Avoiding “sticker” shock

Long-term administration of a coumarin drug is currently the only option for chronic oral anticoagulation therapy in the United States. Coumarins are effective and inexpensive, but close monitoring and frequent dose adjustments are usually required to achieve the desired antithrombotic effect and avoid bleeding. Monitoring requires venipuncture, performing a prothrombin time, and calculating the International Normalized Ratio (INR). Studies have established that an INR of 2 to 3 (or 2.5 to 3.5 in patients with mechanical heart valves) best balances safety and efficacy. The task of the clinician is to determine the dose of drug that consistently produces an INR in the therapeutic range. Current recommendations call for testing to begin 2 to 3 days after therapy is initiated. Once the INR is therapeutic, the monitoring frequency is decreased to 2 to 3 times weekly for 1 to 2 weeks, and if the INR remains in the therapeutic range, eventually to once every 4 weeks for as long as the patient is on the drug.²  Patients whose INRs remain in the therapeutic range on repeated testing are considered to have achieved stable anticoagulation.

Can the frequency of monitoring be decreased in patients on stable anticoagulation? This question has been addressed by Witt et al in this issue of Blood. They conducted an observational study of more than 6000 patients followed by a centralized clinical pharmacy anticoagulation service. Forty-one percent had all INR values within the therapeutic reference interval for a continuous 6-month interval (stable patients), and the remainder (59%) served as comparator patients. Patients had at least 1 INR determination every 8 weeks. In the comparator group, only 47% of INR values were in the therapeutic range, and this group had a significantly higher rate of bleeding (2.8% vs 0.8%; P < .001). Stable patients were less likely to have a goal INR ≥ 3.0 (OR = 0.48, 95% CI 0.38-0.61). Other predictors of a stable INR were age older than 70, and absence of diabetes, heart failure, or other chronic disease. Thus, the authors have identified and characterized patients potentially requiring less frequent anticoagulant monitoring. To confirm that such patients can be safely managed with less testing, they suggest a prospective trial of stably anticoagulated patients in whom the interval between INR determinations is progressively increased to 8 or even 12 weeks.

While such a suggestion is reasonable, the dismal track record of coumarins with respect to bleeding must be recognized. Despite careful monitoring of anticoagulant intensity, the risk of major hemorrhage in several large clinical trials has ranged from 0.3% to 0.5% per year as compared with controls,³  and during long-term anticoagulation, 1.1% per patient-year.⁴  Some of the hemorrhages might have occurred because of unpredictable events such as intercurrent illnesses or trauma. Changes in patient medications and use of nonprescription drugs and alcohol might have precipitated other bleeding events. Perhaps the best approach to limit these complications is a well-informed patient. In this regard, the 10 key issues for patient explanation (see table) are as valid today as they were when published 25 years ago by Errichetti et al.⁵  These issues should be continually reinforced while patients are taking these drugs. Well-educated patients who have achieved stable anticoagulation are candidates for less frequent monitoring and fewer needle sticks.