Stability Of Calibration In The I-Stat System; Validating The Reportable Range - Abbott i-STAT 1 System Manual

StabIlIty of calIbratIon In the i-Stat SyStem

The i-STAT System is a unit-use testing system. Components that cause shifts and drifts in results in multi-
use analyzers: sensors (electrodes), calibration solution, fluid-handling channels and pumps, are housed in
a disposable test cartridge. The sensors are exposed to sample only once, so there is no protein build-up
which is a major cause for deterioration of sensor slope and the need to calibrate and/or verify calibration
on a frequent basis in multi-use analyzers.
The stability and consistency of the manufacturing process allow the slope of the sensors to be
programmed into the analyzer's software. A one-point calibration to set the intercept accounts for any day-
to-day variation in testing conditions. When stored according to directions, the cartridges are stable up to
the expiration date.
The analyzer houses the mechanical and electrical systems necessary to control fluid movement within
the cartridge, control the temperature when measurements are performed at 37°C, measure barometric
pressure, measure electrical signals generated by the sensors and display and transmit results. The
analyzer's functions are factory calibrated to specifications that are programmed into the analyzer along
with acceptability limits, which when exceeded cause the analyzer to display quality check messages or to
display *** rather than results.
The accuracy of results and dependability of the internal quality check system depend upon the ability
of the analyzer to take accurate and sensitive signal readings from the sensors. To check this function,
i-STAT developed an electronic control device. The Electronic Simulator simulates two levels of electronic
signals that stress the analyzer's signal detection function both below and above the reportable ranges.
Injecting signals directly into the analyzer allows very tight control limits to be set. Control limits for liquid
controls are set wide enough to allow for sensor-to-sensor variation. All analyzers that pass the Electronic
Simulator test are equivalent and any variations in results are caused by within and between lot variations in
the cartridges.
The combination of unit-use cartridges, inherently stable electronics of the analyzer, and reliability of the
Electronic Simulator check provides the stability needed for a point-of-care testing system and reduces the
need for frequent stability or calibration verification checks.

ValIdatIng the reportable range

The accuracy of results over the entire reportable range could be assessed by testing the same patient
samples on the new system and on a system with known accuracy and comparing results using an
acceptable difference criteria. However, it is difficult to find samples that cover the low and high ends of the
reportable ranges of many analytes. Ideally, the samples used to validate the ranges should have the same
matrix as the patient samples. Tonometered blood can be used to validate the ranges for PCO and PO .
2 2
But for most analytes, there are no whole blood samples commercially available for this purpose.
Although the i-STAT Calibration Verification Set is aqueous-based, the target values have been determined
over many lots of cartridges and results on these solutions when compared to the target values should
indicate the performance of a particular lot of cartridges.
aSSeSSIng lot-to-lot VarIatIon
Lot-to-lot variation over the entire reportable range for any reagent system could be assessed by testing
calibration verification solutions on old and new lots in parallel. This procedure should be suspended once
it has been determined that lot-to-lot variation is acceptable. Quality controls samples with concentrations
at decision points should always be used to assess new lots or reagent before results are reported.
2
Art: 715209-00E Rev. Date: 06/11/08