Sage Journals: Discover world-class research

Abstract

French

Background:

Despite atrial fibrillation affecting nearly 25% of patients receiving dialysis, conflicting study findings and guideline recommendations continue to make individual treatment decisions challenging.

Objective:

The objective was to examine the efficacy and safety of stroke-prevention interventions in patients with atrial fibrillation receiving dialysis.

Design:

A systematic review of randomized controlled trials and observational studies.

Patients:

Adult patients with non-valvular atrial fibrillation receiving any type of dialysis.

Measurements:

Ischemic strokes or systemic embolism, major bleeding, and all-cause mortality.

Methods:

We searched Medline, Embase, Cochrane Central Register of Controlled Trials, and the gray literature from inception to February 16, 2023. We selected all studies comparing antiplatelet, anticoagulant agents or left atrial appendage closure to placebo, no treatment, or an active comparator (most often vitamin K antagonists). Two reviewers independently screened and included the studies, extracted the data and assessed the risk of bias using the Cochrane Risk of Bias tool (version 1) and the Newcastle-Ottawa Scale. When randomized controlled trials were available, we pooled the risk ratios using fixed-effect meta-analyses. Results from observational studies were described narratively. We appraised the certainty of evidence (CoE) of the findings using GRADE.

Results:

Of the 8526 citations identified, we included 50 studies (4 randomized controlled trials and 46 observational studies) involving 155 058 participants. Our meta-analysis of 4 randomized controlled trials of direct oral anticoagulants (apixaban or rivaroxaban) compared with vitamin K antagonists (486 participants) suggested that the effect of direct oral anticoagulants on ischemic strokes or systemic embolism is very uncertain (risk ratio [RR] = 0.52; 95% confidence interval [CI] = 0.22-1.20; very low CoE). Direct oral anticoagulants may decrease major bleeding (RR = 0.67; 95% CI = 0.44-1.03; low CoE) and may result in little to no difference in mortality (RR = 0.89; 95% CI = 0.70-1.13; low CoE). Our narrative review of the observational evidence suggests that, compared with no treatment, vitamin K antagonists may decrease ischemic strokes or systemic embolism (18 studies, low CoE) and likely increase major bleeding (12 studies, moderate CoE), whereas their effect on mortality is very uncertain (16 studies, very low CoE). Compared with no treatment, the effect of direct oral anticoagulants on all outcomes is very uncertain (2 studies, very low CoE). We also found that the effect of antiplatelet agents, left atrial appendage closure, and heparin-related therapies is very uncertain on all outcomes (very low CoE), except for antiplatelet agents that may increase major bleeding (1 study, low CoE) compared with no treatment.

Limitations:

Most of the included studies were observational and retrospective, resulting in low or very low certainty for nearly all of the outcomes. The included interventions were highly heterogeneous, which precluded meta-analysis of the results in all cases except for direct oral anticoagulants.

Conclusion:

In patients with atrial fibrillation receiving dialysis, vitamin K antagonists may decrease ischemic strokes or systemic embolism compared with no treatment at the cost of an increased risk of major bleeding. Direct oral anticoagulants (apixaban or rivaroxaban) may also lead to less major bleeding than vitamin K antagonists. The overall low to very low certainty of the evidence emphasizes the importance of engaging in shared decision-making when selecting a strategy for stroke prevention for atrial fibrillation in people receiving maintenance dialysis.

PROSPERO registration ID:

CRD42022307009.

Keywords

dialysis atrial fibrillation anticoagulant left atrial appendage systematic review

Introduction

Atrial fibrillation (AF) affects nearly 25% of patients receiving dialysis and is associated with an increased risk of ischemic stroke and mortality.^1
-4 Stroke prevention is complex in this population at higher risk of both ischemic and bleeding events.^4,5

Previous meta-analyses of randomized controlled trials (RCTs) in the general population have suggested warfarin reduces stroke risk by two thirds compared with placebo and direct oral anticoagulants (DOACs) by a further 19% compared with warfarin.^6,7 The DOACs lead to a 52% lower risk of intracranial hemorrhages in the general population and a 21% lower risk of major bleeding in the population with chronic kidney disease (CKD).^7,8 However, patients receiving dialysis were excluded from these large-scale RCTs, thus creating a knowledge gap for patients and clinicians. Recent Canadian, European, and US guidelines have yet to reach a consensus on whether to recommend DOACs, vitamin K antagonists (VKAs) or no anticoagulation for patients with AF receiving dialysis.^9
-12

The VKAs were historically the first anticoagulants used for stroke prevention in AF. However, many studies demonstrated an increase in major bleeding and conflicting results on the reduction of ischemic strokes and systemic embolism in people receiving dialysis.^13,14 Furthermore, VKAs may contribute to vascular calcifications and calciphylaxis, leading many clinicians to be cautious of VKAs in this population.^15
-17 Although warfarin remains the preferred anticoagulant agent (89%) for AF in patients receiving dialysis in Canada, the prescription of apixaban and rivaroxaban has increased substantially in this population despite lacking high-quality evidence.^18
-20 In the meantime, left atrial appendage closure (LAAC) has emerged as a non-pharmacological stroke-prevention intervention, but with minimal evidence in the dialysis setting.²¹

Therefore, the risk-benefit ratio of usual stroke-prevention interventions for patients with AF receiving dialysis remains to be determined. Several systematic reviews addressed this question, but their conclusions were limited by conflicting results, high study heterogeneity, lack of control over essential confounders and, for some of them, the inclusion of patients not receiving dialysis.^13,14,22
-30 Furthermore, an updated systematic review was needed to incorporate new evidence from RCTs on DOACs, to clarify the increasingly numerous observational studies on VKAs and to assess the risk-benefit ratio of other treatment strategies studied in the dialysis setting (eg, LAAC, antiplatelet agents). This systematic review aims to assess the efficacy and safety of all stroke-prevention interventions studied in adults with AF receiving dialysis to help inform shared decision-making with patients.

Methods

We conducted and reported this systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary Table S1).³¹ The protocol was registered on PROSPERO (CRD42022307009).³²

Search Strategy

We searched Medline, Embase, and Cochrane Central Register of Controlled Trials from inception to February 16, 2023. Literature search strategies were adapted from the latest Cochrane search strategies in the fields of AF, anticoagulation and dialysis (Supplementary Table S2).^33
-35 We searched unpublished trials using the International Clinical Trials Registry Platform and ClinicalTrials.gov. We searched the gray literature using Cochrane Kidney, Science.gov, Science.org, MedRxiv, and Google Scholar. Reference lists of eligible studies and recent systematic reviews were also screened for additional citations.

Eligibility Criteria

We formulated the eligibility criteria using the PICO (Population, Intervention, Comparator and Outcome) approach.

Patients

We included studies enrolling adult patients with non-valvular AF receiving any type of dialysis. Studies recruiting a broader population were included if they provided the results for the subset of participants meeting our eligibility criteria.

Interventions

We included any intervention to reduce stroke risks (antiplatelet agents, anticoagulants, LAAC).³⁶

Comparison

In a comprehensive approach, we included studies comparing the interventions to placebo, no treatment or all active comparators.

Outcomes

The outcomes of interest were SSE events (ischemic strokes or systemic embolism), major bleeding, and all-cause mortality. Systemic embolism is generally the primary outcome found in the literature on anticoagulation in AF and includes ischemic strokes, transient ischemic attacks, and arterial embolism. When systemic embolism data were not available, we used data for ischemic strokes. The definition of major bleeding varied across studies. We considered major bleeding that satisfied the International Society on Thrombosis and Haemostasis definition, which includes fatal bleeding, symptomatic bleeding in a critical area or organ, bleeding causing a 20 g/L fall in hemoglobin, or requiring 2 or more blood unit transfusions.³⁷

Other

We included RCTs and observational studies. Studies published in abstract form were included when at least 1 outcome of interest was reported quantitatively. The language was restricted to English and French. We contacted the corresponding authors to clarify the missing data where necessary.

Study Selection

Three authors (O.M., N.M., G.D.) independently screened all titles, abstracts and, when needed, full text for eligibility. Disagreements were resolved through discussion.

Data Extraction and Risk of Bias Assessment

Two authors (all authors, except A.B.-J.) independently extracted the data using a predefined extraction form and assessed the risk of bias. Disagreements were resolved through discussion, and a third author (G.D, N.D.) was available for persisting disagreements. For RCTs, we extracted the number of events and calculated risk ratios (RRs) with 95% confidence intervals (CIs). For observational studies, we extracted the number of events per 100 patient years and the maximally adjusted treatment effect with 95% CIs.

We assessed the risk of bias for RCTs using the Cochrane risk-of-bias tool (version 1).³⁸ We considered an RCT with any domain judged at high risk of bias as having an overall high risk. We assessed the risk of bias for observational studies using the Newcastle-Ottawa Scale (NOS).³⁹ We converted the NOS scores to low, moderate, or high risk of bias using the Agency for Healthcare Research and Quality Standards.⁴⁰

We also used the GRADE approach and its summary of findings table to report the certainty of evidence (CoE).^41
-43 The GRADE assessment was conducted by 2 authors (O.M., G.D.) and approved by all authors. When no pooled estimates were available, we used a narrative description of the size effects in the summary of findings table.^44,45 When RCTs were found, we prioritized their findings in our GRADE assessment but also considered complementary observational studies.⁴⁶

Data Synthesis

We used fixed-effects meta-analyses to pool the results of RCTs. We assessed heterogeneity by visual inspection of the forest plots and the I² statistic. No subgroup analysis was performed due to the small number of studies. Statistical analyses were performed on the RevMan platform v 5.3.

We summarized the results of the observational studies narratively as the very high clinical, methodological, and statistical heterogeneity prevented meta-analysis of the results. We followed the Economic and Social Research Council (ESRC) methodology and the Synthesis Without Meta-analysis (SWiM) guidelines (Supplementary Table S3).^47,48 We considered the observational studies’ findings, their risk of bias, the degree of heterogeneity, their congruence with RCTs’ results, and the presence of potential confounders to reach a consensus statement among authors. For the VKA comparisons, we synthesized the results by providing the number of studies showing benefits vs the number of studies showing harm. We pooled the outcomes for VKAs against no treatment a posteriori to facilitate the interpretation of VKAs’ potential effects and harms and appraise the statistical heterogeneity. We assessed publication bias by visual inspection of the funnel plots whenever a comparison contained more than 10 trials.⁴⁹

Results

Study Selection

We identified 6775 unique citations (8526 before removal of duplicates) (Figure 1). After screening titles and abstracts, we reviewed 302 citations in full text and included 50 studies involving 155 058 participants. We included 1 supplementary RCT whose results were presented at the American Society of Nephrology Annual Meeting after the completion of the search.⁵⁰

Figure 1.

PRISMA study flow chart.

Study Characteristics

Table 1 summarizes the characteristics of the included studies. Overall, we identified 4 RCTs,^50
-53 6 prospective cohort studies,^54
-59 and 40 retrospective cohort studies.^29,60
-98 Six studies were only published as abstracts.^50,62
-66 Eight compared more than 2 groups.^{50,58,59,63,65,67,69,70} Two included only patients receiving peritoneal dialysis,^72,80 and 6 recruited both patients receiving peritoneal dialysis or hemodialysis.^{54,69,85,88,94,96} The mean follow-up duration varied considerably across studies, ranging from 6 months to 5 years.

Table 1.

Characteristics of Included Studies.^a,b

First author (year)	Intervention	Comparator	Design	Follow-up (y)	Treatment, N	Control, N	Age (y)	Female (%)	HD (%)	CHA₂DS₂	CHA₂DS₂-VASc	Antiplatelets (%)
Intervention: VKAs
Chan (2009)	Warfarin	NT	RC	1.6	746	925
Lai (2009)	Warfarin	NT	RC		51	42			100
Winkelmayer (2011)	Warfarin	NT	RC	1.9	237	948	68.9	57.1	96.6
Olesen (2012)	Warfarin	NT	RC	3.2	178	525	66.8	33.6				0
Olesen (2012)	Warfarin + ASA	NT	RC	3.2	45	525	66.8	33.6				100
Khalid (2013)	Warfarin	NT	RC	1.0	34	62	77.2	31.3	100
Chen (2014)	Warfarin	NT	RC	4.1	250	250		57.0	98.0
Wakasugi (2014)	Warfarin	NT	PC	1.8	28	32	68.1	35.0	100			53.3
Shah (2014)	Warfarin	NT	RC		756	870	75.2	39.0				25.0
Friberg (2015)^c	Warfarin	NT	RC		164	583	71.0	32.0	100
Sajeev (2015)	Warfarin	NT	RC		18	53			100		4.5
Shen (2015)	Warfarin	NT	RC	1.4	1838	10 446	61.9	52.8	100			23.9
Chan (2016)	Warfarin	NT	RC	1.5	67	118	69.4	39.5	0		3.1
Yamashita (2016)	Warfarin	NT	PC	2.8	35	55	71.3	39.1		2.5	4.0	47
Garg (2016)	Warfarin	NT	RC	2.1	119	183	77.0	47.0	100			88.4
Mitsuma (2016)	Warfarin	NT	RC		27	55	71.2	28.0	100			36.0
Brancaccio (2016)	Warfarin	NT	RC	1.1	657	657	72.2	39.5	100			28.0
Yodogawa (2016)	Warfarin	NT	RC	3.9	30	54	70.1	29.8	100	1.6		48.8
Wang (2016)	Warfarin	NT	RC	3.4	59	82	61.2	38.3	68.8	2.4	3.8	63.1
Lee (2017)	Warfarin	NT	RC	2.4	589	1767	70.1	59.0	100.0		100% ≥2	58.3
Noseworthy (2017)	VKA	NT	RC		2793	7702
Lai (2017)	Warfarin	NT	RC	4.6	31	137		73.2	100		100% ≥2
Lai (2017)	Warfarin + ASA	NT	RC	4.6	215	137		63.6	100		100% ≥2
Kai (2017)	Warfarin	NT	RC	2.1	888	888	68.9	37.5	100		5.2	26.4
Yoon (2017)	Warfarin	NT	RC	1.3	2274	2274	66.6	41.8	100			≥52.6
Genovesi (2017)	Warfarin	NT	PC	4.0	134	150	74.0	60.8	100			51.4
Sánchez (2018)	Acenocoumarol	NT	RC	2.4	43	31	75.0	33.8	100		3.8	27.0
Voskamp (2018)	VKA	NT	PC	5.0	244	1474	62.2	38.7	64.4
Chang (2019)	Warfarin	NT	RC	3.2	144	779
Tan (2019)	Warfarin	NT	RC	1.0	1651	4114	74.8	56.9	96.6			15.8
Phan (2019)	Warfarin	NT	RC		115	361	64.0	42.2	0		4.3	≥31.7
Waddy (2020)	Warfarin	NT	RC	1.0			69.3	49.2	100		4.1
Agarwal (2020)	Warfarin	NT	RC		6682	16 089	73.5	4.6			5.7	7.3
Königsbrügge (2021)	Phenprocoumon	NT	PC	2.4	61	139	72.1	34.5	100			58.0
Dimitrijevic (2021)	VKA	NT	RC	2.3	69	37	72.0	25.2	100
Kang (2021)	Warfarin	NT	RC		177	547	64.0	36.2	100
Xu (2021)	Warfarin	NT	RC	2.9	16	166	57.8	57.1	100		100% ≥1
Sy (2022)	Warfarin	NT	RC		5960	5960	74.0	4.0	100		6.0
Welander (2022)	Warfarin	NT	RC		405	2843	75.0	29.4			5.0
Brensing (2022)	VKA	NT	RC	2.1	453	723	76.2	37.5	100		4.0
Akbar (2023)	Warfarin	NT	RC	0.9	44	44	52.2	39.8	100
Harel (2023)	Warfarin	NT	RCT	0.5	52	48	71.6	25.0			4.0
Intervention: DOACs
Chan (2015)	Dabigatran	VKA	RC	0.5	281	8064	70.4	38.9	100	2.4		3.2
Chan (2015)	Rivaroxaban	VKA	RC	0.5	244	8064	70.4	38.9	100	2.4		3.2
Noseworthy (2017)	Apixaban	VKA	RC		125	2793
	Rivaroxaban				56
	Dabigatran				36
Siontis (2018)	Apixaban	VKA	RC		2351	23 172	68.2	45.7	94.6		5.2	7.3
Mavrakanas (2020)	Apixaban	NT	RC		521	1561	68.0	46.9	100			23.0
See (2021)	Apixaban	VKA	RC		88	448	74.8	51.5	100		4.6
	Dabigatran				150
	Edoxaban				17
	Rivaroxaban				224
De Vriese (2021)	Rivaroxaban	VKA	RCT	1.9	88	44	80.0	33.3	100		4.7	34.8
Reinecke (2022)	Apixaban	Phenprocoumon	RCT	1.3	48	49	74.7	29.9	100		4.5	34.0
Pokorney (2022)	Apixaban	Warfarin	RCT	0.9	82	72	68	36.4	100		4.0	40.9
Wetmore (2022)	Apixaban	VKA	RC	1.6	4639	12 517	66.2	38.3	100		4.5	17.1
Harel (2023)	Apixaban	NT	RCT	0.5	51	48	71.6	25.0			4
Harel (2023)	Apixaban	Warfarin	RCT	0.5	51	52	71.6	25.0			4
Intervention: antiplatelets
Olesen (2012)	ASA	NT	RC	3.2	153	525	66.8	33.6
Chen (2014)	Antiplatelet	NT	RC	4.1	840	840		50.8	96.0
Chan (2015)	ASA	Warfarin	RC	0.5	6018	8064	71.1	40.5	100	2.4
Chan (2016)	Warfarin	ASA	RC	1.5	67	86	71.5	41.8	0		3.9
Lai (2017)	ASA	NT	RC	4.6	814	137		62.0	100		100% ≥2
Lai (2017)	ASA	Warfarin	RC	4.6	814	31		60.0	100		100% ≥2
Brensing (2022)	ASA +/- CLOP	NT	RC	2.1	456	723	76.0	37.0	100		4.0
Brensing (2022)	ASA +/- CLOP	VKA	RC	2.1	456	453	76.0	37.0	100		4.0
Intervention: LAAC
Genovesi (2021)	VKA	LAAC	PC	1.7	114	92	75.1	29.1			Median: LAAC 4, VKA 4, NT 5	44.2
Genovesi (2021)	NT	LAAC	PC	1.7	148	92	75.2	35.8			Median: LAAC 4, VKA 4, NT 5	67.9
Romero (2022)	LAAC	OAC	RC	1.3	27
Intervention: heparin-related therapies
Königsbrügge (2021)	Enoxaparin	NT	PC	2.4	38	139	73.0	33.9	100		4.0 (median)	67.2
Brensing (2022)	HEP (unspecified) +/- ASA/CLOP	VKA	RC	2.1	162	453	76.0	37.0	100		4.0
Brensing (2022)	HEP (unspecified) +/- ASA/CLOP	NT	RC	2.1	162	723	76.0	37.0	100		4.0

Note. ASA = aspirin; CLOP = clopidogrel; DOAC = direct oral anticoagulation; HEP = heparin; LAAC = left atrial appendage closure; LWMH = low-weight molecular heparin; NT = no treatment; OAC = oral anticoagulation; PC = prospective cohort; RCT = randomized controlled trial; RC = retrospective cohort; VKA = vitamin K antagonists.

When multiple arms were included in 1 study, we combined characteristics for the pertinent arms, whenever possible. If not possible, we presented the characteristics for the full cohort.

If available, we presented the characteristics for the adjusted population after propensity score weighting.

We extracted some data in a previous systematic review.²³

Risk of Bias and Quality Assessment

All the included RCTs were considered at high risk of bias (Supplementary Table S4) for either being open-label, providing incomplete information on allocation concealment or having limited sample sizes. The risk of bias was considered low for 25 (54%) cohort studies^{29,54,55,57
-59,61,67,68,70,75
-77,81,83
-87,89,92,93,95
-97} and moderate to high for the remainder (Supplementary Table S5). Only one third of the studies controlled for both concurrent antiplatelet agents and CHA₂DS₂-VASc or CHADS₂ scores.^{57
-59,67,68,70,73,77,83
-85,87,92,95
-97} Furthermore, observational studies considered at low risk of bias according to the NOS scores remained prone to numerous biases, such as confounding, unexplained heterogeneity, and imprecision of results. Ultimately, we considered the evidence from the included RCTs to be of higher quality than those from the included cohort studies. A funnel plot could only be performed for observational studies of VKAs, and it did not suggest the presence of a publication bias (Supplementary Figure S2).

Findings

Table 2 summarizes the systematic review’s findings, and Table 3 provides the extracted results for individual study outcomes for VKAs and DOACs.

Table 2.

Summary of Findings.

Outcome	Anticipated absolute effect		Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence	Plain language summary
Outcome	Treatment	Comparator	Relative effect (95% CI)	Number of participants (studies)	Certainty of the evidence	Plain language summary
VKAs vs NT
SSE events	In the studies at low risk of bias, 11 reported a decrease (5 with a P < .05) and 7 reported an increase (2 with a P < .05) in SSE events.			36 016(17 CS + 1 RCT)	⊕⊕⊝⊝Low^a	Compared with no treatment, VKAs may decrease SSE events.
Major bleedings	In the studies at low risk of bias, 11 reported an increase (5 with a P < .05) and 1 a non-statistically significant decrease in major bleedings with VKAs.			13 381(11 CS + 1 RCT)	⊕⊕⊕⊝Moderate^b	Compared with no treatment, VKAs likely increase major bleedings.
Mortality	In the studies at low risk of bias, 7 reported a non-statistically significant increase, 8 reported a decrease (5 with a P < .05) and 1 reported no effect.			31 248(15 CS + 1 RCT)	⊕⊝⊝⊝Very low^c	The effect of VKAs on mortality is very uncertain.
DOACs vs NT
SSE events	Two studies found a non-statistically significant decrease in ischemic strokes with apixaban.			2181(1 CS + 1 RCT)	⊕⊝⊝⊝Very low^d	The effect of DOACs on all outcomes is very uncertain.
Major bleedings	The included studies showed conflicting results for this outcome.			2181(1 CS + 1 RCT)	⊕⊝⊝⊝Very low^d
Mortality	Two studies reported a decrease (1 with a P < .05) in mortality with apixaban.			2181(1 CS + 1 RCT)	⊕⊝⊝⊝Very low^d
DOACs vs VKAs
SSE events	2 per 100(1-6)	5 per 100	RR = 0.52(0.22-1.20)	486(4 RCTs)	⊕⊝⊝⊝Very low^e	Compared with VKAs, the effect of DOACs on SSE events is very uncertain.
Major bleedings	11 per 100(5-16)	16 per 100	RR = 0.67(0.44-1.03)	486(4 RCTs)	⊕⊕⊝⊝Low^f	Compared with VKAs, DOACs may decrease major bleedings.
Mortality	27 per 100(21-34)	31	RR = 0.89(0.70-1.13)	486(4 RCTs)	⊕⊕⊝⊝Low^f	Compared with VKAs, DOACs may result in little to no difference in deaths.
Antiplatelet agents vs no NT
SSE events	The included studies showed conflicting results for this outcome.			4488(4 CS)	⊕⊝⊝⊝Very low^g	The effect of antiplatelet agents on SSE events is very uncertain.
Major bleedings	Only 1 study assessed this outcome, showing an increase in major bleedings (HR = 1.63; 95% CI = 1.18-2.26) with antiplatelets.			678(1 CS)	⊕⊕⊝⊝Low	Compared with NT, antiplatelet agents may increase major bleedings.
Mortality	The studies showed small non-significant decreases in mortality rates with antiplatelets.			2130(2 CS)	⊕⊝⊝⊝Very low^h	The effect of antiplatelet agents on mortality is very uncertain.
Antiplatelets vs VKAs
SSE events	The included studies showed conflicting results for this outcome.			15 989(4 CS)	⊕⊝⊝⊝Very lowⁱ	The effect of antiplatelet agents on all outcomes is very uncertain.
Major bleedings	The only study reporting major bleedings showed a decrease with aspirin (RR = 0.78; 95% CI = 0.72-0.84).			14 082(1 CS)	⊕⊝⊝⊝Very low^j
Mortality	One study showed similar death rates between antiplatelets and VKAs. Another one showed a non-significant decrease in deaths with aspirin.			1754(2 CS)	⊕⊝⊝⊝Very low^k
LAAC vs VKAs, OAC or NT
SSE events	In 1 study, LAAC showed a non-significant decrease compared with OAC. The second study reported a significant decrease compared with NT and a non-significant decrease compared with VKA.			381(2 CS)	⊕⊝⊝⊝Very low^l	The effect of LAAC on all outcomes is very uncertain.
Major bleedings	In 1 study, LAAC decreases major bleedings compared with NT and VKA.			354(1 CS)	⊕⊝⊝⊝Very low^l
Mortality	In 2 studies, LAAC was associated with a significant decrease in deaths compared with OAC and VKA. In 1 study, LAAC decreased mortality compared with NT.			381(2 CS)	⊕⊝⊝⊝Very low^l
Heparin-related therapies vs VKAs or NT
SSE events	A non-significant decrease in SSE events was reported for enoxaparin compared with NT. In another study, results were similar between heparin-related therapy, VKA and NT.			1515(2 CS)	⊕⊝⊝⊝Very low^m	The effect of heparin-related therapies on all outcomes is very uncertain.
Major bleedings	A non-significant increase in major bleedings was reported for enoxaparin compared with NT.			177(1 CS)	⊕⊝⊝⊝Very low^m
Mortality	One study reported a significant increase in mortality with enoxaparin compared with NT. The second study showed an increase in mortality with heparin-based therapy vs VKA.			1515(2 CS)	⊕⊝⊝⊝Very low^m

Note. GRADE certainty of the 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. CI = confidence interval; CS = cohort studies; DOAC = direct oral anticoagulant; LAAC = left atrial appendage closure; NT = no treatment; OAC = oral anticoagulation; RCT = randomized controlled trial; RR = relative risk; SSE events = ischemic stroke and/or systemic embolism; VKA = vitamin K antagonist.

^aDowngraded by 1 level for inconsistency (unexplained heterogeneity) and upgraded by 1 level for plausible residual confounding (as VKAs groups were likely at higher risk of SSE events).

^bUpgraded by 1 level for plausible residual confounding (as VKAs groups were at lower risk of major bleedings).

^cDowngraded by 1 level for high risk of bias (insufficient control of confoundings), inconsistency (unexplained heterogeneity), and imprecision (wide CIs in many studies).

^dDowngraded by 1 level for high risk of bias (insufficient details for RCT) and imprecision (wide CIs and/or low numbers of events).

^eDowngraded by 3 levels for high risk of bias (lack of blinding and allocation concealment) and imprecision (low numbers of events and very wide CI).

^fDowngraded by 2 levels for high risk of bias (lack of blinding and allocation concealment) and imprecision (wide CI).

^gDowngraded by 1 level for high risk of bias (insufficient control of confoundings), inconsistency (unexplained heterogeneity), and imprecision (wide CIs in many studies).

^hDowngraded by 1 level for high risk of bias (insufficient control of confoundings) and imprecision (wide CI in the only study reporting CI).

ⁱDowngraded by 1 level for high risk of bias (important selection bias and insufficient control of confoundings), inconsistency (unexplained heterogeneity), and imprecision (wide CIs in many studies).

^jDowngraded by 1 level for high risk of bias (important selection bias and insufficient control of confoundings).

^kDowngraded by 1 level for high risk of bias (important selection bias and insufficient control of confoundings) and imprecision (wide CI in the only study reporting CI).

^lDowngraded by 1 level for high risk of bias (insufficient control of confounding) and imprecision (small sample sizes, low numbers of events, and wide CIs when reported).

^mDowngraded by 1 level for high risk of bias (insufficient control of confounding) and imprecision (wide CI when reported).

Table 3.

Outcomes of Included Studies.

First author (year)	Intervention	Comparator	Incidence of SSE events^a (% or /100 PY)			Incidence of major bleedings^a (% or /100 PY)			Incidence of mortality^a (% or /100 PY)
First author (year)	Intervention	Comparator	Treatment	Control	Relative effect^b (95% CI)	Treatment	Control	Relative effect^b (95% CI)	Treatment	Control	Relative effect^b (95% CI)
Intervention: VKAs
Chan (2009)	Warfarin	NT	5.8	2.3	HR = 1.81 (1.12-2.92)				27.4	25.7	HR = 1.10 (0.94-1.30)
Lai (2009)^c	Warfarin	NT	2.3	13.8	HR = 0.17 (0.07-0.42)
Winkelmayer (2011)	Warfarin	NT	7.4	7.8	HR = 0.92 (0.61-1.37)				42.9	40.2	HR = 1.06 (0.90-1.24)
Olesen (2012)	Warfarin	NT			HR = 0.44 (0.26-0.74)			HR = 1.27 (0.91-1.77)
Olesen (2012)	Warfarin + ASA	NT			HR = 0.82 (0.37-1.80)			HR = 1.71 (0.98-2.99)
Khalid (2013)	Warfarin	NT			HR = 0.44 (0.27-0.73)			HR = 1.72 (1.29-2.30)			HR = 0.22 (0.13-0.40)
Chen (2014)	Warfarin	NT	5.6%	4.0%	HR = 1.02 (0.68-1.54)
Wakasugi (2014)	Warfarin	NT	14.8	8.9	HR = 3.36 (0.67-16.66)	5.3	6.8	HR = 0.85 (0.19-3.64)	14.2	14.2	HR = 1.00 (0.40-2.52)
Shah (2014)	Warfarin	NT			HR = 1.17 (0.79-1.75)			HR = 1.41 (1.09-1.81)
Friberg (2015)^c	Warfarin	NT	1.7	3.0	HR = 0.65 (0.22-1.90)				25.8	34.2	HR = 0.77 (0.58-1.02)
Sajeev (2015)	Warfarin	NT	44.4%	13.2%	(P = .01)				55.6%	26.4%	(P = .02)
Shen (2015)	Warfarin	NT	2.3	3.4	HR = 0.68 (0.47-0.99)				33.0	32.5	HR = 1.01 (0.92-1.11)
Chan (2016)	Warfarin	NT	0%	9.3%	HR = 0.19 (0.06-0.65)
Yamashita (2016)	Warfarin	NT	1.4%	5.6%	HR = 0.26 (0.01-1.52)	4.5%	5.5%	HR = 0.75 (0.17-3.02)	27.4%	16.9%	HR = 1.60 (0.84-3.04)
Garg (2016)	Warfarin	NT	10.9%	11.4%	HR = 0.93 (0.49-1.82)	21.8%	14.2%	HR = 1.53 (0.94-2.51)	81.5%	79.2%	HR = 1.03 (0.91-1.15)
Mitsuma (2016)	Warfarin	NT	11.1	9.1		25.9	16.4		29.6	49.1
Brancaccio (2016)	Warfarin	NT							19.3	21.7	HR = 0.76 (P < .01)
Yodogawa (2016)	Warfarin	NT				6.7	9.3	(P = .08)	20.0	27.8	(P = .68)
Wang (2016)	Warfarin	NT	13.6%	13.4%	(P = 1.000)	37.3%	29.3%	(P = .364)	74.6%	78.0%	(P = .689)
Lee (2017)	Warfarin	NT	0.5	0.4					3.2	2.8
Noseworthy (2017)	Warfarin	NT			HR = 0.42 (0.29-0.61)			HR = 0.74 (0.62-0.89)
Lai (2017)	Warfarin	NT	9.7%	7.3%	HR = 1.07 (0.60-1.93)				45.2%	43.8%	HR = 1.07 (0.60-1.93)
Lai (2017)	Warfarin + ASA	NT	14.4%	7.3%	HR = 1.23 (0.88-1.73)				41.4%	43.8%	HR = 0.82 (0.88-1.73)
Kai (2017)	Warfarin	NT	2.8	4.8	HR = 0.75 (0.53-1.10)				20.6	26.8	HR = 0.78 (0.69-0.89)
Yoon (2017)	Warfarin	NT	7.4%	7.2%	HR = 0.95 (0.78-1.15)
Genovesi (2017)	Warfarin	NT	3.2	4.5	HR = 0.44 (0.16-1.20)	17.0	9.5	HR = 1.16 (0.48-2.82)	22.0	25.0	HR = 0.91 (0.56-1.48)
Sánchez (2018)	Acenocoumarol	NT	14.0%	3.2%		44.2%	9.7%				HR = 0.76 (0.35-1.66)
Voskamp (2018)	VKA	NT							25.0	14.4	HR = 1.2 (1.0-1.5)
Chang (2019)	Warfarin	NT			HR = 3.8 (1.8-8.3)			HR = 2.9 (1.6-5.2)
Tan (2019)	Warfarin	NT	8.2	10.8	HR = 0.88 (0.70-1.11)	41.7	32.3	HR = 1.50 (1.33-1.68)	37.0	52.0	HR = 0.72 (0.65-0.80)
Phan (2019)	Warfarin	NT	6.2	2.4	HR = 2.30 (0.94-5.40)				19.9	21.0	HR = 0.80 (0.53-1.20)
Waddy (2020)	Warfarin	NT			HR = 0.79 (0.66-0.95)						HR = 0.81 (0.72-0.91)
Agarwal (2020)	Warfarin	NT	10.6	10.7	HR = 1.23 (1.16-1.30)	14.5	12.9	HR = 1.36 (1.29-1.44)	46.1	53.7	HR = 0.94 (0.90-0.97)
Königsbrügge (2021)	Warfarin	NT			HR = 1.41 (0.49-4.07)			HR = 2.28 (1.09-4.79)			HR = 1.35 (0.86-2.12)
Dimitrijevic (2021)	VKA	NT			HR = 0.78 (0.69-0.88)			HR = 1.83 (1.34-2.91)			HR = 0.86 (0.71-0.92)
Kang (2021)	Warfarin	NT			HR = 0.99 (0.72-1.36)						HR = 1.15 (0.91-1.46)
Xu (2021)	Warfarin	NT	6.3%	1.8%	HR = 4.12 (0.08-51.36)				68.8%	54.2%	HR = 1.63 (0.86-3.07)
Sy (2022)	Warfarin	NT	3.1	3.0	HR = 1.04 (0.89-1.21)	8.3	8.3	HR = 1.03 (0.93-1.14)			HR = 0.66 (0.64-0.69)
Welander (2022)	Warfarin	NT	1.6	3.3	HR = 0.49 (0.30-0.79)	12.1	10.4	HR = 1.23 (1.00-1.51)	19.3	37.9	HR = 0.49 (0.43-0.56)
Brensing (2022)	VKA	NT	2.9%	2.6%					2.8 y	1.8 y	Median survival
Akbar (2023)	Warfarin	NT	34.1%	18.2%	HR = 0.44 (0.10-1.85)	6.8%	5.7%		40.9%	45.5%	HR = 0.78 (0.49-1.24)
Harel (2023)	Warfarin	NT	0%	2.1%	RR = 0.31 (0.01-7.39)	7.7%	4.2%	RR = 1.85 (0.35-9.63)	17.3%	8.3%	RR = 2.08 (0.68-6.31)
Intervention: DOACs
Chan (2015)	Dabigatran	VKA	10.6	6.2	RR = 1.71 (0.97-2.99)	83.1	35.9	RR = 1.48 (1.21-1.81)
Chan (2015)	Rivaroxaban	VKA	11.2	6.2	RR = 1.80 (0.89-3.64)	68.4	35.9	RR = 1.38 (1.03-1.83)
Noseworthy (2017)	Apixaban	VKA						HR = 0.29 (0.07-1.18)
	Rivaroxaban				HR = 3.54 (0.68-18.33)			HR = 0.69 (0.17-2.87)
	Dabigatran				HR = 3.15 (0.49-20.43)			HR = 1.82 (0.67-4.98)
Siontis (2018)	Apixaban	VKA	12.4	11.8	HR = 0.88 (0.69-1.12)	19.7	22.9	HR = 0.72 (0.59-0.87)	23.7	24.9	HR = 0.85 (0.71-1.01)
Mavrakanas (2020)	Apixaban	NT	3.5	5.0	HR = 0.85 (0.36-1.98)	4.9	1.6	HR = 2.76 (1.38-5.52)			HR = 0.58 (0.43-0.78)
See (2021)	DOACs	VKA	6.7	5.3	HR = 1.21 (0.76-1.92)	7.1	7.2	HR = 0.98 (0.64-1.51)
De Vriese (2021)	Rivaroxaban	VKA	6.8%	15.9%	RR = 0.43 (0.15-1.20)	19.3%	38.6%	RR = 0.50 (0.28-0.88)	64.8%	72.7%	RR = 0.89 (0.70-1.13)
Reinecke (2022)	Apixaban	Phenprocoumon	0%	2.0%	RR = 0.34 (0.01-8.15)	10.4%	12.2%	RR = 0.85 (0.28-2.60)	18.8%	24.5%	RR = 0.77 (0.36-1.65)
Pokorney (2022)	Apixaban	Warfarin	1.2%	2.8%	RR = 0.44 (0.04-4.74)	11.0%	9.7%	RR = 1.13 (0.44-2.88)	25.6%	18.0%	RR = 1.42 (0.77-2.62)
Wetmore (2022)	Apixaban DPL	VKA	2.0	2.1	HR = 0.89 (0.65-1.21)	4.5	6.3	HR = 0.67 (0.55-0.81)	24.4	28.8	HR = 0.85 (0.78-0.92)
Wetmore (2022)	Apixaban LL	VKA	1.9	2.1	HR = 0.85 (0.62-1.17)	4.7	6.3	HR = 0.68 (0.55-0.84)	28.0	28.8	HR = 0.97 (0.89-1.05)
Harel (2023)	Apixaban	NT	2.0%	2.1%	RR = 0.94 (0.06-14.63)	3.9%	4.2%	RR = 0.94 (0.14-6.42)	3.9%	8.3%	RR = 0.47 (0.09-2.45)
Harel (2023)	Apixaban	Warfarin	2.0%	0%	RR = 3.06 (0.13-73.36)	3.9%	7.7%	RR = 0.51 (0.10-2.66)	3.9%	17.3%	RR = 0.23 (0.05-0.998)
Intervention: antiplatelets
Olesen (2012)	ASA	NT			HR = 0.88 (0.59-1.32)			HR = 1.63 (1.18-2.26)
Chen (2014)	Antiplat	NT	6.5%	6.4%	HR = 0.83 (0.57-1.21)
Chan (2015)	ASA	Warfarin	5.0	6.2	RR = 0.81 (0.66-0.99)	35.9	47.1	HR = 0.78 (0.72-0.84)
Chan (2016)	Warfarin	ASA	0%	11.9%	HR = 0.16 (0.04-0.66)
Lai (2017)	ASA	NT	14.6%	7.3%	HR = 2.01 (1.04-3.88)				39.4%	43.8%	HR = 0.84 (0.63-1.12)
Lai (2017)	ASA	Warfarin	14.6%	9.7%	HR = 1.58 (0.50-4.98)				39.4%	45.2%	HR = 0.90 (0.53-1.51)
Brensing (2022)	ASA +/- CLOP	NT	3.7%	2.6%					2.9 y	1.8 y	(Median survival)
Brensing (2022)	ASA +/- CLOP	VKA	3.7%	2.9%					2.9 y	2.8 y	(Median survival)
Intervention: LAAC
Genovesi (2021)	VKA	LAAC	7.0%	2.2%	(P = .092)	23.7%	8.7%	HR = 6.48 (1.32-31.72)	58.8%	18.5%	HR = 2.76 (1.31-5.86)
Genovesi (2021)	NT	LAAC	10.8%	2.2%	(P = .021)	16.2%	8.7%	HR = 4.87 (1.08-21.97)	63.5%	18.5%	HR = 3.09 (1.59-5.98)
Romero (2022)	LAAC	OAC	0%	7.7%					22%	41%
Intervention: heparin-related therapies
Königsbrügge (2021)	Enoxaparin	NT			HR = 0.49 (0.11-2.18)			HR = 1.86 (0.81-4.28)			HR = 1.88 (1.20-2.92)
Brensing (2022)	Hep (unspecified) +/- ASA/CLOP	VKA	1.2%	2.9%					1.7 y	2.8 y	(Median survival)
Brensing (2022)	Hep (unspecified) +/- ASA/CLOP	NT	1.2%	2.6%					1.7 y	1.8 y	(Median survival)

Note. ASA = aspirin; CLOP = clopidogrel; DOAC = direct oral anticoagulant; DPL = dose per label; HEP = heparin; HR = hazard ratio; IRR = incidence rate ratio; LAAC = left atrial appendage closure; LWMH = low-weight molecular heparin; LL = low label; NT = no treatment; OAC = oral anticoagulation; PY = patients-years; RR = risk ratio; VKA = vitamin K antagonist.

Incidence of events was preferably expressed in /100 PY when available, except for the randomized controlled trials (presented in absolute numbers).

We extracted relative effects from the maximally adjusted statistical model.

We extracted some data in a previous systematic review.²³

Vitamin K antagonists

We included 1 RCT (1 comparison, 100 participants)⁵⁰ and 39 observational studies comparing VKAs with no treatment (41 comparisons, 90 492 participants, 31% receiving a VKA).^{60,54
-57,59,61
-65,68
-94,98} Of these 40 studies, 34 studies used warfarin,^{60,50,55
-57,61
-63,68
-80,82
-94,98} 3 did not specify the VKA agent,^54,64,65 1 used acenocoumarol,⁸¹ and another one used phenprocoumon.⁵⁹ One study with a treatment group of both DOACs and VKAs was classified in the VKA treatment group, as 92.8% of the participants were taking warfarin, and no subgroup analysis was provided.⁶³ The participants’ mean age ranged from 52 to 77 years old. Only 13 studies reported the mean or median CHA₂DS₂-VASc score (range = 3.1-6.0),^{50,56,62,65,72,73,80,81,82,86
-89} and 3 reported the mean or median CHA₂DS₂ score (range = 1.6-2.5).^56,88,91 Nineteen studies reported the mean or median use of antiplatelet agents, which ranged from 7.3% to 88.4%.^{55
-57,59,70,73,75,76,79
-81,83
-85,87,88,91
-93} Use of antiplatelet agents was frequent, with 50% of studies reporting a mean or median number of participants under antiplatelet agents of more than 40%. Eleven studies reported mean or median international normalized ratio (INR) (range = 1.5 to 2.3) or time in therapeutic range (TTR) (range = 33.3% to 65%) in the VKA group.^{50,55,56,59,61,79
-81,87,89,91}

There were insufficient events in the single included RCT (SAFE-D) to conclude on the effect of VKAs on the outcomes of interest.⁵⁰ All 39 observational studies were highly heterogeneous for the 3 outcomes assessed (Supplementary Figure S1). We therefore conducted a subgroup analysis of only the studies at low risk of bias for each comparison.

Thirty-seven studies (87 476 participants) reported SSE events.^{60,50,55
-57,59,61
-65,68
-74,76
-90,92
-94,98} The single included RCT was inconclusive on the effect of VKAs on SSE events (0 vs 2.1%; RR = 0.31; 95% CI = 0.01-7.39).⁵⁰ In the 18 studies (36 016 participants) at low risk of bias, 11 reported a decrease^{50,57,70,76,77,84
-87,89,92} (5 with a P < .05)^{70,77,84,86,89} and 7 reported an increase^{55,59,61,68,81,83,93} (2 with a P < .05)^61,68 in SSE events. Visual inspection of the forest plot suggested a downward trend in ischemic events. Because heterogeneity remained high, we further explored these findings within the 13 studies at low risk of bias that adjusted for the CHA₂DS₂-VASc or CHADS₂ scores.^{50,55,57,59,70,76,83,84
-87,89,92} Ten reported a decrease^{50,57,70,76,84
-87,89,92} (4 with a P < .05)^70,84,86,89 and 3 reported a non-statistically significant increase^55,59,83 in SSE events. Therefore, we concluded that VKAs may decrease SSE events (18 studies, low CoE) in people with AF receiving hemodialysis.

Twenty-one studies (59 203 participants) reported major bleeding events.^{50,55
-57,59,63,64,68,70,73,74,76,77,79,81
-83,85,88,89,91} The single included RCT was inconclusive for the effect of VKAs on major bleeding (7.7 vs 4.2%; RR = 1.85; 95% CI = 0.35-9.63).⁵⁰ Among the 12 studies at low risk of bias (13,381 participants), 11 reported an increase^{50,57,59,68,70,76,77,81,83,85,89} (5 with a P < .05)^{59,68,77,83,85} in major bleeding with VKAs. Only 1 study reported a non-statistically significant decrease in major bleeding with VKAs.⁵⁵ Based on this subgroup analysis, we concluded that VKAs likely increase major bleeding (12 studies, moderate CoE).

Thirty-two studies (71 474 participants) reported all-cause mortality.^{60,50,54
-57,61,62,65,71,73
-76,79,80,84,85,87
-91,93,94} The single included RCT was inconclusive on the effect of VKAs on mortality (17.3 vs 8.3%; RR = 2.08; 95% CI = 0.68-6.31).⁵⁰ In the 16 studies (31 248 participants) at low risk of bias, 7 reported a non-statistically significant increase,^{50,54,59,61,76,84,93} 8 reported a decrease^{57,75,77,81,85
-87,89} (5 with a P < .05),^{75,77,86,87,89} and 1 reported no effect.⁵⁵ Therefore, we concluded that the effect of VKAs on mortality is very uncertain (16 studies, very low CoE).

Direct oral anticoagulants

We included 1 RCT⁵⁰ (1 comparison, 99 participants) and 1 observational study⁹⁵ (1 comparison, 2082 participants) comparing apixaban with no treatment, 4 RCTs (4 comparisons, 486 participants) comparing DOACs with VKAs,^50
-53 and 5 observational studies (9 comparisons, 55,205 participants) comparing DOACs with VKAs.^{29,63,67,96,97} The mean age varied between 66 and 80 years old. Patients were at high ischemic stroke risk with a mean CHA₂DS₂-VASc between 4.0 and 5.2. The use of antiplatelet agents was frequent in the RCTs (from 34.0 to 40.9%) and heterogeneous in the cohort studies (from 3.2% to 23.0%). The Valkyrie study randomized participants to a reduced dosing regimen of rivaroxaban (10 mg daily).⁵¹ The RENAL-AF randomized the apixaban group to the usual dosing regimen (5 mg twice daily or 2.5 mg twice daily if at least 2 of the following criteria are met: age ≥ 80 years, body weight ≤ 60 kg, or serum creatinine ≥ 133 μmol/L), AXADIA-AFNET 8 to the reduced dose (2.5 mg twice daily), and SAFE-D permitted both regimens.^50,52,53 The observational studies did not report the TTR for the VKAs group, but the RCTs reported a low TTR (between 44% and 51%).^53
-55

Direct oral anticoagulants vs no treatment

The single included RCT was inconclusive on the effect of apixaban vs placebo on SSE events (2.0% vs 2.1%; RR = 0.94; 95% CI = 0.06-14.63), major bleeding (3.9 vs 4.2%; RR = 0.94; 95% CI = 0.14-6.42), and mortality (3.9 vs 8.3%; RR = 0.47; 95% CI = 0.09-2.45).⁵⁰ The observational study found no significant difference in ischemic strokes with apixaban (hazard ratio [HR] = 0.85; 95% CI = 0.36 to 1.98) compared with no treatment but showed an increase in major bleeding (HR = 2.76; 95% CI = 1.38 to 5.52) and a decrease in mortality (HR = 0.58; 95% CI = 0.43-0.78).⁹⁵ Considering the overall risk of bias and wide CIs, the effects of DOACs vs no treatment on all outcomes remain very uncertain (very low CoE).

Direct oral anticoagulants vs vitamin K antagonists

Four RCTs contributed to the meta-analyses (Figure 2).^50
-53 The evidence comparing VKAs with DOACs regarding SSE events is very uncertain (RR = 0.52; 95% CI = 0.22-1.20; very low CoE). The DOACs may decrease major bleeding (RR = 0.67; 95% CI = 0.44-1.03; low CoE) but may result in little to no difference in mortality (RR = 0.89; 95% CI = 0.70-1.13; low CoE). In these RCTs, only apixaban and rivaroxaban were used in people receiving dialysis.

Figure 2.

Effect of DOACs against VKAs for (A) SSE events, (B) major bleedings, and (C) mortality.

Five observational studies reported SSE events and found no statistically significant differences between DOACs and VKAs.^{29,63,67,96,97} Five observational studies also reported major bleeding.^{29,63,67,96,97} One study reported an increase in major bleeding with dabigatran (HR = 1.48; 95% CI = 1.21-1.81) and rivaroxaban (HR = 1.38; 95% CI = 1.03-1.83).⁶³ However, we noted disproportionately high rates of major bleeding in this study (68.4-83.1 events/100 patients-years compared with 1.6-22.9 in the remaining studies). Two studies reported a significant decrease in major bleeding (HRs of 0.67 and 0.72).^96,97 One study found similar major bleeding risks with multiple DOACs and VKAs.²⁹ Another study showed a signal toward a reduction in major bleeding with apixaban (HR = 0.29; 95% CI = 0.07-1.18) and rivaroxaban (HR = 0.69; 95% CI = 0.17-2.87) and an increase with dabigatran (HR = 1.82; 95% CI = 0.67-4.98).⁶³ Finally, the 2 observational studies reporting mortality data showed a decrease in deaths only with the usual dosing regimen of apixaban in one study (HR = 0.85; 95% CI = 0.78-0.92) and a non-statistically decrease in another study (HR = 0.85; 95% CI = 0.71-1.01), which may be the result of a selection bias.^96,97 Therefore, the observational evidence did not alter our conclusions from the trial-based evidence.

Antiplatelet agents

We included 4 observational studies (4 comparisons, 4488 participants) comparing antiplatelet agents with no treatment^65,69
-71 and 4 observational studies (4 comparisons, 15 989 participants) comparing antiplatelet agents with VKAs.^65,67,71,72 The mean age was 72 years old. Two studies reported mean CHA₂DS₂-VASc scores of 3.9 and 4.0,^65,72 and 1 study reported a mean CHA₂DS₂ score of 2.4.⁶⁷ No studies that included participants on VKAs reported the TTR.

Antiplatelet agents vs no treatment

Compared with no treatment, the effect of antiplatelet agents on SSE events (4 studies)^65,69
-71 and mortality (2 studies)^65,71 was very uncertain (very low CoE), but they may increase the risk of bleeding (1 study, low CoE).⁷⁰ Two studies reported a non-significant decrease in SSE events (HRs of 0.88 and 0.83) with antiplatelet agents.^69,70 One study reported fewer SSE events with aspirin or clopidogrel without providing a measure of association.⁶⁵ One showed a statistically significant increase in ischemic strokes (HR = 2.01; 95% CI = 1.04-3.88).⁷¹ Only 1 study reported major bleeding and showed an increased risk with aspirin (HR = 1.63; 95% CI = 1.18-2.26).⁷⁰ This study was considered sufficiently at low risk of bias to conclude that antiplatelet agents may increase major bleeding. Finally, 2 studies reporting mortality data reported no statistically significant differences in mortality rates.^65,71

Antiplatelet agents vs vitamin K antagonists

Compared with VKAs, the effect of antiplatelet agents on SSE events (4 studies),^65,67,71,72 major bleeding (1 study),⁶⁷ and mortality (2 studies)^65,71 was very uncertain (very low CoE). One study showed a decrease in SSE events (RR = 0.81; 95% CI = 0.66-0.99) and major bleeding (RR = 0.78; 95% CI = 0.72-0.84) with aspirin, but its high incidence of bleeding and lack of adjustment for important confounders call for caution in interpreting its results.⁶⁷ Another study showed an exaggerated decrease in ischemic strokes with VKAs compared with aspirin (HR = 0.16; 95% CI = 0.04-0.66), but the patients taking aspirin were older, sicker, and had higher thromboembolic risks.⁷² Another study showed no statistical differences in ischemic strokes (HR = 1.58; 95% CI = 0.50-4.98) or mortality (HR = 0.90; 95% CI = 0.53-1.51) with aspirin.⁷¹ The last study (abstract-only, NOS at very high risk of bias) showed similar SSE events and mortality rates when taking antiplatelet agents (aspirin or clopidogrel) or VKAs.⁶⁵

Left atrial appendage closure

We included 2 observational studies (3 comparisons, 381 patients) comparing LAAC with no treatment or oral anticoagulation and found that the effect of LAAC on all outcomes was very uncertain (very low CoE).^58,66 The nonrandomized study design, small sample sizes and low number of events limited the results. The first study showed a non-statistically significant decrease in ischemic strokes (0% vs 7.7%) and mortality (22% vs 41%) compared with a historical cohort on oral anticoagulation (choice of anticoagulants unspecified).⁶⁶ The second one reported an increase in SSE events (10.8% vs 2.2%, P = .021), mortality (HR = 3.09; 95% CI = 1.59-5.98) and major bleeding (HR = 4.87; 95% CI = 1.08-21.97) of no treatment compared with LAAC.⁵⁸ Similar increased risks were seen when comparing warfarin with LAAC on mortality (HR = 2.76; 95% CI = 1.31-5.89), major bleeding (HR = 6.48; 95% CI = 1.32-31.72) and SSE events (7.0% vs 2.2%, P = .092).⁵⁸

Heparin-related therapies

We included 2 observational studies—one comparing enoxaparin (4000 units daily on non-hemodialysis days) with no treatment (1 comparison, 177 participants)⁶⁴ and one comparing heparin (unspecified agent) with no treatment and VKAs (2 comparisons, 1338 participants).⁵¹ We found that the effect of heparin-related therapies on all outcomes was very uncertain (very low CoE). The first study reported a statistically significant increase in mortality (HR = 1.88; 95% CI = 1.20-2.92) and non-statistically significant differences for major bleeding (HR = 1.86; 95% CI = 0.81-4.28) and SSE events (HR = 0.49; 95% CI = 0.11-2.18).⁶⁴ The second study, published in abstract-only, reported an increase in mortality with heparin (1.7 years of survival) compared with VKAs (2.8 years) and similar survival times compared with no treatment (1.8 years) but a decrease in SSE events with heparin (1.2%) compared with VKAs (2.9%) and no treatment (2.6%).⁵¹

Discussion

In this systematic review of thromboembolic prevention strategies for adults with AF receiving dialysis, we found that VKAs may be associated with a decrease in SSE events and an increase in major bleeding compared with no treatment. Previous systematic reviews in patients receiving dialysis found no clear association between VKAs and decreased SSE events.^13,22
-24 The discordance with our findings might be explained by several factors: (1) previously published systematic reviews ended their literature search between 2015 and 2019; (2) they included between 9 and 14 studies reporting SSE events with VKAs, whereas our systematic review included 37; and (3) they gave equal weights to all included studies while we gave proportionally more importance to the results of studies at lower risk of bias, including the first RCT of VKAs against placebo. Furthermore, many of the VKAs studies had insufficient power to detect a significant change in SSE events, included a high proportion of participants with low or unreported CHA₂DS₂ or CHA₂DS₂-VASc scores, and reported frequent subtherapeutic INR or TTR among VKAs users. Others have suggested that vascular calcifications, promoted by VKAs, are the main driver of SSE events in this population, which may explain the lack of association between VKAs use and stroke incidence in people with AF receiving dialysis.^14,99,100 We observed a decreased risk in SSE events within the studies at low risk of bias controlling for thromboembolic risk, which may suggest the findings of previous systematic reviews were influenced by studies at higher risk of bias or more prone to confounding. Although we concluded that VKAs may reduce SSE events, an individualized analysis and discussion of the benefits and risks of oral anticoagulation remain critical considering the substantial risk of major bleeding in this population. The major hurdle facing clinicians is to appropriately select patients at high thromboembolic risk, as validation studies in the dialysis setting showed inconsistent predictive value for the CHADS2 or CHA2DS2-VASc (c-statistics ranging from 0.61 to 0.88 with wide CIs).^88,101,102 Given the quality of the evidence and the challenges in balancing risk of ischemic stroke and major bleeding, we can only recommend that clinicians engage in a shared decision-making process using a patient decision aid (such as the one being developed by our team), rely on their clinical judgment to individualize stroke and bleeding risks (without providing absolute thresholds for the CHADS2, CHA2DS2-VASc, or HAS-BLED), and openly discuss with patients the studies’ limitations in the dialysis setting.

Our systematic review suggests that DOACs may be as effective as VKAs to reduce SSE events and may be associated with less major bleeding, but the quality of the evidence was low to very low. To our knowledge, we conducted the first meta-analysis of DOACs against VKAs to include the newest findings from the AXADIA-AFNET 8 and SAFE-D trials.^50,52 Previously published systematic reviews of mostly observational studies concluded that apixaban may decrease^25
-28 or have no effect^29,30 on major bleeding compared with VKAs. Our findings support the inclusion of DOACs as an SSE event prevention strategy in the shared decision-making process with patients. The choice of DOAC in dialysis should be limited to apixaban and rivaroxaban, considering the absence of data for edoxaban and the worrying rates of major bleeding with dabigatran.⁶⁷ A reduced dosing regimen of rivaroxaban (10 mg daily) was used in the Valkyrie study and produced similar pharmacokinetics endpoints in dialysis compared with the standard dose in healthy patients.^51,103 When selecting apixaban, the question of the dosing regimen remains controversial, as no dosing regimen was clearly superior in RENAL-AF, AXADIA-AFNET 8, and SAFE-D.^50,52,53 Pharmacokinetic studies of repeated apixaban doses in hemodialysis inconsistently showed supratherapeutic levels with the usual dosage.^53,104

This systematic review found the level of evidence for antiplatelet agents, heparin-related therapies, and LAAC to be very low. These therapies should, therefore, not be routinely used for stroke prevention in people with AF receiving dialysis. Our findings align with previous guidelines and findings from systematic reviews in the general population to prioritize oral anticoagulation over antiplatelet therapy.^6,9
-12 The LAAC appears to be a potentially interesting alternative to oral anticoagulation in closely selected patients with prohibitively high thromboembolic and bleeding risks. Still, the risks of LAAC in the dialysis population (eg, pericardial effusion, device embolization, procedure-related thrombus) remain to be defined.¹⁰⁵ A prospective observational study (WATCH-HD) comparing LAAC with no treatment is underway.¹⁰⁶

To our knowledge, this is the first systematic review assessing all stroke-prevention interventions for patients with AF receiving dialysis. Its strengths include an extensive search strategy, independent double data extraction, and risk of bias assessment. We included subgroup analyses of data at low risk of bias for the VKAs group and only conducted meta-analyses when appropriate. However, this systematic review also has some significant limitations. The included studies were mostly retrospective cohort studies at moderate to high risk of bias, including classification mistakes (eg, in using procedure codes), confounding by indication (eg, patients receiving anticoagulation at higher thromboembolic risk), and residual confounding (eg, TTR for warfarin and over-the-counter use of aspirin). Indeed, only a third of the observational studies controlled for both antiplatelet agents and CHA₂DS₂-VASc or CHADS₂ score, thus limiting our confidence in the effect size reported in many studies. We also encountered high statistical heterogeneity for the effect of VKAs on SSE events, which prevented robust conclusions on their benefit over no treatment in people receiving dialysis. Two upcoming RCTs comparing anticoagulation with placebo (DANWARD and AVKDIAL) may help answer this question.^107,108 Finally, as per protocol, we did not compare the different DOACs’ dosing regimens, preventing us from providing definite conclusions on the optimal dosing.

Conclusion

Our systematic review showed that VKAs may reduce the risk of SSE events in people with AF receiving dialysis at the cost of an increase in major bleeding. The DOACs (apixaban or rivaroxaban) may be as effective as VKAs in reducing SSE events with possibly fewer bleeding events. However, all but 1 comparison had a low to very low CoE, reminding clinicians to openly discuss values and preferences with patients to guide clinical decisions. Until well-designed RCTs comparing DOACs, warfarin, and placebo help inform clinicians and patients, engaging in a shared decision-making process is the 1 intervention we can confidently recommend for managing AF-related ischemic events in people receiving hemodialysis.

Supplemental Material

sj-docx-1-cjk-10.1177_20543581241290316 – Supplemental material for Stroke Prevention in Patients With Atrial Fibrillation Receiving Dialysis: A Systematic Review

Supplemental material, sj-docx-1-cjk-10.1177_20543581241290316 for Stroke Prevention in Patients With Atrial Fibrillation Receiving Dialysis: A Systematic Review by Olivier Massé, Gabriel Dallaire, Noémie Maurice, Yu Hong, Claudia Mei Mercurio, Catherine Tremblay, Nicolas Dugré and Amélie Bernier-Jean in Canadian Journal of Kidney Health and Disease

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Olivier Massé

Gabriel Dallaire

Supplemental Material

Supplemental material for this article is available online.

References

Findlay

Thomson

Fulton

, et al. Risk factors of ischemic stroke and subsequent outcome in patients receiving hemodialysis. Stroke. 2015;46(9):2477-2481.

Königsbrügge

Posch

Antlanger

, et al. Prevalence of atrial fibrillation and antithrombotic therapy in hemodialysis patients: cross-sectional results of the Vienna InVestigation of AtriaL Fibrillation and Thromboembolism in Patients on HemoDIalysis (VIVALDI). PLoS ONE. 2017;12(1):e0169400.

United States Renal Data System. 2021 USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States. Bethesda, MD: National Institutes of Health. National Institute of Diabetes and Digestive and Kidney Diseases; 2021.

Zimmerman

Sood

Rigatto

Holden

Hiremath

Clase

. Systematic review and meta-analysis of incidence, prevalence and outcomes of atrial fibrillation in patients on dialysis. Nephrol Dial Transplant. 2012;27(10):3816-3822.

Masson

Kelly

Craig

Lindley

Webster

. Risk of stroke in patients with ESRD. Clin J Am Soc Nephrol. 2015;10(9):1585-1592.

Hart

Pearce

Aguilar

. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146(12):857-867.

Ruff

Giugliano

Braunwald

, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomized trials. Lancet. 2014;383(9921):955-962.

Kimachi

Furukawa

Kimachi

Goto

Fukuma

Fukuhara

. Direct oral anticoagulants versus warfarin for preventing stroke and systemic embolic events among atrial fibrillation patients with chronic kidney disease. Cochrane Database Syst Rev. 2017;11(11):1-51.

Andrade

Aguilar

Atzema

, et al. The 2020 Canadian Cardiovascular Society/Canadian Heart Rhythm Society comprehensive guidelines for the management of atrial fibrillation. Can J Cardiol. 2020;36(12):1847-1948.

10.

Lip

GYH

Banerjee

Boriani

, et al. Antithrombotic therapy for atrial fibrillation: CHEST guideline and expert panel report. Chest. 2018;154(5):1121-1201.

11.

Hindricks

Potpara

Dagres

, et al. 2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): the task force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42(5):373-498.

12.

January

Wann

Calkins

, et al. AHA/ACC/HRS Focused Update of the 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines and the heart rhythm society in collaboration with the society of thoracic surgeons. Circulation. 2019;140(2):e125-e151.

13.

Randhawa

Vishwanath

Rai

, et al. Association between use of warfarin for atrial fibrillation and outcomes among patients with end-stage renal disease: a systematic review and meta-analysis. JAMA Netw Open. 2020;3(4):1-13.

14.

Wong

Odutayo

Emdin

, et al. Meta-analysis of anticoagulation use, stroke, thromboembolism, bleeding, and mortality in patients with atrial fibrillation on dialysis. Am J Cardiol. 2016;117(12):1934-1941.

15.

Hayashi

Takamatsu

Kanno

, et al. A case-control study of calciphylaxis in Japanese end-stage renal disease patients. Nephrol Dial Transplant. 2012;27(4):1580-1584.

16.

Levy

Grewal

. Vitamin K deficiency: an emerging player in the pathogenesis of vascular calcification and an iatrogenic consequence of therapies in advanced renal disease. Am J Physiol Renal Physiol. 2020;319:F618-F623.

17.

Benjamin

Nisha

. Should oral anticoagulation be used in ESKD patients on hemodialysis with atrial fibrillation?: CON. Kidney360. 2021;2(9):1409-1411.

18.

Chan

Giugliano

Patel

, et al. Nonvitamin K anticoagulant agents in patients with advanced chronic kidney disease or on dialysis with AF. J Am Coll Cardiol. 2016;67(24):2888-2899.

19.

Halperin

Lee

Liew

, et al. Anticoagulation for patients with atrial fibrillation and end-stage renal disease on dialysis: a national survey. Can J Cardiol. 2021;37(6):924-928.

20.

United States Renal Data System. 2018 USRDS annual data report: epidemiology of kidney disease in the United States. Published 2018. Accessed October 20, 2022. https://www.usrds.org/annual-data-report

21.

Zhang

, et al. Efficacy and safety of left atrial appendage occlusion in atrial fibrillation patients with chronic kidney disease: a systematic review and meta-analysis. Rev Cardiovasc Med. 2020;21(3):443-451.

22.

Nochaiwong

Ruengorn

Awiphan

Dandecha

Noppakun

Phrommintikul

. Efficacy and safety of warfarin in dialysis patients with atrial fibrillation: a systematic review and meta-analysis. Open Heart. 2016;3(1):1-14.

23.

Van Der Meersch

De Bacquer

De Vriese

. Vitamin K antagonists for stroke prevention in hemodialysis patients with atrial fibrillation: a systematic review and meta-analysis. Am Heart J. 2017;184:37-46.

24.

Harel

Chertow

Shah

, et al. Warfarin and the risk of stroke and bleeding in patients with atrial fibrillation receiving dialysis: a systematic review and meta-analysis. Can J Cardiol. 2017;33(6):737-746.

25.

Feldberg

Patel

Farrell

, et al. A systematic review of direct oral anticoagulant use in chronic kidney disease and dialysis patients with atrial fibrillation. Nephrol Dial Transplant. 2019;34(2):265-277.

26.

Kuno

Takagi

Ando

, et al. Oral anticoagulation for patients with atrial fibrillation on long-term dialysis. J Am Coll Cardiol. 2020;75(3):273-285.

27.

Abdullah

Ullah

Jafar

, et al. Safety and efficacy of apixaban, rivaroxaban, and warfarin in end-stage renal disease with atrial fibrillation: a systematic review and meta-analysis. Cardiovasc Revasc Med. 2021;30:26-32.

28.

Chen

Huang

, et al. Efficacy and safety of direct oral anticoagulants vs warfarin in patients with chronic kidney disease and dialysis patients: a systematic review and meta-analysis. Clin Drug Investig. 2021;41(4):341-351.

29.

See

Lee

Chao

, et al. Effectiveness and safety of direct oral anticoagulants in an Asian population with atrial fibrillation undergoing dialysis: a population-based cohort study and meta-analysis. Cardiovasc Drugs Ther. 2021;35(5):975-986.

30.

Chen

Cao

Zheng

, et al. Effect of rivaroxaban or apixaban in atrial fibrillation patients with stage 4-5 chronic kidney disease or on dialysis. Cardiovasc Drugs Ther. 2021;35(2):273-281.

31.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

32.

Masse

Maurice

Dallaire

, et al. Stroke prevention in patients with atrial fibrillation and dialysis: a systematic review. PROSPERO 2022 CRD42022307009. Accessed October 7, 2024. https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022307009

33.

Bernier-Jean

Beruni

Bondonno

, et al. Exercise training for adults undergoing maintenance dialysis. Cochrane Database Syst Rev. 2022;1(1):1-304.

34.

Valembois

Audureau

Takeda

Jarzebowski

Belmin

Lafuente-Lafuente

. Antiarrhythmics for maintaining sinus rhythm after cardioversion of atrial fibrillation. Cochrane Database Syst Rev. 2019;9(9):1-200.

35.

Wang

Ouyang

Yang

Song

Yang

Anderson

. Anticoagulants for acute ischaemic stroke. Cochrane Database Syst Rev. 2021;10(10):1-93.

36.

Katsanos

Kamel

Healey

Hart

. Stroke prevention in atrial fibrillation: looking forward. Circulation. 2020;142(24):2371-2388.

37.

Schulman

Kearon

; Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost. 2005;3(4):692-694.

38.

Higgins

JPT

Altman

Gøtzsche

, et al. The Cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:1-9.

39.

Wells

Shea

O’Connell

, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Published 2013. Accessed October 7, 2024. https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp

40.

Viswanathan

Ansari

Berkman

, et al. Assessing the risk of bias of individual studies in systematic reviews of health care interventions. Methods Guide for Effectiveness and Comparative Effectiveness Reviews. Rockville, MD: Agency for Healthcare Research and Quality; 2012.

41.

Guyatt

Oxman

Vist

, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924-926.

42.

Guyatt

Oxman

Santesso

, et al. GRADE guidelines: 12. Preparing summary of findings tables-binary outcomes. J Clin Epidemiol. 2013;66(2):158-172.

43.

Santesso

Glenton

Dahm

, et al. GRADE guidelines 26: informative statements to communicate the findings of systematic reviews of interventions. J Clin Epidemiol. 2020;119:126-135.

44.

Murad

Mustafa

Schünemann

Sultan

Santesso

. Rating the certainty in evidence in the absence of a single estimate of effect. Evid Based Med. 2017;22(3):85-87.

45.

Morgan

Yepes-Nuñez

Ewusie

, et al. Improving grading of recommendations assessment, development, and evaluation evidence tables part 4: a three-arm noninferiority randomized trial demonstrates improved understanding of content in summary of findings tables with a new format. J Clin Epidemiol. 2023;154:125-135.

46.

Cuello-Garcia

Santesso

Morgan

, et al. GRADE guidance 24 optimizing the integration of randomized and non-randomized studies of interventions in evidence syntheses and health guidelines. J Clin Epidemiol. 2022;142:200-208.

47.

Popay

Roberts

Sowden

, et al. Guidance on the conduct of narrative synthesis in systematic reviews: a product from the ESRC methods programme. Lancaster, UK: Lancaster University; 2006.

48.

Campbell

McKenzie

Sowden

, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ. 2020;368:1-6.

49.

Sterne

Sutton

Ioannidis

, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002.

50.

Harel

. Strategies for the management of atrial fibrillation in PatiEnts receiving dialysis (SAFE-D) [abstract]. ABSTRACT: TH-PO1140. Paper presented at the Proceedings of the Kidney Week 2023 Annual Meeting; November 2023; Philadelphia, PA. Accessed November 22, 2023. https://www.asn-online.org/education/kidneyweek/2023/program-abstract.aspx?controlId=3979056

51.

De Vriese

Caluwé

Van Der Meersch

De Boeck

De Bacquer

. Safety and efficacy of vitamin K antagonists versus rivaroxaban in hemodialysis patients with atrial fibrillation: a multicenter randomized controlled trial. J Am Soc Nephrol. 2021;32(6):1474-1483.

52.

Reinecke

Engelbertz

Bauersachs

, et al. A randomized controlled trial comparing apixaban to the vitamin K-antagonist phenprocoumon in patients on chronic hemodialysis: the AXADIA-AFNET 8 study. Circulation. 2023;147(4):296-309.

53.

Pokorney

Chertow

Al-Khalidi

, et al. Apixaban for patients with atrial fibrillation on hemodialysis: a multicenter randomized controlled trial. Circulation. 2022;146(23):1735-1745.

54.

Voskamp

PWM

Rookmaaker

Verhaar

Dekker

Ocak

. Vitamin K antagonist use and mortality in dialysis patients. Nephrol Dial Transplant. 2018;33(1):170-176.

55.

Wakasugi

Kazama

Tokumoto

, et al. Association between warfarin use and incidence of ischemic stroke in Japanese hemodialysis patients with chronic sustained atrial fibrillation: a prospective cohort study. Clin Exp Nephrol. 2014;18(4):662-669.

56.

Yamashita

Takagi

Hamatani

, et al. Clinical characteristics and outcomes of dialysis patients with atrial fibrillation: the Fushimi AF registry. Heart Vessels. 2016;31(12):2025-2034.

57.

Genovesi

Rebora

Gallieni

, et al. Effect of oral anticoagulant therapy on mortality in end-stage renal disease patients with atrial fibrillation: a prospective study. J Nephrol. 2017;30(4):573-581.

58.

Genovesi

Porcu

Slaviero

, et al. Outcomes on safety and efficacy of left atrial appendage occlusion in end stage renal disease patients undergoing dialysis. J Nephrol. 2021;34(1):63-73.

59.

Königsbrügge

Meisel

Beyer

, et al. Anticoagulation use and the risk of stroke and major bleeding in patients on hemodialysis: from the VIVALDI, a population-based prospective cohort study. J Thromb Haemost. 2021;19(12):2984-2996.

60.

Friberg

Benson

Lip

GYH

. Balancing stroke and bleeding risks in patients with atrial fibrillation and renal failure: the Swedish atrial fibrillation cohort study. Eur Heart J. 2015;136(5):297-306.

61.

Chan

Lazarus

Thadhani

Hakim

. Warfarin use associates with increased risk for stroke in hemodialysis patients with atrial fibrillation. J Am Soc Nephrol. 2009;20(10):2223-2233.

62.

Sajeev

Roberts

Teh

, et al. Clinical outcomes of atrial fibrillation (AF) and anticoagulation within the dialysis cohort. Heart Lung Circ. 2015;24(suppl 3):S378.

63.

Noseworthy

Yao

Tangri

Shah

Nath

. Anticoagulant use and associated outcomes in patients with atrial fibrillation and advanced kidney disease. Eur Heart J. 2017;38(suppl 1):1-2.

64.

Dimitrijevic

Salinger

Tasic

Stojanovic

Paunovic

Mitic

. Clinical impact of oral anticoagulation therapy on mortality, embolic and hemorrhagic events in chronic hemodialysis patients with atrial fibrillation. Nephrol Dial Transplant. 2021;36(suppl 1):1.

65.

Brensing

Ochsmann

Omran

Lonnemann

Reichel

Duttlinger

. Matched control study on applied long-term anti-thrombotic therapies and clinical outcome of atrial fibrillation in HD-patients based on German network data. Nephrol Dial Transplant. 2022;37(suppl 3):1-2.

66.

Romero

Diaz

Purkayastha

, et al. Outcomes of left atrial appendage closure in patients with atrial fibrillation and end-stage renal disease on dialysis. Heart Rhytm. 2022;19(5):S439.

67.

Chan

Edelman

Wenger

Thadhani

Maddux

. Dabigatran and rivaroxaban use in atrial fibrillation patients on hemodialysis. Circulation. 2015;131(11):972-979.

68.

Chang

Yeh

, et al. Efficacy and safety of oral anticoagulants in patients with atrial fibrillation and stages 4 or 5 chronic kidney disease. Am J Med. 2019;132(11):1335-1343.

69.

Chen

Lin

Yang

Hwang

Chen

Lin

. Anti-platelet or anti-coagulant agent for the prevention of ischemic stroke in patients with end-stage renal disease and atrial fibrillation—a nation-wide database analyses. Int J Cardiol. 2014;177(3):1008-1011.

70.

Olesen

Lip

GYH

Kamper

, et al. Stroke and bleeding in atrial fibrillation with chronic kidney disease. N Engl J Med. 2012;367(7):625-635.

71.

Lai

Hsieh

Chou

, et al. Hazards of antithrombotic therapy on hemodialysis patients with atrial fibrillation and high thromboembolic risk: a Taiwanese population-based cohort study. Int J Clin Exp Med. 2017;10:13982-13991.

72.

Chan

Huang

Yip

, et al. Ischaemic stroke in patients with atrial fibrillation with chronic kidney disease undergoing peritoneal dialysis. Europace. 2016;18(5):665-671.

73.

Agarwal

Potukuchi

Sumida

, et al. Clinical outcomes of warfarin initiation in advanced chronic kidney disease patients with incident atrial fibrillation. JACC Clin Electrophysiol. 2020;6(13):1658-1668.

74.

Akbar

Febrianora

Iqbal

. Warfarin usage in patients with atrial fibrillation undergoing hemodialysis in Indonesian population. Curr Probl Cardiol. 2023;48(5):1-15.

75.

Brancaccio

Neri

Bellocchio

, et al. Patients’ characteristics affect the survival benefit of warfarin treatment for hemodialysis patients with atrial fibrillation. A historical cohort study. Am J Nephrol. 2016;44(4):258-267.

76.

Garg

Chen

Haines

. Atrial fibrillation and chronic kidney disease requiring hemodialysis-does warfarin therapy improve the risks of this lethal combination? Int J Cardiol. 2016;222:47-50.

77.

Khalid

Qureshi

Alirhayim

Garikapati

Patsias

. Impact of restarting warfarin therapy in renal disease anticoagulated patients with gastrointestinal hemorrhage. Ren Fail. 2013;35(9):1228-1235.

78.

Lai

Aronow

Kalen

, et al. Incidence of thromboembolic stroke and of major bleeding in patients with atrial fibrillation and chronic kidney disease treated with and without warfarin. Int J Nephrol Renovasc Dis. 2009;2:33-37.

79.

Mitsuma

Matsubara

Hatada

, et al. Clinical characteristics of hemodialysis patients with atrial fibrillation: the RAKUEN (Registry of atrial fibrillation in chronic kidney disease under hemodialysis from Niigata) study. J Cardiol. 2016;68(2):148-155.

80.

Phan

Yang

Shen

Lee

. Effect of warfarin on ischemic stroke, bleeding, and mortality in patients with atrial fibrillation receiving peritoneal dialysis. Am J Cardiovasc Drugs. 2019;19(5):509-515.

81.

Sánchez Soriano

Albero Molina

Chamorro Fernández

, et al. Long-term prognostic impact of anticoagulation on patients with atrial fibrillation undergoing hemodialysis. Nefrologia (Engl Ed). 2018;38(4):394-400.

82.

Wenziger

Marroquin

Kalantar-Zadeh

Kovesdy

Streja

. Warfarin use, stroke, and bleeding risk among pre-existing atrial fibrillation US veterans transitioning to dialysis. Nephron. 2022;146(4):360-368.

83.

Shah

Avgil Tsadok

Jackevicius

, et al. Warfarin use and the risk for stroke and bleeding in patients with atrial fibrillation undergoing dialysis. Circulation. 2014;129(11):1196-1203.

84.

Shen

Montez-Rath

Lenihan

Turakhia

Chang

Winkelmayer

. Outcomes after warfarin initiation in a cohort of hemodialysis patients with newly diagnosed atrial fibrillation. Am J Kidney Dis. 2015;66(4):677-688.

85.

Tan

Bae

Segal

, et al. Warfarin use and the risk of stroke, bleeding, and mortality in older adults on dialysis with incident atrial fibrillation. Nephrology (Carlton). 2019;24(2):234-244.

86.

Waddy

Solomon

Becerra

, et al. Racial/ethnic disparities in atrial fibrillation treatment and outcomes among dialysis patients in the United States. J Am Soc Nephrol. 2020;31(3):637-649.

87.

Kai

Bogorad

Nguyen

, et al. Warfarin use and the risk of mortality, stroke, and bleeding in hemodialysis patients with atrial fibrillation. Heart Rhythm. 2017;14(5):645-651.

88.

Wang

Sathananthan

Marshall

Kerr

Hood

. Relationships between anticoagulation, risk scores and adverse outcomes in dialysis patients with atrial fibrillation. Heart Lung Circ. 2016;25(3):243-249.

89.

Welander

Renlund

Dimény

Holmberg

Själander

. Efficacy and safety of warfarin in patients with non-valvular atrial fibrillation and CKD G3-G5D. Clin Kidney J. 2022;15(6):1169-1178.

90.

Abeyaratne

Wong

YHS

Majoni

. Benefit and harm of anticoagulation in the prevention of thromboembolic stroke for non-valvular atrial fibrillation in haemodialysis patients: a top end of Northern Australia study. Intern Med J. 2023;53(4):517-524.

91.

Yodogawa

Mii

Fukui

, et al. Warfarin use and incidence of stroke in Japanese hemodialysis patients with atrial fibrillation. Heart Vessels. 2016;31(10):1676-1680.

92.

Yoon

Noh

Jhee

, et al. Warfarin use in patients with atrial fibrillation undergoing hemodialysis: a nationwide population-based study. Stroke. 2017;48(9):2472-2479.

93.

Lee

Liu

, et al. Association of warfarin with congestive heart failure and peripheral artery occlusive disease in hemodialysis patients with atrial fibrillation. J Chin Med Assoc. 2017;80(5):277-282.

94.

Winkelmayer

Liu

Setoguchi

Choudhry

. Effectiveness and safety of warfarin initiation in older hemodialysis patients with incident atrial fibrillation. Clin J Am Soc Nephrol. 2011;6(11):2662-2668.

95.

Mavrakanas

Garlo

Charytan

. Apixaban versus no anticoagulation in patients undergoing long-term dialysis with incident atrial fibrillation. Clin J Am Soc Nephrol. 2020;15(8):1146-1154.

96.

Siontis

Zhang

Eckard

, et al. Outcomes associated with apixaban use in patients with end-stage kidney disease and atrial fibrillation in the United States. Circulation. 2018;138(15):1519-1529.

97.

Wetmore

Weinhandl

Yan

Reyes

Herzog

Roetker

. Apixaban dosing patterns versus warfarin in patients with nonvalvular atrial fibrillation receiving dialysis: a retrospective cohort study. Am J Kidney Dis. 2022;80(5):569-579.

98.

Kang

Choi

Kwon

, et al. Clinical outcomes among hemodialysis patients with atrial fibrillation: a Korean nationwide population-based study. Kidney Res Clin Pract. 2021;40(1):99-108.

99.

Yang

Chou

Schweitzer

Hanon

. Warfarin in haemodialysis patients with atrial fibrillation: what benefit? Europace. 2010;12(12):1666-1672.

100.

Danziger

. Vitamin K-dependent proteins, warfarin, and vascular calcification. Clin J Am Soc Nephrol. 2008;3(5):1504-1510.

101.

de Jong

Ramspek

van

der

Endt

VHW

, et al. A systematic review and external validation of stroke prediction models demonstrates poor performance in dialysis patients. J Clin Epidemiol. 2020;123:69-79.

102.

Chao

Liu

Wang

, et al. Incidence and prediction of ischemic stroke among atrial fibrillation patients with end-stage renal disease requiring dialysis. Heart Rhythm. 2014;11(10):1752-1759.

103.

De Vriese

Caluwé

Bailleul

, et al. Dose-finding study of rivaroxaban in hemodialysis patients. Am J Kidney Dis. 2015;66(1):91-98.

104.

Mavrakanas

Samer

Nessim

Frisch

Lipman

. Apixaban pharmacokinetics at steady state in hemodialysis patients. J Am Soc Nephrol. 2017;28(7):2241-2248.

105.

Vallurupalli

Sharma

Al’Aref

Devabhaktuni

Dhar

. Left atrial appendage closure: an alternative to anticoagulation for stroke prevention in patients with kidney disease. Kidney360. 2021;3(2):396-402.

106.

ClinicalTrials.gov. Left atrial appendage occlusion with WATCHMAN^® device in patients with non-valvular atrial fibrillation and end-stage chronic kidney disease on hemodialysis (WATCH-HD). ClinicalTrials.gov identifier: NCT03446794. Updated June 24, 2022. https://clinicaltrials.gov/study/NCT03446794?id=NCT03446794&rank=1. Accessed November 22, 2023.

107.

ClinicalTrials.gov. The Danish warfarin-dialysis study-safety and efficacy of warfarin in patients with atrial fibrillation on dialysis (DANWARD). ClinicalTrials.gov identifier: NCT03862859. Updated December 19, 2023. https://clinicaltrials.gov/study/NCT03862859?id=NCT03862859&rank=1.Accessed February 9, 2024.

108.

ClinicalTrials.gov. Oral anticoagulation in haemodialysis patients (AVKDIAL). ClinicalTrials.gov identifier: NCT02886962. Updated August 21, 2023. https://clinicaltrials.gov/study/NCT02886962?id=NCT02886962&rank=1. Accessed February 9, 2024.