Dexmedetomidine for the prevention of postoperative atrial fibrillation: A systematic review and meta-analysis

Abstract

Objective

Postoperative atrial fibrillation, a prevalent complication following cardiac surgery, significantly increases patient morbidity and healthcare expenditures. Although dexmedetomidine has been proposed as a potential strategy to mitigate the risk of postoperative atrial fibrillation risk, existing evidence from randomized controlled trials has shown conflicting results.

Methods

To evaluate the efficacy of perioperative dexmedetomidine in adult cardiac surgery patients, we conducted a systematic search across four major electronic databases: (a) PubMed; (b) Embase; (c) the Cochrane Library; and (d) Web of Science. Our search sought all relevant randomized controlled trials on this specific intervention and patient population. We utilized random-effects models to synthesize treatment effects, expressed as risk ratios with 95% confidence intervals. Furthermore, trial sequential analysis and the Grading of Recommendations Assessment, Development, and Evaluation framework were applied to determine the robustness and certainty of the evidence, respectively.

Results

Our analysis included 22 randomized controlled trials, involving a total of 4063 patients. Dexmedetomidine administration significantly lowered the risk of postoperative atrial fibrillation risk compared with control groups (risk ratio 0.80, 95% confidence interval: 0.70–0.93; I² = 28%). Subgroup analysis specifically indicated a significant reduction in postoperative atrial fibrillation risk in placebo-controlled trials, while no notable difference was observed when dexmedetomidine was compared with other active sedative agents. In addition, dexmedetomidine was associated with a reduced risk of postoperative delirium (risk ratio 0.62, 95% confidence interval: 0.44–0.87; I² = 56%).

Conclusions

Perioperative dexmedetomidine appears to reduce the risk of both postoperative atrial fibrillation risk and delirium in adult cardiac surgery patients, particularly compared with placebo. However, its efficacy relative to alternative active prophylactic agents remains unclear, necessitating rigorous comparative effectiveness trials.

Introduction

Postoperative atrial fibrillation (POAF) is the most common arrhythmic complication after cardiac surgery, affecting approximately 20%–50% of patients, depending on surgical type and patient characteristics.^1,2 Clinically, the onset of POAF is associated with prolonged intensive care unit (ICU) stays, an elevated risk of stroke, and a substantial financial burden. Furthermore, the subsequent initiation of anticoagulant therapy to mitigate stroke risk is associated with an increased risk of postoperative hemorrhage, which may complicate recovery and affect clinical outcomes. ³ Although the perioperative management of POAF has improved, clinicians still face a scarcity of preventive strategies that are both effective and well-tolerated, underscoring the need to explore novel interventions.^4,5 Consequently, identifying safe and effective prophylactic strategies to reduce the incidence of POAF is a priority in perioperative medicine.

Dexmedetomidine is a highly selective α₂-adrenoceptor agonist with sedative, sympatholytic, and anti-inflammatory properties. ⁶ Its unique pharmacologic profile suggests that it modulates postoperative outcomes. In addition to its primary sedative applications in intensive care, preclinical and clinical studies suggest that dexmedetomidine reduces sympathetic overactivity and mitigates systemic inflammation, thereby contributing to POAF pathogenesis.^7,8

Although several randomized controlled trials (RCTs) and meta-analyses have investigated the role of dexmedetomidine in POAF prevention, conclusions remain inconsistent.^9–11 Some studies have reported a significant reduction in POAF incidence,^10,12 whereas others have reported no significant benefit.^13,14 Recent guidelines and reviews have highlighted the need for updated synthesis of evidence, particularly given recently published RCTs and the need to evaluate outcomes such as postoperative delirium. ¹³

To address these evidence gaps, we conducted a systematic review and meta-analysis of RCTs to evaluate the efficacy of dexmedetomidine in preventing POAF in adult patients undergoing cardiac surgery. We incorporated trial sequential analysis (TSA) to control for random error and assess the robustness of the findings, and applied the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework to evaluate the certainty of evidence.

Methods

Protocol and registration

This systematic review and meta-analysis was performed following the guidelines set forth in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement 2020, ensuring comprehensive and transparent reporting. ¹⁵ To maintain research integrity and prevent duplication, our detailed study protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO, Registration No. CRD420251243477) prior to commencing any data extraction or analysis.

Search strategy and selection criteria

To identify all relevant studies, a comprehensive and systematic literature search was conducted across four prominent electronic databases: (a) PubMed; (b) Embase; (c) the Cochrane Library; and (d) Web of Science. The search encompassed all records from their inception dates up to 30 November 2025. Our search strategy combined controlled vocabulary terms, specifically Medical Subject Headings (MeSH) in PubMed and Emtree terms in Embase, with carefully selected free-text keywords pertinent to “dexmedetomidine,” “cardiac surgery,” “anesthesia,” and “atrial fibrillation.” This approach was designed to maximize search sensitivity while maintaining specificity. No restrictions were imposed on language, thereby broadening the scope of potentially eligible studies. The precise search syntax and comprehensive terms employed for each database are detailed in the Supplementary Materials to facilitate reproducibility.

We established the inclusion criteria for RCTs based on the following:

Population (P). Studies must have enrolled adult patients (aged ≥18 years) undergoing cardiac surgery under general anesthesia.

Intervention (I) and Comparator (C). The trials were required to compare perioperatively administered intravenous dexmedetomidine against either a placebo, standard care, or another active sedative or pharmacologically active agent.

Outcome (O). The primary O of interest, POAF, must have been reported.

Conversely, trials were systematically excluded if they involved pediatric Ps, non-cardiac surgical procedures, or lacked a clearly defined control group for comparative analysis.

Study selection and data extraction

The selection of eligible studies was performed independently by two reviewers. Initially, titles and abstracts of all retrieved records were screened to identify potentially relevant articles. Subsequently, the full texts of these candidate articles were retrieved and independently assessed against the predefined inclusion and exclusion criteria. Any discrepancies or disagreements arising during this critical screening phase were resolved first through discussion to reach a consensus between the two reviewers. If consensus could not be reached, a third independent reviewer was consulted for adjudication.

Data from the included studies were extracted in duplicate by the same two reviewers, using a standardized, pre-piloted data extraction form to ensure consistency and accuracy. The extracted variables encompassed a broad range of study characteristics, including the first author's name, publication year, detailed patient demographics, specific type of cardiac surgery performed, precise dexmedetomidine administration protocols (dosage and timing), characteristics of the C groups, and all reported O measures. The primary O consistently evaluated was the incidence of POAF, as defined by the original trials themselves. Secondary Os, such as the incidence of postoperative delirium, were also systematically collected. In instances where critical data were missing or unclear, we endeavored to contact the corresponding authors of the original studies to obtain the necessary information.

Risk of bias and certainty of evidence assessment

To ensure the methodological quality and reliability of our synthesis, the risk of bias for each individual included trial was independently assessed by two reviewers. This evaluation was performed using the Cochrane Risk of Bias 2.0 tool, which systematically appraises bias arising from the randomization process, deviations from intended Is, missing O data, O measurement, and selective reporting. Each study was subsequently categorized into one of three risk levels: (a) low risk of bias; (b) some concerns; or (c) high risk of bias. Beyond individual study bias, the overall certainty of evidence for the primary O was assessed using the GRADE framework. This comprehensive assessment considered various critical domains, including the risk of bias, consistency across studies, indirectness of evidence, imprecision of effect estimates, and the potential for publication bias.

Statistical analyses

All statistical analyses were performed using Review Manager (RevMan) version 5.3 and Stata version 17.0. For all dichotomous Os, treatment effects were quantified as risk ratios (RRs) with their corresponding 95% confidence intervals (CIs). Recognizing the inherent variability in clinical and methodological characteristics across different trials (such as patient Ps, surgical procedures, and dexmedetomidine administration protocols), we opted for a random-effects model for all meta-analyses. To quantify the degree of inter-study variance, the I² statistic was employed, with higher values indicating greater heterogeneity. Pre-specified subgroup analyses were conducted to explore whether the effect of dexmedetomidine on POAF varied based on the type of C (placebo versus other active agents). To test the robustness of our primary O, sensitivity analyses were systematically performed by sequentially excluding individual studies and re-evaluating the pooled effect. Publication bias, a critical concern in meta-analysis, was assessed through visual inspection of funnel plots for asymmetry. For Os with more than 10 included studies, Egger's test was additionally utilized to statistically evaluate for significant small-study effects.

To guard against the risk of random errors and determine if the cumulative evidence had reached a conclusive threshold, TSA was performed using the TSA software (version 0.9.5.10 Beta). The required information size for TSA was calculated assuming a relative risk reduction of 19.3%, two-sided alpha (α) of 5%, and statistical power of 80%.

Results

Study selection and characteristics

Our initial comprehensive database search identified 1872 unique records. Following a deduplication process, 1245 titles and abstracts underwent initial screening. Of these, 38 full-text articles were deemed potentially eligible and subjected to detailed review. Ultimately, 22 RCTs satisfied our predefined inclusion criteria and were incorporated into the quantitative synthesis.^{7,9,10,12–14,16–31} A complete overview of the study selection process, adhering to the PRISMA guidelines, is graphically depicted in Figure 1.

Figure 1.

PRISMA 2020 flow diagram of the study selection process. A total of 1872 records were identified from four databases. After deduplication, 1245 records were screened; 38 full-text articles were assessed for eligibility, and 22 RCTs were included in the quantitative synthesis.

The 22 included RCTs spanned a publication period from 1997 to 2025, collectively enrolling 4063 participants. Individual study sample sizes exhibited considerable variation, ranging from 20 to 794 patients. Most of these investigations involved patients undergoing coronary artery bypass grafting, valve surgery, or a combination of cardiac procedures. Dexmedetomidine dosing regimens and administration timings varied significantly across the included trials. Control groups primarily consisted of either placebo (often normal saline) or other sedative agents, with propofol being the most frequently employed C. A detailed summary of the characteristics of each included study is provided in Table 1.

Table 1.

Characteristics of the included randomized controlled trials.

Study	Sample size (DEX/control)	ASA	Type of surgery	Dexmedetomidine regimen	Control group	Comparator type	POAF assessment
Alhadidy, Alansary and Elghareeb 202512	124 / 124	II–III	CABG	0.7 µg/kg/h	Placebo (normal saline)	Placebo	New-onset AF within 7 days postoperatively
Balkanay et al. 201516	60 / 28	NR	CABG	0.04 µg/kg/h	Placebo (normal saline)	Placebo	During postoperative hospitalization (5 days)
Corbett et al. 200517	43 / 46	NR	CABG	1.0 µg/kg loading dose, then 0.4 µg/kg/h	Propofol	Active drug	Continuous ECG monitoring during ICU stay
Djaiani et al. 201618	91 / 92	NR	Mixed cardiac surgery	0.4 µg/kg loading dose, then 0.2–0.7 µg/kg/h	Propofol	Active drug	Continuous ECG monitoring during ICU stay
Dong et al. 202119	251 / 257	NR	Mixed cardiac surgery	0.5 µg/kg/h	Placebo (normal saline)	Placebo	NR
Herr, Sum-Ping and England 200320	148 / 147	NR	CABG	1.0 µg/kg loading dose, then 0.4 µg/kg/h	Propofol	Active drug	Continuous ECG monitoring during ICU stay (24 h)
Jadhav, Mankhair and Chakole 202321	10 / 10	III–IV	CABG	0.5 µg/kg/h	Magnesium sulfate	Active drug	Intraoperative and postoperative AF
Jalonen et al. 19977	40 / 40	NR	CABG	50 ng/kg/min	Placebo (normal saline)	Placebo	New-onset AF within 6 days postoperatively
Karaman et al. 2015222	31 / 33	< IV	CABG	0.2–1.0 µg/kg/h	Propofol	Active drug	Within 6 h after surgery
Kim et al. 202123	26 / 25	NR	Aortic surgery	1.0 µg/kg loading dose, then 0.5 µg/kg/h	Placebo (normal saline)	Placebo	New-onset atrial fibrillation
Liu et al. 201624	29 / 32	II–III	Valve surgery	0.2–1.5 µg/kg/h	Propofol	Active drug	Continuous ECG monitoring during ICU stay
Liu et al. 20169	44 / 44	II–III	Mixed cardiac surgery	≤1.5 µg/kg/h	Propofol	Active drug	Within 96 h postoperatively
Park et al. 201425	67 / 75	NR	Mixed cardiac surgery	0.5 µg/kg loading dose, then 0.2–0.8 µg/kg/h	Remifentanil	Active drug	New-onset AF within 3 days postoperatively
Preveden et al. 202326	60 / 60	NR	Mixed cardiac surgery	0.2–0.7 µg/kg/h	Propofol	Active drug	Continuous ECG monitoring during ICU stay
Qu et al. 202313	188 / 206	NR	Mixed cardiac surgery	1.0 µg/kg (max 80 µg)	Placebo (normal saline)	Placebo	NR
Ren et al. 201327	81 / 81	NR	CABG	0.2–0.5 µg/kg/h	Placebo (normal saline)	Placebo	New-onset AF within 3 days postoperatively
Safaric et al. 202514	40 / 20	NR	Aortic valve replacement	0.5 or 1.0 µg/kg/h	Placebo (normal saline)	Placebo	Continuous ECG monitoring during ICU stay
Shehabi et al. 200928	152 / 147	NR	Mixed cardiac surgery	0.1–0.7 µg/kg/h	Morphine	Active drug	New-onset AF within 12 days postoperatively
Soltani et al. 201710	38 / 38	NR	CABG	0.5 µg/kg/h	Placebo (normal saline)	Placebo	NR
Subramaniam et al. 201929	59 / 61	NR	CABG	0.5–1.0 µg/kg	Propofol	Active drug	New-onset AF within 10 days postoperatively
Turan et al. 202030	398 / 396	I–V	Mixed cardiac surgery	0.2 µg/kg/h	Placebo (normal saline)	Placebo	New-onset AF within 5 postoperative days
Zi et al. 202031	62 / 61	I–IV	CABG	0.2–1.0 µg/kg/h	Propofol	Active drug	New-onset AF during hospitalization

ASA: American Society of Anesthesiologists; CABG: coronary artery bypass grafting; DEX: dexmedetomidine; POAF: postoperative atrial fibrillation; NR: not reported; ECG: electrocardiography.

Risk of bias assessment

Adhering to the Cochrane Risk of Bias 2.0 tool methodology, our assessment revealed that nine studies were deemed to have an overall low risk of bias. Ten studies presented some concerns regarding bias, primarily attributable to insufficient clarity in the randomization process or incomplete reporting of O data. Three studies were categorized as having a high risk of bias, largely due to discernible deviations from their intended Is. A comprehensive visual summary of the risk of bias assessment for all included studies is presented in Figure 2.

Figure 2.

Risk of bias summary of the 22 included randomized controlled trials assessed using the Cochrane risk of bias 2.0 tool.

Primary O: incidence of POAF

The pooled analysis of all 22 RCTs showed a statistically significant reduction in the incidence of POAF within the dexmedetomidine group compared with that in the control groups (RR = 0.80, 95% CI: 0.70–0.93, p = 0.003), with a low observed level of heterogeneity (I² = 28%, Figure 3). Further stratified analysis demonstrated a significant reduction in POAF in trials where dexmedetomidine was compared against placebo (RR = 0.76, 95% CI: 0.64–0.92; I² = 23%). Conversely, no significant difference in the POAF incidence was detected when dexmedetomidine was benchmarked against other active sedative agents (RR = 0.91, 95% CI: 0.73–1.12; I² = 20%) (Figure 4).

Figure 3.

Forest plot of the meta-analysis for the primary outcome (postoperative atrial fibrillation (POAF)). Risk ratios (RRs) with 95% confidence intervals (CIs) are shown for each of the 22 included studies. The diamond represents the pooled RR (0.80, 95% CI: 0.70–0.93; I² = 28%), indicating a statistically significant reduction in the POAF with dexmedetomidine compared with the control groups, with low heterogeneity across studies.

Figure 4.

Subgroup analysis of POAF incidence stratified by C type. The placebo controlled subgroup showed a significant reduction in the POAF with dexmedetomidine (RR 0.76, 95% CI: 0.64–0.92; I² = 23%). In contrast, the subgroup comparing dexmedetomidine with other active sedative agents did not demonstrate a significant difference (RR 0.91, 95% CI: 0.73–1.12; I² = 20%).

POAF: postoperative atrial fibrillation

Sensitivity analyses confirmed the stability of this finding; systematically excluding individual studies did not materially alter the pooled effect estimate (Figure 5).

Figure 5.

Sensitivity analysis for the primary outcome (POAF). The forest plot shows the pooled risk ratio after sequentially omitting each individual study. The horizontal lines represent the 95% confidence intervals of the pooled estimate after removal of the corresponding study.

POAF: postoperative atrial fibrillation

Secondary O

The incidence of postoperative delirium was reported in 13 of the included trials. Dexmedetomidine administration was associated with a significantly lower risk of postoperative delirium compared with control groups (RR = 0.62, 95% CI: 0.44–0.87; p  < 0.05). However, this finding was accompanied with moderate heterogeneity (I² = 56%) (Figure 6). This observed variability likely reflects differences in delirium definitions, assessment methodologies, and the timing of O evaluation across the diverse trials.

Figure 6.

Forest plot of the secondary outcome (postoperative delirium) from 13 trials that reported this outcome. Dexmedetomidine was associated with a significantly lower risk of delirium compared with control groups.

TSA and publication bias

For the primary O of POAF incidence, the TSA revealed that the cumulative Z-curve successfully crossed both conventional significance boundary and trial sequential monitoring boundary, although the cumulative information size did not reach the precalculated required threshold of 4213 participants (Figure 7). These results indicate that the evidence supporting the primary O is robust and unlikely to be reversed by future studies, effectively protecting against random error. Furthermore, visual inspection of the funnel plot provided no indication of substantial asymmetry (Figure 8), and Egger's test, employed for Os with more than 10 studies, did not detect significant small-study effects (p = 0.53), suggesting a low probability of publication bias.

Figure 7.

Trial sequential analysis (TSA) for the primary outcome (POAF). The cumulative Z-curve crossed both the conventional significance boundary and the trial sequential monitoring boundary before reaching the required information size.

POAF: postoperative atrial fibrillation

Figure 8.

Funnel plot for publication bias assessment of the primary outcome (POAF). Each dot represents an individual study, with the effect size (log risk ratio) on the horizontal axis and its standard error on the vertical axis.

POAF: postoperative atrial fibrillation

Certainty of evidence (GRADE)

Based on the comprehensive GRADE assessment, the overall certainty of evidence for the primary O (POAF incidence) was rated as moderate. This moderate rating was primarily due to concerns regarding the risk of bias identified across several included studies (Table 2), which led to a single-level downgrading.

Table 2.

GRADE summary of findings.

Outcome	No. of studies (participants)	Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bias	Overall certainty (GRADE)	Summary of effect
Postoperative atrial fibrillation	22 RCTs	Serious a	Not serious	Not serious	Not serious	Not serious	Moderate ⊕⊕⊕◯	RR 0.80 (95% CI: 0.70–0.93)
Postoperative delirium	13 RCTs	Serious b	Serious	Not serious	Not serious	Not serious	Low ⊕⊕◯◯	RR 0.62 (95% CI: 0.44–0.87)

^aDowngraded by one level due to concerns regarding risk of bias in several included trials, primarily related to unclear randomization procedures or deviations from intended interventions.

^bDowngraded by one level due to risk of bias and by one level due to substantial heterogeneity among included studies.

CI: confidence interval; RCT: randomized controlled trial; RR: risk ratio; GRADE: Grading of Recommendations Assessment, Development, and Evaluation.

Discussion

This meta-analysis indicates that perioperative dexmedetomidine may reduce post-cardiac surgery complications. Our synthesis of the current evidence suggests its efficacy not only in reducing the incidence of POAF but also in attenuating postoperative delirium, with these benefits being most pronounced when evaluated against placebo.

Several previous meta-analyses have examined the association between dexmedetomidine and POAF; however, their findings were limited by smaller sample sizes, lack of TSA, or the absence of GRADE certainty assessment.^11,32 For instance, a 2020 meta-analysis by Liu et al. included 13 RCTs and reported a beneficial effect of dexmedetomidine on POAF; however, they did not employ the TSA to evaluate the conclusiveness of the evidence. ¹¹ Our study updates the existing evidence by incorporating 22 RCTs, including some trials published after 2020,^{14,21,30,31,33} thereby providing more precise effect estimates. Furthermore, we applied both TSA and the GRADE framework to assess the robustness and certainty of the pooled results, which most previous meta-analyses have not performed. The subgroup analysis according to C type (placebo vs. active agents) also offers a more nuanced interpretation than prior syntheses. Collectively, these methodological enhancements strengthen the validity of our conclusions and help resolve previously conflicting evidence.

Subgroup analysis showed that the effect of dexmedetomidine on POAF differed by C type. Although its benefit was evident when compared with placebo, this advantage was not observed when benchmarked against other active sedative agents, particularly propofol. This pattern suggests that the observed reduction in POAF does not stem from a uniquely superior anti-arrhythmic property of dexmedetomidine itself, but reflects the advantages it offers over more passive or less optimally managed sedation strategies. This interpretation is consistent with prior meta-analyses comparing dexmedetomidine and propofol, which reported reductions in ventricular arrhythmias without a consistent effect on POAF incidence. ³² Consequently, our findings suggest that dexmedetomidine is more effective as part of a comprehensive perioperative strategy than as a standalone prophylactic agent for POAF. This nuance is vital for guiding clinical practice.

The observed reduction in POAF likely stems from the multifactorial pharmacological properties of dexmedetomidine. First, its central α₂-adrenergic agonism suppresses sympathetic outflow and reduces circulating catecholamine levels. This sympatholytic effect mitigates catecholamine-driven atrial excitability and triggered activity, representing a plausible mechanism for the prevention of perioperative tachydysrhythmias. ³⁴ Second, the agent's well-documented anti-inflammatory effects can mitigate the systemic inflammatory response frequently associated with cardiopulmonary bypass and surgical stress, both of which are critical contributors to atrial electrical instability. ³⁵ Third, experimental and clinical data suggest that dexmedetomidine may exert direct electrophysiological effects on the atrial myocardium itself, further contributing to enhanced electrical stability. ³³ Together, these mechanisms may explain the reduction in POAF observed in this analysis.

The reduction in postoperative delirium is another finding with clinical relevance. Delirium and POAF are known to share overlapping pathophysiological pathways, which include excessive sympathetic activation, neuroinflammation, and oxidative stress.^36,37 Therefore, the dual benefit of dexmedetomidine aligns seamlessly with contemporary perioperative strategies that advocate for opioid-sparing anesthesia and judicious avoidance of delirium-promoting sedatives. These findings support the use of dexmedetomidine in selected high-risk cardiac surgery patients,^38,39 which may improve recovery and cognitive Os.

Despite these findings, certain limitations warrant careful consideration when interpreting our results. First, clinical heterogeneity persisted across the included trials, notably concerning variations in dexmedetomidine dosing regimens, timing of administration, specific types of cardiac surgery performed, and even the precise definitions of POAF. Although statistical methodologies, including random-effects models and subgroup analyses, were used to address these variations, the influence of residual, unmeasured heterogeneity cannot be entirely excluded. This suggests that the observed average effect does not apply to all clinical settings. Second, most of the included studies were single-center trials, which inherently limits the generalizability of our findings to broader, multi-institutional settings. Finally, although no significant publication bias was detected, the possibility of unpublished negative studies cannot be excluded.

Conclusion

Perioperative dexmedetomidine reduces the risk of both POAF and delirium in adult cardiac surgery patients, particularly when compared with placebo. However, its efficacy relative to alternative active prophylactic agents remains unclear, and further comparative effectiveness trials are needed.