Audit of anticoagulant therapy and acute hospital admissions

We conducted an audit on the contribution of failure of control of anticoagulant therapy to acute hospital admissions. Over a period of 3 months there were 1480 acute admissions. One-hundred-and-twelve (7.6%) of these patients were on anticoagulant therapy. One-hundred-and-three of these 112 patients were evaluated, 74 patients had international normalized ratios (INRs) in the therapeutic range, of whom four (5.4%) bled from causes unlikely to be due directly to anticoagulant therapy. Twenty-nine patients were over-anticoagulated. Of these, 17 (59%) were admitted with bleeding symptoms, which may have been a consequence of high INR, while one had a very high INR but no bleeding. Eleven more patients were admitted for reasons unrelated to anticoagulant therapy but were found to have over-therapeutic INRs, which may have influenced their subsequent hospital management. The only clear difference between the bleeding and nonbleeding groups was age. Reasons for over-anticoagulation were poor patient compliance in 31%, influence of other medications in 17, congestive heart failure in 28%, and unknown in 24%. In conclusion, 22/1480 hospital admissions (1.5%) were due to warfarin complications and 16/21 bleeding patients had INRs > 4.5. These are admissions that could potentially be avoided with better anticoagulation control.