COST-EFFECTIVENESS OF SCREENING FOR ASYMPTOMATIC CAROTID ARTERY STENOSIS

Purpose: Since the ACAS and ACST trials have shown that for certain groups Duplex Ultrasound (DUS) screening for asymptomatic carotid artery stenosis and subsequent endarterectomy in case of high grade stenosis is beneficial, the question whether population based screening is warranted has raised much discussion. Despite valid evidence still being unavailable, recently, a US guideline was issued stating that this would be the case. We set out to address the issue directly using modelling and individual patient data cohort estimates of prevalence.

Method: A Markov model (Monte Carlo Simulation) simulating the histories of cohorts of patients according to prevalence distribution of grade of stenosis (<50, 50-69 and 70-99), age (55, 65 and 75 years), gender and co-morbidity was developed reflecting National survival statistics and stroke occurrence. Costs, effects in terms of stroke and overall survival and utility estimates were literature based. Initially, hypothetical prevalence thresholds according to patient characteristics (age, gender, co-morbidity) were identified for screening to become cost-effective (iCER< €20,000/QALY). In parallel pooled individual patient data from several international population based cohort studies were obtained to estimate the prevalence of severe asymptomatic carotid artery stenosis according to the above patient characteristics.

Result: Screening 55 year old men and women appeared cost-effective for a prevalence over 1%. For 65 year old men this was the case for a prevalence over 5%, while for women already from 1% onward screening appeared cost-effective. For 75 year old groups similar results were predicted. Initial analyses on the pooled data (N=17.818) revealed that the prevalence in the general population, especially for men, fell well below the thresholds required for screening to become cost-effective in the general population.

Conclusion: These results corroborate the notion that only in case of high prevalence and sufficiently long life-expectancy screening might be warranted. Presently, such subgroups cannot be accurately identified. This is part of ongoing research. We presently maintain a prudent conclusion that screening of the general population to detect and subsequently operate on asymptomatic carotid artery stenosis should not be introduced.