cobas c111 system

Data Analysis
Version 1.1
cobas c111

Table of contents

Document information ........................................................................................................5

Where to find information ..................................................................................................6
Conventions used in this manual .......................................................................................6
Photometric measurements .................................................................................................7
Absorbance ......................................................................................................................8
Calculation of absorbance .......................................................................................8
Cuvette blank correction .........................................................................................8
Volume correction ....................................................................................................9
Absorbance checks .................................................................................................10
Rate .................................................................................................................................11
Calculation model Endpoint .................................................................................12
Calculation model Kinetic .....................................................................................13
Reaction direction ..................................................................................................15
Antigen Excess check .............................................................................................16
Test Range check (level rate) .................................................................................18
Calculation of mean rate ........................................................................................18
Quantitative calibration ...............................................................................................19
Calibration mode Linear Interpolation ...............................................................19
Calibration mode Linear Regression ...................................................................20
Calibration mode Logit/Log 4 ...............................................................................21
Calibration mode Logit/Log 5 ...............................................................................22
Calibration mode Exponential 5 ...........................................................................23
Calibration checks ..................................................................................................23
Results .............................................................................................................................25
Quantitative result calculation ..............................................................................25
Corrections ..............................................................................................................27
Calculation of replicate mean ...............................................................................28
Result calculation checks .......................................................................................29
Laboratory conversion ...........................................................................................31
Quality control ....................................................................................................................32
Quality control methods ..............................................................................................32
Precision ...................................................................................................................32
Quality control checks ..................................................................................................33
ISE measurements ...............................................................................................................34
Test principle .................................................................................................................35
Nernst equation ......................................................................................................35
Calculation of Sol 1 factor .....................................................................................35
Calibration .....................................................................................................................36
Slope calculation (2-point calibration) ................................................................36
Slope out of Range check .......................................................................................36
Sol 1 Factor check ...................................................................................................37
Results .............................................................................................................................38
Calculation of concentration ................................................................................38
Calculation checks ..................................................................................................38

Roche Diagnostics
Data Analysis · Version 1.1 3
 cobas c111

Roche Diagnostics
4 Data Analysis · Version 1.1
cobas c111 Document information

Document information

Revision history Manual version Software version Revision date Changes

1.0 n/a February 2008 First publication.
1.1 n/a March 2012 Error in formula for
ISE results corrected.

Edition notice This document is for users of cobas c111 instruments. Every effort has been made
to ensure that all the information contained in this manual is correct at the time of
printing. However, Roche Diagnostics GmbH reserves the right to make any changes
necessary without notice as part of ongoing product development.

Intended use The information contained in this document helps users of cobas c111 instruments
and the cobas c111 Development Channel Programming Software understand the
measurement principles used on cobas c111 instruments, and it describes the
calculations and checks executed when performing measurements.
The cobas c111 instrument is a continuous random-access analyzer intended for in
vitro determination of clinical chemistry and electrolyte parameters in serum,
plasma, urine or whole blood (HbA1c).

Copyright © 2012, Roche Diagnostics GmbH. All rights reserved.

Trademarks The following trademarks are acknowledged: COBAS, COBAS C, and LIFE NEEDS
ANSWERS are trademarks of Roche.
All other trademarks are the property of their respective owners.

Instrument approvals The cobas c111 instrument meets the protection requirements laid down in IVD
Directive 98/79/EC and the European Standard EN 591. Furthermore, our
instruments are manufactured and tested according to the following international
standards:
o EN/IEC 61010-1 2nd Edition
o EN/IEC 61010-2-101 1st Edition
This documentation meets the European Standard EN 591.
Regulatory compliance is demonstrated by the following labels:

Complies with European Union (EU) Directive 98/79/EC.

Issued by Underwriters Laboratories, Inc. (UL) for Canada and the

US.
C ® US

Roche Diagnostics
Data Analysis · Version 1.1 5
Where to find information cobas c111

Where to find information

This document covers the measurement principles used with cobas c111
instruments, and it describes the calculations and checks executed when performing
sample, calibration, and control measurements.
e For information on working with cobas c111 instruments, see the cobas c111
Instrument Operator’s Manual.
e For information on working with the cobas c111 Development Channel Programming
Software, see the cobas c111 Development Channel Operator’s Manual.

Conventions used in this manual

Visual cues are used to help locate and interpret information quickly. This section
explains the formatting conventions used in this document.

Symbols The following symbols are used:

Symbol Used for

e Cross-reference

Tip

Roche Diagnostics
6 Data Analysis · Version 1.1
cobas c111 Photometric measurements

Photometric measurements

This chapter explains the calculations and checks that are applied to photometric
measurements.
The following flow chart shows the individual steps that are carried out with
calibration, recalibration, sample and control measurements.

Photometric measurements

Calculation of absorbance
Absorbance

Cuvette blank correction

Volume correction

Absorbance checks

Rate Calculation models

Calculation checks

Reaction direction

Antigen Excess checks

Calibration Calculation of mean rate

Calibration modes

Calibration checks

Recalibration

Results Result calculation

Result calculation checks

Conversions

Figure 1 Chart of steps in photometric measurements

Roche Diagnostics
Data Analysis · Version 1.1 7
Photometric measurements cobas c111
Absorbance

Absorbance

This section explains the absorbance calculations and the associated checks.

Calibration Control Sample

Calculation of absorbance l l l
Cuvette blank correction
Monochromatic measurement l l l
Bichromatic measurement l l l
Volume correction l l l
Absorbance checks
High Absorbance check l l l
Reagent Range check l
High Activity check l l
Low Activity check l l
Table 1 Calculations and checks for absorbance measurements

Calculation of absorbance
Calculation is carried out after every measurement by the absorbance photometer.
There are up to 40 absorbance measurements (cycles) consisting of 1 or 2 out of 12
available wavelengths.
The absorbance signal is corrected using the following values:

measured absorbance +  cuvette blank correction + volume correction

 (mono/bichromatic) 

The values obtained from photometric measurements are converted to absorbance

values. This conversion is carried out for all points measured at the requested
wavelength.

I cuv   n 
A raw cuv   n  = – log ------------------------
I air   

A raw cuv (,n) Raw absorbance of each point (n) at the requested wavelength 
I cuv (,n) Intensity of each point (n) at the requested wavelength 
I air    Intensity of air (without cuvette) at the requested wavelength 

Cuvette blank correction

Cuvette blank correction is carried out in every absorbance calculation.
All absorbencies are corrected using a blank, that is the absorbance of a water-filled
cuvette at the corresponding wavelength.

Roche Diagnostics
8 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Absorbance

Monochromatic measurement The blank measurement is calculated using the following formula:
calculation
A blank    = A raw cuv (,1) + A H 2O  

The absorbance is normalized using the following formula:

AN cuv   n  = A raw cuv   n  – A blank   

A raw cuv (,1) Raw absorbance of point 1 of an empty cuvette at the requested
wavelength 
A blank    Blank raw absorbance at wavelength (Corresponds to a water-filled
cuvette.)
A raw cuv (, n) Raw absorbance of each point (n) at the requested wavelength 
A H 2O   Absorbance of water at the requested wavelength. This value is
calculated from the difference of a cuvette filled with water and an
empty cuvette (air/water calibration system check).
AN cuv (,n) Normalized absorbance of each point (n) at the requested
wavelength 

Bichromatic measurement AN cuv   n  =  A raw cuv   1 n  –  A blank   1   –  A raw cuv   2 n  – A blank   2  

calculation
A raw cuv   n  Raw absorbance value of all points of an empty cuvette at the
requested wavelength.
A blank    Absorbance value of a cuvette filled with water at the requested
wavelength.
AN cuv   n  Normalized absorbance values of a cuvette.
1 Main wavelength.
2 Reference wavelength.

Volume correction
Volume correction is carried out for every absorbance calculation.
All absorbencies are corrected using the volume ratio ( initial volume  final volume ).
This is necessary since, with each addition of liquid, the reaction mixture becomes
more diluted, while the path length of the light and the measuring window remain
constant. Volume correction is carried out for all points.
The following equation is used for volume correction:

AN cuv   ,n   V cuv  n 
A  , n  = ---------
------------------------------------
VF cuv

A   n  Volume-corrected absorbance of each point (n).

AN cuv   n  Normalized absorbance of each point (n) at the wavelength .
V cuv  n  Starting volume of point n of a cuvette before the last addition is
made.
VF cuv Final volume of cuvette after the last addition.

Roche Diagnostics
Data Analysis · Version 1.1 9
Photometric measurements cobas c111
Absorbance

Absorbance checks

High Absorbance check This check is carried out for every absorbance calculation.
All absorbencies used for the calculation of all measuring points of all requested
wavelengths between the first and last point may not exceed 2.0 absorbance units.
The first and last points refer to:
o Calc first Calc last (Endpoint and Kinetic)
o HighAct first HighAct last (High/Low Activity)
o Ag first Ag last (Antigen Excess)

A raw cuv   n   2.0

A raw cuv   n  Absorbance of each point (n) (between the first and last points) at
requested wavelength 

The High Abs flag is displayed if a value between the first and last points exceeds 2.0.

Reagent Range check This check is carried out for every absorbance calculation for the calibrator with the
lowest concentration.
For the reagent blank, all absorbencies required for calculations are checked to ensure
they do not exceed the preset reagent range limits.
RR low  A cuv  n   RE high

A cuv  n  Absorbance of the reagent blank of each point used in the

calculations.
RR low Reagent range lower limit (test parameter).

RR high Reagent range upper limit (test parameter).

The < Reag Rng or > Reag Rng flag is displayed if the absorbance of the reagent in a
cuvette is not within the specified range.

High Activity check This check is used to identify large differences in the absorbance after the start of the
reaction.
With ascending reaction curve:
A cuv  last  – A cuv  first   HLL

With descending reaction curve:

A cuv  first  – A cuv  last   HLL

A cuv  first  Absorbance at the time of the high activity first point (test
parameter).
A cuv  last  Absorbance at the time of high activity last point (test parameter).

HLL High activity limit (test parameter).

The High Act flag and the value of the upper test range limit are displayed in the
result list if the value does not meet the conditions.

Roche Diagnostics
10 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Rate

Low Activity check This check is used to identify zero pipetting of sample and reagents.
With ascending reaction curve:
A cuv  last  – A cuv  first   HLL

With descending reaction curve:

A cuv  first  – A cuv  last   HLL

A cuv  first  Absorbance at the time of the low activity first point (test parameter).

A cuv  last  Absorbance at the time of the low activity last point (test parameter).

HLL Low activity limit (test parameter).

The Low Act flag and the value of the lower test range limit are displayed in the result
list if the value exceeds the lower test limit.

Rate

This section explains the rate calculations and the associated checks.

Calibration Control Sample

Calculation model Endpoint
Absorbance difference l l l
Endpoint Unstable check l l l
Calculation model Kinetic
Limits of kinetic range l l l
Absorbance difference per minute l l l
Nonlinear check l l l
Kinetic readings check l l l
Reaction direction l l l
Antigen Excess check l l
Test range check (level rate) l l
Calculation of mean rate l
Table 2 Calculation and checks for rate determination

Roche Diagnostics
Data Analysis · Version 1.1 11
Photometric measurements cobas c111
Rate

Calculation model Endpoint

Calculation of absorbance The calculation of the absorbance difference is carried out for calibrators, controls,
difference and samples.
Two points are used to determine the rate (A) of the reaction. The rate is used to
calculate the concentration of the sample, based on the kind of calibration mode
used.

Absorbance (A)

A (last)

ΔA

A (first)

Time (T)
T0 Tn

RC cuv = A  last  – A  first 

RC cuv Rate in A

A(last) Absorbance at the last calculation point (last calculation point, test
parameter). See the cobas c111 Method Sheet.
A(first) Absorbance at the fist calculation point (first calculation point, test
parameter). See the cobas c111 Method Sheet.

Endpoint Unstable check This check applies only to the calculation model Endpoint. It is carried out to
determine if the last five measuring points used for determinations are stable.
A  high  – A  low   EP

A  high  Highest absorbance obtained from the last five measuring points.

A  low  Lowest absorbance obtained from the last five measuring points.

EP Endpoint unstable limit in absorbance (dAbs. test parameter).

The EP Unstab flag is displayed if the absorbance difference is not within the
specified range.

Roche Diagnostics
12 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Rate

Calculation model Kinetic

The calculation model Kinetics is applied to calibrations, quality controls and
samples. A linear regression analysis is carried out between the first and last points as
specified in the cobas c111 Method Sheet. The slope of regression line is the rate
(A per minute) of the reaction. The rate is used to calculate the concentration of the
sample based on the kind of calibration mode applied on the calibration curve.

Limits of kinetic range With the Kinetic calculation model, an inspection of the reaction is carried out to
determine if the beginning and end of the kinetic range is within the defined
absorbance range. If this is not the case the measurement range is reduced
accordingly.
Determining the measurement range used for calculating the kinetics:
The first and last absorbance values ( A(first) , A(last) ) in the kinetic range are used as
the first and last measuring points ( MPkin first , MP kin last ) when calculating the
kinetics. All points must fulfill the following conditions:
RL low  A  n   RL high

RL low Absorbance of the lower reaction limit (test parameter).

Minimum value = 0.0.
An Absorbance of each point in the kinetic range.

RL high Absorbance of the upper reaction limit (test parameter).

Maximum value = 2.0.

Roche Diagnostics
Data Analysis · Version 1.1 13
Photometric measurements cobas c111
Rate

To avoid values that are too low when calculating the kinetics, you can perform the
following checks with floating limits (test parameters):
With ascending reaction curve:
fl. RL low = RL low +  A cuv  sample  – A cuv  cal ref   and fl. RL high = RL high

fl. RL low  A  n   fl. RL high

With descending reaction curve:

fl. RL high = RL high +  A cuv  sample  – A cuv  cal ref   and fl. RL low = RL low

fl. RL low  A  n   fl. RL high

If the result for  A cuv  sample  – A cuv  cal   is negative, set the value for this
expression to zero. In such a case, the floating reaction limit becomes a fixed reaction
limit.

fl. RL low Corrected Reaction Limit Low (Floating RL Low, test parameter).

fl. RL high Corrected Reaction Limit High (Floating RL High, test parameter).

A cuv  sample  Sample absorbance value of a cuvette (test parameter).

A cuv  cal RF  Calibrator absorbance value of a reference calibrator cuvette (test

parameter).

Calculation of the absorbance This calculation is performed with the Kinetics calculation model.
difference per minute
With the specified points (at least 2) a Passing/Bablok analysis is carried out to
determine the absorbance difference per minute (Kinetic).

Nonlinear check This check is used with the Kinetics calculation model.
The linearity of the curve is determined by analyzing the points which lie within the
limits specified for the linear region. A Passing/Bablok analysis is carried out on each
half of the curve to determine the straight lines. The slopes of these straight lines are
determined using the differences A1 and A2.
If there is an even number of points within the linear range the two halves of the
curve overlap at two points.
Conditions for running the test:
Rate  0.006abs/min

Test conditions:
A 1  min – A 2  min  lin

A 1 Absorbance change (ascending) of the first half of the curve.

A 2 Absorbance change (ascending) of the second half of the curve.

lin Linearity in  (A/min) (linearity test parameter).

The Non Linear flag is displayed if the conditions are not fulfilled.

Roche Diagnostics
14 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Rate

Kinetic Readings check This check is used with the Kinetics calculation model.
The number of measuring points used to calculate the kinetics is checked to ensure
that the defined number (within the linear region of the curve) is present.
 MP kin last – MP kin first + 1   KR

MP kin last Last measuring point used for the kinetics calculation.

MP kin first First measuring point used for the kinetics calculation.

KR Number of kinetic readings (test parameter).

The < Kin Read flag is displayed if the number of points needed for the kinetics
calculation is less than the number defined, but the result is still calculated.

Reaction direction
This check is carried out for every absorbance calculation.
The rate is corrected according to the expected reaction direction. The reaction
direction is preset (see cobas c111 Method Sheet).
Reaction direction increasing:
R cuv = RC cuv   +1 

Reaction direction decreasing:

R cuv = RC cuv   -1 

R cuv Rate for calibrators, controls, or samples.

RC cuv Rate after absorbance difference calculation. (See also Calculation

model Endpoint on page 12 and Calculation model Kinetic on
page 13.)

Roche Diagnostics
Data Analysis · Version 1.1 15
Photometric measurements cobas c111
Rate

Antigen Excess check

This check is carried out with controls and samples. It can be used to determine an
antigen excess. The binding of an antigen to a specific antibody produces cross-
linked complexes which scatter light.
In the presence of excess antibody, the absorbance increases with increasing antigen
concentration. With antigen excess, the absorbance will decrease with increasing
antigen concentration so that the same A refers to two different concentrations.
The point at which the signal begins to decrease is the beginning of the antigen excess
region of the Heidelberger Curve. The Antigen (AG) Rate should be determined with
the Antigen Excess check first.

300 Antibody excess Antigen excess

250
C
200
Rate

150 B

100

50

0
0 10 20 30 40 50 60 70 80 g/L

Antigen Concentration

Determination of antigen (AG) The test parameters AG First, AG Last and the Endpoint calculation model are
rates required for determining antigen rates.
The AG rate of the standards is determined by the mean of the replicate AG rates.

Roche Diagnostics
16 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Rate

Carrying out the antigen excess The following figure shows a Rate/AG-Rate curve for the standards. The graph was
check created using linear interpolation.

The rates of the samples are determined using the corresponding antigen rates of the
standards from the Rate/AG-Rate curve.
For rates outside the range between STD-5 and STD-1, the following applies:
o If they are below the lowest standard the antigen rate is assigned to the lowest
standard.
o If they are above the highest standard the antigen excess check is not carried out.

AR std  F AG  % 
A cuv  -------------------------------------- and AR cuv  AR std – offset
100  % 

AR std Antigen rate of the standard.

F AG Test-specific factor.

AR cuv Antigen rate of a sample.

offset Defined rate (for cobas c111 approximately 0.004 Abs.).

The AG Excess flag is displayed if the checks are not fulfilled.

Roche Diagnostics
Data Analysis · Version 1.1 17
Photometric measurements cobas c111
Rate

Test Range check (level rate)

This check determines whether the rate values from the nonlinear calibration modes
are within the monotonic curve area.
R WP low  R samp-n  RWP high

R samp-n Rate of controls or samples.

R WP low Rate of the lowest deflection point, determined from Monotonic

Check (test parameter).
R WP high Rate of the highest deflection point, determined from Monotonic
Check (test parameter).

If the rate is not in the monotonic curve area, no valid end results are obtained.
The > Test Rng flag is displayed if the upper test range limit is exceeded.
The < Test Rng flag is displayed if the value falls below the lower test range limit.

Calculation of mean rate

This check is carried out for calibrators only.
The test parameter Replicate may be set to Single or Duplicate. If set to Duplicate, the
calibrators (1…6) are analyzed twice. The individual rates of the replicates are
averaged to provide a mean rate for the calibrators.
The mean is determined from the replicate results.

 R samp-n  1  +  + R samp-n  r  
RM samp-n = ----------------------------------------------------------------------------
r

RM samp-n Mean rate of calibrators.

RM samp-n  r  Rate of calibrators (equivalent to R cuv ).

r Number of replicates.

Roche Diagnostics
18 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Quantitative calibration

Quantitative calibration

This section explains the calibration modes and the associated checks. The following
table gives an overview:

Calibration Control Sample

Calibration modes
Linear Interpolation l
Linear Regression l
Logit/Log 4 l
Logit/Log 5 l
Exponential 5 l
Calibration checks
Curve Direction check l
Curve Non Monotonic check l
Calculation Error check l
Curve Out of Range check l
Table 3 Calibration modes and checks

The calibration yields parameters which enable the calibration curve to be calculated.
The calibration curve can be used to convert the rate of the sample into the
concentration of the analyte being measured.
For determining the calibration curve, the following mathematical modes are used:

Mode Number of calibrations Curve parameters

Linear Interpolation 3…6 none
Linear Regression 0…6 R0 , F
Logit/Log 4 4…6 R0, Kc, A, B
Logit/Log 5 5…6 R0, Kc, A, B, C
Exponential 5 5…6 R0, Kc, A, B, C
Table 4 Calibration modes and their parameters

For the Linear Regression mode, if the number of calibrators = 0, the preset curve
parameters R0 and F are used. If the number of calibrators = 1, the preset curve
parameters R0 or F are used.

Calibration mode Linear Interpolation

Linear interpolation analysis is carried out on the rates of up to six calibrators to plot
a calibration curve.

Roche Diagnostics
Data Analysis · Version 1.1 19
Photometric measurements cobas c111
Quantitative calibration

Calibration mode Linear Regression

Linear regression analysis is carried out on the rates of all calibrators to find the line
of “best fit”. If only one calibrator has been set, either the slope of the linear
regression is fixed and the offset of the curve is determined from this calibrator, or
the rate of one calibrator is fixed and the slope determined.

Rates

individual
data
R0

Conc

Calculation of curve parameter R0 (intercept):

2
   CT std-n     RM std-n   –    CT std-n  RM std-n     CT std-n  
R 0 = ------------------------------------------------------------------------------------------------------------------------------------------------------------------
2 2
   CT std-n   N  –    CT std-n  

Calculation of curve parameter F (factor):

2 2
   CT std-n   –    CT std-n   n 
F = -----------------------------------------------------------------------------------------------------------------------------------------
   CT std-n     RM std-n   –    CT std-n  RM std-n   n 

R0 , F Curve parameters.

CT std-n Calibrator values.

RM std-n Mean calibrator rate.

n Number of calibrators.

Roche Diagnostics
20 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Quantitative calibration

Calibration mode Logit/Log 4

The Logit/Log 4 mode is used to calculate nonlinear calibration curves. Between four
and six calibrators can be entered.

Rates
An

Ax

A3

A2

A1

C1 C2 C3 Cx Cn
Concentration

The curve parameters are calculated with a maximum of 150 iterations, using the
following equation:

Kc
RM std-n = R 0 + -------------------------------------------------------------------------
1 + exp  –  A + B  ln CT std-n  

RM std-n Mean calibrator rate.

CT std-n Calibrator values.

R0 , Kc , A , B Curve parameters.

Roche Diagnostics
Data Analysis · Version 1.1 21
Photometric measurements cobas c111
Quantitative calibration

Calibration mode Logit/Log 5

The Logit/Log 5 mode is used to calculate nonlinear calibration curves. Five or six
calibrators can be entered.
The curve parameters are calculated with a maximum of 150 iterations, using the
following equation:

Kc
RM std n = R 0 + --------------------------------------------------------------------------------------------------------
1 + exp  –  A + B  ln CT std n + C  CT std n  

RM std n Mean calibrator rate.

CT std n Calibrator values.

R0 , Kc , A , B Curve parameters.
Rates

An

Ax

A3

A2

A1

C1 C2 C3 Cx Cn
Concentration

Roche Diagnostics
22 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Quantitative calibration

Calibration mode Exponential 5

The Exponential 5 mode is used to calculate nonlinear calibration curves. Five or six
calibrators can be entered.
The curve parameters are calculated with a maximum of 150 iterations, using the
following equation:
2 3
RM std n = R 0 + K  exp  A  ln CT std n + B   ln CT std n  + C   ln CT std n 

RM std n Mean calibrator rate.

CT std n Calibrator values.

R 0 , K c , A , B, C Curve parameters.
Rates

An

A3

Ax

A2

A1

C1 C2 Cx C3 Cn
Concentration

Calibration checks

Curve Direction check This check determines if the direction of the calibration curve is correct. It is used to
generate the Curv Dir flag if calibrators are placed in the wrong order on the
instrument.
With ascending calibration curve:
RM std highest – RM std lowest  0

With descending calibration curve:

RM std lowest – RM std highest  0

RM std highest Mean rate value of the standard with the highest analyte
concentration or activity.
RM std lowest Mean rate value of the standard with the lowest analyte concentration
or activity.

The Curv Dir flag is displayed if the calibration curve direction is not correct.

Roche Diagnostics
Data Analysis · Version 1.1 23
Photometric measurements cobas c111
Quantitative calibration

Curve Non-Monotonic check This check determines if the calibration curve is monotonic within the test range.
The Non Mono flag is displayed if the curve is not monotonic within the defined test
range. This flag is also displayed if the calculation model Linear Interpolation is used
and the test range is defined outside the defined standard target value.

Calculation Error check The Calc Error flag is displayed if the calibration curve cannot be calculated.

Curve out of Range check With this check the calibration curve is inspected to make sure it is within preset
limits. This check varies according to the calibration mode.
o With Linear Interpolation, the check is not carried out.
o With Logit/Log4, Logit/Log5, Exponential 5, the check is carried out for each
standard.
Theoretical rate for each expected standard concentration:
RT std-n = f  CT std-n 

Theoretical minimum rate for each expected standard concentration:

RL std-n = f low  CT std-n 

Theoretical maximum rate for each expected standard concentration:

RH std-n = f high  CT std-n 

RL std-n  RT std-n  RH std-n

RT std-n Theoretical rate of the standard.

CT std-n Expected standard concentration.

RL std-n Theoretical minimum rate of the standard.

f high , f low Test-specific curve parameters (dependent on the calibration mode).

RH std-n Theoretical maximum rate of the standard.

Linear Regression This check establishes whether the calibration parameters R0 and F are within the
accepted range.
R 0 low  R 0  R 0 high

F low  F  F high

R0 , F Parameters in the calculated calibration.

R 0 low , F low Lower parameter limit.

R 0 high , F high Upper parameter limit.

The Out of Rng flag is displayed if the check criteria are not met.

Roche Diagnostics
24 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Results

Results

This section explains the calibration modes and checks used to determine the results.

Calibration Control Sample

Quantitative result calculation
Linear Interpolation l l l
Linear Regression l l l
Logit/Log 4 l l l
Logit/Log 5 l l l
Exponential 5 l l l
Corrections
Volume correction l l l
Compensation correction l l l
Correlation correction l l l
Calculation of replicate mean l
Result calculation checks
Replicate Deviation check l
Stand. Curve Deviation check l
Test Range check (level concentration) l l
Reference Range check l
Laboratory conversion l l l
Table 5 Result calculation models and checks

Quantitative result calculation

The curve parameters of the various calibration modes and the rates of the
calibrations are used to calculate the results.

Linear Interpolation calibration The standards are used to determine the result. If the rate is out of the range of the
mode standard rates, it is not possible to calculate the result. The result is then flagged with
Calc Error.
The standard concentrations right and left of the sample concentration in the
calibration curve are used to calculate the sample result using linear interpolation.

Linear Regression calibration This mode is used with controls and samples.
mode
Results are determined according to the following equation:
CC samp-n =  RR samp-n – R 0   F

CC samp-n Sample or control result in standard units.

RR samp-n Rate of sample or control.

R0 , F Curve parameters.

Roche Diagnostics
Data Analysis · Version 1.1 25
Photometric measurements cobas c111
Results

Logit/Log 4 calibration mode This mode is used with controls and samples.
The results are determined according to the following equation:

 Kc 
 A + ln ---------------------------------- -–1 
 
 RR samp-n – R 0 
 
CC samp-n = exp ----------------------------------------------------------------- 
 –B 
 
 
 
 

CC samp-n Sample or control result in standard units.

RR samp-n Rate of sample or control.

R0 , Kc , A , B Curve parameters.

If the standard curve is not defined for a particular rate, the Calc Range flag is
displayed.

Logit/Log 5 calibration mode This mode is used with controls and samples.
The results are determined according to the following equation:

RR samp-n – R 0
f  x  = A  x + B  x + C  exp  x  – ln ----------------------------------------------------
K c –  RRsamp-n – R 0 

fx Sample or control result in standard units.

RR samp-n Rate of sample or control.

R 0 , K c , A , B, C Curve parameters.

Exponential 5 calibration mode Results are determined according to the following equation:

RR samp-n – R 0
2 3
f  x  = A  x + B  x + C  x – ln -----------------------------------
Kc

fx Sample or control result in standard units.

RR samp-n Rate of sample or control.

R 0 , K c , A , B, C Curve parameters.

Roche Diagnostics
26 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Results

Corrections

Volume correction after dilution Volume correction after dilution is carried out with quantitative analysis for
calibrations, controls and samples.
The calculation of the results may require correction (for example due to dilution) or
additional calculations. Various checks will then be carried out.
If a control or patient sample has been diluted the corresponding correction factor is
applied.
For calibrations, controls and samples in general the results are determined
according to the following equation:
CV samp-n = CC samp-n  F pred

For diluted samples:

CV samp-n = CC samp-n  F pred  F postd

CV samp-n Results with dilution correction of controls or samples in standard

units.
CC samp-n Results after calculation of controls or samples in standard units.

F postd Sample postdilution factor.

F pred Sample predilution factor.

The  flag is displayed if F postd is used.

Compensation correction Compensation correction is carried out after every result calculation. In this way,
results can be adapted to other methods.
With the compensation offset, additive corrections on the result can be made. These
corrections have to be verified experimentally.
C samp-n = CV samp-n + O comp

C samp-n Results of controls or samples.

CV samp-n Results (including dilution correction) of controls or samples.

O comp Compensation offset.

Correlation correction Carried out after every result calculation. In this way, results of different instruments
in the same lab can be correlated.
The correlation factor and offset are the slope and y-intercept, respectively, of the
straight line obtained. These values must be determined experimentally.
C samp-n =  CV samp-n  F corr  + O corr

C samp-n Correlated result of controls or samples in standard units.

CV samp-n Results (including dilution correction) of controls, or samples.

F corr Lab correlation factor (test parameter).

O corr Lab correlation offset (test parameter).

Roche Diagnostics
Data Analysis · Version 1.1 27
Photometric measurements cobas c111
Results

Calculation of replicate mean

This calculation is carried out for calibrators according to the following equation:

1
CM samp-n = ---   C samp-n  1  +  + C samp-n  r  
r

CM samp-n Mean replicate concentration.

C samp-n  r  Replicate concentration of calibrators.

r Number of replicates.

The calculation of concentration of replicate r and the mean determined by linear

interpolation are used for the Replicate Deviation check.
C lin-int-n  r  = f lin-int-n  R std-n  r  

1
CM lin-int-n = ---   C lin-int-n  1  +  + C lin-int-n  r  
r

C lin-int-n  r  Replicate concentration of calibrators.

f lin-int-n Linear interpolation function.

R std-n  r  Rate of replicate r.

CM lin-int-n Mean replicate concentration.

r Number of replicates.

Roche Diagnostics
28 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Results

Result calculation checks

Replicate Deviation check This check is carried out for calibrators only. It examines the scattering of replicates
of calibrators.
RD n = CM lin-n  F + O

The following equation is applied to each replicate:

 CM lin-n – RD n   C lin int-n  r    CM lin-n + RD n 

C lin int-n  r  Replicate results of calibrators.

Linear interpolation is applied for non-linear calibration modes.
Linear regression is applied for linear calibration modes. A
differentiation based on the underlaying calibration mode is not
required.
CM lin-n Replicate mean.

r Number of replicates.

RD n Scattering range.

F, O Test parameters for replicate deviation (Factor, Offset).

The > Repl Dev flag is displayed if a result exceeds the preset deviation.

Standard Curve Deviation check This check is carried out with quantitative analysis for the main calibration. It is used
to inspect the deviation of each calibrator from the curve.
The scattering range within the monotonic range is determined for each calibrator,
using the following formula:
SD = CT std-n  F + O

The following formula applies for each replicate:

 CT std-n – SD   CM std-n  r    CT std-n + SD 

CT std-n Nominal calibrator concentration.

CM std-n Calibrator mean.

r Number of replicates.

F, O Test parameters for standard deviation (Factor, Offset).

The >Std Dev flag is displayed if the replicates are not within the defined range.

Roche Diagnostics
Data Analysis · Version 1.1 29
Photometric measurements cobas c111
Results

Test Range (level concentration) This check is used with quantitative analysis. This check determines wether the
check concentration is within the measuring range (test range).
TR low  CM samp-n  TR high

CM samp-n Control or sample concentration.

TR low Lower limit of the test range (test parameter).

TR high Upper limit of the test range (test parameter).

The following flags are displayed if the result is not in the test range:
o < Test Rng is displayed if the result is below the test range lower limit.
o > Test Rng is displayed if the result exceeds the test range upper limit.

Reference Range check This check is carried out for samples with quantitative analysis. It compares the
sample result with the preset reference ranges.
Each sample concentration is determined using the following equation:
RR low  C samp-n  RR high

RR low Lower limit of reference range (test parameter).

C samp-n Sample results in laboratory units.

RR high Upper limit of reference range (test parameter).

The following flags are displayed if the result is not in the reference range:
o <RR is displayed if the result is below the reference range.
o >RR is displayed if the result is above the reference range.

Roche Diagnostics
30 Data Analysis · Version 1.1
cobas c111 Photometric measurements
Results

Laboratory conversion
Results can be adjusted by using a laboratory conversion factor.

Laboratory units The conversion is carried out after every result calculation.
The sample units can be converted into any desired unit by means of a conversion
factor.
Conversion to another unit is carried out for replicates, control and sample results,
and the means of calibrators.
For replicates of calibrators:
CL samp-n  r  = C samp-n  F unit

For control, sample or mean of calibrators:

CML samp-n  r  = CM samp-n  F unit

CL samp-n  r  Replicate results of calibrators in laboratory units.

C samp-n Results in standard units of controls and samples or replicates of

calibrators.
CML samp-n Control and sample results or mean of calibrators in laboratory units.
Only calibrators can be run in duplicates and a mean must then be
calculated. For QC and sample results no mean calculation is
required, since no duplicates or triplicates are run.
CM samp-n Control and sample results or mean of calibrators in standard units.

F unit Factor in laboratory units (test parameter).

r Number of replicates.

Roche Diagnostics
Data Analysis · Version 1.1 31
Quality control cobas c111
Quality control methods

Quality control

The cobas c111 instrument allows the performance of real-time quality control,
which consists of the actual control measurements and the preset quality control
rules.
e For information on configuring and performing QC on the cobas c111instrument,
see the cobas c111 Instrument Operator’s Manual.

Quality control methods

Quality Control is used to detect the following types of errors:

o Random errors are usually method-dependent and are caused by the scattering of
measurements.
o Systematic errors are usually instrument-dependent and are caused by, for
example, inaccurate pipetting.
Ideally, quality control methods identify random and systematic errors which are
indications of the precision and accuracy of the analytical system. Quality control
materials used on cobas c111instrument can only be defined as precision.

Precision
The common quantitative criteria for imprecision are the standard deviation and the
coefficient of variation.

2
  xi – xm 
s = ---------------------------
n – 1

s
CV  %  = ------  100
xm

s Standard deviation.
xi Value of sample i.

xm Arithmetic mean value of all samples.

n Number of samples.

CV Coefficient of variation.

Roche Diagnostics
32 Data Analysis · Version 1.1
cobas c111 Quality control
Quality control checks

Quality control checks

For the assessment of the precision the reproducibility of a measurement is

investigated.
The following precision checks (rules) are used to assess the system performance.
Violation of a rule is indicated by a flag.

Rule/Flag Comment
R 1(2.5s) This rule checks whether the precision is within the predefined limits.
A flag is generated if one value is outside ±2.5s.
R 1(3s) This rule checks for random errors.
A flag is generated if one value is outside ±3s.
R 2(2s) This rule checks for systematic errors.
A flag is generated if two values are outside ±2s.
Table 6 Quality control rules and flags

Roche Diagnostics
Data Analysis · Version 1.1 33
ISE measurements cobas c111
Quality control checks

ISE measurements

This section explains the calculations and checks used in electrolyte measurements.
The diagram below shows the individual steps carried out in an ISE calibration or an
ISE sample or control measurement:

ISE main calibration using ISE

Solution 1 and 2

Slope calculation
Slope out or Range check
(2-point calibration)

Calculation of Sol 1 Factor Sol 1 Factor check

ISE sample measurement

ISE control measurement

Sample and calibrator

measurement

Result calculation Test Range check

Correction

Laboratory conversion

Reference Range check

Figure 2 Charts for steps in ISE measurements

Roche Diagnostics
34 Data Analysis · Version 1.1
cobas c111 ISE measurements
Test principle

Test principle

An Ion-Selective Electrode (ISE) makes use of the unique properties of certain

membrane materials to develop an electrical potential (electromotive force, EMF) for
the measurements of ions in solution. The electrode has a selective membrane in
contact with both the test solution and an internal filling solution. The internal filling
solution contains the test ion at a fixed concentration. Because of the particular
nature of the membrane, the test ions will closely associate with the membrane on
each side. The membrane EMF is determined by difference in concentration of the
test ion in the test solution and the internal filling solution. The EMF for a specific
ion in solution develops according to the Nernst equation.

Nernst equation
Electrolyte activity is measured by determining the potential difference between the
measurement and reference electrodes. The relationship between the potential
difference and the ionic activity is given by the Nernst equation.
If the ionic strength of the calibrating solution chosen is equal to that of the sample,
the electrode system can be calibrated with acceptable accuracy in terms of the ionic
concentration instead of the activity.
The potential measured is directly related to the concentration.
E = E 0 + S  log a i

E Potential difference between the measurement and reference

electrodes.
E0 Standard potential of the electrode assembly. (Depends on the
construction of the electrode.)
S Electrode slope (Nernst Factor).

ai Ionic activity of the electrolyte being measured.

Calculation of Sol 1 factor

E sol1-1 – E ISEcal-1 E sol1-2 – E ISEcal-2
C sol1  10 ---------------------------------------- - + C sol1  10 -----------------------------------------
S S
C sol1 calc = ---------------------------------------------------------------------------------------------------------------------------------------
2
C sol1 targ
Sol1Factor = --------------------
-
C sol1 calc

S Slope in mV/dec

C sol1 Known concentration of ISE Solution 1.

E sol1 Electrode potential of ISE Solution 1 in mV.

E ISE cal Electrode potential of the ISE calibrator in mV.

C sol1 targ Known concentration of ISE Solution 1.

C sol1 calc Calculated concentration of ISE Solution 1.

Sol1Factor ISE Solution 1 factor.

Roche Diagnostics
Data Analysis · Version 1.1 35
ISE measurements cobas c111
Calibration

Calibration

To calibrate the electrodes the slope of the measurement system must be calculated.

Slope calculation (2-point calibration)

This calculation is performed for sodium, potassium or chloride. The slope of the
electrodes is determined by measuring two different standard solutions. It is
calculated with the following equation (based on the Nernst equation):

E sol 1 – E sol 2
S = -------------------------------
C sol 1
log ------------ -
C sol 2

S Slope in mV/dec.

E sol 1 Electrode potential of ISE Solution 1 in mV.

E sol 2 Electrode potential of ISE Solution 2 in mV.

C sol 1 Known concentration of ISE Solution 1.

C sol 2 Known concentration of ISE Solution 2.

Slope out of Range check

This check is used in a calibration to determine if the slope lies within the defined
range.
The slope must fulfill the following conditions:
S low  S  S high

S low Lower limit of slope range in mV/dec.

S Slope in mV/dec.

S high Upper limit of slope range in mV/dec.

The Out of Rng flag is displayed if the check criteria are not fulfilled.

Roche Diagnostics
36 Data Analysis · Version 1.1
cobas c111 ISE measurements
Calibration

Sol 1 Factor check

To check if a new ISE Solution 1 factor is correct, the rate values for the second ISE
Solution 1 measurement and its subsequent calibrator measurement of the main
calibration are used. With these rate values, together with the slope and the Sol 1
Factor, the concentration of each ion (sodium, potassium, and chloride) in the ISE
Solution 1 is calculated.
C low  C calc  C high

C low Concentration calculated for ISE Solution 1 using the Sol 1 Factor
determined during the main calibration.
C calc Lower limit of the recalculated ISE Solution 1 concentration.

C high Upper limit of the recalculated ISE Solution 1 concentration.

The limits for all ISE modes are:

Ion Clow Chigh

Sodium 148.8 mmol/L 151.2 mmol/L
Potassium 4.94 mmol/L 5.06 mmol/L
Chloride 113.9 mmol/L 116.1 mmol/L
Table 7 Limits of recalculated ISE solutions

The Sol 1 F Dev flag is displayed if the check criteria are not met.
In addition, the Sol 1 factor must fulfill the following condition:
0.85  Sol 1 factor  1.15

The Sol 1 F Dev flag is displayed if the check criterion is not fulfilled.

Roche Diagnostics
Data Analysis · Version 1.1 37
ISE measurements cobas c111
Results

Results

Each ISE measurement consists of the measurement of the sample and the
measurement of the corresponding ISE calibrator. The electrode potential results
from the respective sample or calibrator measurement.

Calculation of concentration
The sodium, potassium, and chloride concentration is determined using the
following equation:
E samp QC – E ISE cal indir/urine
C samp QC = C sol 1  10 -----------------------------------------------------------------  F
S

C samp QC Ionic concentration of sample or control.

C sol 1 Nominal concentration of ISE Solution 1.

E samp QC Electrode potential of sample or control in mV.

E ISE cal indir/urine Electrode potential of ISE calibrator in mV.

F ISE Solution 1 factor.

S Slope in mV/dec.

Calculation checks

Test range e See Test Range (level concentration) check on page 30.

Correlation correction e See Correlation correction on page 38.

Laboratory conversion e See Laboratory conversion on page 38.

Reference range e See Reference Range check on page 30.

Roche Diagnostics
38 Data Analysis · Version 1.1