Primary stroke prevention
An artificial intelligence-based model for prediction of atrial fibrillation from single-lead sinus rhythm electrocardiograms facilitating screening
Tove Hygrell, Fredrik Viberg, Erik Dahlberg, Peter H Charlton, Katrin Kemp Gudmundsdottir, Jonathan Mant, Josef Lindman Hörnlund, Emma Svennberg
PMID: 36881777. DOI: 10.1093/europace/euad036
Abstract Aims Screening for atrial fibrillation (AF) is recommended in the European Society of Cardiology guidelines. Yields of detection can be low due to the paroxysmal nature of the disease. Prolonged heart rhythm monitoring might be needed to increase yield but can be cumbersome and expensive. The aim of this study was to observe the accuracy of an artificial intelligence (AI)-based network to predict paroxysmal AF from a normal sinus rhythm single-lead ECG.
Methods and results A convolutional neural network model was trained and evaluated using data from three AF screening studies. A total of 478 963 single-lead ECGs from 14 831 patients aged ≥65 years were included in the analysis. The training set included ECGs from 80% of participants in SAFER and STROKESTOP II. The remaining ECGs from 20% of participants in SAFER and STROKESTOP II together with all participants in STROKESTOP I were included in the test set. The accuracy was estimated using the area under the receiver operating characteristic curve (AUC). From a single timepoint ECG, the artificial intelligence-based algorithm predicted paroxysmal AF in the SAFER study with an AUC of 0.80 [confidence interval (CI) 0.78-0.83], which had a wide age range of 65-90+ years. Performance was lower in the age-homogenous groups in STROKESTOP I and STROKESTOP II (age range: 75-76 years), with AUCs of 0.62 (CI 0.61-0.64) and 0.62 (CI 0.58-0.65), respectively.
Conclusion An artificial intelligence-enabled network has the ability to predict AF from a sinus rhythm single-lead ECG. Performance improves with a wider age distribution.
Feasibility of screening for atrial fibrillation in a domiciliary setting:
Opportunistic one-time screening at preventive home visits in municipalities
Poulsen P-B. , Hemmingsen U., Melgaard T-A, Buch Elleby H., Wedell-Wedellsborg D., Dybro L., Lund I-M., Dixen U., Frost L.
Scandinavian Cardiovascular Journal, 2022, VOL. 56, NO. 1, 243–246
Current evidence base for atrial fibrillation (AF) screening is insufficient. An important finding in the STROKESTOP study was that non-participants had significantly worse outcomes. In a group of potentially non-participants feasibility of opportunistic screening in a domiciliary setting with municipality preventive home visits to citizens >75 years was investigated. Handheld ECG device was used by trained municipality caregivers followed by cardiologist assessment. Eighty-five percent consented to being screened, and seven of 477 screened were found with AF. Opportunistic screening in preventive home visits had a high participation rate and was feasible. Randomized trials are needed before making any firm conclusions.
Clinical outcomes in systematic screening for atrial fibrillation (STROKESTOP):
A multicentre, parallel group, unmasked, randomised controlled trial
Svennberg E., Friberg L., Frykman V., Al-Khalili F., Engdahl J. & Rosenqvist M.,
The Lancet 2021, August 29
Presentation of the STROKESTOP study online at EHRA 2021 by Emma Svennberg
The STROKESTOP study was the world’s first randomised clinical trial with endpoints of reduction of mortality and morbidity, as a comparison of systematic screening for atrial fibrillation among elderly individuals with Zenicor-ECG, compared with no screening. The results showed a small net benefit compared with the standard of care, indicating that screening is safe and beneficial in older populations.
Stepwise mass screening for atrial fibrillation using N-terminal B-type natriuretic peptide: the STROKESTOP II study
Gudmunsdottir, K.K., Fredriksson, T., Svennberg, E., Al-Khalili, F., Friberg, L., Frykman, V., Hijazi, Z., Rosenqvist, M. & Engdahl, J.
Europace 2019, Sept 17
The STROKESTOP 2 study is a randomised controlled trial designed to answer the question if atrial fibrilliation screening using a stepwise procedure stratisfying participants to high or low risk group using NT-Pro BNP followed by ECG screening with one ECG (low risk group) and two week ECG screening (high risk group) will lower stroke incidents compared to standard care. This study reports on participation, risk stratification and AF detection of the two study arms.
Feasibility and outcomes of atrial fibrillation screening using intermittent electrocardiography in a primary healthcare setting: A cross-sectional study
Ghazal F., Theobald H., Rosenqvist M., Al-Khalili F.
PLOS One 2018, May 24
All persons between 70-74 years at a health center in Stockholm were invited to screen for two weeks with Zenicor ECG, with a combination of opportunistic and systematic recruitment. Participation was very high, of a total of 415 persons, 324 (78.1%) participated in the screening . The study found 5.5% previously unknown atrial fibrillation. The proportion of newly diagnosed persons who started with stroke preventive treatmentwas high, 88%.
A prospective 5-year follow-up after population-based systematic screening for atrial fibrillation
Engdahl J., Holmén A., Rosenqvist M., Strömberg U.
Europace 2018, April 24
This is the first study ever reporting results from a long-term follow-up after an atrial fibrillation (AF) screening program. After five years the incidences of ischaemic stroke showed a significant beneficial effect in the screened group. The results also showed a high adherence to anticoagulant treatment, and that the natural course of screening detected AF was similar to clinical detected AF.
Mass Screening for Untreated Atrial Fibrillation: The STROKESTOP Study (Main study)
Svennberg, E., Engdahl, J., Al-Khalili, F., Friberg, L., Frykman, V. & Rosenqvist, M.
Circulation. 2015. Jun 23;131(25):2176-84
In a globally unique study at Karolinska Institutet, 13.000 people were offered screening for atrial fibrillation with Zenicor-EKG. Five percent of the participants were identified as new candidates for treatment with oral anticoagulants. Health economical calculations based on the results from the study, show that systematic mass screening with Zenicor-ECG for atrial fibrillation is cost-effective.
Safe automatic 1-lead ECG screening for atrial fibrillation
Svennberg E., Stridh M., Engdahl J., Al-Khalili F., Friberg L., Frykman V. & Rosenqvist M.
Europace 2016, Oct 6
In this the Zenicor-system algorithm for interpretation support was validated against manual ECG interpretation by cardiologist on 30 seconds ECG strips from lead I. A total of 80,149 ECGs from 3,209 patients from the STROKESTOP I study were analyzed. AF was diagnosed in 84 patients of which the algorithm detected all 84 (100%).
Stepwise Screening of Atrial Fibrillation in a 75-Year Old Population: Implications for Stroke Prevention
Engdahl J., Andersson L., Mirskaya M., Rosenqvist M.
Circulation (2013;127:930-937)
In a pilot for the major STROKESTOP study, around 1.300 75-year-olds in Halland were offered screening for atrial fibrillation. Through a two-week investigation with Zenicor-EKG, 7.2% newly detected atrial fibrillations were diagnosed in the screened population, and the number of people treated with anticoagulants was doubled.
Knowledge resource: Health economics evaluation of primary preventive atrial fibrillation screening with thumb ECG
Tandvårds- och läkemedelsförmånsverket (The Dental and Pharmaceutical Benefits Agency) (2014)
Link to study (Only available in Swedish)
As part of a Government commission to evaluate medical technology, based on figures from the StrokeStop study, TLV (The Dental and Pharmaceutical Benefits Agency) has carried out a health economics assessment of primary preventive fibrillation screening. The report shows that such screening using a thumb ECG is cost-effective.
Cost-effectiveness of mass screening for untreated atrial fibrillation using intermittent ECG recording
Aronsson, M., Svennberg, E., Rosenqvist, M., Endahl, J., Al-Khalili, F., Friberg, L., Frykman-Kull, V. & Levin, L-Å.
Europace. 2015. Apr 13. 1023-1029
In a globally unique study at Karolinska Institutet, 13.000 people were offered screening for atrial fibrillation with Zenicor-ECG. Five percent of the participants were identified as new candidates for treatment with oral anticoagulants. According to this Health economical study, based on the results from the STROKESTOP study, systematic mass screening for atrial fibrillation with Zenicor-ECG is cost-effective.
Screening for atrial fibrillation with baseline and intermittent ECG recording in an out-of-hospital population (FLIP)
Hendrikx T., Hörnsten R., Rosenqvist M., Sandström H.
BMC Cardiovascular Disorders (2013, 13:41)
In the FLIP project (Flimmer i Primärvården – atrial fibrillation in primary care) carried out in Västerbotten County, 898 patients with at least one risk factor for stroke (CHADS2 >1) were invited to take part in a four-week investigation with Zenicor-EKG. They took an ECG reading twice a day for 28 days. A total of 4 percent of patients were diagnosed with a previously undetected atrial fibrillation and could be administered oral anticoagulant medication for stroke prevention treatment.
Systematic screening for atrial fibrillation in a 65-year-old population with risk factors for stroke: data from the Akershus Cardiac Examination 1950 study.
Berge, T., Brynildsen, J., Larssen, HKN., Onarheim, S., Jenssen, GR., Ihle-Hansen, H., Christophersen, IE., Myrstad, M., Røsjø, H., Smith, P. & Tveit, A.
Europace, 2017. Oktober 31.
The ACE-1950 study took place in Akershus area around Oslo in Norway where 1,600 65-year olds with one additional risk-factor where screened with two weeks intermittent Zenicor-ECG to detect previously known atrial fibrillation. A total of 7,6% of the examined population had atrial fibrillation of which 0,9% where newly discovered.
Geographic and socio-demographic differences in uptake of population-based screening for atrial fibrillation: The STROKESTOP I study.
Engdahl J, Holmén A, Svennberg E, Friberg L, Frykman-Kull V, Al-Khalili F, Rosenqvist M & Strömberg U.
Int J Cardiol. 2016 Jul 29;222:430-435. [Epub ahead of print]
In an analysis of the persons choosing not to participate in the systematic screening study; STROKESTOP I, the conclusion was drawn that participation was influenced by socioeconomic and geographic conditions. This will provide an important input on how screening for AF should be organized to gain as many as possible. Geographic variation was especially noted in Stockholm that only had one screening centre.
Uptake of atrial fibrillation screening aiming at stroke prevention: geo-mapping of target population and non-participation
Engdahl J., Holmén A., Rosenqvist M. & Strömberg U.
BMC Public Health (2013, 13:715)
In an analysis of people who chose not to participate in the StrokeStop pilot screening, the conclusion could be drawn that non-participants mainly came from areas with a higher number of risk factors, a high proportion of immigrants and an older population on average. This provides key data on how a larger screening initiative should be organised in order to benefit as many people as possible.
