Table 3.
Stroke, bleeding and mortality outcomes reported in included studies
| Author Year | Adjustment for confounding | Outcome | Outcome definition | Study groups | Number of patients with ESRD and AF | Number of outcome events | Rates per 100 PY (95 % CI) | Adjusted HR (95 % CI) |
|---|---|---|---|---|---|---|---|---|
| Chan 2009e [30] | Adjusted for and PS matched on: CHADS2, gender, race, Charlson comorbidity index, entry date, access, body mass index, facility standardized mortality ratio, cardiovascular drugs, dialysis adequacy, baseline lab values, heparin dosage and heparin regimen | Any stroke | Hospitalization and death for stroke and TIAs identified from diagnoses obtained from hospital discharge summaries and cause of death from medical records | W (warfarin) C (no warfarin) |
746 925 |
NR NR |
7.1 (5.7, 8.7) 2.9 (2.0, 4.4) |
1.74 (1.11, 2.72)c |
| Ischemic stroke | W (warfarin) C (no warfarin) |
746 925 |
NR NR |
NR NR |
1.81 (1.12, 2.92) | |||
| Hemorrhagic stroke | W (warfarin) C (no warfarin) |
746 925 |
NR NR |
NR NR |
2.22 (1.01, 4.91) | |||
| Major bleeding | Hospitalization for bleeding | W (warfarin) C (no warfarin) |
746 925 |
97 107 |
0.09 (NR) 0.07 (NR) |
1.04 (0.73, 1.46) | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
746 925 |
333 425 |
27.4 (NR) 25.7 (NR) |
1.10 (0.93, 1.30) | ||
| Lai 2009e [32] | Adjusted for: gender, age, GFR, hemodialysis/transplant, aspirin, stroke history, heart and artery disease, smoking, hypertension, diabetes, dyslipidemia | Ischemic stroke | Thromboembolic stroke | W (warfarin) | HD: 51; GFR: 78 |
HD: 5; GFR: 8 |
HD: 10 %b
GFR: 10 % |
NR |
| C (no warfarin) | HD: 42; GFR: 54 |
HD: 16 GFR: 20 |
HD: 38 % GFR: 37 % |
|||||
| Wizemann 2010 [33] | Adjusted for: age category, sex, race, years with ESRD, study phase, history of stroke, comorbid conditions, permanent pacemaker implanted, history of cardiac arrest, left ventricular hypertrophy, valvular heart disease | Any stroke | Hospitalized for stroke or if they died with cause of death listed as ‘cerebrovascular accident (including intracranial hemorrhage) | W (warfarin) C (no warfarin) |
509 2736 |
NR NR |
NR NR |
age ≤ 65: 1.29 (0.45, 3.68) age 66–75: 1.35 (0.69, 2.63) age > 75 years: 2.17 (1.04, 4.53) |
| Winkelmayer 2011 [22] | PS matched on: age, gender, race, state, dialysis vintage, dialysis type, comorbidity, vascular access surgery, length of stay, number of hospital days in prior year, number of medications used in prior year, H2 blocker or proton pump inhibitor use, prior nursing home stay, hematocrit, erythropoietin | Any stroke | Ischemic and hemorrhagic stroke identified by ICD-9 | W (warfarin) C (no warfarin) |
237 948 |
38 150 |
9.7 8.7 |
1.08c (0.76, 1.55) |
| Ischemic stroke | ICD-9: 433.x1, 434.x1, 436 | W (warfarin) C (no warfarin) |
237 948 |
29 135 |
7.4 7.8 |
0.92c (0.61, 1.37) | ||
| Hemorrhagic stroke | ICD-9: 430–432 | W (warfarin) C (no warfarin) |
237 948 |
11 121 |
2.6 7.5 |
2.38 (1.15, 4.96) | ||
| Gastrointestinal bleed | Previously validated claims-based algorithm | W (warfarin) C (no warfarin) |
237 948 |
48 216 |
13.4 13.6 |
0.96c (0.70, 1.31) | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
237 948 |
181 750 |
42.9 40.2 |
1.06c (0.90, 1.24) | ||
| Olesen 2012 [23] | Adjusted for: changes in renal status or antithrombotic treatment during follow up, risk factors in the CHA2DS2-VASc score, antithrombotic treatment, year of inclusion | Any stroke | Hospitalization or death from stroke or systemic thromboembolism (peripheral-artery embolism, ischemic stroke and transient ischemic attack) | W (warfarin) C (no warfarin) |
NR NR |
NR NR |
NR NR |
0.44 (0.26, 0.74) |
| Major bleeding | Bleeding (gastrointestinal, intracranial, urinary tract and airway bleeding) | W (warfarin) C (no warfarin) |
NR NR |
NR NR |
NR NR |
1.27 (0.91, 1.77) | ||
| Khalid 2013 [34] | PS adjusted for: age, gender, race, fresh frozen plasma, vitamin K, Charlson comorbidity index, cancer, INR, TTR, heart failure, aspirin, NSAIDs and clopidogrel use, ICU stay, blood transfusions | Ischemic stroke | thromboembolism | W (restarted warfarin) C (did not restart warfarin) |
34 62 |
NR NR |
NR NR |
0.44 (0.27, 0.73) |
| Gastrointestinal bleed | NR | W (restarted warfarin) C (did not restart warfarin) |
34 62 |
NR NR |
NR NR |
1.72 (1.29, 2.30) | ||
| All-cause mortality | - | W (restarted warfarin) C (did not restart warfarin) |
34 62 |
NR NR |
NR NR |
0.22 (0.13, 0.40) | ||
| Wakasugi 2014 [29] | Adjusted for: CHADS2 score; PS matched on: age, gender, BMI, duration of dialysis, cause of ESRD, vascular access, medical history, medication, comorbidity, blood pressure, lab data, mobility | Ischemic stroke | New ischemic stroke (fatal or nonfatal) not including TIA | W (warfarin) C (no warfarin) |
28 32 |
8 5 |
14.8 (6.4, 29.2) 8.9 (2.9, 20.8) |
3.36c (0.67, 16.66) |
| Major bleeding | fatal bleeding or bleeding that required hospitalization | W (warfarin) C (no warfarin) |
28 32 |
3 4 |
5.3 (1.1, 15.5) 6.6 (1.8, 17.0) |
0.85d (0.19, 3.64) | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
28 32 |
9 9 |
14.2 (6.5, 26.9) 14.2 (6.5, 26.9) |
1.00d (0.40, 2.52) | ||
| Bonde 2014 [24] | Adjusted for: aspirin treatment, stroke/TE comorbidity, concomitant medication, CHA2DS2-VASc score, bleeding comorbidity, HAS-BLED score | Any stroke | Hospitalization/death from stroke/TE (i.e., peripheral arterial embolism, ischemic stroke and transient ischemic attack) | W (warfarin) C (no warfarin) |
260 882 |
41a,f
186a,f |
NR NR |
NR |
| All-cause mortality | - | W (warfarin) C (no warfarin) |
NR NR |
NR NR |
NR NR |
Low/intermediate risk: 1.36 (0.96, 1.94) | ||
| High risk: 0.85 (0.72, 0.99) | ||||||||
| Carrero 2014 [25] | Adjusted for and PS matched on: age, sex, eGFR, comorbidities, patient presentation characteristics at admission, hospital course, discharge medication, center effect, left ventricular ejection fraction | Ischemic stroke | Hospitalizations due to ischemic stroke identified from claims | W (warfarin) C (no warfarin) |
66 412 |
0 16 |
NR NR |
MF |
| Major bleeding | Readmission due to hemorrhagic stroke, gastrointestinal bleeding, bleeding causing anemia identified from claims | W (warfarin) C (no warfarin) |
66 412 |
4 34 |
9.1 13.5 |
0.46c (0.11, 1.89) | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
66 412 |
32 222 |
NR NR |
0.72 (0.46, 1.14) | ||
| Chen 2014 [26] | Adjusted for and PS matched on: age, gender, dialysis method, diabetes, risk factors, comorbidities, concomitant medication | Any stroke | Ischemic stroke, transient ischemic accident or hemorrhagic stroke identified by ICD9 codes | W (warfarin) C (no warfarin) |
294 (250 analyzed) 2983 (250 analyzed) |
NR NR |
5.1%b
6.6 % |
1.017c (0.673, 1.537) |
| Friberg 2014e [13] | Adjusted for: renal failure, age, sex, year of inclusion, duration since first AF diagnosis, previous thrombo-embolism, venous thrombo-embolism, intracranial bleeding, anaemia, coagulopathy or platelet defect, vascular disease, heart failure, pericarditis, other valvular disease, pacemaker, comorbidity, baseline use of medication | Ischemic stroke | Principal or first secondary diagnosis code for ischemic stroke | W (warfarin) C (no warfarin) |
3766 9669 |
NR NR |
2.7 (2.3, 3.1) 4.6 (4.2, 4.9) |
0.687 (0.584, 0.807)h |
| Any stroke | Codes related to arterial or venous thromboembolism regardless of coding position | W (warfarin) C (no warfarin) |
3766 9669 |
NR NR |
6.2 (5.6, 6.8) 9.2 (8.7, 9.8) |
0.779 (0.697, 0.870)h | ||
| Intracranial bleeding | Principal or first secondary diagnosis codes for haemorrhage | W (warfarin) C (no warfarin) |
3766 9669 |
NR NR |
1.0 (0.8, 1.2) 0.7 (0.6, 0.9) |
1.557 (1.149, 2.110)h | ||
| Any bleeding | Codes related to intracranial, gastrointestinal and other bleeding | W (warfarin) C (no warfarin) |
3766 9669 |
NR NR |
9.6 (8.9, 10.4) 9.8 (9.3, 10.4) |
1.096 (0.995, 1.206)h | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
3766 9669 |
NR NR |
24.7 (23.5, 25.8) 41.7 (40.7, 42.8) |
0.747 (0.708, 0.788)h | ||
| Shah 2014 [27] | PS adjusted for: age, sex, specific components of CHADS2 or HAS-BLED score | Ischemic stroke | ischemic cerebrovascular disease including transient ischemic attack (TIA) and retinal infarct | W (warfarin) C (no warfarin) |
756 870 |
52 55 |
3.37 (NR) 2.91 (NR) |
1.17c (0.79, 1.75) |
| Any bleeding | intracerebral bleeding, gastrointestinal bleeding, intraocular bleeding, hematuria and unspecified location of bleeding | W (warfarin) C (no warfarin) |
756 870 |
149 126 |
10.88 (NR) 7.31 (NR) |
1.41c (1.09, 1.81) | ||
| Genovesi 2015 [31] | Adjusted for: age and dialytic age, gender, antiplatelet therapy and hypertension status at recruitment, permanent AF and bleedings/haemorrhagic strokes, diabetes, ischaemic stroke, ischaemic heart disease and heart failure as time-dependent covariates | Any stroke | cerebrovascular event defined as ischemic or hemorrhagic by computed tomographic scan or nuclear magnetic resonance | W (warfarin) C (no warfarin) |
134 156 |
8 9 |
3.7 (NR) 3.7 (NR) |
0.12 (0.00, 3.59) |
| Major bleeding | hemorrhagic episode requiring hospitalization or blood transfusion, or causing a hemoglobin plasma level reduction >2 g/dl | W (warfarin) C (no warfarin) |
134 156 |
38 29 |
17.6 (NR) 11.8 (NR) |
3.96 (1.15, 13.68) | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
134 156 |
51 64 |
NR NR |
0.96 (0.59, 1.56) | ||
| Chan KE 2015e [20] | Adjusted for: age, sex, race, diabetes, vintage, catheter vascular access, blood pressure, albumin, hemoglobin, thrombocytopenia, erythropoietin dose, heparin dose, antiplatelet use, Charlson comorbidity score, bleeding index score, recent minor or major bleeding event | Ischemic stroke | embolic stroke or arterial embolism, within 2 years of medication initiation | W (warfarin) A (aspirin) D (dabigatran) R (rivaroxaban) |
8064 6018 281 244 |
244 168 13 8 |
6.2 (NR) 5.0 (NR) 10.6 (NR) 11.2 (NR) |
W vs. A: 1.23g (1.01, 1.52) W vs. D: 1.71g (0.97, 2.99) W vs R: 0.55g (0.27, 1.12) |
| Major bleeding | a hemorrhagic event resulting in hospitalization or death | W (warfarin) A (aspirin) D (dabigatran) R (rivaroxaban) |
8064 6018 281 244 |
NR NR NR NR |
47.1 (NR) 35.9 (NR) 83.1 (NR) 68.4 (NR) |
W vs. A: 1.28g (1.19, 1.39) W vs. D: 0.68g (0.55, 0.83) W vs. R: 0.72g (0.55, 0.97) |
||
| Chan PH 2015 [21] | Adjusted for: age, gender, hypertension, diabetes, smoker, heart failure, coronary artery disease, prior stroke/TIA, prior ICH, CHA2DS2-VASc, HAS-BLED | Ischemic stroke | a neurological deficit of sudden onset that persisted for more than 24 h in the absence of primary hemorrhage or other cause and confirmed by CT or MRI | W (warfarin) A (aspirin) C (no antithrombotic therapy) |
67 86 118 |
0 9 11 |
NR NR NR |
T vs. A: 0.16 (0.04, 0.66) T vs. C: 0.19 (0.06, 0.65) |
| Shen 2015 [28] | Included in inverse-probability treatment weighting analysis: age, sex, race, ethnicity, dialysis vintage, geographic location, comorbid conditions, indicators of health services use, baseline medication use, AF diagnosis characteristics | Any stroke | Any stroke or stroke death identified from claims-based algorithms | W (warfarin) C (no warfarin) |
1838 10446 |
116 765 |
4.4 (NR) 5.3 (NR) |
0.83c (0.61, 1.12) |
| Ischemic stroke | Identified from claims-based algorithms | W (warfarin) C (no warfarin) |
1838 10446 |
63 503 |
2.3 (NR) 3.4 (NR) |
0.68c (0.47, 0.99) | ||
| Gastrointestinal bleeding | events requiring hospitalization or with gastrointestinal bleeding as reported cause of death | W (warfarin) C (no warfarin) |
1838 10446 |
153 833 |
5.9 (NR) 5.9 (NR) |
1.00c (0.69, 1.44) | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
1838 10446 |
832 4595 |
33.0 (NR) 32.5 (NR) |
1.01c (0.92, 1.11) | ||
| Wang 2015 [14] | Adjusted for: cerebrovascular disease, congestive heart failure, body mass index, age, warfarin and aspirin, beta-blocker, ischaemic heart disease, peripheral vascular disease | Any stroke | Ischemic stroke and other arterial embolism | W (warfarin) C (no warfarin) |
59 82 |
8 11 |
13.6%b
13.4 % |
1.01 (0.380, 2.70)h |
| Ischemic stroke | focal neurological deficit lasting >24 h with radiological evidence on computed tomography or MRI | W (warfarin) C (no warfarin) |
59 82 |
5 10 |
8.5%b
12.2 % |
0.667 (0.215, 2.06)h | ||
| Any bleed | Intracranial bleed required radiological confirmation, while gastrointestinal, dialysis site and other bleeds required having a blood transfusion to be counted | W (warfarin) C (no warfarin) |
59 82 |
22 24 |
37.3 %b
29.3 % |
1.44 (0.706, 2.92)h | ||
| All-cause mortality | - | W (warfarin) C (no warfarin) |
59 82 |
44 64 |
74.6%b
78.0 % |
0.825 (0.376, 1.81)h | ||
| Yodogawa 2015 [35] | Adjusted for: CHADS2 score | Any stroke | First hospital admission for stroke | W (warfarin) C (no warfarin) |
30 54 |
2 5 |
NR NR |
1.07 (0.20, 5.74) |
| Findlay 2016 [15] | No adjustment | Any stroke | Clinical diagnosis of stroke, presence of ischemic or hemorrhagic stroke on brain imaging, any stroke-related death | W (warfarin) C (no warfarin) |
118 175 |
17 20 |
14.4 %b
11.4 % |
1.024h (0.536, 1.959)h |
| All-cause mortality | - | W (warfarin) C (no warfarin) |
118 175 |
NR NR |
NR NR |
0.671h (0.505, 0.891)h | ||
| Tanaka 2016 [16] | Adjusted for: age, sex, ACE/ARB, diabetes, history of CAD, heart failure, AD, eGFR, β-blocker, hemoglobin, calcium levels, phosphate levels and history of cerebral infarction and ICH | All-cause mortality | - | W (warfarin) C (no warfarin) |
46 47 |
11 10 |
23.9%b
21.3 % |
0.7117h (0.2475, 2.0463)h |
AF atrial fibrillation, HD hemodialysis, PD peritoneal dialysis, CKD chronic kidney disease, ESRD end stage renal disease, RRT renal replacement therapy, CI confidence interval, NR not reported, MF model failed to converge, CAD coronary artery disease, ICH intracranial hemorrhage, GFR glomerular filtration rate, AD aortic disease
aData were abstracted from online supplement
bReference paper reported cumulative proportion instead of rates
cEffect measures listed in the table were from the ITT analysis or propensity score adjusted/matched analysis
dReference table reported unadjusted HR
eQuality of warfarin treatment (i.e. INR or TTR) information provided
fNumbers were combined from CHA2DS2-VASc score = 0, = 1, and > = 2 subgroups
g Reference paper reported RR instead of HR. Effect measure and 95 % CI was calculated by taking the reciprocal of the reported RR
h Effect measure and/or 95 % CIs were obtained from personal communication with the study author