The University of Southampton
University of Southampton Institutional Repository

A randomised controlled trial and cost-effectiveness study of systematic screening (targeted and total population screening) versus routine practice for the detection of atrial fibrillation in people aged 65 and over. The SAFE study

Hobbs, F.D.R., Fitzmaurice, D.A., Mant, J., Murray, E., Jowett, S., Bryan, S., Raftery, J., Davies, M. and Lip, G. (2005) A randomised controlled trial and cost-effectiveness study of systematic screening (targeted and total population screening) versus routine practice for the detection of atrial fibrillation in people aged 65 and over. The SAFE study Health Technology Assessment, 9, (40), [93pp].

Record type: Article


Background Atrial fibrillation (AF) is a major risk factor for stroke. This risk can be reduced through treatment with antithrombotic therapy, with a risk reduction of up to 68% observed with warfarin therapy. Guidelines for treatment of AF recommend ages 65 years and over as an indication for treatment with antithrombotic therapy in the presence of AF. This raises the question of whether screening for AF would be a useful policy, and if so what would be the best method for screening. There are no good data on the prevalence of AF in the UK. One small UK study (four practices, n = 3001) demonstrated that systematic nurse-led screening detected more cases than opportunistic case finding; however, most of those cases detected were already diagnosed. Two further single practice-based studies investigated the role of practice nurses in the screening process and whole population screening, but were too small to be meaningful.
Objectives To evaluate the incremental cost-effectiveness of targeted, population and opportunistic screening with prompts compared with routine clinical practice. To evaluate the relative cost-effectiveness of different methods of recording and interpreting the ECG within a screening programme. To identify the prevalence and incidence of AF in patients aged 65 years and over.
Methods This multicentred randomised controlled trial involved patients aged 65 years and over from 50 primary care centres across the West Midlands. These purposefully selected general practices were randomly allocated to 25 intervention practices and 25 control practices. GPs and practice nurses in the intervention practices received education on the importance of AF detection and ECG interpretation. Patients in the intervention practices were randomly allocated to systematic (n = 5000) or opportunistic screening (n = 5000). Prospective identification of pre-existing risk factors for AF within the screened population enabled comparison between targeted screening of people at higher risk of AF and total population screening. AF detection rates in systematically screened and opportunistically screened populations in the intervention practices were compared with AF detection rate in 5000 patients in the control practices. The screening period was 12 months.
Results The total number of patients included in each arm was: control 4936, opportunistic screening 4933 and systematic screening 4933. Baseline prevalence of AF was 7.2%, with a higher prevalence in males (7.8%) and patients aged 75 years and over (10.3%). The control population demonstrated higher baseline prevalence (7.9%) than either the systematic (6.9%) or opportunistic (6.9%) intervention population. In the control population 47 new cases were detected (incidence 1.04% per year). In the opportunistic arm 243 patients without a baseline diagnosis of AF were found to have an irregular pulse, with 177 having an ECG, yielding 31 new cases (incidence 0.69% per year). A further 44 cases were detected outside the screening programme (overall incidence 1.64% per year). In the systematic arm 2357 patients had an ECG yielding 52 new cases (incidence 1.1% per year). Of these, 31 were detected by targeted screening and a further 21 by total population screening. A further 22 cases were detected outside the screening programme (overall incidence 1.62% per year).
In terms of ECG interpretation, computerised decision support software (CDSS) gave a sensitivity of 87.3%, a specificity of 99.1% and a positive predictive value (PPV) of 89.5% compared with the gold standard (cardiologist reporting). GPs and practice nurses performed less well. The only difference in performance between intervention populations and controls was that practice nurses from the control arm performed less well than intervention practice nurses on interpretation of limb-lead (PPV 38.8% versus 20.8%) and single-lead (PPV 37.7% versus 24.0%) ECGs.
The within-trial economic evaluation results showed the lowest incremental cost to be for the opportunistic arm, with an incremental cost-effectiveness ratio of £337 for each additional case detected compared to the control arm. Opportunistic screening dominated both more intensive screening strategies. Model-based analyses showed small differences in cost and quality-adjusted life-years for different methods and intensities of screening, but annual opportunistic screening resulted in the lowest number of ischaemic strokes and greatest proportion of cases of AF diagnosed. Probabilistic sensitivity results indicated that there was a probability of approximately 60% that screening from the age of 65 was cost-effective in both men and women.
Conclusions The prevalence of AF in this population was found to be 7.2%. The incidence ranged from 1.04 to 1.64% per annum. Within the trial, in terms of a screening programme, the only strategy that improved on routine practice was opportunistic screening, at a cost of £337 per additional case detected. Model-based analyses indicated that there was a probability of approximately 60% of annual opportunistic screening being cost effective. Use of CDSS may be considered for analysis of ECGs for detection of AF.

Full text not available from this repository.

More information

Published date: October 2005
Keywords: stroke, diagnosis, role, risk factors, pulse, probability, patients, agents, arm, education, design, incidence, population, nurses, aged, prevalence, risk, male, methods, atrial fibrillation, software


Local EPrints ID: 61814
ISSN: 1366-5278
PURE UUID: 74858b27-2b8e-497d-a060-310dcc07ce59

Catalogue record

Date deposited: 10 Sep 2008
Last modified: 17 Jul 2017 14:21

Export record

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton:

ePrints Soton supports OAI 2.0 with a base URL of

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.