Abstract
Background:
We aimed to compare the clinical impact of treatment with a rhythm control strategy to a rate control strategy in postcardiac surgery atrial fibrillation patients.
Methods:
A comprehensive search of MEDLINE, Embase, Cochrane Central Register of Controlled Trials, CINAHL, Web of Science, Scopus, ProQuest Dissertations, and ClinicalTrials.gov was conducted from inception to October 2025. Our meta-analysis included randomized controlled trials (RCTs) comparing therapeutic rhythm control interventions with rate control interventions. We used the Cochrane risk-of-bias tool to appraise the quality of included RCTs, the GRADE framework to evaluate the strength of the evidence, and adhered to the PRISMA guidelines for reporting.
Results:
Eight RCTs (n = 894 patients) met the inclusion criteria. There was no difference in hospital length of stay (4 RCTs) between rhythm control and rate control [MD: −0.41 days (95% CI: −3.23, 2.42)]. An aggressive rhythm control strategy (ibutilide, procainamide, propafenone, or electric cardioversion) was associated with higher odds of in-hospital conversion to sinus rhythm [OR: 4.01 (95% CI: 1.30, 12.39)] and a higher risk of medication-related adverse events (hypotension, bradycardia, and syncope) [RR: 3.05 (95% CI: 1.05, 8.89)].
Conclusion:
Among postcardiac surgery patients with new-onset atrial fibrillation, there was no evidence that a rhythm control treatment strategy resulted in better outcomes than a rate control strategy.
Systematic Review Registration:
https://www.crd.york.ac.uk/PROSPERO/view/CRD42019128559, identifier CRD42019128559.
Introduction
Postoperative atrial fibrillation (POAF) is a common complication after cardiac surgery that can lead to significant morbidity and mortality (). Its incidence is highest after combined valve surgery with coronary artery bypass grafting (CABG) (50%), followed by valve surgery in isolation (30%–40%), followed by CABG in isolation (15%–25%) (). Patient characteristics that increase the likelihood of developing new-onset POAF include advanced age, reduced left ventricular systolic function, male gender, pre-existing hypertension, and renal dysfunction (). POAF after cardiac surgery is currently thought to be driven by multiple factors: the inflammatory response of surgery, neurohormonal activation, and the presence of a structural or metabolic substrate for its development ().
Despite its largely self-limiting course, new-onset POAF after cardiac surgery is an independent predictor of serious adverse outcomes, both short-term and long-term. Short-term adverse outcomes include hemodynamic instability, acute congestive heart failure (CHF), pulmonary edema, and increased hospital length of stay (LOS). Long-term adverse outcomes include increased risk of stroke and mortality (, ).
Treatment strategies for POAF include a rhythm-control approach, focusing on conversion to normal sinus rhythm (NSR), or a rate-control approach. A rhythm-control strategy may hasten in-hospital reversion to NSR, reduce the incidence of persistent atrial fibrillation, and obviate the need for anticoagulation. Conversely, a rate-control approach may help avoid side effects and drug interactions associated with rhythm-control medications and protect against rapid ventricular rates until spontaneous reversion to NSR occurs (, ).
A previously published systematic review comparing rhythm vs. rate control in patients with POAF after cardiac surgery identified important gaps in the literature. Our systematic review and meta-analysis aims to address these gaps. The previous review noted a paucity of randomized trials, heterogeneity in interventions, variability in follow-up duration, and a lack of evaluation of long-term clinical outcomes beyond the early postoperative period (). Our study employed a more comprehensive search strategy, expanded the scope of relevant outcomes, and conducted additional analyses not undertaken in the previous study.
Methods
The protocol for this systematic review and meta-analysis was registered and is publicly available in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42019128559). We reported all findings in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines statement ().
Outcomes
We examined the impact of implementing a therapeutic rhythm control strategy vs. a rate control strategy in postcardiac surgery patients with new-onset atrial fibrillation. Postoperative atrial fibrillation in this population has been shown to prolong hospital length of stay (LOS). We therefore compared the impact of each treatment strategy on hospital LOS (primary outcome). Secondary outcomes included the odds of in-hospital conversion to normal sinus rhythm (NSR) and the risk of medication-related hemodynamic side effects (hypotension, bradycardia, and syncope). Given the negative inotropic effects of rhythm and rate control medications, we also examined the effect of these two strategies on the risk of developing congestive heart failure (CHF) or pulmonary edema during the index operative admission (Table 1—Outcome Definitions). Long-term secondary outcomes included the odds of thromboembolic events and death during the study follow-up period (Figure 1—Analytical Framework).
Table 1
| Outcome | Definition | Measurement unit |
|---|---|---|
| Hospital length of stay | Postoperative or postrandomization hospital length of stay for the index cardiac surgery admission | Days |
| In-hospital conversion to normal sinus rhythm | Number of patients converted to normal sinus rhythm at latest reported time-point during index cardiac surgery admission | Frequency |
| Medication adverse effects | Number of patients developing any of the following: hypotension, bradycardia, or syncope during study follow-up | Frequency |
| Incidence of congestive heart failure & pulmonary edema | Number of patients developing new-onset heart failure or pulmonary edema during study follow-up | Frequency |
| Thromboembolic events | Number of patients developing any thromboembolic event (cerebrovascular or non-cerebrovascular) during study follow-up | Frequency |
| Mortality | Number of postcardiac surgery deaths during study follow-up | Frequency |
Outcome definitions for systematic review and meta-analaysis.
Figure 1
Search strategy
We searched Medline & Medline in-process (Ovid), PubMed-NLM, Embase (Ovid), CENTRAL, Cumulative Index of Nursing and Allied Health Literature (CINAHL), Web of Science, Scopus, ProQuest Dissertations and Theses Global from inception to October 2025 using a pre-defined search strategy (Supplementary Appendix 2). We also searched the clinical trial registry (www.ClinicalTrials.gov) for ongoing and completed but unpublished studies. This search was conducted by a professional research librarian using both controlled vocabulary and sensitive keyword search terms. Related subject headings (controlled vocabulary) for each database were also identified and included in the search. We did not impose any language restrictions on the search.
Study selection
We sought full-text published RCT manuscripts using a predefined Population/Intervention/Comparator/Outcome (PICO) framework. Our population was adult patients (≥18 years) with POAF after cardiac surgery. The intervention of interest was rhythm control, defined as attempts to convert to normal sinus rhythm by any method, whether pharmacological (e.g., amiodarone, propafenone, ibutilide, sotalol) or by synchronized electrical conversion. Our comparator was rate-control interventions, defined as pharmacological strategies aimed at reducing heart rate using beta-blockers, calcium channel blockers, or digoxin. The primary outcome was postoperative or post-randomization hospital LOS in days. Secondary outcomes included in-hospital rates of conversion to NSR at the latest reported time-point, antiarrhythmic medication adverse effects (specifically hypotension, bradycardia, and syncope), rates of thrombo-embolic events (cerebrovascular and non-cerebrovascular), CHF/pulmonary edema, and all-cause mortality during the follow-up period. We excluded studies comparing two rate-control strategies or two rhythm-control strategies, case series, case reports, observational studies, studies examining prophylactic interventions, and pediatric surgery studies.
Study screening
All articles were screened independently by two reviewers (MD & MT). Any discrepancies at the screening stage were resolved under the supervision of a third reviewer (AH) through discussion and consensus.
Data extraction and quality assessment
Two reviewers (MD and CT) independently extracted data from eligible RCTs using predefined data extraction forms. For missing study-level data, we corresponded with three primary investigators. Two authors did not respond, while the third replied that he no longer had access to study data due to his relocation. When feasible, patient-level numerical data was extracted from plots using a web-based data extraction tool, WebPlotDigitizer (Ankit Rohatgi, WebPlotDigitizer Version 4.2, San Francisco, California, USA), which is known for high intercoder reliability and validity (). The accuracy of the extracted data was then verified by a third reviewer (AH).
Two reviewers (MD and MT) independently assessed the risk of bias of the included studies using the Revised Cochrane Risk of Bias Tool for Randomized Trials 2 (RoB 2). Any disagreements were resolved through discussion and adjudication by a third reviewer (AH) (). The certainty of evidence for each outcome was assessed using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach. This framework evaluates the quality of evidence based on study limitations, inconsistency, indirectness, imprecision, and publication bias. The overall certainty of evidence was categorized as high, moderate, low, or very low.
Statistical analysis
Dichotomous outcomes were extracted as events/non-events, and continuous outcomes were extracted as means and standard deviations. When only medians and interquartile ranges were available, the methods described by Wan et al. were used to estimate means and standard deviations (). Anticipating numerous sources of clinical heterogeneity, including diverse cardiac surgical procedures, the type and dosing of pharmacological agents, and patient demographics, we used a DerSimonian & Laird random-effects model to pool effect-size data (). Effect summaries were presented as Forest plots, and statistical heterogeneity was assessed using Cochran's Q test and I2. Heterogeneity was considered “high” when Cochran's Q test yielded a statistically significant p-value (p ≤ 0.05) or when I2 was > 75% (). When substantial heterogeneity was observed, subgroup analyses were conducted to further explore potential causes. Publication bias was assessed by examining funnel plot asymmetry using Egger's regression test (). The robustness of GRADE assessments was tested using sensitivity (leave-one-out) analyses. Statistical analyses were performed using Review Manager (RevMan) Version 5.4 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2020) and R version 4.0.0 (R Foundation for Statistical Computing, Vienna, Austria) with the Metafor package for meta-analysis.
Results
Study characteristics
A total of 2,671 records were identified through database searches. Study screening and selection proceeded through multiple phases, as shown in the PRISMA flow diagram (Figure 2). After removing duplicates and excluding ineligible articles, eight RCTs (n = 894 patients) met the inclusion criteria. Most studies had small-to-medium sample sizes (29–150 patients), with only one study enrolling 523 patients. Rhythm control interventions included using any of sotalol, amiodarone, procainamide, propafenone, ibutilide, flecainide, and synchronized electric cardioversion. Rate control interventions included metoprolol, digoxin, diltiazem, and verapamil. All studies reported the incidence of conversion to normal sinus rhythm and medication-related hypotension. Follow-up durations ranged from 12 h to 60 days. Only four studies reported hospital LOS. Characteristics of the included studies are shown in Table 2.
Figure 2
Table 2
| Study details | Study design | Location | Inclusion criteria and exclusion Criteria | Group | Intervention | Number enrolled | Male/Female | Age mean (SD) | Follow-up duration | Study findings |
|---|---|---|---|---|---|---|---|---|---|---|
| Campbell et al. 1985 | Single-center randomized trial | Australia | Inclusion criteria Post-cardiac surgery Atrial Arrhythmias and Ventricular response >120 Exclusion Criteria
| Rhythm control | IV Sotalol | 20 | M: 19, F:1 | 60.5 (9.1) | 12 h |
|
| Rate control | IV Digoxin (IV disopyramide was used only when digoxin had failed) | 20 | M:15, F:5 | 63.5 (5.2) | ||||||
| Cochrane et al. 1994 | Single-center randomized trial | Australia | Inclusion criteria Open heart surgery patients with new onset AF that persisted >20 min, with SBP ≥ 85 mmHg and no inotropic support Exclusion Criteria
| Rhythm control | IV Amiodarone (If reversion to NSR had not occurred, then digoxin was added) | 15 | M:11, F:4 | 60.2 (no SD) | 24 h |
|
| Rate control | IV Digoxin (If reversion to NSR had not occurred, then amiodarone was added) | 15 | M:10, F:5 | 65.8 (no SD) | ||||||
| Gillinov et al. 2016 | Multicenter randomized trial | US and Canada (23 Centres) | Inclusion criteria Adult with hemodynamically stable new-onset POAF (after elective surgeries) that persisted for more than 60 min or recurrent episodes of atrial fibrillation during the index hospitalization (≤7 days after surgery) Exclusion Criteria Patients with a prior history of atrial fibrillation were excluded to avoid making changes to their established preoperative medication regimen for atrial fibrillation and anticoagulation. | Rhythm control | Amiodarone (with or without rate control agent. For persistent AF > 24–8 h, DC cardioversion was recommended) | 261 | M:199, F:62 | 68.4 (8.4) | 60 days |
|
| Rate control | Rate control agents (BB or CCB), with a goal of achieving a resting heart rate of less than 100 bpm. Switching to rhythm control was allowed for hemodynamic or symptom control. | 262 | M:197, F:65 | 69.2 (9.8) | ||||||
| Hejmls et al. 1992 | Single-center randomized trial | Denmark | Inclusion criteria AF after open heart surgery Exclusion Criteria
| Rhythm control | IV Procainamide followed by oral maintenance | 15 | M:13, F:2 | Median: 65 (IQR: 46–74) | 12 h |
|
| Rate control | IV Digoxin followed by oral maintenance | 15 | M:11, F:4 | Median: 60 (IQR: 17–72) | ||||||
| Kamali et al. 2017 | Single-center randomized trial | Iran | Inclusion criteria CABG patients (Age 45–80 years old) with no prior history of arrhythmia or on antiarrhythmic medications Exclusion Criteria
| Rhythm control | IV Amiodarone | 75 | M:42, F:33 | — | 24 h |
|
| Rate control | Metoprolol | 75 | M:37, F:38 | — | ||||||
| Lee et al. 2000 | Single-center randomized trial | Canada | Inclusion criteria 18 years and older who had atrial fibrillation for at least 1 h and had no history of paroxysmal atrial fibrillation Exclusion Criteria
| Rhythm control | Sotalol, propafenone, or procainamide, with or without electric cardioversion | 27 | M:21, F:6 | 67 (7) | 60 days |
|
| Rate control | IV diltiazem, BB or digoxin | 23 | M:18, F:5 | 70 (5) | ||||||
| Soucier et al. 2003 | Randomized trial | USA (2 Centres) | Inclusion criteria Hemodynamically stable new-onset AF patients hospitalized on the telemetry step-down units after open-heart surgery. AF duration was between 3 and 72 h. Exclusion Criteria
| Rhythm control | IV ibutilide oral propafenone | 30 | M:24, F:6 | Ibutilide: 76 (6), Propafenone: 70 (9) | 14 days |
|
| Rate control | Digoxin, beta-blockers and/or calcium channel-blocking agents | 12 | M:9, F:3 | 76 (7) | ||||||
| Wafaa et al. 1989 | Single-center randomized trial | UK | Inclusion criteria 18–80 years old patients who had CAGB complicated by atrial tachyarrhythmia (AF, A. Flutter, and AT) within 96 h post-op and lasting >15 min with ventricular response >120 bpm Exclusion Criteria
| Rhythm control | Flecanide IV bolus followed infusion for 24 h. Verapamil 10 mg was given if, after 45 min, the patient failed to revert to SR with VR <100 | 15 | M:15, F:0 | 61 (8) | 24 h |
|
| Rate control | Digoxin IV 0.5. Verapamil 10 mg was given if, after 45 min, the patient failed to revert to SR with VR <100 | 14 | M: 11, F:3 | 66 (5) |
Characteristics of included studies.
Risk of bias assessment
Most studies adequately described the methods used for randomization and allocation concealment, except for the study by Cochrane et al. (). “Some concerns” were identified with blinding of interventions in 4 RCTs (–). All studies were judged to be at low risk of missing outcome data or ascertainment bias. Six RCTs, however, were considered at “some risk” of bias from selective outcome reporting (–). In summary, 4 RCTs were deemed at an overall “high risk” of bias (, –), while the remaining four had “some concerns” (, –) (Figure 3).
Figure 3
Outcomes
Postoperative or post-randomization hospital length of stay
A total of 4 RCTs (n = 765) evaluated the hospital LOS, and an all-studies-included analysis revealed no difference between rhythm control and rate control interventions [MD: −0.41 days (95% CI: −3.23, 2.42), p = 0.78, I2 = 98%] [Field, (, –)]. Given the observed heterogeneity, we performed an exploratory subgroup analysis, categorizing rhythm control strategies as “aggressive” vs. “conventional”. This categorization was developed post hoc to explore potential sources of statistical heterogeneity. The subgroups were found to be different (test for subgroup differences p = 0.03), and the direction of effect size estimates became concordant when studies exploring aggressive rhythm control strategies were analyzed separately from those exploring less aggressive strategies (amiodarone). The effect estimates trended towards a shorter hospital length of stay with aggressive rhythm control, compared to rate control [MD: −2.59 (95% CI: −5.92, 0.74), p = 0.13, I2 = 91%]. Conversely, less aggressive rhythm control with amiodarone exhibited a trend towards longer hospital length of stay compared to rate control [MD: 1.63 (95% CI: −0.42, 3.69), p = 0.12, I2 = 96%]. Despite providing insight into the underlying mechanisms for heterogeneity, these associations did not attain statistical significance (Figure 4).
Figure 4
In-hospital conversion to normal sinus rhythm
Eight studies (n = 894 patients) reported rates of in-hospital conversion to NSR (–). The effect estimate for all studies combined revealed no difference between a rhythm control and a rate control strategy [OR for conversion to sinus rhythm: 1.98 (95% CI: 0.75, 5.20), p = 0.17, I2 = 74%]. There was, however, a remarkable variation in the employed rhythm control modalities; while some studies adopted a more aggressive rhythm control approach (e.g., procainamide, propafenone, or electric cardioversion), others used a less aggressive approach with amiodarone. It was therefore prudent to analyze the aggressive rhythm control and amiodarone subgroups again separately. Higher odds of conversion to normal sinus rhythm with aggressive rhythm control strategies were observed compared to rate control [OR: 4.01 (95% CI: 1.30, 12.39), p = 0.02, I2 = 45%]. The odds of conversion to sinus rhythm with amiodarone only were no different from rate control [OR: 0.89 (95% CI: 0.22–3.60, p = 0.86, I2 = 82%)] (Figure 5). The test for subgroup differences, however, revealed no significant difference between the subgroups (p = 0.10).
Figure 5
Medication adverse effects (hypotension, bradycardia, and syncope)
All included studies (n = 894 patients) reported the incidence of medication-related hemodynamic adverse events (bradycardia, hypotension, and/or syncope). As a composite outcome, these adverse events were comparable between rhythm and rate control strategies [RR: 1.47 (95% CI: 0.55, 3.92), p = 0.44, I2 = 52%] (–). However, subgroup analysis revealed a higher proportion of hypotension, bradycardia, or syncope with an aggressive rhythm control strategy (30.8%, SD: 4.4%) compared with a rate control strategy (7.1%, SD: 2.8%). This increased incidence of hypotension, bradycardia, and/or syncope with an aggressive rhythm control strategy was statistically significant [RR: 3.05 (95% CI: 1.05–8.89); p = 0.04; I2 = 31%]. Conversely, amiodarone was not associated with a higher risk of hemodynamic adverse effects compared with rate control [RR: 0.50 (95% CI: 0.18, 1.34), p = 0.17, I2 = 0%] (Figure 6). There was a significant between-subgroup difference in effect sizes (p = 0.01).
Figure 6
Congestive heart failure/pulmonary edema
Compared with rate control interventions, rhythm control showed no significant difference in the rate of new-onset CHF/pulmonary edema across the three studies (n = 615) reporting this outcome [OR: 1.09 (95% CI: 0.46, 2.61, p = 0.84, I2 = 0)] (, , ).
Thromboembolic events
Only two studies had sufficient follow-up to report this outcome (n = 565). The pooled effect showed no difference in the rate of thromboembolic events between the two groups [OR: 1.48 (95% CI: 0.31, 7.16), p = 0.62, I2 = 23%] (, ). Given the limited number of studies, the pooled effect estimate should be interpreted with caution.
Mortality
Only three RCTs reported mortality rates for both strategies in our patient population (n = 723). The overall effect was similar; adopting either strategy conferred no difference in mortality [OR: 1.51 (95% CI: 0.64–3.54, p = 0.35, I2 = 0%)] (Figure 7) (, , ).
Figure 7
Assessment of publication bias
Contour-enhanced funnel plots were generated (Figure 8), and Egger's regression test was used to assess funnel plot asymmetry. There was no evidence of funnel plot asymmetry for hospital length of stay (p = 0.73), conversion to sinus rhythm (p = 0.60), hypotension/bradycardia/syncope (p = 0.48), congestive heart failure/pulmonary edema (p = 0.88), or mortality (p = 0.92). However, given the limited number of included studies, Egger's regression results should be interpreted with caution. Notably, our trial registry search did not identify any registered unpublished trials. This suggests a true paucity of studies examining patients with post-cardiac surgery atrial fibrillation rather than a “file drawer effect”.
Figure 8
Strength of evidence using the GRADE criteria
The strength of the evidence was very low regarding the effect of either strategy on hospital length of stay; however, only 4 RCTs met inclusion criteria for this outcome. Conversion to normal sinus rhythm was more likely with a rhythm control strategy, with moderate certainty of evidence. The evidence for medication-related adverse effects (hypotension, bradycardia, or syncope) with a rhythm control strategy was deemed of very low certainty, and this result was robust to sensitivity analysis. The effect of either strategy on the incidence of heart failure, thromboembolic events, and mortality was of low to very low certainty. Details of the GRADE assessment for all study outcomes are listed in Table 3.
Table 3
| Summary of findings: | ||||||
|---|---|---|---|---|---|---|
| Rhythm control compared to Rate control for Postoperative Atrial Fibrillation in Cardiac Surgery | ||||||
| Patient or population: Postoperative Atrial Fibrillation in Cardiac Surgery | ||||||
| Setting: RCTs | ||||||
| Intervention: Rhythm control | ||||||
| Comparison: Rate control |
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect (95% CI) | No of participants (studies) | Certainty of the evidence (GRADE) | Comments | |
|---|---|---|---|---|---|---|
| Risk with Rate control | Risk with Rhythm control | |||||
| Hospital length of stay (LOS) assessed with: Days | The mean hospital length of stay was 8.4 days | MD: 0.41 days lower (3.23 lower to 2.42 higher) | — | 765 (4 RCTs) | ⊕◯◯◯ VERY LOWa,b,c | Neither strategy is superior in reducing the LOS. More recent trials indicate a decreasing trend of LOS for rate control trials, perhaps reflective of acceptance of a higher target heart rate. |
| In-hospital conversion to sinus rhythm (NSR) assessed with: Assessment of heart rhythm | 823 per 1,000 | 902 per 1,000 (777–960) | OR: 1.98 (0.75–5.20) | 894 (8 RCTs) | ⊕⊕⊕◯ MODERATEa | Exclusion of the Kamali study yielded an OR of 3.66 (1.88–7.12). This implied that rhythm control therapy had a higher success rate at converting patients with POAF into sinus rhythm. |
| Thromboembolic events (TE) assessed with Stroke, pulmonary or venous embolism | 15 per 1,000 | 22 per 1,000 (5–98) | OR: 1.48 (0.31–7.16) | 565 (2 RCTs) | ⊕⊕◯◯ LOWa,c | No significant difference in the rate of thromboembolic events with either strategy. Rhythm control may present an effective option when anticoagulant use is associated with a higher risk of bleeding. |
| Heart failure (HF) assessed with: New CHF or pulmonary edema | 37 per 1,000 | 40 per 1,000 (17–91) | OR: 1.09 (0.46–2.61) | 615 (3 RCTs) | ⊕⊕◯◯ LOWa,c | No significant differences in the rate of heart failure with either strategy. |
| Drug adverse effects (AE) assessed with: Hypotension, bradycardia & syncope | 46 per 1,000 | 68 per 1,000 (25–180) | OR: 1.47 (0.55–3.92) | 894 (8 RCTs) | ⊕◯◯◯ VERY LOWa,c | No significant differences in drug-related adverse effects with either strategy. This result was robust to sensitivity analysis. |
| Mortality (Deaths) assessed with the number of deaths | 28 per 1,000 | 42 per 1,000 (18–93) | OR: 1.51 (0.64–3.54) | 723 (3 RCTs) | ⊕◯◯◯ VERY LOWa,c | As with studies in noncardiac surgery, rhythm control does not afford mortality benefit in cardiac surgery. |
Summary of findings & strength of evidence per the grading of recommendations, assessment, development and evaluations (GRADE) framework.
The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI, confidence interval; MD, mean difference; OR, odds ratio.
GRADE Working Group grades of evidence.
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect.
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect.
downgraded for risk of bias.
downgraded for inconsistency (unexplained heterogeneity).
downgraded for imprecision.
Discussion
Our systematic review and meta-analysis found no significant difference in hospital LOS between rhythm and rate control strategies. Aggressive rhythm control interventions (ibutilide, procainamide, propafenone, and electric cardioversion) were more likely to achieve in-hospital conversion to sinus rhythm than rate control, but they were also associated with a higher risk of medication-related side effects. Amiodarone in-hospital conversion rates were no different from rate control. Given the limited power and insufficient follow-up periods, the impact of either treatment strategy on long-term outcomes remains unclear.
Despite higher conversion rates to sinus rhythm with aggressive rhythm control, this did not translate into shorter hospital stays in the two studies that examined this outcome. The higher rates of medication-related hypotension, bradycardia, and syncope with aggressive rhythm control are biologically plausible, as agents such as procainamide and propafenone are known to have strong negative inotropic effects (). Our results therefore support close monitoring and possibly pre-emptive intervention when such aggressive rhythm control agents are used in the cardiac surgery population.
There was no evidence that amiodarone was superior to rate control in achieving normal sinus rhythm in postcardiac surgery patients. In addition, the use of amiodarone was not associated with a shorter hospital LOS compared to rate control; rather, there was a non-significant trend towards a longer hospital stay. Amiodarone has a much slower onset than other antiarrhythmics, which may render it an unfavorable choice for an often self-terminating arrhythmia (, ). The incidence of medication-related adverse events (composite of hypotension, bradycardia, and syncope) with amiodarone in this meta-analysis was no different from rate control. This was not surprising, as amiodarone has long been known to be a hemodynamically well-tolerated agent compared to other antiarrhythmics ().
Thromboembolic events were reported in only 2 studies, with both treatment strategies (rate vs. rhythm control) appearing comparable (, ). Given that rate control is almost always combined with anticoagulation, however, this lack of difference would be expected. Similarly, our review did not find a mortality benefit of any treatment strategy over the other. These results are consistent with other meta-analyses in the noncardiac surgery population (). Despite the pathophysiological triggers and underlying mechanisms of POAF in cardiac surgery patients being very distinct from the general atrial fibrillation (AF) population, it is quite possible that either management strategy would still carry no impact on mortality. In addition, we did not find any difference between the two strategies regarding the development of congestive heart failure or pulmonary edema as a complication of treatment and/or the AF (, ).
Our qualitative literature review supports an individualized patient approach to POAF treatment. Patients with hemodynamic instability clearly attributable to new-onset POAF may warrant synchronized electrical cardioversion; however, a higher recurrence rate remains problematic if the underlying drivers are not corrected concomitantly (). It is therefore important to expeditiously address any underlying drivers of atrial fibrillation and to consider antiarrhythmic therapy before cardioversion to achieve a more sustained effect. In patients with POAF who are hemodynamically stable, our review suggests that either a rate-control strategy or the use of amiodarone is a reasonable option until spontaneous reversion to sinus rhythm occurs. If spontaneous reversion does not occur within 48–72 h, anticoagulation should be initiated in the absence of contraindications (). Patients with relative or absolute contraindications to anticoagulation warrant an initial pharmacologic rhythm-control approach. If refractory, electrical cardioversion at the 48-hour mark (before left atrial clot formation) may be reasonable. Given the high incidence of recurrence, continued pharmacologic therapy post-cardioversion may be considered ().
Preventive strategies in the preoperative setting may also help reduce the incidence of POAF. These include optimization of electrolyte imbalances, continuation or initiation of beta-blockers where appropriate, and the use of prophylactic antiarrhythmic agents in high-risk patients. In addition, addressing modifiable risk factors such as volume status and systemic inflammation may further reduce the likelihood of POAF development. Future studies should explore the integration of such preventive strategies with postoperative management approaches (, ).
Although there was no evidence that either treatment strategy was superior, this meta-analysis calls for further research with well-designed randomized trials focused on clinically meaningful outcomes (hospital length of stay, rates of thromboembolic events, and mortality). Any further research in this domain must include longer follow-up periods. In addition, more aggressive rhythm-control strategies (such as propafenone, ibutilide, and cardioversion) showed some promise in terms of conversion efficacy and warrant further investigation.
Our study has several strengths. It is the first systematic review and meta-analysis to examine the short- and long-term effects of either treatment strategy for new-onset POAF. We employed a comprehensive search strategy across multiple databases. We included only data from RCTs. We followed a robust methodology, used subgroup analyses to assess heterogeneity where appropriate, and graded the strength of our findings. Finally, our exploratory evaluation of secondary outcomes was hypothesis-generating. Our analysis aimed to avoid overstating or inflating any possible type I error. These factors emphasize the validity of our results.
However, our current work has a few limitations. First, our search retrieved a small number of RCTs (k = 8), highlighting the paucity of clinical trials in this domain. Moreover, only two RCTs had sample sizes exceeding 100 patients (, ). Second, the classification of “aggressive” vs. “conventional” rhythm control strategies was developed post hoc to assess potential sources of heterogeneity and should therefore be viewed as exploratory. This approach was necessitated by considerable clinical heterogeneity in treatment approaches among the included studies. The lack of patient-level data precluded the use of more powerful analytic techniques. Additionally, effect estimates for some outcomes (e.g., thromboembolic events) were derived from a very limited number of studies. Producing robust effect estimates and exploring heterogeneity under such constraints may not be feasible. Similarly, variability in follow-up duration across trials in our meta-analysis may limit the generalizability of important outcomes such as mortality. Furthermore, several subgroup analyses were based on a small number of studies, thereby reducing statistical power and limiting the reliability of subgroup-specific conclusions. Subgroup analysis findings should therefore be considered hypothesis-generating.
Conclusion
This meta-analysis found no evidence that a rhythm-control strategy reduced hospital LOS compared with a rate-control strategy in postcardiac-surgery patients. Although an aggressive rhythm-control strategy achieved higher rates of conversion to sinus rhythm, it was associated with a higher risk of medication-related hypotension and did not translate into fewer complications (thromboembolic events, heart failure, and mortality). Consequently, rate control with anticoagulation appears to be as effective as rhythm control for managing POAF after cardiac surgery.
Statements
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.
Author contributions
MD: Conceptualization, Investigation, Writing – original draft, Writing – review & editing. MT: Investigation, Methodology, Software, Writing – original draft. CT: Methodology, Software, Validation, Writing – original draft. HS: Formal analysis, Resources, Writing – original draft. HA: Formal analysis, Resources, Validation, Writing – original draft, Writing – review & editing. SM: Software, Validation, Visualization, Writing – original draft. SK: Investigation, Writing – original draft. WA: Resources, Visualization, Writing – original draft. AH: Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcvm.2026.1820175/full#supplementary-material
Abbreviations
POAF, postoperative atrial fibrillation; CHF, congestive heart failure; CABG, coronary artery bypass grafting; CI, confidence interval; GRADE, grading of recommendations, assessment, development and evaluations; NSR, normal sinus rhythm; MD, mean difference; PICO, population, intervention, comparator, outcome; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; PROSPERO, International Prospective Register of Systematic Reviews; OR, odds ratio; RoB, risk of bias; RCTs, randomized controlled trials.
References
1.
GreenbergJWLancasterTSSchuesslerRBMelbySJ. Postoperative atrial fibrillation following cardiac surgery: a persistent complication. Eur J Cardiothorac Surg. (2017) 52(4):665–72. 10.1093/ejcts/ezx039
2.
CreswellLLSchuesslerRBRosenbloomMCoxJL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg. (1993) 56(3):539–49. 10.1016/0003-4975(93)90894-N
3.
MathewJP. A multicenter risk Index for atrial fibrillation after cardiac surgery. JAMA. (2004) 291(14):1720. 10.1001/jama.291.14.1720
4.
ArankiSFShawDPAdamsDHRizzoRJCouperGSVanderVlietMet al. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation. (1996) 94(3):390–7. 10.1161/01.CIR.94.3.390
5.
HashimotoKIlstrupDMSchaffHV. Influence of clinical and hemodynamic variables on risk of supraventricular tachycardia after coronary artery bypass. J Thorac Cardiovasc Surg. (1991) 101(1):56–65.
6.
MaiselWHRawnJDStevensonWG. Atrial fibrillation after cardiac surgery. Ann Intern Med. (2001) 135(12):1061–73. 10.7326/0003-4819-135-12-200112180-00010
7.
DunningJTreasureTVersteeghMNashefSAM, EACTS Audit and Guidelines Committee. Guidelines on the prevention and management of de novo atrial fibrillation after cardiac and thoracic surgery. Eur J Cardiothorac Surg. (2006) 30(6):852–72. 10.1016/j.ejcts.2006.09.003
8.
AhmedMBelley-CotéEPQiuYBelesiotisPTaoBWolfAet al. Rhythm vs. Rate control in patients with postoperative atrial fibrillation after cardiac surgery: a systematic review and meta-analysis. J Clin Med. (2023) 12(13):4534. 10.3390/jcm12134534
9.
MoherDLiberatiATetzlaffJAltmanDG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. (2009) 6(7):e1000097. 10.1371/journal.pmed.1000097
10.
DrevonDFursaSRMalcolmAL. Intercoder reliability and validity of WebPlotDigitizer in extracting graphed data. Behav Modif. (2017) 41(2):323–39. 10.1177/0145445516673998
11.
WanXWangWLiuJTongT. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. (2014) 14:135. 10.1186/1471-2288-14-135
12.
DerSimonianRLairdN. Meta-analysis in clinical trials. Control Clin Trials. (1986) 7(3):177–88. 10.1016/0197-2456(86)90046-2
13.
HigginsJPTThompsonSG. Quantifying heterogeneity in a meta-analysis. Stat Med. (2002) 21(11):1539–58. 10.1002/sim.1186
14.
DuvalSTweedieR. Trim and fill: a simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics. (2000) 56(2):455–63. 10.1111/j.0006-341x.2000.00455.x
15.
CochraneADSiddinsMRosenfeldtFLSalamonsenRMcConaghyLMarascoSet al. A comparison of amiodarone and digoxin for treatment of supraventricular arrhythmias after cardiac surgery. Eur J Cardiothorac Surg. (1994) 8(4):194–8. 10.1016/1010-7940(94)90114-7
16.
GillinovAMBagiellaEMoskowitzAJRaitenJMGrohMABowdishMEet al. Rate control versus rhythm control for atrial fibrillation after cardiac surgery. N Engl J Med. (2016) 374(20):1911–21. 10.1056/NEJMoa1602002
17.
HjelmsE. Procainamide conversion of acute atrial fibrillation after open-heart surgery compared with digoxin treatment. Scand J Thorac Cardiovasc Surg. (1992) 26(3):193–6. 10.3109/14017439209099077
18.
KamaliASanatkarASharifiMMoshirE. Evaluation of amiodarone versus metoprolol in treating atrial fibrillation after coronary artery bypass grafting. Interv Med Appl Sci. (2017) 9(2):51–5. 10.1556/1646.9.2017.2.11
19.
LeeJKKleinGJKrahnADYeeRZarnkeKSimpsonCet al. Rate-control versus conversion strategy in postoperative atrial fibrillation: a prospective, randomized pilot study. Am Heart J. (2000) 140(6):871–7. 10.1067/mhj.2000.111104
20.
SoucierRSilvermanDAbordoMJaagosildPAbioseAMadhusoodananKPet al. Propafenone versus ibutilide for post operative atrial fibrillation following cardiac surgery: neither strategy improves outcomes compared to rate control alone (the PIPAF study). Med Sci Monit. (2003) 9(3):PI19–23.
21.
WafaSSWardDEParkerDJCammAJ. Efficacy of flecainide acetate for atrial arrhythmias following coronary artery bypass grafting. Am J Cardiol. (1989) 63(15):1058–64. 10.1016/0002-9149(89)90078-7
22.
CampbellTJGavaghanTPMorganJJ. Intravenous sotalol for the treatment of atrial fibrillation and flutter after cardiopulmonary bypass. Comparison with disopyramide and digoxin in a randomised trial. Br Heart J. (1985) 54(1):86–90. 10.1136/hrt.54.1.86
23.
Funck-BrentanoCKroemerHKLeeJTRodenDM. Propafenone. N Engl J Med. (1990) 322(8):518–25. 10.1056/NEJM199002223220806
24.
ClemoHFWoodMAGilliganDMEllenbogenKA. Intravenous amiodarone for acute heart rate control in the critically ill patient with atrial tachyarrhythmias. Am J Cardiol. (1998) 81(5):594–8. 10.1016/s0002-9149(97)00962-4
25.
SamuelsLEHolmesECSamuelsFL. Selective use of amiodarone and early cardioversion for postoperative atrial fibrillation. Ann Thorac Surg. (2005) 79(1):113–6. 10.1016/j.athoracsur.2004.06.049
26.
LarbuissonRVennemanIStielsB. The efficacy and safety of intravenous propafenone versus intravenous amiodarone in the conversion of atrial fibrillation or flutter after cardiac surgery. J Cardiothorac Vasc Anesth. (1996) 10(2):229–34. 10.1016/s1053-0770(96)80243-6
27.
SethiNJFeinbergJNielsenEESafiSGluudCJakobsenJC. The effects of rhythm control strategies versus rate control strategies for atrial fibrillation and atrial flutter: a systematic review with meta-analysis and trial sequential analysis. PLoS One. (2017) 12(10):e0186856. 10.1371/journal.pone.0186856
28.
BidarEBramerSMaesenBMaessenJGSchottenU. Post-operative atrial fibrillation—pathophysiology, treatment and prevention. J Atr Fibrillation. (2013) 5(6):781. 10.4022/jafib.781
29.
FrendlGSodicksonACChungMKWaldoALGershBJTisdaleJEet 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(3):e153–193. 10.1016/j.jtcvs.2014.06.036
30.
MitchellLB, CCS Atrial fibrillation guidelines committee. Canadian Cardiovascular society atrial fibrillation guidelines 2010: prevention and treatment of atrial fibrillation following cardiac surgery. Can J Cardiol. (2011) 27(1):91–7. 10.1016/j.cjca.2010.11.005
31.
CrystalEConnollySJSleikKGingerTJYusufS. Interventions on prevention of postoperative atrial fibrillation in patients undergoing heart surgery: a meta-analysis. Circulation. (2002) 106(1):75–80. 10.1161/01.cir.0000021113.44111.3e
Summary
Keywords
cardiac surgery, meta-analysis, new-onset atrial fibrillation, postoperative atrial fibrillation, rate control, rhythm control, systematic review
Citation
Dairi MS, Tarabzoni M, Tarola C, Sehmbi H, Alwafi H, Alghamdi SM, Khan S, Alotaibi WT and Hegazy AF (2026) Rhythm vs. rate control for treatment of postoperative atrial fibrillation after cardiac surgery: a systematic review and meta-analysis of randomized controlled trials. Front. Cardiovasc. Med. 13:1820175. doi: 10.3389/fcvm.2026.1820175
Received
28 February 2026
Revised
16 June 2026
Accepted
29 June 2026
Published
14 July 2026
Volume
13 - 2026
Edited by
Hendrik Tevaearai Stahel, University Hospital of Bern, Switzerland
Reviewed by
Massimo Baudo, Lankenau Institute for Medical Research, United States
Fan Maitri Aldian, Airlangga University, Indonesia
Updates
Copyright
© 2026 Dairi, Tarabzoni, Tarola, Sehmbi, Alwafi, Alghamdi, Khan, Alotaibi and Hegazy.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ahmed F. Hegazy ahmed.hegazy@mail.harvard.edu
ORCID Sariya Khan orcid.org/0009-0003-9809-872X Waleed Talal Alotaibi orcid.org/0009-0002-0927-414X Ahmed F. Hegazy orcid.org/0000-0001-9998-8968
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.