Effectiveness of anticoagulation among patients discharged from hospital on warfarin

Objectives: To examine the effectiveness of anticoagulation among patients discharged from hospital on warfarin in the care of general practitioners (GPs). Design and setting: A historical cohort with questionnaires to patients discharged from a major metropolitan teaching hospital and their GPs. Participants: Patients discharged between 1 February 1995 and 31 January 1996 identified from hospital pharmacy records as being prescribed warfarin, and their treating GPs. Main outcome measures: Frequency of testing and levels of international normalised ratio of prothrombin time (INR) within six months of discharge; level of INR aimed at by GP; complication rates; and patient knowledge about anticoagulation. Results: Replies were received from 242 (68%) patients and pathology records were examined for 195 (81%) of these. The median gap between INR measures was seven days. The median of the median INR level for each patient was 2.4 (rising to 2.7 in patients with an artificial heart valve); 24% of observed patient time was spent at an INR level of less than 2.0, 54% between 2.0 and 2.9, 18% between 3.0 and 3.9 and 4% at an INR level of 4.0 or more. There were five confirmed major complications (equivalent to 5 per 100 patient-years). Twenty-seven per cent of patients answered at least eight of the 10 knowledge questions correctly: education level predicted knowledge, but there was no relationship between knowledge and INR level. Conclusions: Among this unselected group of patients whose anticoagulation was managed by GPs, there was a high frequency of laboratory testing, INR levels were controlled safely and complication rates were comparable with those in published reports.