Objective: Assessing the quality of anticoagulant (AC) treatment in primary health care with regard to safety.
Design: Surveys of patients on AC treatment during 1999.
Setting: Community health centres (CHCs) in the northeastern region of Stockholm County.
Subjects: Nine hundred and fifty-seven patients, from 16 CHCs.
Main outcome measures: Rate of bleeding and thromboembolic events during AC treatment and rate of values within the recommended treatment interval. The reporting of prothrombin-time (PT) results changed during the year from PT values to international normalised ratio (INR).
Results: A total of 48 bleeding events in 44 patients were noted, i.e. 6.8 per 100 patient-years. Of these, ten were major bleedings, 1.4 per 100 patient-years, including three fatal bleedings, 0.4 per 100 patient-years. Six thrombo-embolic events were noted during treatment, i.e. 0.8 per 100 patient-years. Bleeding events were more common at INR values greater than 2.80 than at values of 2.80, [relative risk (RR) 3.30, 95% confidence interval 1.90-5.71]. Of all the noted PT values, 65% were within the recommended intervals (the most common being PT 15-25%) and of all noted INR values 60% (the most common being INR 2.1-3.0). No differences in the rate of bleeding or the number of thrombo-embolic events between the periods of PT and INR results were found.
Conclusions: The rate of complications was low and AC treatment in primary health care seems to be as safe as in hospital clinics.