Precision of CoaguChek®S monitors – reading discrepancies
Evaluation studies have shown CoaguChek® S to be extremely reliable in terms of precision and accuracy.. The reason why readings (INR/Quick) differ from other systems can be explained in terms of the differing factor sensitivities for the various reagents (thromboplastins) used, and variations in the calibrated values of individual batches of reagent. Variations of up to ? 20 % are found in INR values.
As a rule of thumb:
The greater the INR, the greater the possible variations, e.g.:
- INR < 2.5 possible variations 0.1 – 0.5
- INR 2.5 to 4.5 possible variations 0.5 – 0.9
- INR > 4.5 possible variations 1.0 – 2.0
These are due to reagent differences reported in the literature, even within the INR range (Schmitz et al., 1995). As more experience is gained with INR it is becoming clear that shifts in readings resulting from different reagent factor sensitivities cannot be fully compensated, even by using INR instead of % Quick (Ng et al., 1993). There are in some cases shifts of as much as 1.0 INR and more between various reagents, with the difference increasing as INR becomes higher (Cunningham et al., 1994; Schmitz et al., 1995). For instance, where a difference in readings is obtained using two reagents around 1.0 INR, the patient may be outside his optimum therapeutic range with reagent A, while he may be within it with reagent B.
Much as one might appreciate the concern and distrust of patients in the results they have obtained with CoaguChek® S, the regrettable fact is that these individual variations cannot be avoided. They are inherent in the PT measuring system. The view of the experts, therefore, is that when checking and, if necessary, adjusting their Marcumar dose, patients should always use the same measuring system. CoaguChek® S is a very sensitive system. Even minor decreases in the activity of clotting factors will lead to longer coagulation times and hence to higher INR and lower Quick values. Comparing the CoaguChek® S reading with a less sensitive laboratory reagent, which does not detect these changes in factor activity or detects them less well, may result in considerable differences in the measured coagulation time. This feature is not specific to the CoaguChek System; it applies among all laboratory reagents, too (Roussi et al., 1994; Gosselin et al., 1997). Reagents with a low ISI (International Sensitivity Index) are more sensitive than reagents with a high ISI and measure longer coagulation times (Hirsh et al., 1998). Likewise, between two such reagents in the laboratory there may be differences for one and the same patient of more than a whole INR unit (Adcock and Duff, 2000). Even with reagents of the same factor sensitivity (ISI = 10.0), there may be INR differences of up to 0.5 units (Ng et al., 1998). All of the monitors presently available for self-testing exhibit differences as compared to laboratory reagents. Most tend towards higher INR values (Gosselin et al., 2000; Oral Anticoagulation Monitoring Study Group, 2001). However, self-testing is still superior in helping to adjust the patient to within his optimum therapeutic range, so making a major contribution to reliable therapy of the patient.
In conclusion, please allow us to provide some additional advice in order to clarify any discrepant results there might be between CoaguChek and the laboratory.
1. Avoid possible sources of error when self-testing:
–Never take blood repeatedly from the same puncture site.
–Never squeeze blood out of the puncture site.
–Do not prick the finger until you are ready to apply the drop of blood immediately
to the strip (within 15 seconds).
2. Never compare individual values.
To evaluate differences you always need several readings (at least three independent replicates).
3. The laboratory and reference method most suitable for CoaguChek is the Hepato-Quick laboratory determination.
4. Residues of alcohol/soap on the skin may cause false-positive drifts.
5. We observe drifts especially when haematocrit values are below 32% or over 52%.
6. Lupus antibodies in the blood sample (which should, if necessary, be clarified through analysis) may be associated with a positive INR drift.
7. Always consider the possibility of INR being affected by other drugs taken simultaneously.
Lastly, remember that the proper INR range and the testing method to be used should be checked and defined in close association with the treating physician, who will have the individual’s case history.
Author:Dr. Norbert Weis, Roche Diagnostics (July 2005)