Cost effectiveness of primary stroke prevention in atrial fibrillation: Swedish national perspective.

Abstract

Objective--To assess the potential effects of primary prevention with anticoagulants or aspirin in atrial fibrillation on Swedish population.

Design--Analysis of cost effectiveness based on the following assumptions: about 83 000 people have atrial fibrillation in Sweden, of whom 22 000 would be potential candidates for treatment with anticoagulants and 55 000 for aspirin treatment; the annual 5% stroke rate is reduced by 64% (with anticoagulants) and 25% (with aspirin); incidence of intracranial haemorrhage of 0.3%, 1.3%, or 2.0% per year; direct and indirect costs of a stroke of Kr180 000 and Kr90 000; estimated annual cost of treatment is Kr5030 for anticoagulants and Kr100 for aspirin.

Setting--Total Swedish population.

Main outcome measures--Direct and indirect costs of stroke saved, number of strokes prevented, and cost of preventive treatment.

Results--Depending on the rate of haemorrhagic complications 34 to 83 patients would need to be treated annually with anticoagulants to prevent one stroke; 83 patients would need to be treated with aspirin. Giving anticoagulant treatment only would reduce costs by Kr60 million if the incidence of intracranial haemorrhage were 0.3% but would imply a net expense if the complication rate exceeded 1.3%. The total savings from giving anticoagulant (22 000 patients) and aspirin (55 000 patients) treatment would be Kr175 million per year corresponding to [pound] 2 million per million inhabitants each year.

Conclusions--Treatment with anticoagulants and, if contraindications exist, with aspirin is cost effective provided that the risk of serious haemorrhage complications due to anticoagulants is kept low.

Introduction

Non-valvar atrial fibrillation is associated with an increased risk of ischaemic stroke.[1] Four randomised studies have recently reported that treatment with anticoagulants significantly reduces the risk of stroke.[2-5] One of these studies also showed a positive effect with aspirin.[3] No analysis has considered the effects on a whole society if this treatment was adopted.

We looked at the medical and economic consequences for the population of Sweden (8.5 million inhabitants) of introducing primary prevention with anticoagulant treatment as standard treatment for patients with atrial fibrillation. We assessed the economic effects of using anticoagulants or aspirin alone and of using anticoagulants in selected patients and aspirin in the others.

Methods

We based our calculations of the number of people eligible for treatment on prevalences and risks reported in previous studies. The Framingham study showed that the prevalence of atrial fibrillation increased from 0.5% in 50-59 year old people to 8.8% in 80-89 year olds.[1] Other population based studies, including a report from Sweden, have shown increasing prevalence rates of the same magnitude in different age groups.[6 7] We therefore used the Framingham data to estimate the total number of patients with chronic atrial fibrillation in Sweden. The incidence of stroke in non-valvar atrial fibrillation patients …