ONLINE APPENDIX
Additional methods
Statistical analyses
Prognostic value of the individual data elements of CABG related bleeding (BARC type 4 bleeding and
TIMI CABG related bleeding)
The prognostic value of CABG related bleeding and data elements of CABG-related bleeding was
assessed separately in patients who underwent CABG during the index admission and who did not
suffer a fatal bleeding event (n=61). Patients without CABG related bleeding were considered the
reference category. Adjusted hazard ratios were calculated using Cox regression analyses adjusting for
predictors of one year mortality in this specific subset of patients. These predictors were established
using a stepwise Cox proportional hazards model, including the time until CABG and the variables in
tables 1 and 2 with a significant relationship with mortality by univariable analysis in this specific subset
of patients.
Prognostic value of the individual data elements of non CABG related bleeding
The prognostic value of the individual data elements of non-CABG related bleeding (BARC type 3 and
TIMI non CABG major and minor bleeding) was assessed in patients who did not undergo CABG during
the index admission and who did not suffer a fatal bleeding event. We used univariable and
multivariable Cox regression analyses to calculate unadjusted and adjusted hazard ratios for mortality.
Multivariable adjustments were made for the previously identified independent predictors of one year
mortality. The same methods were applied to calculate unadjusted and adjusted hazard ratios for the
individual data elements of the ISTH and GUSTO bleeding classification. However, since these
classifications have no separate definition for CABG related bleeding, the Cox regression analyses were
performed in the complete cohort of patients who did not suffer a fatal bleeding event.
Definition
BARC
Type 0
Type 1
Type 2
Type 3
Type 3a
Type 3b
No bleeding
Bleeding that is not actionable and does not cause the patient to seek unscheduled performance of studies,
hospitalization, or treatment by a healthcare professional; may include episodes leading to self-discontinuation of
medical therapy by the patient without consulting a healthcare professional.
Any overt, actionable sign of hemorrhage (eg, more bleeding than would be expected for a clinical circumstance,
including bleeding found by imaging alone) that does not fit the criteria for type 3, 4, or 5 but does meet at least one of
the following criteria: (1) requiring nonsurgical, medical intervention by a healthcare professional, (2) leading to
hospitalization or increased level of care, or (3) prompting evaluation.
Overt bleeding plus hemoglobin drop of 3 to <5 g/dL* (provided hemoglobin drop is related to bleed)
Any transfusion with overt bleeding.
Overt bleeding plus hemoglobin drop ≥5 g/dL* (provided hemoglobin drop is related to bleed).
Cardiac tamponade.
Bleeding requiring surgical intervention for control (excluding dental/nasal/skin/hemorrhoid).
Bleeding requiring intravenous vasoactive agents.
Type 3c
Intracranial hemorrhage (does not include microbleeds or hemorrhagic transformation, does include intraspinal).
Subcategories confirmed by autopsy or imaging or lumbar puncture.
Intraocular bleed compromising vision.
Type 4: CABG-related
bleeding
Perioperative intracranial bleeding within 48 h
Reoperation after closure of sternotomy for the purpose of controlling bleeding.
Transfusion of ≥5 U whole blood or packed red blood cells within a 48-h period.
Chest tube output ≥2L within a 24-h period.
Type 5: fatal bleeding
Type 5a
Type 5b
TIMI
Supplemental table 1. Bleeding defined by BARC, TIMI and GUSTO and ISTH Criteria
Criteria
Definite fatal bleeding; overt bleeding or autopsy or imaging confirmation.
Probable fatal bleeding; no autopsy or imaging confirmation but clinically suspicious.
Major
Minor
GUSTO
Severe
Moderate
Intracranial or clinically significant overt signs of hemorrhage associated with a hemoglobin decrease greater than 5
g/L*
The diagnosis of intracranial bleeding required confirmation by computed tomography or magnetic resonance imaging
of the head.
Observed blood loss and a decrease in hemoglobin level of 3 to 5 g/dL*
Either intracranial hemorrhage or bleeding that causes hemodynamic compromise and requires intervention
Bleeding that requires blood transfusion but does not result in hemodynamic compromise
ISTH
Major
1 Fatal bleeding, and/or
2 Symptomatic bleeding in a critical area or organ, such as intracranial, intraspinal, intraocular, retroperitoneal,
intraarticular or pericardial, or intramuscular with compartment syndrome, and/or
3 Bleeding causing a fall in hemoglobin level of 20 g/L (1.24 mmol/L) or more, or leading to transfusion of two
or more units of whole blood or red cells.
*
Corrected for transfusion (1 U packed red blood cells or 1 U whole blood=1 g/dL hemoglobin).
Supplemental table 2. Multivariable predictors of one year all-cause mortality after
PPCI for STEMI*
HR
95% CI
p-value
Multivessel disease without CTO
1.47
1.07
2.04
0.019
Multivessel disease with CTO
1.88
1.34
2.63
< 0.001
Age (per 10 years increment)
1.28
1.11
1.48
0.001
9
Thrombocyte count (x 10 /L)
< 150
2.57
1.67
3.96
< 0.001
150 – 400
> 400
1.25
0.72
2.16
0.423
9
Leukocyte count > 11 x 10 /L
2.08
1.57
2.77
< 0.001
†
Anemia
2.32
1.74
3.09
< 0.001
Creatinine clearance < 60 ml/min/1.73 m2
1.60
1.14
2.24
0.006
Family history of coronary artery disease
0.44
0.30
0.63
< 0.001
History of malignant disease
1.53
1.08
2.16
0.016
Cardiogenic shock
2.03
1.38
2.98
< 0.001
Current smoker
0.73
0.53
1.02
0.062
IABP
1.77
1.24
2.55
0.002
PCI access site
Femoral
Radial access
2.23
1.32
3.78
0.003
Other or combinations
4.79
2.29
10.01
< 0.001
*
Determined with multivariable stepwise backward selection Cox regression analysis including
variables with a significant univariate analysis.
†
Anemia was defined as baseline hemoglobin less than 13 g/dL for males and less than 12 g/dL for
females.
PPCI indicates primary percutaneous coronary intervention; STEMI: ST-segment elevation myocardial
infarction; HR: hazard ratio; CI: confidence interval; CTO indicates chronic total occlusion; IABP:
intraaortic balloon pump
Supplemental table 3. Thirty day all-cause mortality according to in hospital bleeding status for the different bleeding definitions
Unadjusted
Adjusted†
Bleeding definition
No. of patients (%) Mortality (n/N (%*))
HR
95% CI
p-value
HR
95% CI
p-value
ISTH
Minor or no bleeding
Major bleeding
1616 (80.7)
380 (19.0)
6.3 (102/1616)
15.3 (58/380)
1
2.48
…
1.80
…
3.42
…
< 0.001
1
1.18
…
0.84
…
1.66
…
0.34
Mild or no bleeding
Moderate bleeding
Severe bleeding
1814 (90.6)
127 (6.3)
55 (2.7)
6.9 (125/1814)
17.3 (22/127)
23.6 (13/55)
1
2.58
3.66
…
1.64
2.07
…
4.07
6.49
…
< 0.001
< 0.001
1
0.70
1.54
…
0.43
0.84
…
1.14
2.81
…
0.15
0.16
Minimal or no bleeding
Minor bleeding
Major bleeding
CABG related
1710 (85.4)
159 (7.9)
100 (5.0)
27 (1.3)
6.1 (105/1710)
16.4 (26/159)
24.0 (24/100)
18.5 (5/27)
1
2.74
4.20
3.25
…
1.78
2.69
1.32
…
4.21
6.54
7.96
…
< 0.001
< 0.001
0.010
1
1.16
1.76
1.38
…
0.73
1.10
0.55
…
1.82
2.82
3.45
…
0.53
0.018
0.49
1601 (80.0)
88 (4.4)
170 (8.5)
110 (5.5)
27 (1.3)
6.3 (101/1601)
3.4 (3/88)
15.3 (26/170)
22.7 (25/110)
18.5 (5/27)
1
0.53
2.48
3.85
3.16
…
0.17
1.61
2.48
1.29
…
1.67
3.81
5.96
7.75
…
0.28
< 0.001
< 0.001
0.012
1
0.43
1.01
1.60
1.30
…
0.14
0.64
1.01
0.52
…
1.37
1.60
2.54
3.26
…
0.15
0.95
0.047
GUSTO
TIMI
BARC
Type 0 or 1
Type 2
Type 3a
Type 3b + c*
Type 4
*
Kaplan-Meier estimates.
†
Calculated using Cox proportional hazards models adjusting for the variables in supplemental table 2.
TIMI indicates Thrombolysis In Myocardial Infarction; ISTH: International Society of Thrombosis and Haemostasis; GUSTO: Global Utilization of
Streptokinase and Tissue Plasminogen Activator for Occluded Arteries; BARC: Bleeding Academic Research Consortium. Other abbreviations as in
supplemental table 2.
0.57
Supplemental table 4. Discrimination of Cox regression models for thirty day mortality in STEMI patients
Model
Harrell's C index
95% CI
p-value*
NRI
p-value*
IDI
p-value*
Multivariable
0.851
0.821 - 0.881
Multivariable + TIMI
0.855
0.826 - 0.884
0.13
0.355
< 0.001
0.98
-0.0000840
Multivariable + ISTH
0.853
0.824 - 0.882
0.19
0.374
< 0.001
0.052
-0.00230
Multivariable + GUSTO
0.852
0.823 - 0.882
0.67
0.479
< 0.001
0.75
0.000920
Multivariable + BARC
0.856
0.827 - 0.885
0.084
0.181
0.0098
0.0050
-0.00331
*
p-value for comparison with the multivariable model including the independent baseline predictors of mortality (shown in supplemental table 2).
C index indicates concordance index; NRI: net reclassification index; IDI: integrated discrimination improvement; other abbreviations as in tables 2 and 3.