- Email: [email protected]
Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia
HLC 1918 1–11
ORIGINAL ARTICLE
Heart, Lung and Circulation (2015) xx, 1–11 1443-9506/04/$36.00 http://dx.doi.org/10.1016/j.hlc.2015.06.826
1 2 3 4 5
6
Q1
Q2
7 8 9 10 11 12 13 14 15 16 17
22
[TD$FIRSNAME]John[TD$FIRSNAME.] [TD$SURNAME]Wlodarczyk[,TD$SURNAME.] PhD a, [TD$FIRSNAME]Andrew E.[TD$FIRSNAME.] [TD$SURNAME]Ajani[TD$SURNAME.], FRACP b,c*, [TD$FIRSNAME]Dante[TD$FIRSNAME.] [TD$SURNAME]Kemp[TD$SURNAME.], MS a, [TD$FIRSNAME] Nick[TD$FIRSNAME.] [TD$SURNAME]Andrianopoulos[TD$SURNAME.], MBiostat b, [TD$FIRSNAME]Angela L.[TD$FIRSNAME.] [TD$SURNAME]Brennan[TD$SURNAME.], CCRN b, [TD$FIRSNAME]Stephen J.[TD$FIRSNAME.] [TD$SURNAME]Duffy[TD$SURNAME.], PhD, FRACP d, [TD$FIRSNAME]David J.[TD$FIRSNAME.] [TD$SURNAME]Clark[TD$SURNAME.], FRACP e, [TD$FIRSNAME]Christopher M.[TD$FIRSNAME.] [TD$SURNAME]Reid[TD$SURNAME.], PhD b a
John Wlodarczyk Consulting Services, Newcastle, NSW, Australia Monash University, Melbourne, Vic, Australia and Monash Centre of Cardiovascular Research & Education in Therapeutics, Department of Epidemiology and Preventive Medicine, Melbourne, Vic, Australia c Royal Melbourne Hospital, Melbourne, Vic, Australia d Alfred Hospital, Melbourne, Vic, Australia e Austin Hospital, Melbourne, Vic, Australia b
Received 23 February 2015; received in revised form 10 June 2015; accepted 14 June 2015; online published-ahead-of-print xxx
18 19 20 21
Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia
Background
Major bleeding is a serious complication of percutaneous coronary intervention (PCI). We set out to investigate the impact of major bleeding and its impact on hospitalisation and long-term mortality.
Method
We examined seven years of registry data encompassing 16,860 PCI procedures.
Results
Between 2005 and 2011 major bleeding increased from 1.3% to 3.4%. In patients with ST elevated myocardial infarction (STEMI), the rate increased from 2.3% to 6.4%. The increase remained significant after adjusting for patient and procedural characteristics (OR=1.09/year, p=0.001). Bleeding risk was highest in patients presenting with out-of-hospital cardiac arrest and cardiogenic shock (CS). Women, STEMI patients, those aged over 70yrs or weighing <60 kg were at higher risk. Glycoprotein IIb/IIIa-inhibitor use more than doubled the risk of bleeding (OR=2.28, p=<0.001). Mortality rates at one year were 4.18% overall and 7.9% in STEMI. Bleeding was a strong predictor of mortality after adjusting for potential confounders (HR=2.92, 95% CI: 2.08, 4.09). Bleeding significantly increased length of stay (med four days vs seven days) and rehospitalisation at 12 months (OR=1.36, 95% CI: 1.08, 1.70).
Conclusions
Major bleeding rates post-PCI appear to be increasing in Australia. Bleeding increases hospitalisation and is associated with poor clinical outcomes.
Keywords
Percutaneous coronary intervention Bleeding Hospitalisation Mortality PCI STEMI
Introduction Q3
Bleeding is a common complication of PCI that not only has an immediate impact on the well-being of the patient and the
need for additional resources in hospital, but also contributes to poor long-term outcomes including increased risk of death [1–8]. The risk of bleeding is likely to depend on the individual risk profile of the patient being treated, the
*Corresponding author at: Email: [email protected] © 2015 Published by Elsevier Inc on behalf of Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ).
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
23 24 25 26
HLC 1918 1–11
2
27
J. Wlodarczyk et al.
36
interventions undertaken and the anti-coagulant regimens being used [5,9,10]. Since there is a trade-off between avoiding ischaemic events and avoiding bleeding, knowing the risks for bleeding events and the outcomes following bleeding can help target strategies to reduce the incidence of bleeding [9–11]. In this study we present the incidence, predictors and outcomes of bleeding following PCI in a cohort of Australian patients from the Melbourne Interventional Group (MIG) registry.
37
Methods
38
Study Population
39
66
We undertook a retrospective analysis of MIG Registry data comprising 16,860 PCI procedures from six major Victorian hospitals between 2005 and 2011. Human Research Ethics Committee approval was sought and obtained. Details of participation and data collection have been published previously. In brief, baseline demographics, clinical, angiographic, and procedural characteristics of consecutive patients undergoing PCI are prospectively recorded on case report forms using standardised definitions for all fields [12]. The protocol has been approved by the ethics committee in each participating hospital, including the approval for the use of ‘‘optout’’ consent [13–15]. In-hospital outcomes and complications were recorded at the time of discharge. Follow-up was undertaken at 30-days and 12-months, either by telephone or record review, using a standardised questionnaire [12–15]. Clinical events were validated by reviewing medical records, where available. The registry is coordinated by the Centre for Cardiovascular Research & Education in Therapeutics, a research body within the Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia. An independent audit is conducted annually at all enrolling sites by an investigator not affiliated with that institution. A number of verifiable fields from 5% of all procedures entered from each site annually are randomly selected and audited. In the most recent audit undertaken, 27 fields were reviewed and data accuracy was 98%, which is comparable to other large registries [15,16].
67
Clinical Endpoints
68
A major bleeding complication was defined as any bleed that occurred during or after the catheter laboratory visit until discharge, and required a transfusion, and/or prolonged the hospital stay, and/or caused a drop in haemoglobin > 3.0 gm/dl. This definition is similar to the major bleed definition used in major trials [17–19]. Bleeding site was categorised as percutaneous entry site, retroperitoneal, or ‘other’, which includes gastrointestinal, genital or urinary, other sites, or unknown. Time-to-death was defined as the difference between the procedure date and the date of death up to 365 days. Length of stay was defined as the difference between date of
28 29 30 31 32 33 34 35
40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
69 70 71 72 73 74 75 76 77 78 79
admission and date of discharge from admitting hospital (or date of death). Readmissions to hospital were recorded for patients who survived the initial hospitalisation up to one year after discharge, including the primary readmission reason.
80
Statistical Analysis
85
Logistic regression was used to identify predictors of bleeding and re-admission to hospital. Cox models were used to evaluate the effect of bleeding on length of stay in hospital and mortality. SAS software, version 9.3 was used for all analyses. Candidate variables (including age, sex, co-morbidities, disease type, procedure characteristics and anti-coagulation strategies) were initially evaluated in univariate models. Variables which were statistically significant (p<0.05) were included in a multivariate model. Backward selection was used to select variables significant at the p<0.1 level for the final models. Two-way interactions were included if the p-value for the interaction was less than 0.01. There was evidence that the mortality risk from bleeding differed in CS patients. Since CS was an extremely strong predictor of mortality we ran separate models for CS and non-CS patients.
86
Results
101
A total of 16,860 procedures from six hospitals were included in the analysis. Baseline characteristics are shown in Table 1. The mean age at the time of procedure was 64.611.9 years and three quarters of the patients were male. Clinically, 28% of the cohort presented with STEMI, with a further 37% presenting with NSTEACS. Major bleeding occurred in 2.4% of the procedures. 38.9% of bleeds were at the percutaneous entry site, 8.3% were retro-peritoneal and the remaining 46.5% were classified as ‘other’. Among patients with an out-of-hospital cardiac arrest (OHCA), only 10% (4/40) of bleeds were located at the percutaneous entry site. Thirty-three per cent (132/405) of bleeds required a blood transfusion. Five per cent (22/405) of bleeds were associated with a pseudo-aneurysm and four of these required surgery. Bleeding rates increased significantly during the study period (Table 2). Bleeding rates were highest in older patients and females. Patients presenting with chronic stable angina had the lowest rates of bleeding (1.2%) while the highest rates were in patients presenting with STEMI (4.4%). Bleeding and mortality rates for co-morbidities and anticoagulant treatments are shown in Table 3. Procedural characteristics are summarised in Appendix 2.
102
Factors Associated with Bleeding
124
Age, gender, weight, OHCA, cardiogenic shock (CS), glycoprotein IIb/IIIa-inhibitor (GPI) use, and ACS type were found to be predictive of bleeding complications (Table 4). The risk of bleeding was lowest in 50-69 year-olds. The rates were higher in patients under 50 and those over 70. The interaction between OHCA and gender was significant.
125
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
81 82 83 84
87 88 89 90 91 92 93 94 95 96 97 98 99 100
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
126 127 128 129 130
HLC 1918 1–11
3
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Q4
Table 1 Summary of Outcomes by Patient Characteristics from each Procedure Category
All Procedures
N
16860
Major Bleeds
Deaths at 1 Year
n (%a)
n (%a)
405 (2.4)
704 (4.2)
34 (1.8)
Age (years) Mean SDb
64.6 11.9
Under 50
1900
46 (2.4)
50 to 59
3845
65 (1.7)
83 (2.2)
60 to 69
4868
81 (1.7)
164 (3.4)
70 to 79
4382
143 (3.3)
217 (5.0)
80 or Over Gender
1865
70 (3.8)
206 (11.1)
Male Female
12684
249 (2.0)
487 (3.8)
4180
156 (3.7)
217 (5.2)
166 (8.3)
Weight (kg) Mean SDb
82.0 16.4
60 or under
1998
77 (3.9)
60 to 70
2795
82 (2.9)
148 (5.3)
70 to 80 80 to 90
4143 3707
100 (2.4) 80 (2.2)
159 (3.8) 124 (3.4)
90 to 100
2448
37 (1.51)
76 (3.1)
Over 100
1769
29 (1.6)
31 (1.8)
I
7897
187 (2.4)
256 (3.2)
II
2824
49 (1.7)
126 (4.5)
III
1446
23 (1.6)
58 (4.0)
IV Missing
1706 2987
53 (3.1) 93 (3.1)
124 (7.3) 140 (4.7)
Out-of-hospital Cardiac Arrest
341
40 (11.7)
113 (33.1)
Cardiogenic Shock
490
72 (14.7)
218 (44.5)
NYHAc Class
Angina or ACS Type Atypical Chest Pain Chronic Stable Angina
131
6 (2.6) 82 (1.6)
1818
25 (1.4)
47 (2.6)
NSTEMI STEMI
4392 4705
97 (2.2) 206 (4.4)
186 (4.2) 373 (7.9)
676
9 (1.3)
10 (1.5)
Percentages for bleeds and deaths are based on number of procedures in each row.
b c
9 (3.9) 59 (1.2)
UAP
None/Missing a
229 5040
Standard Deviation (SD).
NYHA – New York Heart Association.
CS. However, in patients with CS, mortality rates were not significantly increased by bleeding (CS and bleeding event 44.4% vs CS and no bleeding event 44.5%, p = 0.08). Other predictors of mortality were increasing age, low weight (<60 kg), out-of-hospital cardiac arrest and STEMI, in addition to several co-morbidities (Table 5).
141
136
Among patients who did not experience an OHCA, females were at higher risk of bleeding than males. However, males were at a higher risk than females in the group that did experience an OHCA. Patients receiving GPI were more than twice as likely to have a bleeding event. The c-statistic for the bleeding prediction model was 0.761.
137
Clinical Outcomes of Bleeding
147 148
138 139
Mortality Bleeding was a strong predictor of mortality at one year (hazard ratio = 2.74, 95% CI [1.30, 3.51]) in patients without
Length of Stay Bleeding was a significant predictor of length of stay, with a time to discharge HR=0.58, (95% CI [0.53, 0.64]) indicating that a patient with a bleeding complication was less likely to be discharged than a patient without a bleed. The median
132 133 134 135
140
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
142 143 144 145 146
149 150 151
HLC 1918 1–11
4
J. Wlodarczyk et al.
Table 2 Proportions of Procedures with a Bleeding Complication over Time in All Patients and STEMI Patients Year
All Procedures
STEMI Cohort
(n=16860)
(n=4705)
Number of Procedures
Bleeding Complications
Number of Procedures
n(%)
Bleeding Complications n(%)
2005 2006
2562 2866
34 (1.3) 53 (1.9)
513 746
12 (2.3) 21 (2.8)
2007
2534
71 (2.8)
710
30 (4.2)
2008
2494
71 (2.9)
686
33 (4.8)
2009
2116
52 (2.5)
650
29 (4.5)
2010
2183
53 (2.4)
710
37 (5.2)
2011
2105
71 (3.4)
690
44 (6.4)
Bleeding trend over time: OR=1.090 per year 95% CI: 1.035, 1.149, p=0.001.
Table 3 Summary of Outcomes by Comorbidities and Treatments Category
All Procedures
Major Bleeds
Deaths up to 1 Year
n (%a)
n (%a)
405 (2.4%)
704 (4.2)
4149
99 (2.4)
252 (6.1)
968
27 (2.8)
68 (7.0) 127 (3.2)
N
16860
Diabetes Yes Treated with Insulin Smoking History Current
3989
86 (2.2)
Prior
7282
155 (2.1)
321 (4.4)
Never
5472
157 (2.9)
228 (4.2)
Missing Congestive Heart Failure – Prior 2 Weeks
117
7 (6.0)
28 (23.9)
837
52 (6.2)
166 (19.8)
Peripheral Vascular Disease
1165
40 (3.4)
124 (10.6)
Cerebrovascular Disease
1129
42 (3.7)
101 (9.0)
Chronic Lung Disease Atrial Fibrillation
1683 520
32 (1.9) 30 (5.8)
121 (7.2) 80 (15.4)
GPI IIb/IIIa Useb Heparin Bivalirudin
5059
228 (4.5)
297 (5.9)
16355
401 (2.5)
686 (4.2)
335
4 (1.2)
12 (3.6)
Clopidogrel No
a
153 154 155 156 157
25 (4.7)
7539
153 (2.0)
212 (2.8)
64 (12.0)
During/After
8786
227 (2.6)
428 (4.9)
Percentages for bleeds and deaths are based on number of procedures in each row.
b
152
535
Prior
GPI was administered intravenously. Intra-coronary administration was rarely used in the participating hospitals.
length of stay increased from four days to seven days and mean length of stay from 4.74.8 days to 11.214.9 days.
regression showed bleeding significantly increased the odds of re-admission (OR=1.36, 95% CI: [1.08, 1.70], p = 0.009).
158
Re-admission to Hospital The proportion of patients re-admitted to hospital in the year after discharge was higher in patients with a major bleed (37.5% (152/405) vs. 31.8% (5,357/16,860)). Logistic
Discussion
160
We examined seven years of registry data in a broad PCI population including elective, urgent and emergency
161
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
159
162
HLC 1918 1–11
5
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Table 4 Bleeding Complication Logistic Regression Model Effect
Odds Ratio
Age Under 50
1
95% Confidence Limits
P-value
50 to 59
0.72
0.49
1.06
0.094
60 to 69
0.68
0.47
1.00
0.048
70 to 79
1.38
0.97
1.97
0.076
80 or Over
1.34
0.89
2.01
0.159
Out-of-hospital Cardiac Arrest (OHCA) and Gender Interaction No OHCA and Male
1
OHCA and Female OHCA and Male
1.95 2.74
0.87 1.75
4.40 4.31
0.107 <0.001
No OHCA and Female
1.75
1.38
2.23
<0.001
1.79
2.90
<0.001
GPI Use No
1.00
Yes
2.28
Angina or ACS Type STEMI
1
NSTEMI UAP
0.82 0.62
0.63 0.40
1.08 0.97
0.160 0.037
ACP
2.30
1.12
4.73
0.023
CSA
0.63
0.44
0.90
0.010
None/Missing
0.69
0.34
1.39
0.296
2.74
5.27
<0.001
Cardiogenic Shock No
1.00
Yes
3.80
Weight 60 Kg or under
1
60 to 70 Kg
0.78
0.54
1.11
0.159
70 to 80 Kg
0.76
0.53
1.08
0.120
80 to 90 Kg
0.73
0.50
1.07
0.107
90 to 100 Kg
0.56
0.36
0.89
0.013
Over 100 Kg
0.65
0.40
1.06
0.085
Missing
0.55
0.30
1.00
0.050
Year of Admission 2011
1
2005
0.43
0.28
0.66
<0.001
2006
0.60
0.41
0.87
0.007
2007
0.92
0.65
1.31
0.650
2008
0.88
0.62
1.25
0.475
2009
0.71
0.49
1.03
0.074
2010
0.72
0.50
1.05
0.084
Percutaneous Entry Location Femoral
1
Brachial Entry
3.78
1.47
9.74
0.006
Radial Entry
0.45
0.23
0.88
0.019
0.98
1.63
0.078
Transfer Patient No
1
Yes
1.26
OHCA by gender interaction p<0.001.
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
6
J. Wlodarczyk et al.
Table 5 Time to Death up to One Year Cox model – Procedures without Cardiogenic Shock Parameter
Hazard Ratio
Bleeding Complication No Bleed
1
Bleed
2.74
95% Confidence Limits
P-value
1.95
3.85
<0.001
0.552
Age Under 50
1
50 to 59
1.17
0.70
1.97
60 to 69
1.64
1
2.68
0.050
70 to 79
2.14
1.30
3.51
0.003
80 or Over Weight
3.97
2.39
6.59
<0.001
60 Kg or Under
1
60 to 70 Kg
0.87
0.64
1.16
0.336
70 to 80 Kg
0.65
0.48
0.87
0.004
80 to 90 Kg
0.55
0.40
0.77
0.000
90 to 100 Kg
0.62
0.44
0.89
0.010
Over 100 Kg
0.35
0.21
0.59
<0.001
1.01
0.64
1.59
0.957
4.13
8.29
<0.001
Missing Out-of-hospital Cardiac Arrest No
1
Yes
5.85
Atrial Fibrillation No
1
Yes
1.61
1.17
2.22
0.003
Never Current
1 0.94
0.70
1.27
0.706
Prior
1.35
1.10
1.67
0.005
Missing
3.13
1.54
6.35
0.002
Smoking History
Family history of CAD No
1
Yes
0.74
0.60
0.91
0.004
Missing
1.55
0.88
2.73
0.126
1.57
2.69
<0.001
1.46
2.42
<0.001
0.99
1.70
0.059
1.18
1.90
0.001
Congestive heart failure - prior 2 weeks No 1 Yes
2.06
Peripheral vascular disease No
1
Yes
1.88
Cerebrovascular disease No Yes Chronic Lung Disease
1 1.30
No
1
Yes
1.49
Diabetes No
1
Yes
1.61
1.32
1.96
<0.001
1 1.22
0.97
1.53
0.089
0.29
1.64
0.398
Hypertension No Yes Angina or ACS Type STEMI
1
ACP
0.69
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
7
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Table 5. (continued). Parameter
Hazard Ratio
95% Confidence Limits
CSA
0.49
0.30
0.81
P-value 0.005
NSTEMI
0.61
0.49
0.77
<0.001
UAP None/Missing
0.51 0.37
0.36 0.18
0.71 0.76
<0.001 0.007
Percutaneous Entry Location Femoral
1
Brachial Entry
2.37
1.10
5.09
0.027
Radial Entry
0.94
0.62
1.43
0.783
0.41
1.02
0.059
Elective Presentation
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
No
1
Yes
0.65
patients, ranging in risk from atypical chest pain and stable angina to STEMI and cardiogenic shock. We found that major bleeding rates post-PCI are strongly associated with increased mortality and have been increasing over time. In addition, bleeding increases both the duration of hospitalisation and risk of re-hospitalisation. We explored several potential causes for the increase in bleeding rates including study design, clinical, and procedural variables. The risk of a selection bias was minimal as participation in the registry was extremely high (>97%) [20]. The bleeding definition used in the registry is similar to that used by other groups, including the National Cardiovascular Data Registry (NCDR1), TIMI major, GUSTO severe [21]. Importantly, it was pre-specified and used consistently, thus ensuring the comparability of bleeding rates over time. Further, an exploratory analysis showed consistent trends at each participating hospital indicating that the increases weren’t due to changes at a local hospital level. Therefore, it is unlikely that the increase in reported bleeding was due to patient selection, ascertainment, or changing practices at individual hospitals. Consistent with several studies, we found GPI use to be a strong independent predictor of bleeding [3,5,10]. GPI use in STEMI increased slightly from 64% to 69% during the study period while in NSTEMI GPI use fell from 30% to 22%. The increase in bleeding rates was still evident after adjusting for GPI use. However, the type of GPI used and dose of heparin were not recorded in the registry, and therefore changes in anti-coagulation use cannot be ruled out as contributing factors. There were increases in the proportion of patients with CS and OHCA (Appendix 1) [23–25]. Although these patients were at high risk of bleeding, they only made a small contribution to the overall increase in the bleeding rates. Bleeding risk was also high in STEMI, the elderly, women and patients with low body weight. A tendency towards treatment of frail patients might explain the increase in bleeding rates. However, the age, sex and weight distributions of patients remained stable during the study.
It is likely that the increase was due to a combination of patient and clinical factors. Further research is needed to determine the cause of the increase in bleeding rates. This study confirms the strong association between bleeding and mortality that has been noted previously [2–6,8]. Overall, the mortality rate increases two- to three-fold in patients who bled, and 10% (70/704) of patients who died had a major bleed. While the majority of the increased risk is in-hospital, the adverse impact of bleeding on survival is maintained for at least one year. The finding that bleeding does not increase mortality risk in CS patients could be due to the fact that CS patients were already at very high risk of death. Furthermore, death occurred early in CS patients (median time to death of two days), and these patients might not have survived long enough to be at risk for bleeding. This study also confirms not only that bleeding is associated with a doubling of hospital stay, which is consistent with published studies, but also that patients who experienced major bleeding were at significantly increased risk of readmission to hospital [1,24,26,27]. Indicating that in addition to clinical benefits, strategies to reduce bleeding might have important economic benefits to patients and institutions. There are a number of limitations to this study. An observational study cannot establish cause and effect. Some of the observed associations could be due to selection bias, confounding or effect modification. For instance, anti-coagulant choice could be based on factors associated with bleeding, or patients at higher risk of death could also be more likely to bleed. However, the consistency of results after multivariate modelling suggests that the associations are robust. The major strengths of this study are the quality of the registry data set which recruited almost all patients undergoing PCI at the participating centres, used consistent methodology, and had high follow-up rates to one year. The registry population appears to be representative of PCI patients in Australia, and similar to those in other major cohorts [8,9,22]. The rates and types of major bleeding are comparable with those reported recently in other countries [1,8,9,22].
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241
HLC 1918 1–11
8
242
J. Wlodarczyk et al.
251
However, caution should be used when generalising our results to other populations. These results might not be applicable where different PCI access methods or different anti-coagulation strategies are used. There were few patients treated with bivalirudin in the registry (n=335 (2.0%)), thus we were also unable to explore the relative impact of this anti-coagulant on bleeding. Also the vast majority of patients in the registry had femoral arterial access for their procedure, however radial patients had a significantly lower bleed risk which is consistent with previous reports [2].
252
Conclusion
253
261
The likelihood of major bleeding following PCI in this cohort has been increasing steadily in recent years with rates doubling over a seven-year period to 2011. This suggests that greater efforts are required to avoid this serious complication of PCI in Australia. These results identify high risk patients for bleeding that could benefit most from different anti-coagulation strategies which has the potential to reducing the burden to healthcare providers and improve patient outcomes.
262
Financial Disclosure
263
The Melbourne Interventional Group acknowledges funding from Abbott, Astra-Zeneca, Biotronik, Boston Scientific, Bristol-Myers Squibb, Cordis Johnson & Johnson, CSL, Medtronic, MSD, Pfizer, Sanofi-Aventis, Servier, ScheringPlough and The Medicines Company. These companies do not have access to data. Dr. Duffy’s work is supported by a NHMRC Program Grant to the Alfred and Baker Medical Unit. Prof Reid’s work is supported a NHMRC Senior Research Fellowship and NHMRC Program Grant. Dr. Wlodarczyk is the principle of a consulting company contracted by The Medicines Company to perform
243 244 245 246 247 248 249 250
254 255 256 257 258 259 260
264 265 266 267 268 269 270 271 272 273
the analyses found in this report and to prepare the manuscript.
274
Acknowledgements
276
We would like to thank Bishoy Rizkilla, Medical Director for The Medicines Company for his assistance in developing this project. We would also like to thank the MIG team:
277
MIG Data Management Centre, CCRE, Monash University: Professor Chris Reid, Dr Nick Andrianopoulos, Ms Angela Brennan.
275
278 279
280
MIG Steering Committee: Professor Chris Reid, A/Professor Andrew Ajani, Dr Stephen Duffy, Dr David Clark, A/Professor Gishel New, Dr Chin Hiew, Dr Nick Andrianopoulos, Dr Ernesto Oqueli, Ms Angela Brennan.
281
The following investigators, data managers and institutions participated in the MIG Database: Alfred Hospital: SJ Duffy, JA Shaw, A Walton, A Dart, A Broughton, J Federman, C Keighley, C Hengel, KH Peter, J O’Brien, H Bala, S Nanayakkara, R Vandernet, R Huntington; Austin Hospital: DJ Clark, O Farouque, M Horrigan, J Johns, L Oliver, J Brennan, R Chan, G Proimos, T Dortimer, B Chan, V Nadurata, R Huq, D Fernando, K Charter, L Brown, A AlFiadh, H Sugumar, R Spencer, J Ramchand, P Scott; Box Hill Hospital: G New, L Roberts, M Freeman, A Teh, M Rowe, G Proimos, Y Cheong, C Goods, D Fernando, A Baradi, D Jackson, J Sajeev; Frankston Hospital: R Lew, G Szto, R Teperman, R Huq; Geelong Hospital: C Hiew, M Sebastian, T Yip, M Mok, K Rankin, C Machado, J Dyson, B McDonald, L Duff; Monash University: H Krum, C Reid, N Andrianopoulos, AL Brennan, V Chand, D Dinh, BP Yan; Royal Melbourne Hospital: AE Ajani, R Warren, D Eccleston, J Lefkovits, R Iyer, R Gurvitch, W Wilson, M Brooks, M Yudi, S Biswas, J Yeoh, C Cheshire, N Gaikwad.
285 286
282 283 284
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
Appendix 1. Changes in Patient Characteristics and Anti-Coagulation Use over Time Year
All Procedures
2005
2006
2007
2008
2009
2010
2011
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
2562
2866
2534
2494
2116
2183
2105
16860
Age Under 50
300 (11.7)
342 (11.9)
284 (11.2)
308 (12.3)
243 (11.5)
230 (10.5)
193 (9.2)
1900 (11.3)
50 to 59
536 (20.9)
673 (23.5)
540 (21.3)
595 (23.9)
496 (23.4)
515 (23.6)
490 (23.3)
3845 (22.8)
60 to 69
708 (27.6)
789 (27.5)
752 (29.7)
700 (28.1)
601 (28.4)
670 (30.7)
648 (30.8)
4868 (28.9)
70 to 79
732 (28.6)
769 (26.8)
688 (27.2)
611 (24.5)
543 (25.7)
494 (22.6)
545 (25.9)
4382 (26)
80 or Over
286 (11.2)
293 (10.2)
270 (10.7)
280 (11.2)
233 (11)
274 (12.6)
229 (10.9)
1865 (11.1)
Sex Male
1849 (72.2) 2140 (74.7) 1939 (76.5) 1897 (76.1) 1619 (76.5) 1631 (74.7) 1606 (76.3) 12681 (75.2)
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
9
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Appendix 1. (continued). Year
2005
2006
2007
2008
2009
2010
2011
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
Weight 60 Kg or under
245 (9.6)
248 (8.7)
183 (7.2)
198 (7.9)
156 (7.4)
188 (8.6)
175 (8.3)
60 to 70 Kg
476 (18.6)
457 (15.9)
418 (16.5)
411 (16.5)
337 (15.9)
360 (16.5)
336 (16)
1393 (8.3) 2795 (16.6)
70 to 80 Kg
625 (24.4)
711 (24.8)
625 (24.7)
590 (23.7)
547 (25.9)
563 (25.8)
482 (22.9)
4143 (24.6)
80 to 90 Kg
532 (20.8)
664 (23.2)
567 (22.4)
557 (22.3)
445 (21)
468 (21.4)
474 (22.5)
3707 (22)
90 to 100 Kg Over 100 Kg
366 (14.3) 214 (8.4)
423 (14.8) 256 (8.9)
351 (13.9) 240 (9.5)
355 (14.2) 289 (11.6)
312 (14.7) 213 (10.1)
331 (15.2) 250 (11.5)
310 (14.7) 307 (14.6)
2448 (14.5) 1769 (10.5)
Missing
104 (4.1)
107 (3.7)
150 (5.9)
94 (3.8)
106 (5)
23 (1.1)
21 (1)
605 (3.6)
NYHA I
957 (45.2) 1029 (47.1) 1154 (54.8)
7897 (46.8)
II
453 (17.7)
465 (16.2)
489 (19.3)
321 (12.9)
432 (20.4)
329 (15.1)
2824 (16.7)
III
114 (4.4)
260 (9.1)
166 (6.6)
365 (14.6)
254 (12)
191 (8.7)
IV
60 (2.3)
246 (8.6)
231 (9.1)
185 (7.4)
289 (13.7)
401 (18.4)
294 (14)
1706 (10.1)
405 (15.8) 17 (0.7)
787 (27.5) 23 (0.8)
540 (21.3) 36 (1.4)
612 (24.5) 55 (2.2)
184 (8.7) 72 (3.4)
233 (10.7) 64 (2.9)
226 (10.7) 74 (3.5)
2987 (17.7) 341 (2)
51 (2)
71 (2.5)
61 (2.4)
78 (3.1)
87 (4.1)
76 (3.6)
490 (2.9)
77 (3)
51 (1.8)
Missing Out-of-Hospital Cardiac Arrest Cardiogenic Shock
1530 (59.7) 1108 (38.7) 1108 (43.7) 1011 (40.5)
66 (3)
335 (15.9) 96 (4.6)
1446 (8.6)
Angina or ACS Type Atypical Chest Pain Chronic Stable Angina
792 (30.9) 1019 (35.6)
27 (1.1)
18 (0.7)
6 (0.3)
782 (30.9)
729 (29.2)
605 (28.6)
21 (1) 572 (26.2)
29 (1.4)
229 (1.4)
541 (25.7)
5040 (29.9)
UAP
481 (18.8)
291 (10.2)
228 (9)
219 (8.8)
202 (9.5)
205 (9.4)
192 (9.1)
1818 (10.8)
NSTEMI
560 (21.9)
672 (23.4)
645 (25.5)
761 (30.5)
594 (28.1)
594 (27.2)
566 (26.9)
4392 (26)
STEMI None/Missing
513 (20) 139 (5.4)
746 (26) 87 (3)
710 (28) 142 (5.6)
686 (27.5) 81 (3.2)
650 (30.7) 59 (2.8)
710 (32.5) 81 (3.7)
690 (32.8) 87 (4.1)
4705 (27.9) 676 (4)
GPI
655 (25.6)
816 (28.5)
750 (29.6)
795 (31.9)
688 (32.5)
678 (31.1)
677 (32.2)
5059 (30)
Clopidogrel - Prior
877 (34.2) 1310 (45.7) 1236 (48.8) 1282 (51.4)
966 (45.7) 1002 (45.9)
866 (41.1)
7539 (44.7)
1105 (52.2) 1108 (50.8) 1051 (49.9)
8786 (52.1)
Anti-Coagulant Usage
Clopidogrel - During or After
1608 (62.8) 1495 (52.2) 1247 (49.2) 1172 (47)
Heparin
2495 (97.4) 2745 (95.8) 2446 (96.5) 2430 (97.4) 2061 (97.4) 2128 (97.5) 2050 (97.4) 16355 (97)
Bivalirudin Aspirin Prasugrel Ticagrelor
3 (0.1)
97 (3.4)
94 (3.7)
58 (2.3)
31 (1.5)
2527 (98.6) 2822 (98.5) 2504 (98.8) 2460 (98.6) 2095 (99) 0 (0) 0 (0) 0 (0) 0 (0) 3 (0.1) 0 (0)
0 (0)
0 (0)
0 (0)
30 (1.4)
22 (1)
335 (2)
2163 (99.1) 2068 (98.2) 16639 (98.7) 59 (2.7) 228 (10.8) 290 (1.7)
0 (0)
0 (0)
4 (0.2)
4 (0)
Appendix 2. Summary of Disease Extent and Procedural Characteristics Number of Procedures (n=16860) n(%)
Major Bleeds n (%a)
Deaths up to 1 Year n (%a)
Disease Extent Single Vessel Disease
6760 (40.1)
8 (2.1)
193 (2.9)
Multi Vessel Disease
9710 (57.6)
155 (2.3)
502 (5.2)
390 (2.3)
242 (2.5)
9 (2.3)
1
13902 (82.5)
327 (2.4)
2
2627 (15.6)
67 (2.6)
331 (2.0)
11 (3.3)
6066 (36.0)
140 (2.3)
Missing Number of Lesions Treated
3+ Right Coronary Artery Lesion Left Main Lesion
193 (1.1)
4 (2.1)
559 (4) 122 (4.6) 23 (6.9transf) 215 (3.5) 34 (17.6)
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
10
J. Wlodarczyk et al.
Appendix 2. (continued). Number of Procedures
Major Bleeds
Deaths up to 1 Year
(n=16860) n(%)
n (%a)
n (%a)
Left Coronary Artery Lesion
6733 (39.9)
194 (2.9)
322 (4.8)
Left Circumflex Lesion
4006 (23.8)
67 (1.7)
152 (3.8)
Elective Presentation
6221 (36.9)
74 (1.2)
100 (1.6)
Transfer Patient
2962 (17.6)
88 (3)
123 (4.2)
397 (2.4)
58 (14.6)
161 (40.6)
75 (0.4)
5 (6.7)
12 (16)
Intra-Aortic Balloon Pump Percutaneous Entry Location Brachial Entry Radial Entry Femoral Missing
945 (5.6)
9 (1)
27 (2.9)
15839 (93.9)
391 (2.5)
665 (4.2)
1 (0.0)
Stent Used in Procedure
0 (0)
0 (0)
15869 (94.1)
370 (2.3)
631 (4)
BMSb only
8539 (50.7)
225 (2.6)
408 (4.8)
DESc only Both BMS and DES
6893 (40.9) 437 (2.6)
127 (1.8) 18 (4.1)
202 (2.9) 21 (4.8)
991 (5.9)
35 (3.5)
73 (7.4)
Type of Stent
None (PCI only) Closure Device Seal
a
43 (3.6)
25 (2.1)
280 (1.7)
6 (2.1)
4 (1.4)
Other
160 (1.0)
4 (2.5)
2 (1.3)
None
15210 (90.2)
352 (2.3)
673 (4.4)
Percentages for bleeds and deaths are based on number of procedures in each row.
b c
1210 (7.2)
Suture
BMS: Bare metal stent.
DES: Drug eluting stent.
Appendix 3. Summary of Percutaneous Entry Locations by Year Year
2005 n (%)
2006 n (%)
2007 n (%)
2008 n (%)
2009 n (%)
2010 n (%)
2011 n (%)
Total n (%)
All Procedures
2562
2866
2534
2494
2116
2183
2105
16860
Brachial
15 (0.6)
14 (0.5)
10 (0.4)
12 (0.5)
Radial Femoral
59 (2.3) 2488 (97.1)
125 (4.4) 2727 (95.2)
121 (4.8) 2403 (94.8)
77 (3.1) 2405 (96.4)
Missing
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
0
0
0
References [1] Happe LE, Rao SV, Horblyukc R, Franklin M, Lunacseka OE, Menditto L. Consequences of major bleeding in hospitalized patients with non-ST segment elevationacute coronary syndromes receiving injectable anticoagulants. Cur Med Research and Opinion 2009;25:413–20. [2] Rao SV, Kaul PR, Liao L, Armstrong PW, Ohman EM, Granger CB, et al. Association between bleeding, blood transfusion, and costs among patients with non–ST-segment elevation acute coronary syndromes. Am Heart J 2008;155:369–74. [3] Kinnaird TD, Stabile E, Mintz GS, Lee CW, Canos DA, Gevorkian N, et al. Incidence, predictors, and prognostic implications of bleeding and blood transfusion following percutaneous coronary interventions. Am J Cardiol 2003;82:930–5. [4] Mrdovic I, Savic L, Asanin M, Cvetinovic N, Brdar N, Djuricic N, et al. Sex-related analysis of short- and long-term clinical outcomes and
0
8 (0.4) 101 (4.8) 2007 (94.9) 0
5 (0.2) 167 (7.7) 2010 (92.1) 1 (0.0)
11 (0.5) 295 (14) 1799 (85.5) 0
75 (0.4) 945 (5.6) 15839 (93.9) 1 (0.0)
bleeding among patients treated with primary percutaneous coronary intervention: an evaluation of the RISK-PCI data. Can J Cardiol 2013;29:1097–103. [5] Mehran R, Pocock S, Nikolsky E, Dangas GD, Clayton T, Claessen BE, et al. Impact of bleeding on mortality after percutaneous coronary intervention results from a patient-level pooled analysis of the REPLACE-2 (randomized evaluation of PCI linking angiomax to reduced clinical events), ACUITY (acute catheterization and urgent intervention triage strategy), and HORIZONS-AMI (harmonizing outcomes with revascularization and stents in acute myocardial infarction) trials. JACC Cardiovasc Interv 2011;4:654–64. [6] Boden H, Velders MA, van der Hoeven BL, Cannegieter SC, Schalij MJ. In-hospital major bleeding and its clinical relevance in patients with ST elevation myocardial infarction treated with primary percutaneous coronary intervention. Am J Cardiol 2013;112:1533–9. [7] Rao SV, Dai D, Subherwal S, Weintraub WS, Brindis RS, Messenger JC, et al. Association between periprocedural bleeding and long-term
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
HLC 1918 1–11
11
Incidence, Predictors and Outcomes of Major Bleeding in Patients
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
outcomes following percutaneous coronary intervention in older patients. JACC Cardiovasc Interv 2012;5:958–65. Chhatriwalla AK, Amin AP, Kennedy KF, House JA, Cohen DJ, Rao SV, et al. Association between bleeding events and in-hospital mortality after percutaneous coronary intervention. JAMA 2013;309:1022–9. Mehta SK, Frutkin AD, Lindsey JB, House JA, Spertus JA, Rao SV, et al. Bleeding in patients undergoing percutaneous coronary intervention: The development of a clinical risk algorithm from the National Cardiovascular Data Registry. Circ Cardiovasc Interv 2009;2:222–9. Pocock SJ, Mehran R, Clayton TC, Nikolsky E, Parise H, Fahy M, et al. Prognostic modelling of individual patient risk and mortality impact of ischemic and hemorrhagic complications: Assessment from the ACUITY Trial. Circulation 2010;121:43–51. Hanna EB, Rao SV, Manoukian SV, Saucedo JF. The evolving role of glycoprotein IIb/IIIa inhibitors in the setting of percutaneous coronary intervention strategies to minimize bleeding risk and optimize outcomes. JACC Cardiovasc Interv 2010;3:1209–19. Chan W, Clark DJ, Ajani AE, Yap CH, Andrianopoulos N, Brennan AL, et al. Progress towards a National Cardiac Procedure Database – Development of the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) and Melbourne Interventional Group (MIG) Registries. Heart Lung Circ 2011;20:10–8. Ajani AE, Szto G, Eccleston D, Clark DJ, Lefkovits J, Chew DP, et al. The foundation and launch of the Melbourne interventional group: a collaborative interventional cardiology project. Heart Lung Circ 2006;15:44–7. Ajani AE, Duffy SJ, Clark DJ, Lefkovits J, Warren R, Gurvitch R, et al. Outcomes after percutaneous coronary intervention in contemporary Australian practice: insights from a large multicentre registry. MJA 2008;189(8):423–8. Andrianopoulos N, Dinh D, Duffy SJ, Clark DJ, Brennan A, Chan W, et al. Quality Control activities associated with registries in Interventional Cardiology and Surgery. Heart Lung Circ 2011;20:180–6. Lagerqvist B, James SK, Stenestrand U, Lindback J, Nilsson T, Wallentin L. Long-term outcomes with drug-eluting stents versus bare-metal stents in Sweden. N Engl J Med 2007;356:1009–19. Lincoff AM, Kleiman NS, Kereiakes DJ, Feit F, Kleiman NS, Jackman JD, et al. Long-term efficacy of bivalirudin and provisional glycoprotein IIb/ IIIa blockade vs heparin and planned glycoprotein IIb/IIIa blockade
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
during percutaneous coronary revascularization: REPLACE-2 randomized trial. JAMA 2004;292:696–703. Stone GW, Bertrand M, Colombo A, Dangas G, Farkouh ME, Feit F, et al. Acute Catheterization and Urgent Intervention Triage strategy (ACUITY) trial: study design and rationale. Am Heart J 2004;148:764–75. Stone GW, Witzenbichler B, Guagliumi G, Peruga JZ, Brodie BR, Dudek D, et al. Bivalirudin during primary PCI in acute myocardial infarction. N Engl J Med 2008;358:2218–30. Centre of Research Excellence in Patient Safety, Melbourne Interventional Group Interventional Cardiology [Internet], Melbourne Vic, Centre of Research Excellence in Patient Safety; 2006 [Updated: 2010 September 06; cited 2015 May 29]. Available from: http://www. registries.org.au/registries/melb_interventional_cardiology.html. Mehran R, Rao SV, Bhatt DL, Gibson CM, Caixeta A, Eikelboom J, et al. Standardized Bleeding Definitions for Cardiovascular Clinical Trials: A Consensus Report From the Bleeding Academic Research Consortium. Circulation 2011;123:2736–47. Chew DP, French J, Briffa TG, Hammett CJ, Ellis CJ, Ranaslnghe I, et al. Acute coronary syndrome care across Australia and New Zealand: the SNAPSHOT ACS study. MJA 2013;199:1–7. Brennan AL, Andrianopoulos N, Duffy SJ, Reid CM, Clark DJ, Loane P, et al. Trends in door-to-balloon time and outcomes following primary percutaneous coronary intervention for ST-elevation myocardial infarction: An Australian perspective. Internal Medicine Journal 2014;44:471–7. Yan BP, Ajani AE, Clark DJ, Duffy SJ, Andrianopoulos N, Brennan AL, et al. Recent trends in Australian percutaneous coronary intervention practice: insights from the Melbourne International Group Registry. MJA 2011;195:122–7. Spencer RJ, Andrianopoulos N, Scott P, Scott P, Farouque O, Duffy SJ, et al. Increasing risk profile amongst patients undergoing PCI in Australia. Global Heart 2014;9(1):e202. Aronow HD, Peyser PA, Eagle KA, Bates ER, Werns SW, Russman PL, et al. Predictors of length of stay after coronary stenting. Am Heart J 2001;42:799–805. Rao SV, Ou FS, Wang TY, Roe MT, Brindis RG, Rumsfeld JS, et al. Trends in the prevalence and outcomes of radial and femoral approaches to percutaneous coronary intervention. A report from the National Cardiovascular Data Registry. J Am Coll Cardiol Intv 2008;1:379–86.
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
ORIGINAL ARTICLE
Heart, Lung and Circulation (2015) xx, 1–11 1443-9506/04/$36.00 http://dx.doi.org/10.1016/j.hlc.2015.06.826
1 2 3 4 5
6
Q1
Q2
7 8 9 10 11 12 13 14 15 16 17
22
[TD$FIRSNAME]John[TD$FIRSNAME.] [TD$SURNAME]Wlodarczyk[,TD$SURNAME.] PhD a, [TD$FIRSNAME]Andrew E.[TD$FIRSNAME.] [TD$SURNAME]Ajani[TD$SURNAME.], FRACP b,c*, [TD$FIRSNAME]Dante[TD$FIRSNAME.] [TD$SURNAME]Kemp[TD$SURNAME.], MS a, [TD$FIRSNAME] Nick[TD$FIRSNAME.] [TD$SURNAME]Andrianopoulos[TD$SURNAME.], MBiostat b, [TD$FIRSNAME]Angela L.[TD$FIRSNAME.] [TD$SURNAME]Brennan[TD$SURNAME.], CCRN b, [TD$FIRSNAME]Stephen J.[TD$FIRSNAME.] [TD$SURNAME]Duffy[TD$SURNAME.], PhD, FRACP d, [TD$FIRSNAME]David J.[TD$FIRSNAME.] [TD$SURNAME]Clark[TD$SURNAME.], FRACP e, [TD$FIRSNAME]Christopher M.[TD$FIRSNAME.] [TD$SURNAME]Reid[TD$SURNAME.], PhD b a
John Wlodarczyk Consulting Services, Newcastle, NSW, Australia Monash University, Melbourne, Vic, Australia and Monash Centre of Cardiovascular Research & Education in Therapeutics, Department of Epidemiology and Preventive Medicine, Melbourne, Vic, Australia c Royal Melbourne Hospital, Melbourne, Vic, Australia d Alfred Hospital, Melbourne, Vic, Australia e Austin Hospital, Melbourne, Vic, Australia b
Received 23 February 2015; received in revised form 10 June 2015; accepted 14 June 2015; online published-ahead-of-print xxx
18 19 20 21
Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia
Background
Major bleeding is a serious complication of percutaneous coronary intervention (PCI). We set out to investigate the impact of major bleeding and its impact on hospitalisation and long-term mortality.
Method
We examined seven years of registry data encompassing 16,860 PCI procedures.
Results
Between 2005 and 2011 major bleeding increased from 1.3% to 3.4%. In patients with ST elevated myocardial infarction (STEMI), the rate increased from 2.3% to 6.4%. The increase remained significant after adjusting for patient and procedural characteristics (OR=1.09/year, p=0.001). Bleeding risk was highest in patients presenting with out-of-hospital cardiac arrest and cardiogenic shock (CS). Women, STEMI patients, those aged over 70yrs or weighing <60 kg were at higher risk. Glycoprotein IIb/IIIa-inhibitor use more than doubled the risk of bleeding (OR=2.28, p=<0.001). Mortality rates at one year were 4.18% overall and 7.9% in STEMI. Bleeding was a strong predictor of mortality after adjusting for potential confounders (HR=2.92, 95% CI: 2.08, 4.09). Bleeding significantly increased length of stay (med four days vs seven days) and rehospitalisation at 12 months (OR=1.36, 95% CI: 1.08, 1.70).
Conclusions
Major bleeding rates post-PCI appear to be increasing in Australia. Bleeding increases hospitalisation and is associated with poor clinical outcomes.
Keywords
Percutaneous coronary intervention Bleeding Hospitalisation Mortality PCI STEMI
Introduction Q3
Bleeding is a common complication of PCI that not only has an immediate impact on the well-being of the patient and the
need for additional resources in hospital, but also contributes to poor long-term outcomes including increased risk of death [1–8]. The risk of bleeding is likely to depend on the individual risk profile of the patient being treated, the
*Corresponding author at: Email: [email protected] © 2015 Published by Elsevier Inc on behalf of Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ).
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
23 24 25 26
HLC 1918 1–11
2
27
J. Wlodarczyk et al.
36
interventions undertaken and the anti-coagulant regimens being used [5,9,10]. Since there is a trade-off between avoiding ischaemic events and avoiding bleeding, knowing the risks for bleeding events and the outcomes following bleeding can help target strategies to reduce the incidence of bleeding [9–11]. In this study we present the incidence, predictors and outcomes of bleeding following PCI in a cohort of Australian patients from the Melbourne Interventional Group (MIG) registry.
37
Methods
38
Study Population
39
66
We undertook a retrospective analysis of MIG Registry data comprising 16,860 PCI procedures from six major Victorian hospitals between 2005 and 2011. Human Research Ethics Committee approval was sought and obtained. Details of participation and data collection have been published previously. In brief, baseline demographics, clinical, angiographic, and procedural characteristics of consecutive patients undergoing PCI are prospectively recorded on case report forms using standardised definitions for all fields [12]. The protocol has been approved by the ethics committee in each participating hospital, including the approval for the use of ‘‘optout’’ consent [13–15]. In-hospital outcomes and complications were recorded at the time of discharge. Follow-up was undertaken at 30-days and 12-months, either by telephone or record review, using a standardised questionnaire [12–15]. Clinical events were validated by reviewing medical records, where available. The registry is coordinated by the Centre for Cardiovascular Research & Education in Therapeutics, a research body within the Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia. An independent audit is conducted annually at all enrolling sites by an investigator not affiliated with that institution. A number of verifiable fields from 5% of all procedures entered from each site annually are randomly selected and audited. In the most recent audit undertaken, 27 fields were reviewed and data accuracy was 98%, which is comparable to other large registries [15,16].
67
Clinical Endpoints
68
A major bleeding complication was defined as any bleed that occurred during or after the catheter laboratory visit until discharge, and required a transfusion, and/or prolonged the hospital stay, and/or caused a drop in haemoglobin > 3.0 gm/dl. This definition is similar to the major bleed definition used in major trials [17–19]. Bleeding site was categorised as percutaneous entry site, retroperitoneal, or ‘other’, which includes gastrointestinal, genital or urinary, other sites, or unknown. Time-to-death was defined as the difference between the procedure date and the date of death up to 365 days. Length of stay was defined as the difference between date of
28 29 30 31 32 33 34 35
40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
69 70 71 72 73 74 75 76 77 78 79
admission and date of discharge from admitting hospital (or date of death). Readmissions to hospital were recorded for patients who survived the initial hospitalisation up to one year after discharge, including the primary readmission reason.
80
Statistical Analysis
85
Logistic regression was used to identify predictors of bleeding and re-admission to hospital. Cox models were used to evaluate the effect of bleeding on length of stay in hospital and mortality. SAS software, version 9.3 was used for all analyses. Candidate variables (including age, sex, co-morbidities, disease type, procedure characteristics and anti-coagulation strategies) were initially evaluated in univariate models. Variables which were statistically significant (p<0.05) were included in a multivariate model. Backward selection was used to select variables significant at the p<0.1 level for the final models. Two-way interactions were included if the p-value for the interaction was less than 0.01. There was evidence that the mortality risk from bleeding differed in CS patients. Since CS was an extremely strong predictor of mortality we ran separate models for CS and non-CS patients.
86
Results
101
A total of 16,860 procedures from six hospitals were included in the analysis. Baseline characteristics are shown in Table 1. The mean age at the time of procedure was 64.611.9 years and three quarters of the patients were male. Clinically, 28% of the cohort presented with STEMI, with a further 37% presenting with NSTEACS. Major bleeding occurred in 2.4% of the procedures. 38.9% of bleeds were at the percutaneous entry site, 8.3% were retro-peritoneal and the remaining 46.5% were classified as ‘other’. Among patients with an out-of-hospital cardiac arrest (OHCA), only 10% (4/40) of bleeds were located at the percutaneous entry site. Thirty-three per cent (132/405) of bleeds required a blood transfusion. Five per cent (22/405) of bleeds were associated with a pseudo-aneurysm and four of these required surgery. Bleeding rates increased significantly during the study period (Table 2). Bleeding rates were highest in older patients and females. Patients presenting with chronic stable angina had the lowest rates of bleeding (1.2%) while the highest rates were in patients presenting with STEMI (4.4%). Bleeding and mortality rates for co-morbidities and anticoagulant treatments are shown in Table 3. Procedural characteristics are summarised in Appendix 2.
102
Factors Associated with Bleeding
124
Age, gender, weight, OHCA, cardiogenic shock (CS), glycoprotein IIb/IIIa-inhibitor (GPI) use, and ACS type were found to be predictive of bleeding complications (Table 4). The risk of bleeding was lowest in 50-69 year-olds. The rates were higher in patients under 50 and those over 70. The interaction between OHCA and gender was significant.
125
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
81 82 83 84
87 88 89 90 91 92 93 94 95 96 97 98 99 100
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
126 127 128 129 130
HLC 1918 1–11
3
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Q4
Table 1 Summary of Outcomes by Patient Characteristics from each Procedure Category
All Procedures
N
16860
Major Bleeds
Deaths at 1 Year
n (%a)
n (%a)
405 (2.4)
704 (4.2)
34 (1.8)
Age (years) Mean SDb
64.6 11.9
Under 50
1900
46 (2.4)
50 to 59
3845
65 (1.7)
83 (2.2)
60 to 69
4868
81 (1.7)
164 (3.4)
70 to 79
4382
143 (3.3)
217 (5.0)
80 or Over Gender
1865
70 (3.8)
206 (11.1)
Male Female
12684
249 (2.0)
487 (3.8)
4180
156 (3.7)
217 (5.2)
166 (8.3)
Weight (kg) Mean SDb
82.0 16.4
60 or under
1998
77 (3.9)
60 to 70
2795
82 (2.9)
148 (5.3)
70 to 80 80 to 90
4143 3707
100 (2.4) 80 (2.2)
159 (3.8) 124 (3.4)
90 to 100
2448
37 (1.51)
76 (3.1)
Over 100
1769
29 (1.6)
31 (1.8)
I
7897
187 (2.4)
256 (3.2)
II
2824
49 (1.7)
126 (4.5)
III
1446
23 (1.6)
58 (4.0)
IV Missing
1706 2987
53 (3.1) 93 (3.1)
124 (7.3) 140 (4.7)
Out-of-hospital Cardiac Arrest
341
40 (11.7)
113 (33.1)
Cardiogenic Shock
490
72 (14.7)
218 (44.5)
NYHAc Class
Angina or ACS Type Atypical Chest Pain Chronic Stable Angina
131
6 (2.6) 82 (1.6)
1818
25 (1.4)
47 (2.6)
NSTEMI STEMI
4392 4705
97 (2.2) 206 (4.4)
186 (4.2) 373 (7.9)
676
9 (1.3)
10 (1.5)
Percentages for bleeds and deaths are based on number of procedures in each row.
b c
9 (3.9) 59 (1.2)
UAP
None/Missing a
229 5040
Standard Deviation (SD).
NYHA – New York Heart Association.
CS. However, in patients with CS, mortality rates were not significantly increased by bleeding (CS and bleeding event 44.4% vs CS and no bleeding event 44.5%, p = 0.08). Other predictors of mortality were increasing age, low weight (<60 kg), out-of-hospital cardiac arrest and STEMI, in addition to several co-morbidities (Table 5).
141
136
Among patients who did not experience an OHCA, females were at higher risk of bleeding than males. However, males were at a higher risk than females in the group that did experience an OHCA. Patients receiving GPI were more than twice as likely to have a bleeding event. The c-statistic for the bleeding prediction model was 0.761.
137
Clinical Outcomes of Bleeding
147 148
138 139
Mortality Bleeding was a strong predictor of mortality at one year (hazard ratio = 2.74, 95% CI [1.30, 3.51]) in patients without
Length of Stay Bleeding was a significant predictor of length of stay, with a time to discharge HR=0.58, (95% CI [0.53, 0.64]) indicating that a patient with a bleeding complication was less likely to be discharged than a patient without a bleed. The median
132 133 134 135
140
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
142 143 144 145 146
149 150 151
HLC 1918 1–11
4
J. Wlodarczyk et al.
Table 2 Proportions of Procedures with a Bleeding Complication over Time in All Patients and STEMI Patients Year
All Procedures
STEMI Cohort
(n=16860)
(n=4705)
Number of Procedures
Bleeding Complications
Number of Procedures
n(%)
Bleeding Complications n(%)
2005 2006
2562 2866
34 (1.3) 53 (1.9)
513 746
12 (2.3) 21 (2.8)
2007
2534
71 (2.8)
710
30 (4.2)
2008
2494
71 (2.9)
686
33 (4.8)
2009
2116
52 (2.5)
650
29 (4.5)
2010
2183
53 (2.4)
710
37 (5.2)
2011
2105
71 (3.4)
690
44 (6.4)
Bleeding trend over time: OR=1.090 per year 95% CI: 1.035, 1.149, p=0.001.
Table 3 Summary of Outcomes by Comorbidities and Treatments Category
All Procedures
Major Bleeds
Deaths up to 1 Year
n (%a)
n (%a)
405 (2.4%)
704 (4.2)
4149
99 (2.4)
252 (6.1)
968
27 (2.8)
68 (7.0) 127 (3.2)
N
16860
Diabetes Yes Treated with Insulin Smoking History Current
3989
86 (2.2)
Prior
7282
155 (2.1)
321 (4.4)
Never
5472
157 (2.9)
228 (4.2)
Missing Congestive Heart Failure – Prior 2 Weeks
117
7 (6.0)
28 (23.9)
837
52 (6.2)
166 (19.8)
Peripheral Vascular Disease
1165
40 (3.4)
124 (10.6)
Cerebrovascular Disease
1129
42 (3.7)
101 (9.0)
Chronic Lung Disease Atrial Fibrillation
1683 520
32 (1.9) 30 (5.8)
121 (7.2) 80 (15.4)
GPI IIb/IIIa Useb Heparin Bivalirudin
5059
228 (4.5)
297 (5.9)
16355
401 (2.5)
686 (4.2)
335
4 (1.2)
12 (3.6)
Clopidogrel No
a
153 154 155 156 157
25 (4.7)
7539
153 (2.0)
212 (2.8)
64 (12.0)
During/After
8786
227 (2.6)
428 (4.9)
Percentages for bleeds and deaths are based on number of procedures in each row.
b
152
535
Prior
GPI was administered intravenously. Intra-coronary administration was rarely used in the participating hospitals.
length of stay increased from four days to seven days and mean length of stay from 4.74.8 days to 11.214.9 days.
regression showed bleeding significantly increased the odds of re-admission (OR=1.36, 95% CI: [1.08, 1.70], p = 0.009).
158
Re-admission to Hospital The proportion of patients re-admitted to hospital in the year after discharge was higher in patients with a major bleed (37.5% (152/405) vs. 31.8% (5,357/16,860)). Logistic
Discussion
160
We examined seven years of registry data in a broad PCI population including elective, urgent and emergency
161
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
159
162
HLC 1918 1–11
5
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Table 4 Bleeding Complication Logistic Regression Model Effect
Odds Ratio
Age Under 50
1
95% Confidence Limits
P-value
50 to 59
0.72
0.49
1.06
0.094
60 to 69
0.68
0.47
1.00
0.048
70 to 79
1.38
0.97
1.97
0.076
80 or Over
1.34
0.89
2.01
0.159
Out-of-hospital Cardiac Arrest (OHCA) and Gender Interaction No OHCA and Male
1
OHCA and Female OHCA and Male
1.95 2.74
0.87 1.75
4.40 4.31
0.107 <0.001
No OHCA and Female
1.75
1.38
2.23
<0.001
1.79
2.90
<0.001
GPI Use No
1.00
Yes
2.28
Angina or ACS Type STEMI
1
NSTEMI UAP
0.82 0.62
0.63 0.40
1.08 0.97
0.160 0.037
ACP
2.30
1.12
4.73
0.023
CSA
0.63
0.44
0.90
0.010
None/Missing
0.69
0.34
1.39
0.296
2.74
5.27
<0.001
Cardiogenic Shock No
1.00
Yes
3.80
Weight 60 Kg or under
1
60 to 70 Kg
0.78
0.54
1.11
0.159
70 to 80 Kg
0.76
0.53
1.08
0.120
80 to 90 Kg
0.73
0.50
1.07
0.107
90 to 100 Kg
0.56
0.36
0.89
0.013
Over 100 Kg
0.65
0.40
1.06
0.085
Missing
0.55
0.30
1.00
0.050
Year of Admission 2011
1
2005
0.43
0.28
0.66
<0.001
2006
0.60
0.41
0.87
0.007
2007
0.92
0.65
1.31
0.650
2008
0.88
0.62
1.25
0.475
2009
0.71
0.49
1.03
0.074
2010
0.72
0.50
1.05
0.084
Percutaneous Entry Location Femoral
1
Brachial Entry
3.78
1.47
9.74
0.006
Radial Entry
0.45
0.23
0.88
0.019
0.98
1.63
0.078
Transfer Patient No
1
Yes
1.26
OHCA by gender interaction p<0.001.
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
6
J. Wlodarczyk et al.
Table 5 Time to Death up to One Year Cox model – Procedures without Cardiogenic Shock Parameter
Hazard Ratio
Bleeding Complication No Bleed
1
Bleed
2.74
95% Confidence Limits
P-value
1.95
3.85
<0.001
0.552
Age Under 50
1
50 to 59
1.17
0.70
1.97
60 to 69
1.64
1
2.68
0.050
70 to 79
2.14
1.30
3.51
0.003
80 or Over Weight
3.97
2.39
6.59
<0.001
60 Kg or Under
1
60 to 70 Kg
0.87
0.64
1.16
0.336
70 to 80 Kg
0.65
0.48
0.87
0.004
80 to 90 Kg
0.55
0.40
0.77
0.000
90 to 100 Kg
0.62
0.44
0.89
0.010
Over 100 Kg
0.35
0.21
0.59
<0.001
1.01
0.64
1.59
0.957
4.13
8.29
<0.001
Missing Out-of-hospital Cardiac Arrest No
1
Yes
5.85
Atrial Fibrillation No
1
Yes
1.61
1.17
2.22
0.003
Never Current
1 0.94
0.70
1.27
0.706
Prior
1.35
1.10
1.67
0.005
Missing
3.13
1.54
6.35
0.002
Smoking History
Family history of CAD No
1
Yes
0.74
0.60
0.91
0.004
Missing
1.55
0.88
2.73
0.126
1.57
2.69
<0.001
1.46
2.42
<0.001
0.99
1.70
0.059
1.18
1.90
0.001
Congestive heart failure - prior 2 weeks No 1 Yes
2.06
Peripheral vascular disease No
1
Yes
1.88
Cerebrovascular disease No Yes Chronic Lung Disease
1 1.30
No
1
Yes
1.49
Diabetes No
1
Yes
1.61
1.32
1.96
<0.001
1 1.22
0.97
1.53
0.089
0.29
1.64
0.398
Hypertension No Yes Angina or ACS Type STEMI
1
ACP
0.69
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
7
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Table 5. (continued). Parameter
Hazard Ratio
95% Confidence Limits
CSA
0.49
0.30
0.81
P-value 0.005
NSTEMI
0.61
0.49
0.77
<0.001
UAP None/Missing
0.51 0.37
0.36 0.18
0.71 0.76
<0.001 0.007
Percutaneous Entry Location Femoral
1
Brachial Entry
2.37
1.10
5.09
0.027
Radial Entry
0.94
0.62
1.43
0.783
0.41
1.02
0.059
Elective Presentation
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
No
1
Yes
0.65
patients, ranging in risk from atypical chest pain and stable angina to STEMI and cardiogenic shock. We found that major bleeding rates post-PCI are strongly associated with increased mortality and have been increasing over time. In addition, bleeding increases both the duration of hospitalisation and risk of re-hospitalisation. We explored several potential causes for the increase in bleeding rates including study design, clinical, and procedural variables. The risk of a selection bias was minimal as participation in the registry was extremely high (>97%) [20]. The bleeding definition used in the registry is similar to that used by other groups, including the National Cardiovascular Data Registry (NCDR1), TIMI major, GUSTO severe [21]. Importantly, it was pre-specified and used consistently, thus ensuring the comparability of bleeding rates over time. Further, an exploratory analysis showed consistent trends at each participating hospital indicating that the increases weren’t due to changes at a local hospital level. Therefore, it is unlikely that the increase in reported bleeding was due to patient selection, ascertainment, or changing practices at individual hospitals. Consistent with several studies, we found GPI use to be a strong independent predictor of bleeding [3,5,10]. GPI use in STEMI increased slightly from 64% to 69% during the study period while in NSTEMI GPI use fell from 30% to 22%. The increase in bleeding rates was still evident after adjusting for GPI use. However, the type of GPI used and dose of heparin were not recorded in the registry, and therefore changes in anti-coagulation use cannot be ruled out as contributing factors. There were increases in the proportion of patients with CS and OHCA (Appendix 1) [23–25]. Although these patients were at high risk of bleeding, they only made a small contribution to the overall increase in the bleeding rates. Bleeding risk was also high in STEMI, the elderly, women and patients with low body weight. A tendency towards treatment of frail patients might explain the increase in bleeding rates. However, the age, sex and weight distributions of patients remained stable during the study.
It is likely that the increase was due to a combination of patient and clinical factors. Further research is needed to determine the cause of the increase in bleeding rates. This study confirms the strong association between bleeding and mortality that has been noted previously [2–6,8]. Overall, the mortality rate increases two- to three-fold in patients who bled, and 10% (70/704) of patients who died had a major bleed. While the majority of the increased risk is in-hospital, the adverse impact of bleeding on survival is maintained for at least one year. The finding that bleeding does not increase mortality risk in CS patients could be due to the fact that CS patients were already at very high risk of death. Furthermore, death occurred early in CS patients (median time to death of two days), and these patients might not have survived long enough to be at risk for bleeding. This study also confirms not only that bleeding is associated with a doubling of hospital stay, which is consistent with published studies, but also that patients who experienced major bleeding were at significantly increased risk of readmission to hospital [1,24,26,27]. Indicating that in addition to clinical benefits, strategies to reduce bleeding might have important economic benefits to patients and institutions. There are a number of limitations to this study. An observational study cannot establish cause and effect. Some of the observed associations could be due to selection bias, confounding or effect modification. For instance, anti-coagulant choice could be based on factors associated with bleeding, or patients at higher risk of death could also be more likely to bleed. However, the consistency of results after multivariate modelling suggests that the associations are robust. The major strengths of this study are the quality of the registry data set which recruited almost all patients undergoing PCI at the participating centres, used consistent methodology, and had high follow-up rates to one year. The registry population appears to be representative of PCI patients in Australia, and similar to those in other major cohorts [8,9,22]. The rates and types of major bleeding are comparable with those reported recently in other countries [1,8,9,22].
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241
HLC 1918 1–11
8
242
J. Wlodarczyk et al.
251
However, caution should be used when generalising our results to other populations. These results might not be applicable where different PCI access methods or different anti-coagulation strategies are used. There were few patients treated with bivalirudin in the registry (n=335 (2.0%)), thus we were also unable to explore the relative impact of this anti-coagulant on bleeding. Also the vast majority of patients in the registry had femoral arterial access for their procedure, however radial patients had a significantly lower bleed risk which is consistent with previous reports [2].
252
Conclusion
253
261
The likelihood of major bleeding following PCI in this cohort has been increasing steadily in recent years with rates doubling over a seven-year period to 2011. This suggests that greater efforts are required to avoid this serious complication of PCI in Australia. These results identify high risk patients for bleeding that could benefit most from different anti-coagulation strategies which has the potential to reducing the burden to healthcare providers and improve patient outcomes.
262
Financial Disclosure
263
The Melbourne Interventional Group acknowledges funding from Abbott, Astra-Zeneca, Biotronik, Boston Scientific, Bristol-Myers Squibb, Cordis Johnson & Johnson, CSL, Medtronic, MSD, Pfizer, Sanofi-Aventis, Servier, ScheringPlough and The Medicines Company. These companies do not have access to data. Dr. Duffy’s work is supported by a NHMRC Program Grant to the Alfred and Baker Medical Unit. Prof Reid’s work is supported a NHMRC Senior Research Fellowship and NHMRC Program Grant. Dr. Wlodarczyk is the principle of a consulting company contracted by The Medicines Company to perform
243 244 245 246 247 248 249 250
254 255 256 257 258 259 260
264 265 266 267 268 269 270 271 272 273
the analyses found in this report and to prepare the manuscript.
274
Acknowledgements
276
We would like to thank Bishoy Rizkilla, Medical Director for The Medicines Company for his assistance in developing this project. We would also like to thank the MIG team:
277
MIG Data Management Centre, CCRE, Monash University: Professor Chris Reid, Dr Nick Andrianopoulos, Ms Angela Brennan.
275
278 279
280
MIG Steering Committee: Professor Chris Reid, A/Professor Andrew Ajani, Dr Stephen Duffy, Dr David Clark, A/Professor Gishel New, Dr Chin Hiew, Dr Nick Andrianopoulos, Dr Ernesto Oqueli, Ms Angela Brennan.
281
The following investigators, data managers and institutions participated in the MIG Database: Alfred Hospital: SJ Duffy, JA Shaw, A Walton, A Dart, A Broughton, J Federman, C Keighley, C Hengel, KH Peter, J O’Brien, H Bala, S Nanayakkara, R Vandernet, R Huntington; Austin Hospital: DJ Clark, O Farouque, M Horrigan, J Johns, L Oliver, J Brennan, R Chan, G Proimos, T Dortimer, B Chan, V Nadurata, R Huq, D Fernando, K Charter, L Brown, A AlFiadh, H Sugumar, R Spencer, J Ramchand, P Scott; Box Hill Hospital: G New, L Roberts, M Freeman, A Teh, M Rowe, G Proimos, Y Cheong, C Goods, D Fernando, A Baradi, D Jackson, J Sajeev; Frankston Hospital: R Lew, G Szto, R Teperman, R Huq; Geelong Hospital: C Hiew, M Sebastian, T Yip, M Mok, K Rankin, C Machado, J Dyson, B McDonald, L Duff; Monash University: H Krum, C Reid, N Andrianopoulos, AL Brennan, V Chand, D Dinh, BP Yan; Royal Melbourne Hospital: AE Ajani, R Warren, D Eccleston, J Lefkovits, R Iyer, R Gurvitch, W Wilson, M Brooks, M Yudi, S Biswas, J Yeoh, C Cheshire, N Gaikwad.
285 286
282 283 284
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
Appendix 1. Changes in Patient Characteristics and Anti-Coagulation Use over Time Year
All Procedures
2005
2006
2007
2008
2009
2010
2011
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
2562
2866
2534
2494
2116
2183
2105
16860
Age Under 50
300 (11.7)
342 (11.9)
284 (11.2)
308 (12.3)
243 (11.5)
230 (10.5)
193 (9.2)
1900 (11.3)
50 to 59
536 (20.9)
673 (23.5)
540 (21.3)
595 (23.9)
496 (23.4)
515 (23.6)
490 (23.3)
3845 (22.8)
60 to 69
708 (27.6)
789 (27.5)
752 (29.7)
700 (28.1)
601 (28.4)
670 (30.7)
648 (30.8)
4868 (28.9)
70 to 79
732 (28.6)
769 (26.8)
688 (27.2)
611 (24.5)
543 (25.7)
494 (22.6)
545 (25.9)
4382 (26)
80 or Over
286 (11.2)
293 (10.2)
270 (10.7)
280 (11.2)
233 (11)
274 (12.6)
229 (10.9)
1865 (11.1)
Sex Male
1849 (72.2) 2140 (74.7) 1939 (76.5) 1897 (76.1) 1619 (76.5) 1631 (74.7) 1606 (76.3) 12681 (75.2)
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
9
Incidence, Predictors and Outcomes of Major Bleeding in Patients
Appendix 1. (continued). Year
2005
2006
2007
2008
2009
2010
2011
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
Weight 60 Kg or under
245 (9.6)
248 (8.7)
183 (7.2)
198 (7.9)
156 (7.4)
188 (8.6)
175 (8.3)
60 to 70 Kg
476 (18.6)
457 (15.9)
418 (16.5)
411 (16.5)
337 (15.9)
360 (16.5)
336 (16)
1393 (8.3) 2795 (16.6)
70 to 80 Kg
625 (24.4)
711 (24.8)
625 (24.7)
590 (23.7)
547 (25.9)
563 (25.8)
482 (22.9)
4143 (24.6)
80 to 90 Kg
532 (20.8)
664 (23.2)
567 (22.4)
557 (22.3)
445 (21)
468 (21.4)
474 (22.5)
3707 (22)
90 to 100 Kg Over 100 Kg
366 (14.3) 214 (8.4)
423 (14.8) 256 (8.9)
351 (13.9) 240 (9.5)
355 (14.2) 289 (11.6)
312 (14.7) 213 (10.1)
331 (15.2) 250 (11.5)
310 (14.7) 307 (14.6)
2448 (14.5) 1769 (10.5)
Missing
104 (4.1)
107 (3.7)
150 (5.9)
94 (3.8)
106 (5)
23 (1.1)
21 (1)
605 (3.6)
NYHA I
957 (45.2) 1029 (47.1) 1154 (54.8)
7897 (46.8)
II
453 (17.7)
465 (16.2)
489 (19.3)
321 (12.9)
432 (20.4)
329 (15.1)
2824 (16.7)
III
114 (4.4)
260 (9.1)
166 (6.6)
365 (14.6)
254 (12)
191 (8.7)
IV
60 (2.3)
246 (8.6)
231 (9.1)
185 (7.4)
289 (13.7)
401 (18.4)
294 (14)
1706 (10.1)
405 (15.8) 17 (0.7)
787 (27.5) 23 (0.8)
540 (21.3) 36 (1.4)
612 (24.5) 55 (2.2)
184 (8.7) 72 (3.4)
233 (10.7) 64 (2.9)
226 (10.7) 74 (3.5)
2987 (17.7) 341 (2)
51 (2)
71 (2.5)
61 (2.4)
78 (3.1)
87 (4.1)
76 (3.6)
490 (2.9)
77 (3)
51 (1.8)
Missing Out-of-Hospital Cardiac Arrest Cardiogenic Shock
1530 (59.7) 1108 (38.7) 1108 (43.7) 1011 (40.5)
66 (3)
335 (15.9) 96 (4.6)
1446 (8.6)
Angina or ACS Type Atypical Chest Pain Chronic Stable Angina
792 (30.9) 1019 (35.6)
27 (1.1)
18 (0.7)
6 (0.3)
782 (30.9)
729 (29.2)
605 (28.6)
21 (1) 572 (26.2)
29 (1.4)
229 (1.4)
541 (25.7)
5040 (29.9)
UAP
481 (18.8)
291 (10.2)
228 (9)
219 (8.8)
202 (9.5)
205 (9.4)
192 (9.1)
1818 (10.8)
NSTEMI
560 (21.9)
672 (23.4)
645 (25.5)
761 (30.5)
594 (28.1)
594 (27.2)
566 (26.9)
4392 (26)
STEMI None/Missing
513 (20) 139 (5.4)
746 (26) 87 (3)
710 (28) 142 (5.6)
686 (27.5) 81 (3.2)
650 (30.7) 59 (2.8)
710 (32.5) 81 (3.7)
690 (32.8) 87 (4.1)
4705 (27.9) 676 (4)
GPI
655 (25.6)
816 (28.5)
750 (29.6)
795 (31.9)
688 (32.5)
678 (31.1)
677 (32.2)
5059 (30)
Clopidogrel - Prior
877 (34.2) 1310 (45.7) 1236 (48.8) 1282 (51.4)
966 (45.7) 1002 (45.9)
866 (41.1)
7539 (44.7)
1105 (52.2) 1108 (50.8) 1051 (49.9)
8786 (52.1)
Anti-Coagulant Usage
Clopidogrel - During or After
1608 (62.8) 1495 (52.2) 1247 (49.2) 1172 (47)
Heparin
2495 (97.4) 2745 (95.8) 2446 (96.5) 2430 (97.4) 2061 (97.4) 2128 (97.5) 2050 (97.4) 16355 (97)
Bivalirudin Aspirin Prasugrel Ticagrelor
3 (0.1)
97 (3.4)
94 (3.7)
58 (2.3)
31 (1.5)
2527 (98.6) 2822 (98.5) 2504 (98.8) 2460 (98.6) 2095 (99) 0 (0) 0 (0) 0 (0) 0 (0) 3 (0.1) 0 (0)
0 (0)
0 (0)
0 (0)
30 (1.4)
22 (1)
335 (2)
2163 (99.1) 2068 (98.2) 16639 (98.7) 59 (2.7) 228 (10.8) 290 (1.7)
0 (0)
0 (0)
4 (0.2)
4 (0)
Appendix 2. Summary of Disease Extent and Procedural Characteristics Number of Procedures (n=16860) n(%)
Major Bleeds n (%a)
Deaths up to 1 Year n (%a)
Disease Extent Single Vessel Disease
6760 (40.1)
8 (2.1)
193 (2.9)
Multi Vessel Disease
9710 (57.6)
155 (2.3)
502 (5.2)
390 (2.3)
242 (2.5)
9 (2.3)
1
13902 (82.5)
327 (2.4)
2
2627 (15.6)
67 (2.6)
331 (2.0)
11 (3.3)
6066 (36.0)
140 (2.3)
Missing Number of Lesions Treated
3+ Right Coronary Artery Lesion Left Main Lesion
193 (1.1)
4 (2.1)
559 (4) 122 (4.6) 23 (6.9transf) 215 (3.5) 34 (17.6)
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
HLC 1918 1–11
10
J. Wlodarczyk et al.
Appendix 2. (continued). Number of Procedures
Major Bleeds
Deaths up to 1 Year
(n=16860) n(%)
n (%a)
n (%a)
Left Coronary Artery Lesion
6733 (39.9)
194 (2.9)
322 (4.8)
Left Circumflex Lesion
4006 (23.8)
67 (1.7)
152 (3.8)
Elective Presentation
6221 (36.9)
74 (1.2)
100 (1.6)
Transfer Patient
2962 (17.6)
88 (3)
123 (4.2)
397 (2.4)
58 (14.6)
161 (40.6)
75 (0.4)
5 (6.7)
12 (16)
Intra-Aortic Balloon Pump Percutaneous Entry Location Brachial Entry Radial Entry Femoral Missing
945 (5.6)
9 (1)
27 (2.9)
15839 (93.9)
391 (2.5)
665 (4.2)
1 (0.0)
Stent Used in Procedure
0 (0)
0 (0)
15869 (94.1)
370 (2.3)
631 (4)
BMSb only
8539 (50.7)
225 (2.6)
408 (4.8)
DESc only Both BMS and DES
6893 (40.9) 437 (2.6)
127 (1.8) 18 (4.1)
202 (2.9) 21 (4.8)
991 (5.9)
35 (3.5)
73 (7.4)
Type of Stent
None (PCI only) Closure Device Seal
a
43 (3.6)
25 (2.1)
280 (1.7)
6 (2.1)
4 (1.4)
Other
160 (1.0)
4 (2.5)
2 (1.3)
None
15210 (90.2)
352 (2.3)
673 (4.4)
Percentages for bleeds and deaths are based on number of procedures in each row.
b c
1210 (7.2)
Suture
BMS: Bare metal stent.
DES: Drug eluting stent.
Appendix 3. Summary of Percutaneous Entry Locations by Year Year
2005 n (%)
2006 n (%)
2007 n (%)
2008 n (%)
2009 n (%)
2010 n (%)
2011 n (%)
Total n (%)
All Procedures
2562
2866
2534
2494
2116
2183
2105
16860
Brachial
15 (0.6)
14 (0.5)
10 (0.4)
12 (0.5)
Radial Femoral
59 (2.3) 2488 (97.1)
125 (4.4) 2727 (95.2)
121 (4.8) 2403 (94.8)
77 (3.1) 2405 (96.4)
Missing
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
0
0
0
References [1] Happe LE, Rao SV, Horblyukc R, Franklin M, Lunacseka OE, Menditto L. Consequences of major bleeding in hospitalized patients with non-ST segment elevationacute coronary syndromes receiving injectable anticoagulants. Cur Med Research and Opinion 2009;25:413–20. [2] Rao SV, Kaul PR, Liao L, Armstrong PW, Ohman EM, Granger CB, et al. Association between bleeding, blood transfusion, and costs among patients with non–ST-segment elevation acute coronary syndromes. Am Heart J 2008;155:369–74. [3] Kinnaird TD, Stabile E, Mintz GS, Lee CW, Canos DA, Gevorkian N, et al. Incidence, predictors, and prognostic implications of bleeding and blood transfusion following percutaneous coronary interventions. Am J Cardiol 2003;82:930–5. [4] Mrdovic I, Savic L, Asanin M, Cvetinovic N, Brdar N, Djuricic N, et al. Sex-related analysis of short- and long-term clinical outcomes and
0
8 (0.4) 101 (4.8) 2007 (94.9) 0
5 (0.2) 167 (7.7) 2010 (92.1) 1 (0.0)
11 (0.5) 295 (14) 1799 (85.5) 0
75 (0.4) 945 (5.6) 15839 (93.9) 1 (0.0)
bleeding among patients treated with primary percutaneous coronary intervention: an evaluation of the RISK-PCI data. Can J Cardiol 2013;29:1097–103. [5] Mehran R, Pocock S, Nikolsky E, Dangas GD, Clayton T, Claessen BE, et al. Impact of bleeding on mortality after percutaneous coronary intervention results from a patient-level pooled analysis of the REPLACE-2 (randomized evaluation of PCI linking angiomax to reduced clinical events), ACUITY (acute catheterization and urgent intervention triage strategy), and HORIZONS-AMI (harmonizing outcomes with revascularization and stents in acute myocardial infarction) trials. JACC Cardiovasc Interv 2011;4:654–64. [6] Boden H, Velders MA, van der Hoeven BL, Cannegieter SC, Schalij MJ. In-hospital major bleeding and its clinical relevance in patients with ST elevation myocardial infarction treated with primary percutaneous coronary intervention. Am J Cardiol 2013;112:1533–9. [7] Rao SV, Dai D, Subherwal S, Weintraub WS, Brindis RS, Messenger JC, et al. Association between periprocedural bleeding and long-term
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
HLC 1918 1–11
11
Incidence, Predictors and Outcomes of Major Bleeding in Patients
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
outcomes following percutaneous coronary intervention in older patients. JACC Cardiovasc Interv 2012;5:958–65. Chhatriwalla AK, Amin AP, Kennedy KF, House JA, Cohen DJ, Rao SV, et al. Association between bleeding events and in-hospital mortality after percutaneous coronary intervention. JAMA 2013;309:1022–9. Mehta SK, Frutkin AD, Lindsey JB, House JA, Spertus JA, Rao SV, et al. Bleeding in patients undergoing percutaneous coronary intervention: The development of a clinical risk algorithm from the National Cardiovascular Data Registry. Circ Cardiovasc Interv 2009;2:222–9. Pocock SJ, Mehran R, Clayton TC, Nikolsky E, Parise H, Fahy M, et al. Prognostic modelling of individual patient risk and mortality impact of ischemic and hemorrhagic complications: Assessment from the ACUITY Trial. Circulation 2010;121:43–51. Hanna EB, Rao SV, Manoukian SV, Saucedo JF. The evolving role of glycoprotein IIb/IIIa inhibitors in the setting of percutaneous coronary intervention strategies to minimize bleeding risk and optimize outcomes. JACC Cardiovasc Interv 2010;3:1209–19. Chan W, Clark DJ, Ajani AE, Yap CH, Andrianopoulos N, Brennan AL, et al. Progress towards a National Cardiac Procedure Database – Development of the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) and Melbourne Interventional Group (MIG) Registries. Heart Lung Circ 2011;20:10–8. Ajani AE, Szto G, Eccleston D, Clark DJ, Lefkovits J, Chew DP, et al. The foundation and launch of the Melbourne interventional group: a collaborative interventional cardiology project. Heart Lung Circ 2006;15:44–7. Ajani AE, Duffy SJ, Clark DJ, Lefkovits J, Warren R, Gurvitch R, et al. Outcomes after percutaneous coronary intervention in contemporary Australian practice: insights from a large multicentre registry. MJA 2008;189(8):423–8. Andrianopoulos N, Dinh D, Duffy SJ, Clark DJ, Brennan A, Chan W, et al. Quality Control activities associated with registries in Interventional Cardiology and Surgery. Heart Lung Circ 2011;20:180–6. Lagerqvist B, James SK, Stenestrand U, Lindback J, Nilsson T, Wallentin L. Long-term outcomes with drug-eluting stents versus bare-metal stents in Sweden. N Engl J Med 2007;356:1009–19. Lincoff AM, Kleiman NS, Kereiakes DJ, Feit F, Kleiman NS, Jackman JD, et al. Long-term efficacy of bivalirudin and provisional glycoprotein IIb/ IIIa blockade vs heparin and planned glycoprotein IIb/IIIa blockade
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
during percutaneous coronary revascularization: REPLACE-2 randomized trial. JAMA 2004;292:696–703. Stone GW, Bertrand M, Colombo A, Dangas G, Farkouh ME, Feit F, et al. Acute Catheterization and Urgent Intervention Triage strategy (ACUITY) trial: study design and rationale. Am Heart J 2004;148:764–75. Stone GW, Witzenbichler B, Guagliumi G, Peruga JZ, Brodie BR, Dudek D, et al. Bivalirudin during primary PCI in acute myocardial infarction. N Engl J Med 2008;358:2218–30. Centre of Research Excellence in Patient Safety, Melbourne Interventional Group Interventional Cardiology [Internet], Melbourne Vic, Centre of Research Excellence in Patient Safety; 2006 [Updated: 2010 September 06; cited 2015 May 29]. Available from: http://www. registries.org.au/registries/melb_interventional_cardiology.html. Mehran R, Rao SV, Bhatt DL, Gibson CM, Caixeta A, Eikelboom J, et al. Standardized Bleeding Definitions for Cardiovascular Clinical Trials: A Consensus Report From the Bleeding Academic Research Consortium. Circulation 2011;123:2736–47. Chew DP, French J, Briffa TG, Hammett CJ, Ellis CJ, Ranaslnghe I, et al. Acute coronary syndrome care across Australia and New Zealand: the SNAPSHOT ACS study. MJA 2013;199:1–7. Brennan AL, Andrianopoulos N, Duffy SJ, Reid CM, Clark DJ, Loane P, et al. Trends in door-to-balloon time and outcomes following primary percutaneous coronary intervention for ST-elevation myocardial infarction: An Australian perspective. Internal Medicine Journal 2014;44:471–7. Yan BP, Ajani AE, Clark DJ, Duffy SJ, Andrianopoulos N, Brennan AL, et al. Recent trends in Australian percutaneous coronary intervention practice: insights from the Melbourne International Group Registry. MJA 2011;195:122–7. Spencer RJ, Andrianopoulos N, Scott P, Scott P, Farouque O, Duffy SJ, et al. Increasing risk profile amongst patients undergoing PCI in Australia. Global Heart 2014;9(1):e202. Aronow HD, Peyser PA, Eagle KA, Bates ER, Werns SW, Russman PL, et al. Predictors of length of stay after coronary stenting. Am Heart J 2001;42:799–805. Rao SV, Ou FS, Wang TY, Roe MT, Brindis RG, Rumsfeld JS, et al. Trends in the prevalence and outcomes of radial and femoral approaches to percutaneous coronary intervention. A report from the National Cardiovascular Data Registry. J Am Coll Cardiol Intv 2008;1:379–86.
Please cite this article in press as: Wlodarczyk J, et al. Incidence, Predictors and Outcomes of Major Bleeding in Patients Following Percutaneous Coronary Interventions in Australia. Heart, Lung and Circulation (2015), http://dx.doi.org/ 10.1016/j.hlc.2015.06.826
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412