WHAT IS KNOWN AND OBJECTIVE: There is wide inter-patient and intra-patient variability in the pharmacodynamic profile of warfarin. To determine the prevailing aetiologies of non-therapeutic warfarin anticoagulation episodes among patients currently enrolled in an outpatient anticoagulation clinic and compare the relative frequency in which they occur compared to therapeutic anticoagulation regimens. METHODS: Prospective, observational cohort study set within three pharmacist-managed anticoagulation clinics in a community outpatient health system. Patients were included, if they were seen for an office visit during the 6-month period from September 2006 to March 2007 and evaluated for the presence or absence of 12 investigational factors linked to non-therapeutic anticoagulation results. Multivariate stepwise logistic regression performed to assess predictive value of each factor. RESULTS: A total of 5817 patient-visits were documented producing 2886 (49.6%) non-therapeutic and 2931 (50.4%) therapeutic International Normalized Ratio (INR) readings. The most prevalent aetiologies linked to non-therapeutic INR results included change in dietary vitamin K intake (16.9%, OR 6.4), non-compliance (15.0%, OR 4.9), and initiation of anticoagulant therapy (9.9%, OR 2.3). The factor with the highest predictive value of non-therapeutic INR results was a change in health status (OR 9.5) despite its lower rate of frequency (4.9%). Despite identification of many causative factors in this study, 40.2% of non-therapeutic INR readings had no known aetiology. In the end, the lack of any study factor was a greater predictor of therapeutic anticoagulation (86.2%), than the presence of a study factor was for predicting non-therapeutic INR values (51.4%). WHAT IS NEW AND CONCLUSION: Change in health status was the strongest predictor of non-therapeutic INR levels out of the investigational factors evaluated. Our study demonstrated that there are many aetiologies for non-therapeutic INR values that were not explained by our investigational factors.
WHAT IS KNOWN AND OBJECTIVE: There is wide inter-patient and intra-patient variability in the pharmacodynamic profile of warfarin. To determine the prevailing aetiologies of non-therapeutic warfarin anticoagulation episodes among patients currently enrolled in an outpatient anticoagulation clinic and compare the relative frequency in which they occur compared to therapeutic anticoagulation regimens. METHODS: Prospective, observational cohort study set within three pharmacist-managed anticoagulation clinics in a community outpatient health system. Patients were included, if they were seen for an office visit during the 6-month period from September 2006 to March 2007 and evaluated for the presence or absence of 12 investigational factors linked to non-therapeutic anticoagulation results. Multivariate stepwise logistic regression performed to assess predictive value of each factor. RESULTS: A total of 5817 patient-visits were documented producing 2886 (49.6%) non-therapeutic and 2931 (50.4%) therapeutic International Normalized Ratio (INR) readings. The most prevalent aetiologies linked to non-therapeutic INR results included change in dietary vitamin K intake (16.9%, OR 6.4), non-compliance (15.0%, OR 4.9), and initiation of anticoagulant therapy (9.9%, OR 2.3). The factor with the highest predictive value of non-therapeutic INR results was a change in health status (OR 9.5) despite its lower rate of frequency (4.9%). Despite identification of many causative factors in this study, 40.2% of non-therapeutic INR readings had no known aetiology. In the end, the lack of any study factor was a greater predictor of therapeutic anticoagulation (86.2%), than the presence of a study factor was for predicting non-therapeutic INR values (51.4%). WHAT IS NEW AND CONCLUSION: Change in health status was the strongest predictor of non-therapeutic INR levels out of the investigational factors evaluated. Our study demonstrated that there are many aetiologies for non-therapeutic INR values that were not explained by our investigational factors.