 2 diabetes treated with lifestyle

{18}------------------------------------------------

or metformin only, long life expectancy, or no significant cardiovascular disease. Less stringent A 1C goals (such as 8 % (64 mmol/mol)) may be appropriate for patients with a history of severe hypoglycemia, limited life expectancy, advanced microvascular or macrovascular complications, extensive comorbid conditions, or longstanding diabetes in whom the general goal is difficult to attain despite diabetes self-management education, appropriate glucose monitoring, and effective doses of multiple glucose-lowering agents including insulin.

Source: American Diabetes Association. Standards of Medical Care in Diabetes-2016. Diabetes Care. 2016 Jan:39 Suppl. 1: S1-S112.

EXTERNAL (CLINICAL) TESTING 5.

The following external performance data were provided in support of the substantial equivalence determination:

Precision (see Section 4.1) •
Method Comparison to Reference Method .
Matrix Comparison Anticoagulants .

Method Comparison 5.1.

Method Comparison versus Reference 5.1.1. -

Method Comparison of the cobas b 101 HbA1c assay was performed according to CLSI EP09-A3 at 3 Point of Care sites. Capillary whole blood and venous whole blood (EDTA (K2) and Li-Heparin) from prospective blood sampling were measured on the cobas b 101 in singlicate. No contrived samples were tested. Venous EDTA (K2) whole blood was measured on the Tosoh G8 HPLC Analyzer as the reference method. For capillary whole blood, the range tested was 4.3-11.6% HbA1c. For EDTA (K2) whole blood, the range tested was 4.4-11.4% HbA1c. For Lithium Heparin whole blood, the range tested was 4.3-11.6% HbA1c. The Passing Bablok regression results are listed in the tables below.

{19}------------------------------------------------

Table 9: Capillary

Site	N	Regression Line	Pearson's r
1	125	$y = 1.00x-0.10$	0.99
2	133	$y = 1.00x-0.10$	0.99
3	121	$y = 1.00x-0.20$	0.99

Table 10: EDTA K2

Site	N	Regression Line	Pearson's r
1	125	$y = 1.00x-0.20$	0.99
2	133	$y = 1.00x-0.20$	0.99
3	121	$y = 0.97x+0.04$	0.99

Table 11: Lithium Heparin

Site	N	Regression Line	Pearson's r
1	125	$y = 1.00x-0.20$	0.99
2	130	$y = 1.00x-0.20$	0.99
3	117	$y = 1.00x-0.20$	0.99

5.2. Matrix Comparison

A matrix study was performed using EDTA (K2) whole blood (reference) and EDTA (K3) whole blood. Ninety one samples were collected and tested in singlicate on the cobas b 101. No samples were contrived. The Passing Bablok regression results are shown in the table below.

Table 12: cobas b101 EDTA (K2) WB vs EDTA (K3) WB

N	Regression Line	Pearson's r
91	y = 1.03x-0.00	0.99

{20}------------------------------------------------

SYSTEM DESCRIPTIONS 6.

Modes of Operation: 6.1.

Does the applicant's device contain the ability to transmit data to a computer, webserver, or mobile device?

Yes _ _ _X _ __ or No ________

Does the applicant's device contain the ability to transmit data to a computer, webserver, or mobile device using wireless transmission?

Yes

6.2. Software

FDA has reviewed applicant's Hazard Analysis and software development processes for this line of product types:

Yes X_ or No

The applicant has provided documentation that indicates the device was designed and developed under good software life-cycle processes.

Specimen Identification 6.3.

A barcode reader or keyboard may be used to enter patient information. Results are displayed on the screen and can be printed out by using an optional external printer.

Specimen Sampling and Handling 6.4.

There are no pre-analytics needed. If a fingerstick is performed, the disc allows direct blood application straight from the fingertip to the disc. If a venous blood sample is drawn, the sample is transferred from the collection tube to the disc using a pipette or dropper. The disc is selffilling by capillary forces so that exact pipetting is not necessary. The test is completed in < 6 minutes.

{21}------------------------------------------------

6.5. Calibration

Calibration information is contained on each disc and is specific to each lot of reagent. The instrument automatically reads in the lot-specific calibration data from the barcode information printed on the cobas HbA1c Test disc. There is no calibration needed by the user.

6.6. Quality control

Two levels of controls are provided which are below and above the respective thresholds. Controls are used to check the system performance. The controls are liquid ready to use and include a dropper for easy application of the control liquid to the disc. The HbA1c control is a whole blood based matrix. Target ranges will be assigned per each control lot and checked with each disc lot. The mean value shall meet a defined target range to reflect a normal or pathologic value respectively based on cut-off points which are recommended e.g. by American Diabetes Association (ADA).

7. CLINICAL PERFORMANCE EVALUATION

Not applicable.

8. CONCLUSIONS

Based on the information provided in this submission, the cobas b 101 HbA1c system is substantially equivalent to the predicate device.

Regulation Number and Section

§ 864.7470 Glycosylated hemoglobin assay.

(a)
Identification. A glycosylated hemoglobin assay is a device used to measure the glycosylated hemoglobins (A1a , A1b , and A1c ) in a patient's blood by a column chromatographic procedure. Measurement of glycosylated hemoglobin is used to assess the level of control of a patient's diabetes and to determine the proper insulin dosage for a patient. Elevated levels of glycosylated hemoglobin indicate uncontrolled diabetes in a patient.(b)
Classification. Class II (performance standards).