Although the INR has improved standardization, it is still limited by variations in reagents, instrumentation, and detection methods. Identical samples tested in different PT assays do not give the same results even after conversion to INR. These imperfections reflect differences in PT reagents or techniques used to perform the test. Perhaps the most important study to bring this discrepancy to the attention of physicians in North America was the prospective study of patients with venous thrombosis by Hull and coworkers, showing that patients managed in a similar therapeutic range by PT ratio (but different by INR), using sensitive and insensitive reagent, actually represented a degree of over anticoagulation in the "insensitive" group, leading to a higher rate of hemorrhagic complications (1). These studies and others support the adoption of the INR system of reporting and the use of more sensitive thromboplastins.

Thromboplastin reagents whether recombinant or derived from a natural source vary widely with regard to their responsiveness or sensitivity. The sensitivity of a thromboplastin indicates it’s strength, and a stronger thromboplastin will result in a faster (or shorter) prothrombin. Therefore as the sensitivity index of the reagent increases, the responsiveness is lessened. A thromboplastin with a high ISI (2.0 or higher) is considered a less sensitive, less responsive reagent. A small change in the PT result is magnified in the INR since the ratio is raised to the power of the ISI. A lower ISI (around 1.0) allows for a wider range of PT values and changes or differences in instrumentation systems are less magnified in the INR. Therefore, the precision of the INR is improved. This allows a physician to fine tune warfarin dosage for patients. An additional argument in support of a sensitive thromboplastin is the greater ease of evaluating liver patients. The effects of liver disease are much more easily detected with sensitive low ISI thromboplastins.

In addition to being influenced by the choice of thromboplastin, the INR is sensitive to the instrument utilized in the performance of the prothrombin time. Many commercial vendors provide ISI’s for their reagents in combination with the most widely used instruments. However, it is the responsibility of the laboratory to validate its prothrombin times adequately.

A recent study reported that the INR can be affected by the concentration of citrate anticoagulant (2). It was noted that low ISI thromboplastins yield higher INR values when samples were collected in 3.8% citrate. This type of variation was not seen with less sensitive (high ISI) reagents. To eliminate these variations The National Committee for Clinical Laboratory Standards and the International Society for Thrombosis and Hemostasis recommend that 3.2% citrate be used for specimen collection.

While the INR reporting format improves the management of anticoagulated patients, the INR is inappropriate in three subsets of patients. First it is designated for patients who are stabilized on oral anticoagulant therapy and is not appropriate for those patients who have recently begun their treatment. Second, patients with liver disease should not be monitored by the INR since they frequently lack circulating factors. Lastly, patients who are being routinely screened for clotting factor deficiencies should not be evaluated with the INR.

Clinicians should use the INR in making therapeutic decisions regarding oral anticoagulants (3). Despite the fact that using sensitive thromboplastins for routine screening will increase the frequency of finding modest abnormalities among patients not on anticoagulants, tremendous benefits will be achieved in management of patients on oral anticoagulants.