Screening for atrial fibrillation and treatment with anticoagulant to prevent stroke

Abstract

AIMS AND METHODS: The aim of this thesis was to evaluate the outcomes of two cross-sectional screening projects for atrial fibrillation (AF) in primary care using intermittent ECG three times per day over a 2-week period. The target population in the first project were 70–74-year-old individuals registered in a single primary healthcare centre, while in the second project the target population were patients 65 years of age and older attending four different primary healthcare centres. The outcomes were the detection rates of new AF cases and the initiation rates of oral anticoagulant (OAC) treatment to prevent stroke. In the first screening project, we also investigated the roll of NT-proBNP for detecting new AF cases. In addition, we assessed the safety of the first screening project at a three-year follow-up, as well as the cost-effectiveness of the screening. Finally, we studied the validity of self-pulse palpation for detecting new AF cases in the second screening project in which the participants were instructed in how to take their own pulses simultaneously with intermittent ECG measurements.

RESULTS: Study I. The target population was invited to the screening study when visiting the primary healthcare centre over a ten-month period, while those not in contact with the centre during this ten-month period were invited to participate by letter. Of the 415 eligible individuals, 324 (78.1%) participated in the study. The mean age of the participants was 72 years and 52.2% of them were female. In the target population, 34 (8.2%) patients had previously known AF. Among participants without previously known AF, 16 (5.5%) cases of AF were detected. The final prevalence of AF in the target population was 12%. OAC therapy was initiated in 88% of the patients with newly detected AF. Study II. Plasma NT-proBNP was measured in all patients with previously known AF, all newly detected AF and 53 participants without AF. The median NT-proBNP levels were 697 ng/L, 335 ng/L and 146 ng/L in patients with previously known AF, in patients with newly detected AF and in participants without AF, respectively. After adjustment for several clinical variables, the differences of median NT-proBNP levels were statistically significant between patients with previously known AF and patients with newly detected AF, as well as between patients with newly detected AF and those without AF. The area under the receiver operating characteristic curve for detection of new AF cases was 0.68 (95% CI 0.56 to 0.79), resulting in a cut-off point of 124 ng/L with 75% sensitivity, 45% specificity and 86% negative predictive value. Study III. While the mortality rate among patients with known AF was higher than those with no AF (hazard ratio 3.6, 95% CI 1.5 to 8.7), there was no statistically significant difference in the mortality rate between cases of new AF compared to those cases of no AF (hazard ratio 0.86, 95% CI 0.12 to 6.44). Adherence to OAC was 92%. No stroke or severe bleeding was detected. The incremental cost-effectiveness ratio of screening vs. no screening was EUR 2,389/QALY gained. In a probabilistic sensitivity analysis, the screening showed a 99% probability of being cost-effective compared to no screening at a willingness-to-pay threshold of EUR 20,000/Quality-Adjusted Life-Year (QALY). Study IV. A total of 1,010 patients (mean age 73 years, 61% female) participated in the study and 27 (2.7%, 95% CI 1.8 to 3.9%) new cases of AF were detected. Anticoagulants could be initiated in 26 (96%, 95% CI 81 to 100%) of these cases. A total of 53,782 simultaneous pulse and ECG recordings were registered. AF was verified in 311 ECG recordings, of which the pulse was palpated as irregular in 77 patients (25%, 95% CI 20 to 30% sensitivity per measurement occasion). 15 out of 27 AF cases felt an irregular pulse on at least one occasion (56%, 95% CI 35 to 75% sensitivity per individual). 187 individuals without AF felt an irregular pulse on at least one occasion. These resulted in a specificity of (98%, 95% CI 98 to 98%) and (81%, 95% CI 78 to 83%) per measurement occasion and per individual, respectively.

CONCLUSIONS: The detection rate for new AF in these screening studies was 5.5% among 70–74-year olds and 2.7% among patients aged 65 years and older. Initiation of OAC was high (88% and 96%, respectively) and three-year adherence was high (92%) in the first screening study. The participation rate in the first study was high (78.1%). NT-proBNP would appear to be a useful screening marker for AF detection and AF persistence while the validity of self-pulse palpation for AF detection was low. The screening appears to be safe and cost-effective using traditional cost-effectiveness thresholds (EUR 2,389/QALY gained). Thus, opportunistic screening of AF in primary care using intermittent ECG (with or without NT-proBNP) and initiation of OAC for detected AF cases could be a suitable method for preventing stroke. However, there is a need to evaluate the efficacy of such a screening programme by a large multicentre randomized control study with stroke as the endpoint.