If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The aim of this study was to establish the prevalence and prognostic implication of progressive supraventricular arrhythmias from frequent supraventricular ectopic complexes, isolated, in bi- or trigeminy, to supraventricular tachycardias with different characteristics. In the STROKESTOP I mass-screening study for atrial fibrillation (AF) in 75- and 76-year olds in Sweden, participants registered 30-second intermittent ECG twice daily for two weeks. The ECG-recordings from STROKESTOP I were re-evaluated using an automated algorithm to detect individuals with frequent supraventricular ectopic complexes or runs. Detected episodes were manually re-examined to confirm the findings. The primary endpoint was AF as ascertained from the national Swedish Patient register. Exploratory secondary endpoints were stroke and death. Median follow-up was 4.2 (interquartile range [IQR] 3.8-4.4) years. Of the examined 6,100 participants, 85% were free of significant supraventricular arrhythmia. In the 894 participants that had arrhythmia, frequent supraventricular ectopic complexes were the most common arrhythmia, n = 709 (11.6%) and irregular supraventricular tachycardias were more common than regular. Individuals with the most AF similar supraventricular tachycardias, irregular and lacking p-waves (termed micro-AF), n = 97 (1.6%) had the highest risk of developing AF (hazard ratio 4.3; 95% confidence interval [CI] 2.7-6.8). They also had increased risk of death (hazard ratio 2.0; CI 1.1-3.8). In conclusion, progression of atrial arrhythmias from supraventricular ectopic complexes to more AF-like episodes is associated with development of AF. Extended screening for AF should be considered in individuals with frequent supraventricular activity, especially in those with supraventricular tachycardias with AF characteristics.
According to consensus, atrial fibrillation (AF) is an irregular heart rhythm with absence of p-waves, lasting at least 30 seconds.
2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS).
The 30-second rule is arbitrary and based on consensus. One could argue that AF might be a continuous disease, ranging from occasional bursts of AF-like activity to permanent AF, reflecting the severity of the underlying pathophysiological process rather than a dichotomous condition. It has been reported previously in cohort studies that patients with supraventricular ectopic complexes (SVECs) and supraventricular tachycardias (SVTs) have an increased risk of developing AF.
In this study we aimed to characterise the prevalence of frequent SVECs, SVECs in bigeminy, SVECs in trigeminy and SVTs with different characteristics. We also aimed to determine the association between these arrhythmias and progression to clinically diagnosed AF. As exploratory secondary endpoints we also investigated the relationship between these arrhythmias, stroke and death.
Methods
This is a study using data from a population-based screening study crossmatched with data from national Swedish health registers. Our study population is derived from the systematic screening study for AF, STROKESTOP, conducted 2012-2014 in the Stockholm region and Halland county, Sweden.
In short, during STROKESTOP all residents in Stockholm or Halland that were 75- and 76-years old (born in 1936 or 1937) were randomized 1:1 to be invited to either participate in the study or be a part of a control group. Participants in the screening group, n = 7,173, were asked to perform 30-second intermittent ECG twice daily with a one-lead ambulatory handheld Zenicor II device (Zenicor Medical Systems, Stockholm, Sweden) for two weeks. Within the STROKESTOP study, participants were followed-up in registries with prespecified endpoints.
As we wanted to study progression to AF, patients diagnosed with AF prior to or during the STROKESTOP study were excluded, leaving n = 6,289 eligible for follow-up. The intermittent ECGs of 26 individuals were lost, and there was no registry data available in 151 participants. During SVT characterization 12 individuals were excluded from participation as re-interpretation of their ECGs could not with certainty dismiss AF. Hence a total of 6,100 individuals were included.
ECG-recordings from the included patients were re-evaluated using an automated algorithm (Cardiolund ECG Parser, www.cardiolund.com) to detect individuals with frequent isolated SVECs, SVECs in bigeminy, SVECs in trigeminy and SVTs with different characteristics.
The algorithm identified SVECs by prematurity, post-contraction pauses and the duration of QRS complex. To be considered an isolated SVEC, the supraventricular beat needed to be preceded and followed by normal sinus rhythm. All participants included in the top tenth percentile according to SVEC count were determined to have frequent isolated SVECs. The SVEC count was reported as the total number of isolated SVECs.
Bigeminy and trigeminy were identified as SVECs followed by one versus two sinus beats in more than three repeat sequences. All ECG episodes identified as SVECs in bigeminy or trigeminy by the algorithm were manually re-examined to confirm the findings. SVECs in bigeminy and trigeminy were presented as number of 30-second ECGs with presence of each arrhythmia.
Episodes that had ≥5 beats with rate-to-rate (RR) intervals ≤13% of median RR-interval (based on analysis of RR-intervals within AF episodes used by the clinical decision support system Cardiolund
Safe automatic one-lead electrocardiogram analysis in screening for atrial fibrillation.
Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology.2017; 19: 1449-1453
) and with mean RR-intervals of ≤600 milliseconds (corresponding to the common tachycardia definition of 100 beats per minute) were identified by the algorithm as suspected SVTs. Such an episode was always preceded and followed by normal sinus rhythm. The algorithm quality control included exclusion of candidates with simultaneous local baseline disturbances, and with beats of very different or unrealistic morphologies. The automated algorithm performed quality control on all SVT candidates to reduce the number of false detections caused by disturbances.
All ECGs from patients with SVTs according to the algorithm were first manually assessed on a patient level for the presence of p-waves. Second, ECGs were dichotomised into regular or irregular. Irregularity was assessed as ≥13% beat-to-beat variability within the SVT episode. Third, if the p-wave/QRS ratio was below 20%, the episodes was deemed as lacking p-waves, Figure 1. Some SVT episodes were irregular and AF-like with regards to absence of p-waves but that still did not fulfil the AF time-definition of 30 seconds. These episodes were named micro-AF, according to our previous publication.
Figure 1Supraventricular tachycardia classification. a) Regular SVT with p-waves b) Irregular SVT with p-waves c) Regular SVT without p-waves d) Irregular SVT without p-waves (micro-AF).
Before we used the algorithm in our study database it was validated by us for detection of micro-AF in a similar large dataset from the STROKESTOP II study, in which all micro-AF episodes had previously been manually annotated. The evaluation of the automated algorithm is presented in the result section of this manuscript.
Participants were allocated to mutually exclusive groups according to the most AF-like recorded supraventricular arrhythmia, progressing from normal rhythm, frequent isolated SVECs, SVECs in trigeminy, SVECs in bigeminy to SVTs, Figure 2. The SVT episodes were reported according to classification, length of the episode displaying most abnormality and number of episodes independent of classification. For SVT episodes the classes were divided into ascending AF-similarity from regular with p-waves, irregular with p-waves, regular without p-waves, to irregular without p-waves (micro-AF), which was considered the most AF-like.
Patient registries from the National Board of Health and Welfare and the Cause of Death register were used for endpoint retrieval and baseline medical history, Supplementary Table 1. The primary endpoint was a diagnosis of AF or atrial flutter. The secondary endpoints were ischaemic stroke diagnosis and all-cause mortality. The follow-up period was defined as years until AF diagnosis, death or end of follow-up for AF analysis. For analysis of secondary endpoints, the follow-up period was defined as years until ischaemic stroke, death or end of follow-up. Censoring was made on the day of an event, death, emigration or end of the follow-up period.
Non-normally distributed continuous variables and ordinal variables were reported as median, interquartile range (IQR). Normally distributed continuous variables were reported as mean, confidence interval (CI). Cox regression analysis was used for univariable and multivariable analyses. Variables included in CHA2DS2-VASc score were considered potential confounders and were adjusted for, except from thromboembolism due to risk of co-variation with AF. All tests were two-sided, and a value of p<0.05 was regarded as significant. All analyses were performed using IBM SPSS statistics, version 24 software (IBM SPSS Statistics, IBM Corp, Somers, NY) or Stata versions 14 and 16 (Stata Corp., College Station, TX, USA).
Written informed consent was obtained from each patient during inclusion in STROKESTOP. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. The protocol was approved by the regional ethics committee (DNR 2011-13631-31, 2011/1363-31/3, 2014/1235-32 and 20181836-32).
Results
Our 6,100 participants had performed a mean of 29.0 (CI 22.5-35.5) ECG recordings each during the STROKESTOP study, resulting in a total of 175,074 ECGs. Baseline characteristics for participants in the different groups are presented in Table 1.
Table 1Baseline characteristics
Variable
All (n = 6100)
No arrhythmia (n = 5206)
Frequent SVECs (n = 709)
Regular SVT (+ p-waves) (n = 48)
Irregular SVT (+ p-waves) (n = 39)
Regular SVT (– p-waves) (n = 1)
Irregular SVT (– p-waves) (n = 97)
Women, n (%) Age (years) median (IQR) CHA2DS2-VASc, n median (IQR) Congestive heart failure, n (%) Hypertension, n (%) Diabetes Mellitus, n (%) Thromboembolism/stroke/TIA, n (%) Vascular disease, n (%)
Of participants, n = 5,206 had no significant arrhythmia, n = 709 had frequent SVECs (isolated, in bigeminy or trigemini), n = 48 had regular SVTs with p-waves, n = 39 had irregular SVTs with p-waves, n = 1 had a regular SVT without p-waves and n = 97 had irregular SVTs without p-waves. On average participants in the SVT groups had 1.0 (IQR 1.0-2.0) SVT each. The median SVT duration was 7.0 (IQR 6.0–12.0) beats.
During the median follow-up period of 4.2 (IQR 3.8-4.4) years, there were in total 387 new AF diagnoses, 161 stroke events and 354 deaths. Cumulative incidences and number of events per 1,000 person-years are presented in Table 2, showing a higher incidence of AF, stroke and death in participants with supraventricular activity than in the no arrhythmia group. The cumulative incidence of AF was highest in individuals with the most AF-like activity, micro-AF, 19.6%. The micro-AF group also had the highest cumulative incidence of stroke (4.1%) and death (10.3%).
Table 2Cumulative incidences and events per 1000 person-years
Variable
All (n = 6100)
No arrhythmia (n = 5206)
Frequent SVECs (n = 709)
Regular SVT (+ p-waves) (n = 48)
Irregular SVT (+ p-waves) (n = 39)
Irregular SVT (– P-waves) Micro-AF(n = 97)
AF cases, n (cumulative incidence, %) Incidence of AF per 1000 person-years (IQR) Stroke cases, n (cumulative incidence, %) Incidence of stroke, per 1000 person-years (IQR) Total mortality, n (cumulative incidence, %) Incidence of total mortality per 1000 person-years (IQR)
Participants with more organized atrial arrhythmias such as SVTs, especially those with micro-AF, were found to be more prone to develop AF compared to participants with frequent SVECs, Figure 3. The only individual with a regular SVT lacking p-waves did not have any events during follow-up and was excluded from further analysis.
Figure 3Kaplan-Meier curves showing primary end-point event rate (atrial fibrillation) in participants with different arrhythmias. a) Shows event rates for participants without arrhythmia, with frequent SVECs and SVTs. b) Shows event rates for participants without arrhythmia, with frequent SVECs independent of pattern, regular SVT with p-waves, irregular SVT with p-waves and irregular SVT without p-waves (micro-AF). c) Shows event rates for participants without arrhythmia, with frequent isolated SVECs, SVECs in trigeminy, SVECs in bigeminy, regular SVT with p-waves, irregular SVT with p-waves and irregular SVT without p-waves (micro-AF).
Presence of any SVT was associated with a hazard ratio of 3.1 (CI 2.1-4.6) for AF and frequent SVECs with a hazard ratio of 2.1 (CI 1.6-2.7), after adjustments for age, gender, hypertension, heart failure, vascular disease and diabetes mellitus, p < 0.001. We observed comparable associations with AF for all characterizations of SVECs, hence there was no significant difference in the risk of developing AF for frequent isolated SVECs compared to more organized SVECs such as bigeminy or trigeminy patterns (frequent isolated hazard ratio 2.1 (CI 1.4-3.0), trigeminy 2.4 (CI 1.5-3.9) and bigeminy 1.9 (CI 1.3-2.7), p < 0.001 for all comparisons).
There was no association found between presence of supraventricular activity and stroke, but micro-AF was associated with increased risk of all cause death, with a hazard ratio of 2.0 (1.1-3.8), p = 0.030. These results together with associations with AF for the different SVT classes are shown in Table 3.
Table 3Cox Regression Models for predicted risk of atrial fibrillation, stroke and death in relation to frequent SVECs (including isolated SVECs, trigeminy and bigeminy) and SVTs with different characteristics.
Variable
Frequent SVECs, HR (95% CI)
P
Regular SVT with p-waves, HR(95% CI)
P
Irregular SVT with p-waves, HR(95% CI)
P
Irregular SVT without p-waves, Micro-AF,HR (95% CI)
P
AF –crude AF – adjusted Stroke –crude Stroke – adjusted Death –crude Death –Adjusted
After adjusting for potential confounders (age, gender, hypertension, heart failure, vascular disease and diabetes mellitus), the hazard ratio for AF per SVT count was 1.0 (CI 0.9-1.0), p = 0.607 and duration of SVT per beat 1.0 (CI 0.9-1.1), p = 0.494. Although, the majority of participants had short SVTs and only one SVT registered.
The automated algorithm used for arrhythmia detection was evaluated with regards to micro-AF detection in the STROKESTOP II database. Per ECG it had a 91% sensitivity, 95% specificity and a positive predictive value of 3.1% for micro-AF. It also had a 92% sensitivity, 65% specificity and a positive predictive value of 8.5% per participant for micro-AF detection.
Discussion
In this large screening study for AF with more than 6,000 individuals free from AF and > 175,000 ECGs, we used an automated algorithm to show that frequent supraventricular arrhythmias affect 15 % of the elderly population. Progression of atrial arrhythmias from isolated SVECs to more AF-like activity “micro-AF” is associated with development of subsequent AF. In addition, a significant increase in all-cause death was seen in participants with micro-AF.
In a prior study of a similar cohort, we have seen that detection of short episodes of AF-like activity, termed micro-AF, is a sign of prevalent AF. By extended screening immediately after micro-AF identification AF was diagnosed in 13%.
In this study we aimed to see to what extent participants with micro-AF would have developed clinical AF (using registry diagnosis of AF as a proxy) three years after initial detection.
We found a similar proportion of patients diagnosed with clinical AF after three years as we did when extending ECG-recordings at time of diagnosis, indicating either that 1) no more patients developed AF, 2) it takes time for these patients to become symptomatic enough to seek healthcare or 3) under-diagnosis is common in this group. These patients may be detected earlier with a structured approach with extended long-term ECG monitoring.
To our knowledge there is only one smaller study comparing outcomes for individuals with SVTs with different characteristics. In a population of 377 AF free individuals with mean age 65 years, micro-AF was seen in 5% (n = 19/377). Micro-AF was found to be associated with the highest risk of AF, with a cumulative risk of AF of 47% during the 13-year follow-up period compared to 20.5% in participants with SVTs with regular p-waves. The cumulative incidence of stroke was 11% in participants with micro-AF but was surprisingly higher, 29%, in participants with regular SVTs without p-waves.
Although a trend towards increased stroke risk was seen in our participants with increased supraventricular activity, no statistically significant associations between supraventricular activity and stroke were found. Our study was likely not powered to detect a difference in ischaemic stroke during such short follow-up time.
The prevalence of micro-AF was 1.6% (97/6,100) in our study population from STROKESTOP compared to 6% (221/3,763) in the STROKESTOP II study, another mass-screening study for AF in the same age group using the same screening method.
The difference in prevalence could partly be explained by differences in the research protocols, where in the first STROKESTOP study AF could be diagnosed not only by full-filling the common diagnostic criteria (duration ≥ 30 seconds), but also in participants with two or more episodes of AF-activity lasting >10 seconds. Participants in STROKESTOP only did 26 intermittent ECG recordings on average compared to 49 recordings in the STROKESTOP II study. Participants in the STROKESTOP II study all had an increased NT-proBNP level (>125 ng/L) which might be related to an atrial myopathy that could explain the difference in prevalence. The algorithms used to detect suspected micro-AF episodes were identical in the two STROKESTOP studies.
The study has several limitations. Most study participants were Caucasians, and all were elderly. This may affect the external validity of the study. Individuals participating in screening are commonly healthier than the general population.
This might also apply to our study population, weakening rather than amplifying the statistical associations. Some of the participants may have had AF that was not detected during the screening process in STROKESTOP, which in case may have amplified the statistical associations. As the ECGs assessed in the study are 1-lead ECGs, p-wave analysis can be difficult in some cases, leading to misclassification bias. Hence some SVTs without p-waves might be misclassified. As lack of p-waves has been generally classified as more abnormal, this could lead to weakened associations. Some subclinical cases of both AF and stroke could also have been undiagnosed throughout the follow-up period and therefore not been included in registers. This might lead to a weakening of the associations.
Progression of atrial arrhythmias from SVECs to more AF-like arrhythmic episodes is associated with development of AF. Extended screening for AF should be considered in individuals with frequent supraventricular activity, especially in those with irregular SVTs lacking p-waves (micro-AF).
Credit Author
Tove Hygrell: Conceptualization, Methodology, Formal Analysis, Investigation, Data curation, Writing – Original Draft, Visualization, Funding acquisition. Martin Stridh: Methodology, Software, Resources, Writing – Review & Editing. Leif Friberg: Conceptualization, Formal analysis, Data curation, Writing – Review & Editing, Supervision.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Tove Hygrell reports financial support was provided by Stiftelsen Hjartat. Tove Hygrell reports financial support was provided by Boehringer Ingelheim. Tove Hygrell reports financial support was provided by The Swedish Heart and Lung Association. Tove Hygrell reports financial support was provided by Capio Forskningsstiftelse. Tove Hygrell reports a relationship with Pfizer that includes: speaking and lecture fees. Martin Stridh reports a relationship with Cardiolund AB that includes: equity or stocks. Leif Friberg reports a relationship with Bayer that includes: consulting or advisory. Leif Friberg reports a relationship with Boehringer-Ingelheim that includes: consulting or advisory. Leif Friberg reports a relationship with Sanofi that includes: consulting or advisory. Leif Friberg reports a relationship with BMS, Pfizer that includes: consulting or advisory. Emma Svennberg reports a relationship with Bayer that includes: speaking and lecture fees. Emma Svennberg reports a relationship with Bristol Myers Squibb, Pfizer that includes: speaking and lecture fees. Emma Svennberg reports a relationship with Boehringer-Ingelheim that includes: funding grants and speaking and lecture fees. Emma Svennberg reports a relationship with Merck Sharpe & Dohme that includes: speaking and lecture fees. Emma Svennberg reports a relationship with Sanofi that includes: speaking and lecture fees. Emma Svennberg reports a relationship with Roche Diagnostics that includes: funding grants. Emma Svennberg reports a relationship with Carl Bennet that includes: funding grants. Emma Svennberg reports financial support was provided by Ake Wibergs Stiftelse. Emma Svennberg reports a relationship with Stockholm County Council that includes: funding grants.
Disclosures
This study was supported by Stiftelsen Hjärtat; Boehringer Ingelheim; The Swedish Heart and Lung Association; Capio Forskningsstiftelse and Åke Wibergs Stiftelse. STROKESTOP was founded by Stockholm County Council; the Swedish Heart & Lung Foundation; King Gustav V and Queen Victoria's Freemasons’ Foundation; the Klebergska Foundation; the Tornspiran Foundation; the Scientific Council of Halland Region; the Southern Regional Healthcare Committee; the Swedish Stroke Fund; Boehringer-Ingelheim; Bayer and Pfizer. Emma Svennberg is supported by the Stockholm County Council (Clinical postdoctoral appointment) and has received research funding from the Swedish Society of Medicine. The sponsors had no involvement in the study other than financing.
2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS).
Safe automatic one-lead electrocardiogram analysis in screening for atrial fibrillation.
Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology.2017; 19: 1449-1453
00374-X&id=gr1.jpg){kind=link}
00374-X&id=gr2.jpg){kind=link}
00374-X&id=gr3.jpg){kind=link}
00374-X&id=gr4.jpg){kind=link}