Incidence of postoperative atrial fibrillation recurrence in patients discharged in sinus rhythm after cardiac surgery: a systematic review and meta-analysis

Incidence of postoperative atrial fibrillation recurrence in patients discharged in sinus rhythm... Abstract Postoperative atrial fibrillation (POAF) is associated with increased stroke risk and mortality post-discharge. POAF is often considered transient; however, recurrence is likely under-recognized as symptoms are an unreliable guide. Surveillance post-discharge may identify asymptomatic POAF recurrences in patients discharged in sinus rhythm. Therefore, we performed a systematic review and meta-analysis of studies investigating POAF recurrence post-discharge, in patients with new-onset POAF following cardiac surgery who reverted to sinus rhythm prior to discharge. Two independent reviewers searched medical databases, clinical trial registries, reference lists and the Internet. After screening from 6525 studies, 8 studies were identified (n = 1157 participants, mean age 66 ± 10 years and 73% men). Monitoring methods included the following: telemetry during twice-daily exercise sessions (n = 2), continuous telemetry for 3 weeks (n = 1), daily 20-s electrocardiography (ECG) using wearable event recorder (n = 1), 30-s single-lead ECG, 4 times/day (n = 1) and implanted continuous monitoring (n = 2). The incidence rate of POAF recurrence identified through non-invasive monitoring in the first 4 weeks post-discharge was 28.3% [confidence interval (CI) 23.0–33.6%]; recurring 12 ± 5 days (mean ± SD) post-surgery. The incidence rate identified through implanted continuous monitoring was 61–100% within 2 years. Between 40% and 93% of episodes were asymptomatic. In one small study reporting stroke risk, 8 of 10 patients with recurrence were guideline-indicated (CHA2DS2-VASc score ≥2) for oral anticoagulation for stroke prevention. Monitoring for POAF recurrence post-hospital discharge identifies significant numbers of early asymptomatic recurrences in patients at high risk of stroke who may benefit from anticoagulation for stroke prevention. More intense monitoring is more likely to identify POAF recurrence. Future research is required to investigate the prognostic significance of POAF recurrence, especially stroke and mortality risk. Postoperative, Atrial fibrillation, Screening, Recurrence, Electrocardiogram INTRODUCTION New-onset postoperative atrial fibrillation (POAF) occurs in 16–44% of patients following cardiac surgery prior to discharge [1, 2] and is associated with an increased risk of postoperative complications such as stroke, respiratory failure and pneumonia [3]. An episode of POAF after cardiac surgery is associated with a significantly increased long-term risk of embolic stroke [hazard ratio 2.1, 95% confidence interval (CI) 1.4–3.1] [2] and increased all-cause mortality, with an overall 10-year survival for patients with POAF being 65.5% compared to 75.3% in those without POAF [3]. POAF is often thought to be transient as it often resolves spontaneously, and there is also uncertainty about the need for treatment [4]. Therefore, patients who convert to sinus rhythm prior to discharge are rarely monitored for recurrence. However, these patients with POAF have a 5-fold increased risk of developing long-term atrial fibrillation (AF) [5]. In current practice, routine surveillance for POAF recurrence largely falls to the patient and their general practitioner and generally relies on the identification of AF symptoms. Symptoms occur in only 1 of 3 of all AF episodes [6], and many symptoms experienced such as fatigue, dizziness or shortness of breath are non-specific and easily confused with what might be expected during recovery from surgery. Notably, post-discharge from cardiac surgery, palpitations occur in only 30% of POAF recurrences [7]. Therefore, symptoms are unreliable as a guide to identify POAF recurrence. The need for surveillance of POAF recurrence is advocated by many recent studies [3, 5, 8–10]. It is likely that surveillance post-discharge would identify patients with asymptomatic recurrences of AF who are at increased risk of stroke; however, this is neither routinely implemented nor recommended in guidelines [11, 12]. Therefore, the aim of this systematic review was to focus on patients with new-onset POAF deemed to be in stable sinus rhythm on discharge and to determine the recurrence of POAF identified through active screening post-discharge. The secondary aim was to determine the stroke risk of the patients with recurrence of POAF after discharge. METHODS Search strategy Relevant studies were identified by a search of the medical databases MEDLINE (to February 2017), EMBASE (to February 2017) and Cochrane Library (to February 2017); clinical trials registries; reference lists and the Internet using Google search engine. The keyword search terms were atrial fibrillation AND (postoperative OR post-surgery OR post CABG OR post-surgical OR cardiac surgery OR heart surgery OR valve surgery) AND (detection OR screening OR identification OR incidence OR prevalence OR surveillance OR ECG OR Electrocardiogram OR monitoring OR recurrence OR case finding). Limits were applied for the years 1990 onwards and for studies on humans only. All languages were included. For Stage 1 of the review, 2 independent reviewers (G.M. and N.L.) reviewed all titles and abstracts from the search to identify potentially relevant articles. All relevant references were imported into Endnote X7 bibliographic software and duplicates removed. Where articles reported outcomes for the same study at different time points, the longest follow-up point was used. In Stage 2, full-text articles were obtained for all potentially relevant studies and screened by the reviewers (G.M. and N.L.) against the inclusion and exclusion criteria. Inclusion criteria included the following: cardiac surgery population, with episode of new-onset POAF during hospitalization for cardiac surgery; patients reverted to sinus rhythm prior to hospital discharge and patients were actively screened for AF recurrence post-hospital discharge (no limits placed on time frame). Exclusion criteria included the following: patients with known history of AF prior to surgery; studies did not distinguish any known history of AF prior to surgery; studies did not distinguish the patients who reverted to sinus rhythm prior to discharge and studies only screened in the post-surgical inpatient period. Outcome data were collected independently by two reviewers (G.M. and N.L.) on a predetermined data extraction form. In cases where data were unclear or not reported, the primary study authors were contacted. Quality of reporting and risk of bias in each study were assessed using the Newcastle–Ottawa scale [13]. Outcomes The primary outcome was recurrence of POAF following hospital discharge. Secondary outcomes were reported time frame to first recurrence, symptomatic status of AF recurrences, stroke risk scores (i.e. CHADS2 or CHA2DS2-VASc) [12] and eligibility and prescription of oral anticoagulation (OAC) therapy. Data analysis Meta analyses were performed on outcome data with the use of Comprehensive Meta-Analysis software version 3.3 (Biostat, Inc.). Incidence rates were calculated utilizing random-effects models, and heterogeneity tests were performed. RESULTS Study selection and characteristics A total of 6525 studies were screened against the inclusion and exclusion criteria and 56 full-text articles were reviewed (Fig. 1). Cohen’s kappa coefficient for inter-rater agreement was 0.47 (CI 0.28–0.66) for Stage 1 of the review process and 0.78 (CI 0.58–0.98) for Stage 2. Eight unique studies were ultimately included (Fig. 1), with a combined total of 1157 participants, a mean age of 66 ± 10 years and 73% men. Figure 1: View largeDownload slide Study selection. Figure 1: View largeDownload slide Study selection. The studies originated from Italy [14–17], USA [18, 19], Russia [20] and Australia [7] (Table 1). Study designs varied, including prospective randomized controlled trials [15, 20], prospective cohorts [7, 17–19] and retrospective cohorts [14, 16]. All studies recruited cardiac surgery patients with a transient episode of POAF, in sinus rhythm on discharge, with no history of AF and with stable sinus rhythm on admission. Participants were monitored for AF recurrence using a variation of 12-lead ECG, single-lead ECG, 24-h Holter monitor, telemetry, event recorders and implanted devices (Table 1). Although AF guidelines define AF as any episode lasting ≥30 s [12], AF definitions varied between each study (Table 1). The period of post-discharge monitoring ranged from 2 to 4 weeks, with 3 studies performing additional long-term follow-up of up to 24 months (Table 1). Table 1: Study characteristics Author, year  Country  Study design  Inclusion criteria  Exclusion criteria  Method of monitoring  Time period of monitoring  AF definition  Intermittent monitoring     Ambrosetti et al., [14]  Italy  Prospective registry study  Cardiac surgery, attending cardiac rehabilitation  Nil reported  Monitored during cardiac rehabilitation sessions Diagnostic method not standardized between centers; varied according to local protocol and access to continuous monitoring  13 ± 20 days post-discharge  Clinically documented   Cioffi et al., [15]  Italy  Randomized controlled trial  Cardiac surgery, referred to rehabilitation centre  Severe postoperative complications, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 2 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Cioffi et al., [16] (retrospective arm only)  Italy  Retrospective analysis  CABG or valve surgery, referred to rehabilitation centre  >1/52 in intensive care, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 4 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Funk et al., [18]  USA  Prospective longitudinal  CABG ± valve surgery  Postoperative stay >2/52, communication barrier  20-s ECG once daily, plus extra if symptoms detected Wearable cardiac event recorder (King of Hearts Express recorder, Instromedix)  2 weeks post-discharge  Not defined   Giaccardi et al., [17]  Italy  Prospective cohort  Isolated CABG, ≥65 years, attending cardiac rehabilitation  Cognitive or physical activity impairment, sleep apnoea, previous cardiothoracic surgery  Monitored 24 h/day for the duration of cardiac rehabilitation Electrocardiogram telemetry  3 weeks post-discharge  Standard ECG criteria   Lowres et al., [7]  Australia  Prospective cohort  Cardiac surgery  Communication barrier  30-s ECG 3–4 times/day, plus extra if symptoms detected Hand-held single-lead ECG (AliveCor Heart monitor)  29 ± 5 days post-discharge  30-s rhythm strip  Implanted cardiac monitoring     Lomivorotov et al., [20]  Russia  Randomized controlled study  CABG, >18 years  Unstable cardiac disease, pacemaker, antiarrhythmic or omega-3 treatment  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  24 months postoperatively  ≥30 s  Reviewed at 10 days; then 3, 6, 12 and 24 months   El-Chami et al., [19]  USA  Prospective cohort  CABG, CHADS2 score ≥1  Known AF history, valve surgery  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  25 ± 10 months monitoring  Documented AF requiring treatment  Reviewed monthly  Author, year  Country  Study design  Inclusion criteria  Exclusion criteria  Method of monitoring  Time period of monitoring  AF definition  Intermittent monitoring     Ambrosetti et al., [14]  Italy  Prospective registry study  Cardiac surgery, attending cardiac rehabilitation  Nil reported  Monitored during cardiac rehabilitation sessions Diagnostic method not standardized between centers; varied according to local protocol and access to continuous monitoring  13 ± 20 days post-discharge  Clinically documented   Cioffi et al., [15]  Italy  Randomized controlled trial  Cardiac surgery, referred to rehabilitation centre  Severe postoperative complications, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 2 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Cioffi et al., [16] (retrospective arm only)  Italy  Retrospective analysis  CABG or valve surgery, referred to rehabilitation centre  >1/52 in intensive care, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 4 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Funk et al., [18]  USA  Prospective longitudinal  CABG ± valve surgery  Postoperative stay >2/52, communication barrier  20-s ECG once daily, plus extra if symptoms detected Wearable cardiac event recorder (King of Hearts Express recorder, Instromedix)  2 weeks post-discharge  Not defined   Giaccardi et al., [17]  Italy  Prospective cohort  Isolated CABG, ≥65 years, attending cardiac rehabilitation  Cognitive or physical activity impairment, sleep apnoea, previous cardiothoracic surgery  Monitored 24 h/day for the duration of cardiac rehabilitation Electrocardiogram telemetry  3 weeks post-discharge  Standard ECG criteria   Lowres et al., [7]  Australia  Prospective cohort  Cardiac surgery  Communication barrier  30-s ECG 3–4 times/day, plus extra if symptoms detected Hand-held single-lead ECG (AliveCor Heart monitor)  29 ± 5 days post-discharge  30-s rhythm strip  Implanted cardiac monitoring     Lomivorotov et al., [20]  Russia  Randomized controlled study  CABG, >18 years  Unstable cardiac disease, pacemaker, antiarrhythmic or omega-3 treatment  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  24 months postoperatively  ≥30 s  Reviewed at 10 days; then 3, 6, 12 and 24 months   El-Chami et al., [19]  USA  Prospective cohort  CABG, CHADS2 score ≥1  Known AF history, valve surgery  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  25 ± 10 months monitoring  Documented AF requiring treatment  Reviewed monthly  AF: atrial fibrillation; CABG: coronary artery bypass graft; CHADS2: stroke risk score [C = congestive heart failure, H = hypertension, A = age ≥65, D = diabetes, S2 = prior stroke or TIA]; ECG: electrocardiogram. Quality assessment The overall quality of reporting was moderate. All studies scored 8 of 9 stars on the Newcastle–Ottawa Scale, where 9 stars is the best score [13]; however, against the STROBE reporting guidelines [21], 5 studies did not adequately report patient numbers at each stage, 6 studies did not report efforts to address potential study bias and 2 studies failed to address the limitations of the study (Supplementary Material, Table S1). Recurrence of postoperative atrial fibrillation With non-invasive monitoring in the first 2–4 weeks post-discharge, the incidence rate (cases per 100 persons screened) of POAF recurrence was 28.3% (95% CI 23.0–33.6%; 6 studies, n = 1125) [7, 14–18] according to random-effects analysis (Fig. 2). Heterogeneity between studies was low–moderate, with I2 = 30.0%, where a value of I2 <25% is considered homogeneous and an I2value of 50% indicates moderate heterogeneity [22]. When monitored with an implanted device over a 2-year period, a higher incidence of AF recurrence was identified. The 2 implanted device studies reported an incidence of 60.9% [19] and 100% [20]. Figure 2: View largeDownload slide Forest plot of incidence of recurrent postoperative atrial fibrillation using intermittent monitoring. CI: confidence interval. Figure 2: View largeDownload slide Forest plot of incidence of recurrent postoperative atrial fibrillation using intermittent monitoring. CI: confidence interval. Time frame to first atrial fibrillation recurrence Of the 6 non-invasive monitoring studies, time frame to first recurrence of POAF was reported in only 3 studies (n = 332): all 3 studies noted that AF recurred a mean of 12 days post-surgery (Table 2) [7, 15, 16]. One implanted device study reported that 39% (9/23) of participants had an AF recurrence within 3 months of surgery, but the mean time to first recurrence was 146 ± 250 days [19]. Table 2: Study outcomes Author, year  Age, mean ± SD  Male (%)  Number with POAF  POAF recurrence after discharge, n (%)  Time frame to recurrence  Symptom status of recurrences  Intermittent monitoring   Ambrosetti et al., [14]  67 ± 10  75  623  141 (22.6)       Cioffi et al., [15]  65 ± 9  68  185  60 (32.4)  12 ± 5 days post-surgery  53% asymptomatic  47% symptomatic   Cioffi et al., [16] (retrospective arm)  62 ± 10  72  93  35 (37.6)  12 ± 4 days post-surgery  48% asymptomatic  52% symptomatic and higher ventricular rate   Funk et al., [18]  64 ± 11  73  117  31 (26.5)    69% asymptomatic  17% palpitations  14% atypical symptoms   Giaccardi et al., [17]  74 ± 6  57  65  22 (33.8)       Lowres et al., [7]  69 ± 9  80  42  10 (23.8)  12.5 ± 6 days post-surgery  40% asymptomatic  30% palpitations  22% atypical  11% mixture palpitations ±  atypical  Implanted cardiac monitoring   Lomivorotov et al., [20]  62 ± 5  95  9  9 (100)       El-Chami et al., [19]  69 ± 7  87  23  14 (60.8) (9 found in  < 3 months)  146 ± 250 days  93% asymptomatic  7% symptomatic  Author, year  Age, mean ± SD  Male (%)  Number with POAF  POAF recurrence after discharge, n (%)  Time frame to recurrence  Symptom status of recurrences  Intermittent monitoring   Ambrosetti et al., [14]  67 ± 10  75  623  141 (22.6)       Cioffi et al., [15]  65 ± 9  68  185  60 (32.4)  12 ± 5 days post-surgery  53% asymptomatic  47% symptomatic   Cioffi et al., [16] (retrospective arm)  62 ± 10  72  93  35 (37.6)  12 ± 4 days post-surgery  48% asymptomatic  52% symptomatic and higher ventricular rate   Funk et al., [18]  64 ± 11  73  117  31 (26.5)    69% asymptomatic  17% palpitations  14% atypical symptoms   Giaccardi et al., [17]  74 ± 6  57  65  22 (33.8)       Lowres et al., [7]  69 ± 9  80  42  10 (23.8)  12.5 ± 6 days post-surgery  40% asymptomatic  30% palpitations  22% atypical  11% mixture palpitations ±  atypical  Implanted cardiac monitoring   Lomivorotov et al., [20]  62 ± 5  95  9  9 (100)       El-Chami et al., [19]  69 ± 7  87  23  14 (60.8) (9 found in  < 3 months)  146 ± 250 days  93% asymptomatic  7% symptomatic  POAF: postoperative atrial fibrillation; SD: standard deviation. Table 3: Medications and POAF after discharge Study  Medication/dose  POAF post-discharge, % (n)   P-value  On medication  Not taking medication  β-blocker   Ambrosetti et al., [14]  -  7%  8.9%  0.317   Cioffi et al., [15]  Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  10% (4/42)  47% (25/50)a  <0.0001   Cioffi et al., [16] (retrospective arm)  Atenolol 25mg/day, or Metoprolol 100mg/day, or Bisoprolol 5 mg/day  4.3% (5/116)  21.5% (56/260)  <0.0001   Funk et al., [18]  ‘Preoperatively’  10% (20/201)  21% (21/101)  0.01   Giaccardi et al., [17]  -  26.2% (22/84)  14.8% (11/74)  0.081  Amiodarone   Cioffi et al., [15]  250 ± 50mg/day  50% (29/62)  47% (25/50)a  0.85   Giaccardi et al., [17]  -  28.3% (15/53)  17.1% (18/105)  0.103  Amiodarone plus β-blocker   Cioffi et al., [15]  Amiodarone 250 ± 50mg/day; Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  9% (2/22)  47% (25/50)a  0.001  ACE-I/ARB   Ambrosetti et al., [14]  ACE-I/ARB  6.1%  9.3%  0.08   Cioffi et al., [16] (retrospective arm)  ACE-I  10.7% (3/28)  16.6% (58/348)  0.60   Giaccardi et al., [17]  ACE-I  20.7% (23/111)  21.3% (10/47)  0.937  Study  Medication/dose  POAF post-discharge, % (n)   P-value  On medication  Not taking medication  β-blocker   Ambrosetti et al., [14]  -  7%  8.9%  0.317   Cioffi et al., [15]  Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  10% (4/42)  47% (25/50)a  <0.0001   Cioffi et al., [16] (retrospective arm)  Atenolol 25mg/day, or Metoprolol 100mg/day, or Bisoprolol 5 mg/day  4.3% (5/116)  21.5% (56/260)  <0.0001   Funk et al., [18]  ‘Preoperatively’  10% (20/201)  21% (21/101)  0.01   Giaccardi et al., [17]  -  26.2% (22/84)  14.8% (11/74)  0.081  Amiodarone   Cioffi et al., [15]  250 ± 50mg/day  50% (29/62)  47% (25/50)a  0.85   Giaccardi et al., [17]  -  28.3% (15/53)  17.1% (18/105)  0.103  Amiodarone plus β-blocker   Cioffi et al., [15]  Amiodarone 250 ± 50mg/day; Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  9% (2/22)  47% (25/50)a  0.001  ACE-I/ARB   Ambrosetti et al., [14]  ACE-I/ARB  6.1%  9.3%  0.08   Cioffi et al., [16] (retrospective arm)  ACE-I  10.7% (3/28)  16.6% (58/348)  0.60   Giaccardi et al., [17]  ACE-I  20.7% (23/111)  21.3% (10/47)  0.937  a Control group ‘not’ taking either amiodarone or β-blocker. ACE-I: angiotensin-converting enzyme inhibitor; ARB: angiotensin receptor blocker. Symptomatic status of atrial fibrillation recurrences Five studies reported the symptomatic status of recurrent episodes of AF (Table 2). Recurrences were most commonly asymptomatic, ranging from 40% to 69% of episodes in the non-invasive monitoring studies [7, 15, 16, 18]. In the implanted device study reporting symptomatic status, 93% (13/14) of recurrences were asymptomatic [19]. Medications and atrial fibrillation post-discharge Medications were not reported specifically for the patients with a POAF recurrence who were discharged home in sinus rhythm, and the results were reported for the whole study population. The effect of β-blocker medications were reported in 5 studies. Three studies indicated that β-blockers may reduce the recurrence of AF post-discharge [15, 16, 18], and 2 studies reported no association between β-blockers and AF following discharge [14, 17] (Table 2). The effect of amiodarone medication was reported in only 2 small studies, both reporting no association with reducing AF post-discharge [15, 17] (Table 2). Meta-analysis was not possible due to insufficient reporting of data in the majority of studies, especially in relation to listing the medications and dose. Stroke risk and oral anticoagulation prescription Stroke risk for patients with AF recurrence and eligibility for prescription of OAC were reported in only 1 study (n = 42) [7]. In this study, 8 of the 10 patients with recurrent AF had a CHA2DS2-VASc score ≥2. However, only 3 of 8 patients were on OAC, with 1 of the 3 requiring warfarin because of a mechanical valve. Two studies reported that OAC was routinely prescribed for recurrences of AF that lasted longer than 24 and 48 h; however, these studies neither reported the numbers of patients in which this occurred nor reported their calculated stroke risk scores [15, 16]. Late-onset postoperative atrial fibrillation post-discharge Incidence of late-onset POAF post-hospital discharge was an additional finding reported in 3 studies (n = 2716) [14, 17, 18]. For people with no in-hospital episode of AF, intermittent monitoring in the first 2–4 weeks post-discharge also identified individuals with late-onset POAF post-hospital discharge: incidence rate (cases per 100 persons screened) 3.9% (95% CI, 2.7–5.4%) (Fig. 3). Heterogeneity for these studies was moderate, with I2 = 50.3%, as defined by an I2 value of 50% indicating moderate heterogeneity [22]. Figure 3: View largeDownload slide Forest plot of incidence of new-onset postoperative atrial fibrillation. CI: confidence interval. Figure 3: View largeDownload slide Forest plot of incidence of new-onset postoperative atrial fibrillation. CI: confidence interval. DISCUSSION Statement of findings Monitoring of patients discharged in stable sinus rhythm post-cardiac surgery identifies POAF recurrence in 28.3 cases per 100 persons screened in the first 2–4 weeks post-discharge using non-invasive techniques; implanted devices can identify recurrence in 60–100% of cases, suggesting that in-hospital episodes of new POAF are not necessarily isolated or transient. Importantly, most POAF recurrences are asymptomatic; therefore, these episodes are unlikely to be recognized without the aid of additional monitoring post-discharge. Furthermore, in the one study that calculated risk of stroke, 8 of 10 patients with POAF recurrence had a sufficient risk of stroke (i.e. CHA2DS2-VASc score ≥2) to be considered eligible for OAC. This highlights the potential importance of early identification of POAF recurrence through monitoring in the post-discharge period and appropriate treatment and prescription of OAC to prevent stroke. There are multiple hypotheses as to the mechanism for development of POAF following surgery patients. It has been reported that inflammation and the associated increase in plasma concentrations of C-reactive protein and interleukin 6 in conjunction with cardiac surgery may be associated with the development of AF [23, 24]. Other mechanisms include increased sympathoadrenal activation post-surgery, which prolongs the atrial refractive period and promotes ectopic activity, and oxidative stress following bypass surgery [24, 25]. However, it remains unknown as to why POAF develops in some patients and not in others. Pre-existing risk factors that can lead to the formation of an atrial AF substrate may also influence risk [24]. It is thought that POAF may occur post-surgery when patients reach a ‘pathophysiological threshold’ of pro-arrhythmic factors due to a combination of pre-existing risk factors and post-surgical factors [24, 26]. There is a natural incidence of AF over and above that related to the surgery and the postoperative period, and this is likely to factor in the yield of AF detected from long-term monitoring. But it is possible that the patients who develop POAF post-surgery may be those with some existing AF substrate changes, who are predisposed to developing AF both during the provocation of surgery and spontaneously in the long term. Risk scores have been developed to predict the development of new-onset POAF after cardiac surgery, such as the POAF score [27] and the Brazilian risk score [28]. Although validated for predicting new-onset POAF, these scores have not been extended to predicting recurrence of POAF but may be useful, as recent evidence suggests that new-onset AF is a predictor of long-term AF [5]. There is a 6-fold increase in the likelihood of clinical AF recurrence in patients with an episode of POAF compared to those without POAF; however, late-onset POAF can also occur post-hospital discharge [16]. Incidence of late-onset POAF was a secondary finding of the studies included in our review; however, other research has identified similar incidences of 3.6–8% of new AF post-discharge [10, 29]. It is also probable that many in-hospital episodes of POAF go undetected. A study using Holter monitoring found 13% of all patients had ‘silent’ POAF during the post-cardiac surgery inpatient period, which was asymptomatic and remained undiagnosed at discharge [9]. ‘Silent’ POAF was associated with higher AF recurrence rate at 1-year follow-up [9]. It is therefore difficult to predict which patients are at higher risk of AF recurrence or AF development in the post-discharge period, suggesting perhaps that monitoring might also be indicated for patients without diagnosed POAF during the admission. Identification of an appropriate means of monitoring is becoming increasingly important, given that the number of elderly patients undergoing cardiac surgery is projected to increase [30], and age is a primary risk factor for the development of both AF and stroke [31, 32]. Furthermore, obesity [33] and metabolic syndrome [34] are also associated with a higher incidence of POAF post-cardiac surgery. We are therefore likely to see an increase in the incidence of POAF corresponding to the rise in obesity in the next 20 years, along with an associated increase in stroke, morbidity, mortality and health system costs. The mode of surveillance will influence the number of cases of recurrent AF identified. Implanted devices with continual rhythm recording will pickup much more brief paroxysmal episodes than intermittent sampling, but the clinical significance of these episodes, such as cardiac implanted device-detected atrial high rate episodes may be less than clinically detected AF or AF detected by intermittent brief recordings [35, 36]. Although technology allowing centralized remote monitoring of multiple patients with implanted devices could make monitoring more feasible [37], the likely lower risk in patients detected in this way needs to be taken into consideration. Patient tolerance, acceptance and compliance of an implanted device is an additional consideration, with 50% refusal rate reported from 1 study designed to have an implanted device for 2 years [20]. Despite the clear evidence that POAF is a major cause of morbidity and mortality and that asymptomatic POAF recurrences can be identified in the weeks following discharge, there is a lack of evidence with regard to the long-term effect of anticoagulation on stroke risk in POAF patients and therefore no indication that treatment will be altered as a result of identifying POAF recurrence through monitoring [12]. POAF management guidelines recommend anticoagulation for at least 4 weeks post-cardioversion of patients with POAF lasting >48 h and anticoagulation for ‘high-risk patients’ in which sinus rhythm was restored within 24 h of onset [11]. The results of this study raise questions regarding which patients should be considered ‘high risk’ and which patients are at risk of recurrence or late-onset AF, thus which patients should monitored. Instigating monitoring comes at a cost. However, it could well be cost-effective from the standpoint of stroke prevention. Although the studies in this review utilized devices for monitoring, there are also other low-cost methods to address detection of AF through improved patient education regarding symptom awareness and routine self-monitoring using pulse palpation, and some hand-held ECG devices have a fairly modest cost. To date, screening studies for primary AF detection using new devices, such as hand-held single-lead ECGs, have all been demonstrated to be cost-effective [38–42]. Limitations The main limitation of this review is the small number of studies included and design of the studies. Participants were non-homogeneous with regard to age, cardiac procedure and comorbidities, and the numbers in each trial were small with <100 participants in 5 of 8 studies. Study designs varied significantly between the studies, with only 2 randomized controlled trials, and the method and time frame for monitoring varied greatly between the trials. Furthermore, the majority of trials did not report data for our secondary objectives. Only 5 studies reported symptomatic status of AF episodes, the data regarding reduced recurrence with β-blockers was not conclusive, only 2 small studies investigated the effect of amiodarone and only 1 study with small numbers reported the risk of stroke in patients identified with POAF recurrence. Therefore, there is insufficient strength in the available data to draw conclusions about the risk of stroke in this population, the need for anticoagulation treatment and the effect of preventative treatments such as amiodarone and β-blockers. This highlights a need for future studies to look at amiodarone and its role in preventing recurrence of POAF in this group of patients and for more understanding of how antiarrhythmic medications such as amiodarone can influence prevention of recurrence. Furthermore, although 3 of 5 studies suggested a beneficial effect of β-blockers in the prevention of AF post-discharge, it is noteworthy that most institutions prescribe β-blockers at the time of discharge as a process of continuing preoperative medications and not with the intention of offering antiarrhythmic protection. There is no evidence or uniformity in prescribing one particular β-blocker or any one specific dose to offer adequate antiarrhythmic protection. Further research is indicated to address these issues. Although we know that risk of stroke is increased in patients with an episode of POAF following cardiac surgery [2], and asymptomatic AF episodes in the general population increase stroke risk [36], and risk of stroke from paroxysmal AF is sufficiently high to require anticoagulation [43], this review highlights that there is insufficient evidence to determine the stroke risk of recurrences in patients with POAF who are discharged home and presumed to be in stable sinus rhythm. It is also uncertain whether the increased recurrence rate detected by continuous invasive recordings will have the same prognostic significance as recurrence detected by more brief or intermittent recordings. These data are required to inform recommendations for the length and intensity of monitoring required post-discharge and anticoagulation and preventative treatments for recurrences. Overall, the results of this review indicate a need for large-scale research to look at the prognostic significance of POAF recurrences and late-onset AF episodes post-discharge, in relation to stroke, morbidity and mortality, and therefore treatment implications and recommendations. CONCLUSION Monitoring for POAF recurrence post-hospital discharge identifies significant numbers of asymptomatic recurrences in patients at high risk of stroke, who may benefit from early detection of POAF recurrence and anticoagulation for stroke prevention. More intense monitoring is more likely to identify POAF recurrence. Further large-scale studies are required to investigate the prognostic significance of POAF recurrence in relation to risk of stroke and mortality. SUPPLEMENTARY MATERIAL Supplementary material is available at ICVTS online. Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Nicole Lowres is supported by a NSW Health Early Career Fellowship [H16/52168 to N.L.]. Conflict of interest: Ben Freedman reports grants, personal fees and non-financial support from Bayer Pharma AG outside the submitted work, grants and non-financial support from Boehringer Ingelheim outside the submitted work, grants and personal fees from BMS/Pfizer outside the submitted work, personal fees from Servier outside the submitted work, personal fees from Astra-Zeneca outside the submitted work and consulting fees from Gilead outside the submitted work. Lis Neubeck has received grants and honoraria from BMS/Pfizer outside the submitted work. All other authors have no competing interests to disclose. REFERENCES 1 Helgadottir S, Sigurdsson MI, Ingvarsdottir IL, Arnar DO, Gudbjartsson T. Atrial fibrillation following cardiac surgery: risk analysis and long-term survival. J Cardiothorac Surg  2012; 7: 87. Google Scholar CrossRef Search ADS PubMed  2 Gialdini G, Nearing K, Bhave PD, Bonuccelli U, Iadecola C, Healey JS et al.   Perioperative atrial fibrillation and the long-term risk of ischemic stroke. JAMA  2014; 312: 616– 22. Google Scholar CrossRef Search ADS PubMed  3 Phan K, Ha HS, Phan S, Medi C, Thomas SP, Yan TD. New-onset atrial fibrillation following coronary bypass surgery predicts long-term mortality: a systematic review and meta-analysis. Eur J Cardiothorac Surg  2015; 48: 817– 24. Google Scholar CrossRef Search ADS PubMed  4 Levy D, Kannel WB. Postoperative atrial fibrillation and mortality: do the risks merit changes in clinical practice? J Am Coll Cardiol  2004; 43: 749– 51. Google Scholar CrossRef Search ADS PubMed  5 Lee SH, Kang DR, Uhm JS, Shim J, Sung JH, Kim JY et al.   New-onset atrial fibrillation predicts long-term newly developed atrial fibrillation after coronary artery bypass graft. Am Heart J  2014; 167: 593– 600.e1. Google Scholar CrossRef Search ADS PubMed  6 Healey JS, Connolly SJ, Gold MR, Israel CW, Van Gelder IC, Capucci A et al.   Subclinical atrial fibrillation and the risk of stroke. N Engl J Med  2012; 366: 120– 9. Google Scholar CrossRef Search ADS PubMed  7 Lowres N, Mulcahy G, Gallagher R, Ben Freedman S, Marshman D, Kirkness A et al.   Self-monitoring for atrial fibrillation recurrence in the discharge period post-cardiac surgery using an iPhone electrocardiogram. Eur J Cardiothorac Surg  2016; 50: 44– 51. Google Scholar CrossRef Search ADS PubMed  8 Saxena A, Dinh DT, Smith JA, Shardey GC, Reid CM, Newcomb AE. Usefulness of postoperative atrial fibrillation as an independent predictor for worse early and late outcomes after isolated coronary artery bypass grafting (multicenter Australian study of 19, 497 patients). Am J Cardiol  2012; 109: 219– 25. Google Scholar CrossRef Search ADS PubMed  9 Guenancia C, Pujos C, Debomy F, Malapert G, Laurent G, Bouchot O. Incidence and predictors of new-onset silent atrial fibrillation after coronary artery bypass graft surgery. Biomed Res Int  2015; 2015: 703685. Google Scholar CrossRef Search ADS PubMed  10 Bidar E, Maesen B, Nieman F, Verheule S, Schotten U, Maessen JG. A prospective randomized controlled trial on the incidence and predictors of late-phase postoperative atrial fibrillation up to 30 days and the preventive value of biatrial pacing. Heart Rhythm  2014; 11: 1156– 62. Google Scholar CrossRef Search ADS PubMed  11 Frendl G, Sodickson AC, Chung MK, Waldo AL, Gersh BJ, Tisdale JE et al.   2014 AATS guidelines for the prevention and management of perioperative atrial fibrillation and flutter for thoracic surgical procedures. J Thorac Cardiovasc Surg  2014; 148: e153– 93. Google Scholar CrossRef Search ADS PubMed  12 Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B et al.   2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur J Cardiothorac Surg  2016; 50: e1– e88. Google Scholar CrossRef Search ADS PubMed  13 Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, et al.   The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/nosgen.pdf (3 July 2017, date last accessed). 14 Ambrosetti M, Tramarin R, Griffo R, De Feo S, Fattirolli F, Vestri A et al.   Late postoperative atrial fibrillation after cardiac surgery: a national survey within the cardiac rehabilitation setting. J Cardiovasc Med (Hagerstown)  2011; 12: 390– 5. Google Scholar CrossRef Search ADS PubMed  15 Cioffi G, Cemin C, Russo TE, Pellegrini A, Terrasi F, Ferrario G. Post-discharge recurrences of new-onset atrial fibrillation following cardiac surgery: impact of low-dose amiodarone and beta-blocker prophylaxis. Ital Heart J  2000; 1: 691– 7. Google Scholar PubMed  16 Cioffi G, Mureddu G, Cemin C, Russo TE, Pellegrini A, Terrasi F et al.   Characterization of post-discharge atrial fibrillation following open-heart surgery in uncomplicated patients referred to an early rehabilitation program. Ital Heart J  2001; 2: 519– 28. Google Scholar PubMed  17 Giaccardi M, Macchi C, Colella A, Polcaro P, Zipoli R, Cecchi F et al.   Postacute rehabilitation after coronary surgery: the effect of preoperative physical activity on the incidence of paroxysmal atrial fibrillation. Am J Phys Med Rehabil  2011; 90: 308– 15. Google Scholar CrossRef Search ADS PubMed  18 Funk M, Richards SB, Desjardins J, Bebon C, Wilcox H. Incidence, timing, symptoms, and risk factors for atrial fibrillation after cardiac surgery. Am J Crit Care  2003; 12: 424– 33; quiz 34–5. Google Scholar PubMed  19 El-Chami MF, Merchant FM, Smith P, Levy M, Nelms AG, Merlino J et al.   Management of new-onset postoperative atrial fibrillation utilizing insertable cardiac monitor technology to observe recurrence of AF (MONITOR-AF). Pacing Clin Electrophysiol  2016; 39: 1083– 9. Google Scholar CrossRef Search ADS PubMed  20 Lomivorotov VV, Efremov SM, Pokushalov EA, Romanov AB, Ponomarev DN, Cherniavsky AM et al.   Randomized trial of fish oil infusion to prevent atrial fibrillation after cardiac surgery: data from an implantable continuous cardiac monitor. J Cardiothorac Vasc Anesth  2014; 28: 1278– 84. Google Scholar CrossRef Search ADS PubMed  21 von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Br Med J  2007; 335: 806– 8. Google Scholar CrossRef Search ADS   22 Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med  2002; 21: 1539– 58. Google Scholar CrossRef Search ADS PubMed  23 Wu N, Xu B, Xiang Y, Wu L, Zhang Y, Ma X et al.   Association of inflammatory factors with occurrence and recurrence of atrial fibrillation: a meta-analysis. Int J Cardiol  2013; 169: 62– 72. Google Scholar CrossRef Search ADS PubMed  24 Maesen B, Nijs J, Maessen J, Allessie M, Schotten U. Post-operative atrial fibrillation: a maze of mechanisms. Europace  2012; 14: 159– 74. Google Scholar CrossRef Search ADS PubMed  25 Echahidi N, Pibarot P, O’Hara G, Mathieu P. Mechanisms, prevention, and treatment of atrial fibrillation after cardiac surgery. J Am Coll Cardiol  2008; 51: 793– 801. Google Scholar CrossRef Search ADS PubMed  26 Schotten U, Verheule S, Kirchhof P, Goette A. Pathophysiological mechanisms of atrial fibrillation: a translational appraisal. Physiol Rev  2011; 91: 265– 325. Google Scholar CrossRef Search ADS PubMed  27 Mariscalco G, Biancari F, Zanobini M, Cottini M, Piffaretti G, Saccocci M et al.   Bedside tool for predicting the risk of postoperative atrial fibrillation after cardiac surgery: the POAF score. J Am Heart Assoc  2014; 3: e000752. Google Scholar CrossRef Search ADS PubMed  28 Sa MP, Sa MV, Albuquerque AC, Silva BB, Siqueira JW, Brito PR et al.   Predicting risk of atrial fibrillation after heart valve surgery: evaluation of a Brazilian risk score. Rev Bras Cir Cardiovasc  2012; 27: 117– 22. Google Scholar CrossRef Search ADS PubMed  29 Ahlsson A, Fengsrud E, Bodin L, Englund A. Postoperative atrial fibrillation in patients undergoing aortocoronary bypass surgery carries an eightfold risk of future atrial fibrillation and a doubled cardiovascular mortality. Eur J Cardiothorac Surg  2010; 37: 1353– 9. Google Scholar CrossRef Search ADS PubMed  30 Lee JJ, Park NH, Lee KS, Chee HK, Sim SB, Kim MJ et al.   Projections of demand for cardiovascular surgery and supply of surgeons. Korean J Thorac Cardiovasc Surg  2016; 49: S37– 43. Google Scholar CrossRef Search ADS PubMed  31 Mathew JP, Fontes ML, Tudor IC, Ramsay J, Duke P, Mazer CD et al.   A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA  2004; 291: 1720– 9. Google Scholar CrossRef Search ADS PubMed  32 Chyou JY, Hunter TD, Mollenkopf SA, Turakhia MP, Reynolds MR. Individual and combined risk factors for incident atrial fibrillation and incident stroke: an analysis of 3 million at-risk US patients. J Am Heart Assoc  2015; 4: e001723. Google Scholar CrossRef Search ADS PubMed  33 Hernandez AV, Kaw R, Pasupuleti V, Bina P, Ioannidis JP, Bueno H et al.   Association between obesity and postoperative atrial fibrillation in patients undergoing cardiac operations: a systematic review and meta-analysis. Ann Thorac Surg  2013; 96: 1104– 16. Google Scholar CrossRef Search ADS PubMed  34 Ivanovic B, Tadic M, Bradic Z, Zivkovic N, Stanisavljevic D, Celic V. The influence of the metabolic syndrome on atrial fibrillation occurrence and outcome after coronary bypass surgery: a 3-year follow-up study. Thorac Cardiovasc Surg  2014; 62: 561– 8. Google Scholar CrossRef Search ADS PubMed  35 Van Gelder IC, Healey JS, Crijns HJ, Wang J, Hohnloser SH, Gold MR et al.   Duration of device-detected subclinical atrial fibrillation and occurrence of stroke in ASSERT. Eur Heart J  2017; doi:10.1093/eurheartj/ehx042. 36 Freedman B, Camm J, Calkins H, Healey JS, Rosenqvist M, Wang J et al.   Screening for atrial fibrillation: a report of the AF-SCREEN international collaboration. Circulation  2017; 135: 1851– 67. Google Scholar CrossRef Search ADS PubMed  37 Zoppo F, Facchin D, Molon G, Zanotto G, Catanzariti D, Rossillo A et al.   Improving atrial fibrillation detection in patients with implantable cardiac devices by means of a remote monitoring and management application. Pacing Clin Electrophysiol  2014; 37: 1610– 8. Google Scholar CrossRef Search ADS PubMed  38 Lowres N, Neubeck L, Salkeld G, Krass I, McLachlan AJ, Redfern J et al.   Feasibility and cost effectiveness of stroke prevention through community screening for atrial fibrillation using iPhone ECG in pharmacies. The SEARCH-AF study. Thromb Haemost  2014; 111: 1167– 76. Google Scholar CrossRef Search ADS PubMed  39 Aronsson M, Svennberg E, Rosenqvist M, Engdahl J, Al-Khalili F, Friberg L et al.   Cost-effectiveness of mass screening for untreated atrial fibrillation using intermittent ECG recording. Europace  2015; 17: 1023– 9. Google Scholar CrossRef Search ADS PubMed  40 Jacobs MS, Kaasenbrood F, Postma MJ, van Hulst M, Tieleman RG. Cost-effectiveness of screening for atrial fibrillation in primary care with a handheld, single-lead electrocardiogram device in the Netherlands. Europace  2016. doi:10.1093/europace/euw285. 41 Moran PS, Teljeur C, Harrington P, Smith SM, Smyth B, Harbison J et al.   Cost-effectiveness of a national opportunistic screening program for atrial fibrillation in Ireland. Value Health  2016; 19: 985– 95. Google Scholar CrossRef Search ADS PubMed  42 Hobbs FD, Fitzmaurice DA, Mant J, Murray E, Jowett S, Bryan S et al.   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 Technol Assess  2005; 9:iii–v, ix–x, 1– 74. Google Scholar CrossRef Search ADS   43 Ganesan AN, Chew DP, Hartshorne T, Selvanayagam JB, Aylward PE, Sanders P et al.   The impact of atrial fibrillation type on the risk of thromboembolism, mortality, and bleeding: a systematic review and meta-analysis. Eur Heart J  2016; 37: 1591– 602. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Interactive CardioVascular and Thoracic Surgery Oxford University Press

Incidence of postoperative atrial fibrillation recurrence in patients discharged in sinus rhythm after cardiac surgery: a systematic review and meta-analysis

Loading next page...
 
/lp/ou_press/incidence-of-postoperative-atrial-fibrillation-recurrence-in-patients-7icFTnOb0O
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1569-9293
eISSN
1569-9285
D.O.I.
10.1093/icvts/ivx348
Publisher site
See Article on Publisher Site

Abstract

Abstract Postoperative atrial fibrillation (POAF) is associated with increased stroke risk and mortality post-discharge. POAF is often considered transient; however, recurrence is likely under-recognized as symptoms are an unreliable guide. Surveillance post-discharge may identify asymptomatic POAF recurrences in patients discharged in sinus rhythm. Therefore, we performed a systematic review and meta-analysis of studies investigating POAF recurrence post-discharge, in patients with new-onset POAF following cardiac surgery who reverted to sinus rhythm prior to discharge. Two independent reviewers searched medical databases, clinical trial registries, reference lists and the Internet. After screening from 6525 studies, 8 studies were identified (n = 1157 participants, mean age 66 ± 10 years and 73% men). Monitoring methods included the following: telemetry during twice-daily exercise sessions (n = 2), continuous telemetry for 3 weeks (n = 1), daily 20-s electrocardiography (ECG) using wearable event recorder (n = 1), 30-s single-lead ECG, 4 times/day (n = 1) and implanted continuous monitoring (n = 2). The incidence rate of POAF recurrence identified through non-invasive monitoring in the first 4 weeks post-discharge was 28.3% [confidence interval (CI) 23.0–33.6%]; recurring 12 ± 5 days (mean ± SD) post-surgery. The incidence rate identified through implanted continuous monitoring was 61–100% within 2 years. Between 40% and 93% of episodes were asymptomatic. In one small study reporting stroke risk, 8 of 10 patients with recurrence were guideline-indicated (CHA2DS2-VASc score ≥2) for oral anticoagulation for stroke prevention. Monitoring for POAF recurrence post-hospital discharge identifies significant numbers of early asymptomatic recurrences in patients at high risk of stroke who may benefit from anticoagulation for stroke prevention. More intense monitoring is more likely to identify POAF recurrence. Future research is required to investigate the prognostic significance of POAF recurrence, especially stroke and mortality risk. Postoperative, Atrial fibrillation, Screening, Recurrence, Electrocardiogram INTRODUCTION New-onset postoperative atrial fibrillation (POAF) occurs in 16–44% of patients following cardiac surgery prior to discharge [1, 2] and is associated with an increased risk of postoperative complications such as stroke, respiratory failure and pneumonia [3]. An episode of POAF after cardiac surgery is associated with a significantly increased long-term risk of embolic stroke [hazard ratio 2.1, 95% confidence interval (CI) 1.4–3.1] [2] and increased all-cause mortality, with an overall 10-year survival for patients with POAF being 65.5% compared to 75.3% in those without POAF [3]. POAF is often thought to be transient as it often resolves spontaneously, and there is also uncertainty about the need for treatment [4]. Therefore, patients who convert to sinus rhythm prior to discharge are rarely monitored for recurrence. However, these patients with POAF have a 5-fold increased risk of developing long-term atrial fibrillation (AF) [5]. In current practice, routine surveillance for POAF recurrence largely falls to the patient and their general practitioner and generally relies on the identification of AF symptoms. Symptoms occur in only 1 of 3 of all AF episodes [6], and many symptoms experienced such as fatigue, dizziness or shortness of breath are non-specific and easily confused with what might be expected during recovery from surgery. Notably, post-discharge from cardiac surgery, palpitations occur in only 30% of POAF recurrences [7]. Therefore, symptoms are unreliable as a guide to identify POAF recurrence. The need for surveillance of POAF recurrence is advocated by many recent studies [3, 5, 8–10]. It is likely that surveillance post-discharge would identify patients with asymptomatic recurrences of AF who are at increased risk of stroke; however, this is neither routinely implemented nor recommended in guidelines [11, 12]. Therefore, the aim of this systematic review was to focus on patients with new-onset POAF deemed to be in stable sinus rhythm on discharge and to determine the recurrence of POAF identified through active screening post-discharge. The secondary aim was to determine the stroke risk of the patients with recurrence of POAF after discharge. METHODS Search strategy Relevant studies were identified by a search of the medical databases MEDLINE (to February 2017), EMBASE (to February 2017) and Cochrane Library (to February 2017); clinical trials registries; reference lists and the Internet using Google search engine. The keyword search terms were atrial fibrillation AND (postoperative OR post-surgery OR post CABG OR post-surgical OR cardiac surgery OR heart surgery OR valve surgery) AND (detection OR screening OR identification OR incidence OR prevalence OR surveillance OR ECG OR Electrocardiogram OR monitoring OR recurrence OR case finding). Limits were applied for the years 1990 onwards and for studies on humans only. All languages were included. For Stage 1 of the review, 2 independent reviewers (G.M. and N.L.) reviewed all titles and abstracts from the search to identify potentially relevant articles. All relevant references were imported into Endnote X7 bibliographic software and duplicates removed. Where articles reported outcomes for the same study at different time points, the longest follow-up point was used. In Stage 2, full-text articles were obtained for all potentially relevant studies and screened by the reviewers (G.M. and N.L.) against the inclusion and exclusion criteria. Inclusion criteria included the following: cardiac surgery population, with episode of new-onset POAF during hospitalization for cardiac surgery; patients reverted to sinus rhythm prior to hospital discharge and patients were actively screened for AF recurrence post-hospital discharge (no limits placed on time frame). Exclusion criteria included the following: patients with known history of AF prior to surgery; studies did not distinguish any known history of AF prior to surgery; studies did not distinguish the patients who reverted to sinus rhythm prior to discharge and studies only screened in the post-surgical inpatient period. Outcome data were collected independently by two reviewers (G.M. and N.L.) on a predetermined data extraction form. In cases where data were unclear or not reported, the primary study authors were contacted. Quality of reporting and risk of bias in each study were assessed using the Newcastle–Ottawa scale [13]. Outcomes The primary outcome was recurrence of POAF following hospital discharge. Secondary outcomes were reported time frame to first recurrence, symptomatic status of AF recurrences, stroke risk scores (i.e. CHADS2 or CHA2DS2-VASc) [12] and eligibility and prescription of oral anticoagulation (OAC) therapy. Data analysis Meta analyses were performed on outcome data with the use of Comprehensive Meta-Analysis software version 3.3 (Biostat, Inc.). Incidence rates were calculated utilizing random-effects models, and heterogeneity tests were performed. RESULTS Study selection and characteristics A total of 6525 studies were screened against the inclusion and exclusion criteria and 56 full-text articles were reviewed (Fig. 1). Cohen’s kappa coefficient for inter-rater agreement was 0.47 (CI 0.28–0.66) for Stage 1 of the review process and 0.78 (CI 0.58–0.98) for Stage 2. Eight unique studies were ultimately included (Fig. 1), with a combined total of 1157 participants, a mean age of 66 ± 10 years and 73% men. Figure 1: View largeDownload slide Study selection. Figure 1: View largeDownload slide Study selection. The studies originated from Italy [14–17], USA [18, 19], Russia [20] and Australia [7] (Table 1). Study designs varied, including prospective randomized controlled trials [15, 20], prospective cohorts [7, 17–19] and retrospective cohorts [14, 16]. All studies recruited cardiac surgery patients with a transient episode of POAF, in sinus rhythm on discharge, with no history of AF and with stable sinus rhythm on admission. Participants were monitored for AF recurrence using a variation of 12-lead ECG, single-lead ECG, 24-h Holter monitor, telemetry, event recorders and implanted devices (Table 1). Although AF guidelines define AF as any episode lasting ≥30 s [12], AF definitions varied between each study (Table 1). The period of post-discharge monitoring ranged from 2 to 4 weeks, with 3 studies performing additional long-term follow-up of up to 24 months (Table 1). Table 1: Study characteristics Author, year  Country  Study design  Inclusion criteria  Exclusion criteria  Method of monitoring  Time period of monitoring  AF definition  Intermittent monitoring     Ambrosetti et al., [14]  Italy  Prospective registry study  Cardiac surgery, attending cardiac rehabilitation  Nil reported  Monitored during cardiac rehabilitation sessions Diagnostic method not standardized between centers; varied according to local protocol and access to continuous monitoring  13 ± 20 days post-discharge  Clinically documented   Cioffi et al., [15]  Italy  Randomized controlled trial  Cardiac surgery, referred to rehabilitation centre  Severe postoperative complications, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 2 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Cioffi et al., [16] (retrospective arm only)  Italy  Retrospective analysis  CABG or valve surgery, referred to rehabilitation centre  >1/52 in intensive care, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 4 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Funk et al., [18]  USA  Prospective longitudinal  CABG ± valve surgery  Postoperative stay >2/52, communication barrier  20-s ECG once daily, plus extra if symptoms detected Wearable cardiac event recorder (King of Hearts Express recorder, Instromedix)  2 weeks post-discharge  Not defined   Giaccardi et al., [17]  Italy  Prospective cohort  Isolated CABG, ≥65 years, attending cardiac rehabilitation  Cognitive or physical activity impairment, sleep apnoea, previous cardiothoracic surgery  Monitored 24 h/day for the duration of cardiac rehabilitation Electrocardiogram telemetry  3 weeks post-discharge  Standard ECG criteria   Lowres et al., [7]  Australia  Prospective cohort  Cardiac surgery  Communication barrier  30-s ECG 3–4 times/day, plus extra if symptoms detected Hand-held single-lead ECG (AliveCor Heart monitor)  29 ± 5 days post-discharge  30-s rhythm strip  Implanted cardiac monitoring     Lomivorotov et al., [20]  Russia  Randomized controlled study  CABG, >18 years  Unstable cardiac disease, pacemaker, antiarrhythmic or omega-3 treatment  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  24 months postoperatively  ≥30 s  Reviewed at 10 days; then 3, 6, 12 and 24 months   El-Chami et al., [19]  USA  Prospective cohort  CABG, CHADS2 score ≥1  Known AF history, valve surgery  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  25 ± 10 months monitoring  Documented AF requiring treatment  Reviewed monthly  Author, year  Country  Study design  Inclusion criteria  Exclusion criteria  Method of monitoring  Time period of monitoring  AF definition  Intermittent monitoring     Ambrosetti et al., [14]  Italy  Prospective registry study  Cardiac surgery, attending cardiac rehabilitation  Nil reported  Monitored during cardiac rehabilitation sessions Diagnostic method not standardized between centers; varied according to local protocol and access to continuous monitoring  13 ± 20 days post-discharge  Clinically documented   Cioffi et al., [15]  Italy  Randomized controlled trial  Cardiac surgery, referred to rehabilitation centre  Severe postoperative complications, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 2 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Cioffi et al., [16] (retrospective arm only)  Italy  Retrospective analysis  CABG or valve surgery, referred to rehabilitation centre  >1/52 in intensive care, amiodarone treatment  Monitored during cardiac rehabilitation sessions One lead telemetry: during twice-daily exercise sessions 12-lead ECG: weekly 24-h Holter: 14 days and 10 months post-surgery  21 ± 4 days post-discharge  Documented AF episodes lasting >2 mins  Plus long-term review at 10 months   Funk et al., [18]  USA  Prospective longitudinal  CABG ± valve surgery  Postoperative stay >2/52, communication barrier  20-s ECG once daily, plus extra if symptoms detected Wearable cardiac event recorder (King of Hearts Express recorder, Instromedix)  2 weeks post-discharge  Not defined   Giaccardi et al., [17]  Italy  Prospective cohort  Isolated CABG, ≥65 years, attending cardiac rehabilitation  Cognitive or physical activity impairment, sleep apnoea, previous cardiothoracic surgery  Monitored 24 h/day for the duration of cardiac rehabilitation Electrocardiogram telemetry  3 weeks post-discharge  Standard ECG criteria   Lowres et al., [7]  Australia  Prospective cohort  Cardiac surgery  Communication barrier  30-s ECG 3–4 times/day, plus extra if symptoms detected Hand-held single-lead ECG (AliveCor Heart monitor)  29 ± 5 days post-discharge  30-s rhythm strip  Implanted cardiac monitoring     Lomivorotov et al., [20]  Russia  Randomized controlled study  CABG, >18 years  Unstable cardiac disease, pacemaker, antiarrhythmic or omega-3 treatment  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  24 months postoperatively  ≥30 s  Reviewed at 10 days; then 3, 6, 12 and 24 months   El-Chami et al., [19]  USA  Prospective cohort  CABG, CHADS2 score ≥1  Known AF history, valve surgery  Implanted continuous cardiac monitor (Reveal XT, Medtronic)  25 ± 10 months monitoring  Documented AF requiring treatment  Reviewed monthly  AF: atrial fibrillation; CABG: coronary artery bypass graft; CHADS2: stroke risk score [C = congestive heart failure, H = hypertension, A = age ≥65, D = diabetes, S2 = prior stroke or TIA]; ECG: electrocardiogram. Quality assessment The overall quality of reporting was moderate. All studies scored 8 of 9 stars on the Newcastle–Ottawa Scale, where 9 stars is the best score [13]; however, against the STROBE reporting guidelines [21], 5 studies did not adequately report patient numbers at each stage, 6 studies did not report efforts to address potential study bias and 2 studies failed to address the limitations of the study (Supplementary Material, Table S1). Recurrence of postoperative atrial fibrillation With non-invasive monitoring in the first 2–4 weeks post-discharge, the incidence rate (cases per 100 persons screened) of POAF recurrence was 28.3% (95% CI 23.0–33.6%; 6 studies, n = 1125) [7, 14–18] according to random-effects analysis (Fig. 2). Heterogeneity between studies was low–moderate, with I2 = 30.0%, where a value of I2 <25% is considered homogeneous and an I2value of 50% indicates moderate heterogeneity [22]. When monitored with an implanted device over a 2-year period, a higher incidence of AF recurrence was identified. The 2 implanted device studies reported an incidence of 60.9% [19] and 100% [20]. Figure 2: View largeDownload slide Forest plot of incidence of recurrent postoperative atrial fibrillation using intermittent monitoring. CI: confidence interval. Figure 2: View largeDownload slide Forest plot of incidence of recurrent postoperative atrial fibrillation using intermittent monitoring. CI: confidence interval. Time frame to first atrial fibrillation recurrence Of the 6 non-invasive monitoring studies, time frame to first recurrence of POAF was reported in only 3 studies (n = 332): all 3 studies noted that AF recurred a mean of 12 days post-surgery (Table 2) [7, 15, 16]. One implanted device study reported that 39% (9/23) of participants had an AF recurrence within 3 months of surgery, but the mean time to first recurrence was 146 ± 250 days [19]. Table 2: Study outcomes Author, year  Age, mean ± SD  Male (%)  Number with POAF  POAF recurrence after discharge, n (%)  Time frame to recurrence  Symptom status of recurrences  Intermittent monitoring   Ambrosetti et al., [14]  67 ± 10  75  623  141 (22.6)       Cioffi et al., [15]  65 ± 9  68  185  60 (32.4)  12 ± 5 days post-surgery  53% asymptomatic  47% symptomatic   Cioffi et al., [16] (retrospective arm)  62 ± 10  72  93  35 (37.6)  12 ± 4 days post-surgery  48% asymptomatic  52% symptomatic and higher ventricular rate   Funk et al., [18]  64 ± 11  73  117  31 (26.5)    69% asymptomatic  17% palpitations  14% atypical symptoms   Giaccardi et al., [17]  74 ± 6  57  65  22 (33.8)       Lowres et al., [7]  69 ± 9  80  42  10 (23.8)  12.5 ± 6 days post-surgery  40% asymptomatic  30% palpitations  22% atypical  11% mixture palpitations ±  atypical  Implanted cardiac monitoring   Lomivorotov et al., [20]  62 ± 5  95  9  9 (100)       El-Chami et al., [19]  69 ± 7  87  23  14 (60.8) (9 found in  < 3 months)  146 ± 250 days  93% asymptomatic  7% symptomatic  Author, year  Age, mean ± SD  Male (%)  Number with POAF  POAF recurrence after discharge, n (%)  Time frame to recurrence  Symptom status of recurrences  Intermittent monitoring   Ambrosetti et al., [14]  67 ± 10  75  623  141 (22.6)       Cioffi et al., [15]  65 ± 9  68  185  60 (32.4)  12 ± 5 days post-surgery  53% asymptomatic  47% symptomatic   Cioffi et al., [16] (retrospective arm)  62 ± 10  72  93  35 (37.6)  12 ± 4 days post-surgery  48% asymptomatic  52% symptomatic and higher ventricular rate   Funk et al., [18]  64 ± 11  73  117  31 (26.5)    69% asymptomatic  17% palpitations  14% atypical symptoms   Giaccardi et al., [17]  74 ± 6  57  65  22 (33.8)       Lowres et al., [7]  69 ± 9  80  42  10 (23.8)  12.5 ± 6 days post-surgery  40% asymptomatic  30% palpitations  22% atypical  11% mixture palpitations ±  atypical  Implanted cardiac monitoring   Lomivorotov et al., [20]  62 ± 5  95  9  9 (100)       El-Chami et al., [19]  69 ± 7  87  23  14 (60.8) (9 found in  < 3 months)  146 ± 250 days  93% asymptomatic  7% symptomatic  POAF: postoperative atrial fibrillation; SD: standard deviation. Table 3: Medications and POAF after discharge Study  Medication/dose  POAF post-discharge, % (n)   P-value  On medication  Not taking medication  β-blocker   Ambrosetti et al., [14]  -  7%  8.9%  0.317   Cioffi et al., [15]  Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  10% (4/42)  47% (25/50)a  <0.0001   Cioffi et al., [16] (retrospective arm)  Atenolol 25mg/day, or Metoprolol 100mg/day, or Bisoprolol 5 mg/day  4.3% (5/116)  21.5% (56/260)  <0.0001   Funk et al., [18]  ‘Preoperatively’  10% (20/201)  21% (21/101)  0.01   Giaccardi et al., [17]  -  26.2% (22/84)  14.8% (11/74)  0.081  Amiodarone   Cioffi et al., [15]  250 ± 50mg/day  50% (29/62)  47% (25/50)a  0.85   Giaccardi et al., [17]  -  28.3% (15/53)  17.1% (18/105)  0.103  Amiodarone plus β-blocker   Cioffi et al., [15]  Amiodarone 250 ± 50mg/day; Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  9% (2/22)  47% (25/50)a  0.001  ACE-I/ARB   Ambrosetti et al., [14]  ACE-I/ARB  6.1%  9.3%  0.08   Cioffi et al., [16] (retrospective arm)  ACE-I  10.7% (3/28)  16.6% (58/348)  0.60   Giaccardi et al., [17]  ACE-I  20.7% (23/111)  21.3% (10/47)  0.937  Study  Medication/dose  POAF post-discharge, % (n)   P-value  On medication  Not taking medication  β-blocker   Ambrosetti et al., [14]  -  7%  8.9%  0.317   Cioffi et al., [15]  Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  10% (4/42)  47% (25/50)a  <0.0001   Cioffi et al., [16] (retrospective arm)  Atenolol 25mg/day, or Metoprolol 100mg/day, or Bisoprolol 5 mg/day  4.3% (5/116)  21.5% (56/260)  <0.0001   Funk et al., [18]  ‘Preoperatively’  10% (20/201)  21% (21/101)  0.01   Giaccardi et al., [17]  -  26.2% (22/84)  14.8% (11/74)  0.081  Amiodarone   Cioffi et al., [15]  250 ± 50mg/day  50% (29/62)  47% (25/50)a  0.85   Giaccardi et al., [17]  -  28.3% (15/53)  17.1% (18/105)  0.103  Amiodarone plus β-blocker   Cioffi et al., [15]  Amiodarone 250 ± 50mg/day; Atenolol 40 ± 15mg/day, or Metoprolol 50mg/day, or Bisoprolol 5 mg/day  9% (2/22)  47% (25/50)a  0.001  ACE-I/ARB   Ambrosetti et al., [14]  ACE-I/ARB  6.1%  9.3%  0.08   Cioffi et al., [16] (retrospective arm)  ACE-I  10.7% (3/28)  16.6% (58/348)  0.60   Giaccardi et al., [17]  ACE-I  20.7% (23/111)  21.3% (10/47)  0.937  a Control group ‘not’ taking either amiodarone or β-blocker. ACE-I: angiotensin-converting enzyme inhibitor; ARB: angiotensin receptor blocker. Symptomatic status of atrial fibrillation recurrences Five studies reported the symptomatic status of recurrent episodes of AF (Table 2). Recurrences were most commonly asymptomatic, ranging from 40% to 69% of episodes in the non-invasive monitoring studies [7, 15, 16, 18]. In the implanted device study reporting symptomatic status, 93% (13/14) of recurrences were asymptomatic [19]. Medications and atrial fibrillation post-discharge Medications were not reported specifically for the patients with a POAF recurrence who were discharged home in sinus rhythm, and the results were reported for the whole study population. The effect of β-blocker medications were reported in 5 studies. Three studies indicated that β-blockers may reduce the recurrence of AF post-discharge [15, 16, 18], and 2 studies reported no association between β-blockers and AF following discharge [14, 17] (Table 2). The effect of amiodarone medication was reported in only 2 small studies, both reporting no association with reducing AF post-discharge [15, 17] (Table 2). Meta-analysis was not possible due to insufficient reporting of data in the majority of studies, especially in relation to listing the medications and dose. Stroke risk and oral anticoagulation prescription Stroke risk for patients with AF recurrence and eligibility for prescription of OAC were reported in only 1 study (n = 42) [7]. In this study, 8 of the 10 patients with recurrent AF had a CHA2DS2-VASc score ≥2. However, only 3 of 8 patients were on OAC, with 1 of the 3 requiring warfarin because of a mechanical valve. Two studies reported that OAC was routinely prescribed for recurrences of AF that lasted longer than 24 and 48 h; however, these studies neither reported the numbers of patients in which this occurred nor reported their calculated stroke risk scores [15, 16]. Late-onset postoperative atrial fibrillation post-discharge Incidence of late-onset POAF post-hospital discharge was an additional finding reported in 3 studies (n = 2716) [14, 17, 18]. For people with no in-hospital episode of AF, intermittent monitoring in the first 2–4 weeks post-discharge also identified individuals with late-onset POAF post-hospital discharge: incidence rate (cases per 100 persons screened) 3.9% (95% CI, 2.7–5.4%) (Fig. 3). Heterogeneity for these studies was moderate, with I2 = 50.3%, as defined by an I2 value of 50% indicating moderate heterogeneity [22]. Figure 3: View largeDownload slide Forest plot of incidence of new-onset postoperative atrial fibrillation. CI: confidence interval. Figure 3: View largeDownload slide Forest plot of incidence of new-onset postoperative atrial fibrillation. CI: confidence interval. DISCUSSION Statement of findings Monitoring of patients discharged in stable sinus rhythm post-cardiac surgery identifies POAF recurrence in 28.3 cases per 100 persons screened in the first 2–4 weeks post-discharge using non-invasive techniques; implanted devices can identify recurrence in 60–100% of cases, suggesting that in-hospital episodes of new POAF are not necessarily isolated or transient. Importantly, most POAF recurrences are asymptomatic; therefore, these episodes are unlikely to be recognized without the aid of additional monitoring post-discharge. Furthermore, in the one study that calculated risk of stroke, 8 of 10 patients with POAF recurrence had a sufficient risk of stroke (i.e. CHA2DS2-VASc score ≥2) to be considered eligible for OAC. This highlights the potential importance of early identification of POAF recurrence through monitoring in the post-discharge period and appropriate treatment and prescription of OAC to prevent stroke. There are multiple hypotheses as to the mechanism for development of POAF following surgery patients. It has been reported that inflammation and the associated increase in plasma concentrations of C-reactive protein and interleukin 6 in conjunction with cardiac surgery may be associated with the development of AF [23, 24]. Other mechanisms include increased sympathoadrenal activation post-surgery, which prolongs the atrial refractive period and promotes ectopic activity, and oxidative stress following bypass surgery [24, 25]. However, it remains unknown as to why POAF develops in some patients and not in others. Pre-existing risk factors that can lead to the formation of an atrial AF substrate may also influence risk [24]. It is thought that POAF may occur post-surgery when patients reach a ‘pathophysiological threshold’ of pro-arrhythmic factors due to a combination of pre-existing risk factors and post-surgical factors [24, 26]. There is a natural incidence of AF over and above that related to the surgery and the postoperative period, and this is likely to factor in the yield of AF detected from long-term monitoring. But it is possible that the patients who develop POAF post-surgery may be those with some existing AF substrate changes, who are predisposed to developing AF both during the provocation of surgery and spontaneously in the long term. Risk scores have been developed to predict the development of new-onset POAF after cardiac surgery, such as the POAF score [27] and the Brazilian risk score [28]. Although validated for predicting new-onset POAF, these scores have not been extended to predicting recurrence of POAF but may be useful, as recent evidence suggests that new-onset AF is a predictor of long-term AF [5]. There is a 6-fold increase in the likelihood of clinical AF recurrence in patients with an episode of POAF compared to those without POAF; however, late-onset POAF can also occur post-hospital discharge [16]. Incidence of late-onset POAF was a secondary finding of the studies included in our review; however, other research has identified similar incidences of 3.6–8% of new AF post-discharge [10, 29]. It is also probable that many in-hospital episodes of POAF go undetected. A study using Holter monitoring found 13% of all patients had ‘silent’ POAF during the post-cardiac surgery inpatient period, which was asymptomatic and remained undiagnosed at discharge [9]. ‘Silent’ POAF was associated with higher AF recurrence rate at 1-year follow-up [9]. It is therefore difficult to predict which patients are at higher risk of AF recurrence or AF development in the post-discharge period, suggesting perhaps that monitoring might also be indicated for patients without diagnosed POAF during the admission. Identification of an appropriate means of monitoring is becoming increasingly important, given that the number of elderly patients undergoing cardiac surgery is projected to increase [30], and age is a primary risk factor for the development of both AF and stroke [31, 32]. Furthermore, obesity [33] and metabolic syndrome [34] are also associated with a higher incidence of POAF post-cardiac surgery. We are therefore likely to see an increase in the incidence of POAF corresponding to the rise in obesity in the next 20 years, along with an associated increase in stroke, morbidity, mortality and health system costs. The mode of surveillance will influence the number of cases of recurrent AF identified. Implanted devices with continual rhythm recording will pickup much more brief paroxysmal episodes than intermittent sampling, but the clinical significance of these episodes, such as cardiac implanted device-detected atrial high rate episodes may be less than clinically detected AF or AF detected by intermittent brief recordings [35, 36]. Although technology allowing centralized remote monitoring of multiple patients with implanted devices could make monitoring more feasible [37], the likely lower risk in patients detected in this way needs to be taken into consideration. Patient tolerance, acceptance and compliance of an implanted device is an additional consideration, with 50% refusal rate reported from 1 study designed to have an implanted device for 2 years [20]. Despite the clear evidence that POAF is a major cause of morbidity and mortality and that asymptomatic POAF recurrences can be identified in the weeks following discharge, there is a lack of evidence with regard to the long-term effect of anticoagulation on stroke risk in POAF patients and therefore no indication that treatment will be altered as a result of identifying POAF recurrence through monitoring [12]. POAF management guidelines recommend anticoagulation for at least 4 weeks post-cardioversion of patients with POAF lasting >48 h and anticoagulation for ‘high-risk patients’ in which sinus rhythm was restored within 24 h of onset [11]. The results of this study raise questions regarding which patients should be considered ‘high risk’ and which patients are at risk of recurrence or late-onset AF, thus which patients should monitored. Instigating monitoring comes at a cost. However, it could well be cost-effective from the standpoint of stroke prevention. Although the studies in this review utilized devices for monitoring, there are also other low-cost methods to address detection of AF through improved patient education regarding symptom awareness and routine self-monitoring using pulse palpation, and some hand-held ECG devices have a fairly modest cost. To date, screening studies for primary AF detection using new devices, such as hand-held single-lead ECGs, have all been demonstrated to be cost-effective [38–42]. Limitations The main limitation of this review is the small number of studies included and design of the studies. Participants were non-homogeneous with regard to age, cardiac procedure and comorbidities, and the numbers in each trial were small with <100 participants in 5 of 8 studies. Study designs varied significantly between the studies, with only 2 randomized controlled trials, and the method and time frame for monitoring varied greatly between the trials. Furthermore, the majority of trials did not report data for our secondary objectives. Only 5 studies reported symptomatic status of AF episodes, the data regarding reduced recurrence with β-blockers was not conclusive, only 2 small studies investigated the effect of amiodarone and only 1 study with small numbers reported the risk of stroke in patients identified with POAF recurrence. Therefore, there is insufficient strength in the available data to draw conclusions about the risk of stroke in this population, the need for anticoagulation treatment and the effect of preventative treatments such as amiodarone and β-blockers. This highlights a need for future studies to look at amiodarone and its role in preventing recurrence of POAF in this group of patients and for more understanding of how antiarrhythmic medications such as amiodarone can influence prevention of recurrence. Furthermore, although 3 of 5 studies suggested a beneficial effect of β-blockers in the prevention of AF post-discharge, it is noteworthy that most institutions prescribe β-blockers at the time of discharge as a process of continuing preoperative medications and not with the intention of offering antiarrhythmic protection. There is no evidence or uniformity in prescribing one particular β-blocker or any one specific dose to offer adequate antiarrhythmic protection. Further research is indicated to address these issues. Although we know that risk of stroke is increased in patients with an episode of POAF following cardiac surgery [2], and asymptomatic AF episodes in the general population increase stroke risk [36], and risk of stroke from paroxysmal AF is sufficiently high to require anticoagulation [43], this review highlights that there is insufficient evidence to determine the stroke risk of recurrences in patients with POAF who are discharged home and presumed to be in stable sinus rhythm. It is also uncertain whether the increased recurrence rate detected by continuous invasive recordings will have the same prognostic significance as recurrence detected by more brief or intermittent recordings. These data are required to inform recommendations for the length and intensity of monitoring required post-discharge and anticoagulation and preventative treatments for recurrences. Overall, the results of this review indicate a need for large-scale research to look at the prognostic significance of POAF recurrences and late-onset AF episodes post-discharge, in relation to stroke, morbidity and mortality, and therefore treatment implications and recommendations. CONCLUSION Monitoring for POAF recurrence post-hospital discharge identifies significant numbers of asymptomatic recurrences in patients at high risk of stroke, who may benefit from early detection of POAF recurrence and anticoagulation for stroke prevention. More intense monitoring is more likely to identify POAF recurrence. Further large-scale studies are required to investigate the prognostic significance of POAF recurrence in relation to risk of stroke and mortality. SUPPLEMENTARY MATERIAL Supplementary material is available at ICVTS online. Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. Nicole Lowres is supported by a NSW Health Early Career Fellowship [H16/52168 to N.L.]. Conflict of interest: Ben Freedman reports grants, personal fees and non-financial support from Bayer Pharma AG outside the submitted work, grants and non-financial support from Boehringer Ingelheim outside the submitted work, grants and personal fees from BMS/Pfizer outside the submitted work, personal fees from Servier outside the submitted work, personal fees from Astra-Zeneca outside the submitted work and consulting fees from Gilead outside the submitted work. Lis Neubeck has received grants and honoraria from BMS/Pfizer outside the submitted work. All other authors have no competing interests to disclose. REFERENCES 1 Helgadottir S, Sigurdsson MI, Ingvarsdottir IL, Arnar DO, Gudbjartsson T. Atrial fibrillation following cardiac surgery: risk analysis and long-term survival. J Cardiothorac Surg  2012; 7: 87. Google Scholar CrossRef Search ADS PubMed  2 Gialdini G, Nearing K, Bhave PD, Bonuccelli U, Iadecola C, Healey JS et al.   Perioperative atrial fibrillation and the long-term risk of ischemic stroke. JAMA  2014; 312: 616– 22. Google Scholar CrossRef Search ADS PubMed  3 Phan K, Ha HS, Phan S, Medi C, Thomas SP, Yan TD. New-onset atrial fibrillation following coronary bypass surgery predicts long-term mortality: a systematic review and meta-analysis. Eur J Cardiothorac Surg  2015; 48: 817– 24. Google Scholar CrossRef Search ADS PubMed  4 Levy D, Kannel WB. Postoperative atrial fibrillation and mortality: do the risks merit changes in clinical practice? J Am Coll Cardiol  2004; 43: 749– 51. Google Scholar CrossRef Search ADS PubMed  5 Lee SH, Kang DR, Uhm JS, Shim J, Sung JH, Kim JY et al.   New-onset atrial fibrillation predicts long-term newly developed atrial fibrillation after coronary artery bypass graft. Am Heart J  2014; 167: 593– 600.e1. Google Scholar CrossRef Search ADS PubMed  6 Healey JS, Connolly SJ, Gold MR, Israel CW, Van Gelder IC, Capucci A et al.   Subclinical atrial fibrillation and the risk of stroke. N Engl J Med  2012; 366: 120– 9. Google Scholar CrossRef Search ADS PubMed  7 Lowres N, Mulcahy G, Gallagher R, Ben Freedman S, Marshman D, Kirkness A et al.   Self-monitoring for atrial fibrillation recurrence in the discharge period post-cardiac surgery using an iPhone electrocardiogram. Eur J Cardiothorac Surg  2016; 50: 44– 51. Google Scholar CrossRef Search ADS PubMed  8 Saxena A, Dinh DT, Smith JA, Shardey GC, Reid CM, Newcomb AE. Usefulness of postoperative atrial fibrillation as an independent predictor for worse early and late outcomes after isolated coronary artery bypass grafting (multicenter Australian study of 19, 497 patients). Am J Cardiol  2012; 109: 219– 25. Google Scholar CrossRef Search ADS PubMed  9 Guenancia C, Pujos C, Debomy F, Malapert G, Laurent G, Bouchot O. Incidence and predictors of new-onset silent atrial fibrillation after coronary artery bypass graft surgery. Biomed Res Int  2015; 2015: 703685. Google Scholar CrossRef Search ADS PubMed  10 Bidar E, Maesen B, Nieman F, Verheule S, Schotten U, Maessen JG. A prospective randomized controlled trial on the incidence and predictors of late-phase postoperative atrial fibrillation up to 30 days and the preventive value of biatrial pacing. Heart Rhythm  2014; 11: 1156– 62. Google Scholar CrossRef Search ADS PubMed  11 Frendl G, Sodickson AC, Chung MK, Waldo AL, Gersh BJ, Tisdale JE et al.   2014 AATS guidelines for the prevention and management of perioperative atrial fibrillation and flutter for thoracic surgical procedures. J Thorac Cardiovasc Surg  2014; 148: e153– 93. Google Scholar CrossRef Search ADS PubMed  12 Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B et al.   2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur J Cardiothorac Surg  2016; 50: e1– e88. Google Scholar CrossRef Search ADS PubMed  13 Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, et al.   The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/nosgen.pdf (3 July 2017, date last accessed). 14 Ambrosetti M, Tramarin R, Griffo R, De Feo S, Fattirolli F, Vestri A et al.   Late postoperative atrial fibrillation after cardiac surgery: a national survey within the cardiac rehabilitation setting. J Cardiovasc Med (Hagerstown)  2011; 12: 390– 5. Google Scholar CrossRef Search ADS PubMed  15 Cioffi G, Cemin C, Russo TE, Pellegrini A, Terrasi F, Ferrario G. Post-discharge recurrences of new-onset atrial fibrillation following cardiac surgery: impact of low-dose amiodarone and beta-blocker prophylaxis. Ital Heart J  2000; 1: 691– 7. Google Scholar PubMed  16 Cioffi G, Mureddu G, Cemin C, Russo TE, Pellegrini A, Terrasi F et al.   Characterization of post-discharge atrial fibrillation following open-heart surgery in uncomplicated patients referred to an early rehabilitation program. Ital Heart J  2001; 2: 519– 28. Google Scholar PubMed  17 Giaccardi M, Macchi C, Colella A, Polcaro P, Zipoli R, Cecchi F et al.   Postacute rehabilitation after coronary surgery: the effect of preoperative physical activity on the incidence of paroxysmal atrial fibrillation. Am J Phys Med Rehabil  2011; 90: 308– 15. Google Scholar CrossRef Search ADS PubMed  18 Funk M, Richards SB, Desjardins J, Bebon C, Wilcox H. Incidence, timing, symptoms, and risk factors for atrial fibrillation after cardiac surgery. Am J Crit Care  2003; 12: 424– 33; quiz 34–5. Google Scholar PubMed  19 El-Chami MF, Merchant FM, Smith P, Levy M, Nelms AG, Merlino J et al.   Management of new-onset postoperative atrial fibrillation utilizing insertable cardiac monitor technology to observe recurrence of AF (MONITOR-AF). Pacing Clin Electrophysiol  2016; 39: 1083– 9. Google Scholar CrossRef Search ADS PubMed  20 Lomivorotov VV, Efremov SM, Pokushalov EA, Romanov AB, Ponomarev DN, Cherniavsky AM et al.   Randomized trial of fish oil infusion to prevent atrial fibrillation after cardiac surgery: data from an implantable continuous cardiac monitor. J Cardiothorac Vasc Anesth  2014; 28: 1278– 84. Google Scholar CrossRef Search ADS PubMed  21 von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Br Med J  2007; 335: 806– 8. Google Scholar CrossRef Search ADS   22 Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med  2002; 21: 1539– 58. Google Scholar CrossRef Search ADS PubMed  23 Wu N, Xu B, Xiang Y, Wu L, Zhang Y, Ma X et al.   Association of inflammatory factors with occurrence and recurrence of atrial fibrillation: a meta-analysis. Int J Cardiol  2013; 169: 62– 72. Google Scholar CrossRef Search ADS PubMed  24 Maesen B, Nijs J, Maessen J, Allessie M, Schotten U. Post-operative atrial fibrillation: a maze of mechanisms. Europace  2012; 14: 159– 74. Google Scholar CrossRef Search ADS PubMed  25 Echahidi N, Pibarot P, O’Hara G, Mathieu P. Mechanisms, prevention, and treatment of atrial fibrillation after cardiac surgery. J Am Coll Cardiol  2008; 51: 793– 801. Google Scholar CrossRef Search ADS PubMed  26 Schotten U, Verheule S, Kirchhof P, Goette A. Pathophysiological mechanisms of atrial fibrillation: a translational appraisal. Physiol Rev  2011; 91: 265– 325. Google Scholar CrossRef Search ADS PubMed  27 Mariscalco G, Biancari F, Zanobini M, Cottini M, Piffaretti G, Saccocci M et al.   Bedside tool for predicting the risk of postoperative atrial fibrillation after cardiac surgery: the POAF score. J Am Heart Assoc  2014; 3: e000752. Google Scholar CrossRef Search ADS PubMed  28 Sa MP, Sa MV, Albuquerque AC, Silva BB, Siqueira JW, Brito PR et al.   Predicting risk of atrial fibrillation after heart valve surgery: evaluation of a Brazilian risk score. Rev Bras Cir Cardiovasc  2012; 27: 117– 22. Google Scholar CrossRef Search ADS PubMed  29 Ahlsson A, Fengsrud E, Bodin L, Englund A. Postoperative atrial fibrillation in patients undergoing aortocoronary bypass surgery carries an eightfold risk of future atrial fibrillation and a doubled cardiovascular mortality. Eur J Cardiothorac Surg  2010; 37: 1353– 9. Google Scholar CrossRef Search ADS PubMed  30 Lee JJ, Park NH, Lee KS, Chee HK, Sim SB, Kim MJ et al.   Projections of demand for cardiovascular surgery and supply of surgeons. Korean J Thorac Cardiovasc Surg  2016; 49: S37– 43. Google Scholar CrossRef Search ADS PubMed  31 Mathew JP, Fontes ML, Tudor IC, Ramsay J, Duke P, Mazer CD et al.   A multicenter risk index for atrial fibrillation after cardiac surgery. JAMA  2004; 291: 1720– 9. Google Scholar CrossRef Search ADS PubMed  32 Chyou JY, Hunter TD, Mollenkopf SA, Turakhia MP, Reynolds MR. Individual and combined risk factors for incident atrial fibrillation and incident stroke: an analysis of 3 million at-risk US patients. J Am Heart Assoc  2015; 4: e001723. Google Scholar CrossRef Search ADS PubMed  33 Hernandez AV, Kaw R, Pasupuleti V, Bina P, Ioannidis JP, Bueno H et al.   Association between obesity and postoperative atrial fibrillation in patients undergoing cardiac operations: a systematic review and meta-analysis. Ann Thorac Surg  2013; 96: 1104– 16. Google Scholar CrossRef Search ADS PubMed  34 Ivanovic B, Tadic M, Bradic Z, Zivkovic N, Stanisavljevic D, Celic V. The influence of the metabolic syndrome on atrial fibrillation occurrence and outcome after coronary bypass surgery: a 3-year follow-up study. Thorac Cardiovasc Surg  2014; 62: 561– 8. Google Scholar CrossRef Search ADS PubMed  35 Van Gelder IC, Healey JS, Crijns HJ, Wang J, Hohnloser SH, Gold MR et al.   Duration of device-detected subclinical atrial fibrillation and occurrence of stroke in ASSERT. Eur Heart J  2017; doi:10.1093/eurheartj/ehx042. 36 Freedman B, Camm J, Calkins H, Healey JS, Rosenqvist M, Wang J et al.   Screening for atrial fibrillation: a report of the AF-SCREEN international collaboration. Circulation  2017; 135: 1851– 67. Google Scholar CrossRef Search ADS PubMed  37 Zoppo F, Facchin D, Molon G, Zanotto G, Catanzariti D, Rossillo A et al.   Improving atrial fibrillation detection in patients with implantable cardiac devices by means of a remote monitoring and management application. Pacing Clin Electrophysiol  2014; 37: 1610– 8. Google Scholar CrossRef Search ADS PubMed  38 Lowres N, Neubeck L, Salkeld G, Krass I, McLachlan AJ, Redfern J et al.   Feasibility and cost effectiveness of stroke prevention through community screening for atrial fibrillation using iPhone ECG in pharmacies. The SEARCH-AF study. Thromb Haemost  2014; 111: 1167– 76. Google Scholar CrossRef Search ADS PubMed  39 Aronsson M, Svennberg E, Rosenqvist M, Engdahl J, Al-Khalili F, Friberg L et al.   Cost-effectiveness of mass screening for untreated atrial fibrillation using intermittent ECG recording. Europace  2015; 17: 1023– 9. Google Scholar CrossRef Search ADS PubMed  40 Jacobs MS, Kaasenbrood F, Postma MJ, van Hulst M, Tieleman RG. Cost-effectiveness of screening for atrial fibrillation in primary care with a handheld, single-lead electrocardiogram device in the Netherlands. Europace  2016. doi:10.1093/europace/euw285. 41 Moran PS, Teljeur C, Harrington P, Smith SM, Smyth B, Harbison J et al.   Cost-effectiveness of a national opportunistic screening program for atrial fibrillation in Ireland. Value Health  2016; 19: 985– 95. Google Scholar CrossRef Search ADS PubMed  42 Hobbs FD, Fitzmaurice DA, Mant J, Murray E, Jowett S, Bryan S et al.   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 Technol Assess  2005; 9:iii–v, ix–x, 1– 74. Google Scholar CrossRef Search ADS   43 Ganesan AN, Chew DP, Hartshorne T, Selvanayagam JB, Aylward PE, Sanders P et al.   The impact of atrial fibrillation type on the risk of thromboembolism, mortality, and bleeding: a systematic review and meta-analysis. Eur Heart J  2016; 37: 1591– 602. Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

Journal

Interactive CardioVascular and Thoracic SurgeryOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from Google Scholar, PubMed
Create lists to organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off