1

Coronary perforation (CP) is a rare but serious complica-tion of percutaneous coronary intervention (PCI) with an

estimated incidence of ≈0.5%.1 Entry of arterial blood into the pericardial space can lead to rapid elevation of the pericardial pressure and rapid hemodynamic compromise. Historically, urgent surgical drainage was a standard treatment, but the development of new technologies such as covered stents and

embolization coils and new techniques such as thrombus and fat injection have allowed many perforations to be treated in the catheterisation laboratory without the need for surgical intervention.2,3 However, despite improvements in interven-tional skills and equipment, PCIs are increasingly complex with a higher prevalence of multivessel disease, worsening comorbidities (such as increasing age and renal dysfunction),

Background—As coronary perforation (CP) is a rare but serious complication of percutaneous coronary intervention (PCI) the current evidence base is limited to small series. Using a national PCI database, the incidence, predictors, and outcomes of CP as a complication of PCI were defined.

Methods and Results—Data were prospectively collected and retrospectively analyzed from the British Cardiovascular Intervention Society data set on all PCI procedures performed in England and Wales between 2006 and 2013. Multivariate logistic regressions and propensity scores were used to identify predictors of CP and its association with outcomes. In total, 1762 CPs were recorded from 527 121 PCI procedures (incidence of 0.33%). Patients with CP were more often women or older, with a greater burden of comorbidity and underwent more complex PCI procedures. Factors predictive of CP included age per year (odds ratio [OR], 1.03; 95% confidence intervals, 1.02–1.03; P<0.001), previous coronary artery bypass graft (OR, 1.44; 95% confidence intervals, 1.17–1.77; P<0.001), left main (OR, 1.54; 95% confidence intervals, 1.21–1.96; P<0.001), use of rotational atherectomy (OR, 2.37; 95% confidence intervals, 1.80–3.11; P<0.001), and chronic total occlusions intervention (OR, 3.96; 95% confidence intervals, 3.28–4.78; P<0.001). Adjusted odds of adverse outcomes were higher in patients with CP for all major adverse coronary events, including stroke, bleeding, and mortality. Emergency surgery was required in 3% of cases. Predictors of mortality in patients with CP included age, diabetes mellitus, previous myocardial infarction, renal disease, ventilatory support, use of circulatory support, glycoprotein inhibitor use, and stent type.

Conclusions—Using a national PCI database for the first time, the incidence, predictors, and outcomes of CP were defined. Although CP as a complication of PCI occurred rarely, it was strongly associated with poor outcomes. (Circ Cardiovasc Interv. 2016;9:e003449. DOI: 10.1161/CIRCINTERVENTIONS.115.003449.)

Key Words: atherectomy, coronary ◼ percutaneous coronary intervention ◼ propensity score ◼ stents ◼ stroke

© 2016 American Heart Association, Inc.

Circ Cardiovasc Interv is available at http://circinterventions.ahajournals.org DOI: 10.1161/CIRCINTERVENTIONS.115.003449

Received November 27, 2015; accepted June 23, 2016.From the Department of Cardiology, University Hospital of Wales, Cardiff, United Kingdom (T.K., N.O.-G., R.A.); Keele Cardiovascular Research

Group, Institute of Science and Technology in Medicine and Primary Care and Health Sciences, University of Keele, Stoke-on-Trent and Royal Stoke Hospital, UHNM, Stoke-on-Trent, United Kingdom (C.S.K., M.A.M.); Farr Institute, University of Manchester, United Kingdom (E.K., M.A.M.); Department of Cardiology, Queen Elizabeth Hospital, Edgbaston, Birmingham, United Kingdom (P.L.); and Department of Cardiology, The James Cook University Hospital, Middlesborough, United Kingdom (M.d.B.).

The Data Supplement is available at http://circinterventions.ahajournals.org/lookup/suppl/doi:10.1161/CIRCINTERVENTIONS.115.003449/-/DC1.Correspondence to Tim Kinnaird, MD, Department of Cardiology, University Hospital of Wales, Cardiff, United Kingdom. E-mail tim.Kinnaird2@

wales.nhs.uk

Incidence, Determinants, and Outcomes of Coronary Perforation During Percutaneous Coronary Intervention in

the United Kingdom Between 2006 and 2013An Analysis of 527 121 Cases From the British Cardiovascular Intervention

Society Database

Tim Kinnaird, MD; Chun Shing Kwok, MBBS, MSc, BSc; Evangelos Kontopantelis, PhD; Nicholas Ossei-Gerning, MD; Peter Ludman, MD; Mark deBelder, MD;

Richard Anderson, MD; Mamas A. Mamas, MADPhil; on behalf of the British Cardiovascular Intervention Society and the National Institute for

Cardiovascular Outcomes Research

Coronary Interventions

2 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

and increasingly complex procedures including the treatment of chronic total occlusions (CTO). In light of these temporal changes, it is likely that CPs will continue to occur.

Although there are several published series of CP, the rarity of the complication has limited the literature to small series derived from single-center experience. The largest published series describes 124 events, and the total litera-ture is <1000 cases.4–20 Given the small numbers of events per published series, it has not been possible to examine several key questions relating to the occurrence and out-comes of CP. First, it is unclear from the available evidence what the true incidence of CP is and whether its incidence increasing or decreasing. Second, because of the size of the published series, it is not been possible to define exactly which demographic and procedural factors are predictive of CP. Finally, the treatment and outcomes of CP are not clearly defined.

Therefore, the objectives of the present study were to over-come these limitations, and use for the first time a national PCI database to define the true incidence of CP during PCI, to describe which factors are predictive of CP, and to define the outcomes after CP.

Methods

Study Design, Setting, and ParticipantsWe retrospectively analyzed prospectively collected national data from all patients who underwent PCI in England and Wales between January 2006 and December 2013.

Setting, Data Source, and Study SizeData on PCI practice in the United Kingdom were obtained from the British Cardiovascular Intervention Society (BCIS) data set that re-cords this information prospectively and publishes this information in the public domain as a part of the national transparency agenda.21 The overall data collection process is overseen The National Institute of Cardiovascular Outcomes Research (http://www.ucl.ac.uk/ni-cor/), and in 2013, 98.6% of all PCI procedures performed in the National Health Service hospitals in England and Wales (www.bcis.org.uk/) were recorded on the database. The BCIS-National Institute of Cardiovascular Outcomes Research database contains 113 clini-cal, procedural, and outcomes variables with ≈80 000 new records added each year.22–24 The participants of the database are tracked by the Medical Research Information Services for subsequent mortal-ity using the patients’ National Health Service number (a unique

identifier for any person registered within the National Health Service in England and Wales).

Study DefinitionsWe analyzed all recorded PCI procedures that were undertaken in the England and Wales between January 1, 2006, and December 31, 2013. Patients were categorized according to whether they sustained a CP dur-ing the PCI procedure. In the BCIS data set, the definition of CP is left to the discretion of the operators. The outcomes examined were in-hos-pital major adverse cardiac events (MACE) (a composite of in-hospital mortality and in-hospital myocardial infarction or reinfarction and tar-get vessel revascularization), 30-day mortality, 1-year mortality, 5-year mortality, in-hospital reinfarction, in-hospital emergency coronary artery bypass graft (CABG), in-hospital cardiac tamponade, in-hospital stroke, and in-hospital major bleeding (defined as gastrointestinal bleed, intrace-rebral bleed, retroperitoneal hematoma, blood or platelet transfusion, or an arterial access site complication requiring surgery). Participants with missing information on CP, age, or sex were excluded. A detailed ac-count of the participant inclusion process is shown in Figure 1.

Data AnalysesStatistical analysis was performed using Stata v13.1 (College Station, TX). Multiple imputations using the mi impute command were used to reduce the potential bias from missing data, assuming missing at random mechanisms. We used chained equations to impute the data for all variables with missing information and generated 10 data sets to be used the analyses.

We examined the baseline characteristics of participants by CP status. These variables included age, sex, smoking status, body mass index, family history of coronary heart disease, hypertension, hyper-lipidaemia, diabetes mellitus, previous myocardial infarction, previ-ous stroke, peripheral vascular disease, valvular heart disease, renal disease, previous PCI, previous CABG, left ventricular function, cardiogenic shock, circulatory support, mechanical ventilation, an-tiplatelet therapy, warfarin use, bivalirudin use, glycoprotein IIb/IIIa inhibitors use, vessel attempted for PCI (vein graft, left main, left anterior descending, circumflex, and right coronary), postprocedural TIMI (Thrombolysis in Myocardial Infarction) flow, radial access, stent implanted, rotational atherectomy use, laser angioplasty use, cutting balloon use, presence of a chronic occlusion, surgical cover, year of PCI, and indication (stable angina, non–ST-segment–eleva-tion myocardial infarction, and ST-segment–elevation myocardial infarction). Descriptive measures for all these variables were calcu-lated over time. We tested for associations between each categorical variable and CP using a χ2 test, and for continuous variables we used 1-way ANOVA. In addition, the rates of CP by year of PCI are pre-sented. The outcomes of interest were 30-day mortality, 1-year mor-tality, 5-year mortality, in-hospital bleeding, reinfarction, emergency CABG, stroke, cardiac tamponade, side branch occlusion, and coro-nary dissection for which we initially calculated the crude rates by CP status. To obtain adjusted measures of the associations between CP and the outcomes, we used logistic regressions. In the first set of models, we used multiple logistic regression and included all poten-tial predictors as covariates, to quantify the independent association between perforation and outcomes. In a second set of models, we used the potential predictors to calculate a propensity score on per-foration and weighted simple logistic regressions, of perforation on outcomes, on the inverse of the score (inverse probability treatment weighting). A third set of simple regressions used the propensity score to perform matching with replacement to control for the effect of the covariates and estimate the average treatment effect of CP on outcomes (teffects psmatch in Stata). The average treatment effect is the mean difference in the outcome between patients with perfora-tion and those without. The covariates included in the models were as follows: age, sex, smoking status, body mass index, family histo-ry of coronary artery disease, hypercholesterolaemia, hypertension, diabetes mellitus, previous myocardial infarction, previous stroke, peripheral vascular disease, valvular heart disease, renal disease, previous PCI, previous CABG, left ventricular function, cardiogenic

WhAT IS KNOWN

•Coronary perforation is a rare event, and there are little data about the predictors, outcomes, and tem-poral trends of coronary perforation after percutane-ous coronary intervention.

WhAT ThE STUDy ADDS

• In a national database, the incidence of coronary per-foration was 0.3%.

•Factors associated with perforation included age, previous, and left main or chronic total occlusion interventions.

3 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

shock, ventilator use, circulatory support, antiplatelet therapy, war-farin, glycoprotein IIb/IIIa inhibitor, bivalirudin use, radial access, surgical cover, year, vessel of PCI, chronic occlusion, postprocedure TIMI flow, stent type used, rotational atherectomy, laser angioplasty, cutting balloon, side branch occlusion, coronary dissection, and di-agnosis. The success of the propensity score matching was investi-gated by calculating descriptive statistics of the propensity score in each group and its absolute difference within each matched pair (bal-ance diagnostics). An additional sensitivity analysis was performed by performing multiple logistic regressions with and without adjust-ments for center volume.

ResultsIncidence and Baseline Demographics by Perforation StatusBetween 2006 and 2013, 1762 CPs were recorded from 527 121 PCI procedures giving an overall incidence of 0.33%. The crude numbers of CP increased year on year (Figure 2), reflecting an increase in the total PCI volume in the United Kingdom during the study period. However, the annual incidence varied from 0.29% to 0.36% with a trend upward that did not reach statistical significance (P=0.359). The baseline demographics for patients with and without CP are presented in Table 1. Patients with CP were more often women or older, with a greater burden of comor-bidity including hypertension, hypercholesterolemia, previous myocardial infarction, peripheral vascular disease, and left ven-tricular dysfunction. Importantly, when considering consent for procedures, CP was more likely to occur in stable angina PCI.

Procedural Variables by Perforation StatusThe procedural variables for patients with and without CP are presented in Table 2. Perforation was associated with the use of glycoprotein IIb/IIIa inhibitor, left main disease, circumflex disease, and right coronary disease, postprocedure TIMI flow, type of stent implanted, rotational atherectomy, laser angio-plasty, and chronic occlusion.

Using multivariate analyses covariates found to be asso-ciated with CP were identified and are presented in Table 3. Factors associated with an increased risk of perforation were age, hypercholesterolaemia, previous CABG, left main inter-vention, CTO intervention, use of rotational atherectomy, procedural dissection, side-branch occlusion, and non–ST-segment–elevation myocardial infarction diagnosis. Factors associated with a decreased risk of perforation were male sex, diabetes mellitus, presentation with shock, and use of a cutting balloon.

Clinical Outcomes by Perforation StatusClinical and procedural complications and adverse out-comes recorded in the BCIS database were more frequent in patients with CP (Table 4). In-hospital MACE was sig-nificantly higher in those patients with CP than in those without (26% versus 2%; P<0.001), as was 30-day, 1-year, and 5-year mortality. The 30-day mortality after perforation varied from 6.6% to 15.5% with a significant upward trend (P=0.049) that was also significant for 1-year mortality

Figure 1. Flow diagram of participant inclusion. PCI indicates percutaneous coronary intervention.

Figure 2. Crude numbers and incidence of coronary perforation from 2006 to 2013. PCI indicates percutaneous coronary intervention.

4 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

(P<0.001). We observed similar significant increase in mortality at both 30 days and 1 year for the group that did not sustain CP over time (Table I in the Data Supplement). Overall, national PCI mortality with and without perfora-tion by year is presented in Figure 3. The characteristics of predictive factors for perforation over time are presented in Table II in the Data Supplement (means or percentages and respective confidence intervals) and indicate a signifi-cant increase in the mean complexity of the cases expe-riencing perforation (including patient age, presentation with shock, diabetes mellitus, history of CABG, and left main PCI). Noncoronary end points including in-hospital major bleeding and stroke were also significantly more frequent. In patients with CP, tamponade occurred in 14% of cases with 3% of patients required emergency repara-tive surgery. Using multiple logistic regression analyses and inverse probability-weighting propensity scoring, the adjusted odds of clinical outcomes are presented in Table 5. The predictors of 30-day mortality in patients with a CP are presented in Table 6 and included increasing age, pre-vious diabetes mellitus, previous myocardial infarction, renal disease, use of ventilatory support, use of circulatory support, glycoprotein IIb/IIIa inhibitor use, and stent type. The results of the propensity score matching analysis on 10 imputed data sets using average treatment effects are demonstrated in Table III in the Data Supplement. There were significant increases in in-hospital MACE, 30-day mortality, 1-year mortality, in-hospital cardiac tamponade, and in-hospital bleeding with CP after propensity score matching. The balance diagnostics of the propensity score matching are shown in Table IV in the Data Supplement. In considering the effect of center volume on outcomes, the overall median and interquartile range of center volume was 1295 (617–1771) procedures/y. The respective center volume results for patients who had CPs was lower 1189 (599–1771) procedures/y. Sensitivity analysis considering the effect of an additional adjustment for center volume did not significantly alter the results in the absence of this adjustment (Table V in the Data Supplement).

DiscussionThe current study is the first analysis of CP as a complication of PCI performed from a national angioplasty database. Although there are several previously published series, their small size

Table 1. Baseline Participant Characteristics Variables by Coronary Perforation Status

VariableNo Perforation (n=525 359)

Perforation (n=1762) P Value

Age, y 64.8±11.8 68.9±11.1 <0.001

Male, n (%) 388 723 (74) 1154 (65) <0.001

Smoking status, n (%) <0.001

Never smoked 168 993 (37) 578 (38)

Ex-smoker 181 699 (39) 695 (45)

Current smoker 110 110 (24) 261 (17)

BMI, kg/m2 28.3±5.1 28.0±5.0 0.093

Family history of CAD, n (%) 205 100 (46) 637 (44) 0.123

Hypertension, n (%) 266 498 (52) 1042 (61) <0.001

Hypercholesterolaemia, n (%) 278 443 (55) 1034 (60) <0.001

Diabetes mellitus, n (%) 96 919 (19) 332 (20) 0.598

Previous MI, n (%) 138 973 (28) 585 (37) <0.001

Previous stroke, n (%) 19 230 (4) 88 (5) 0.003

Peripheral vascular disease, n (%)

23 416 (5) 125 (7) <0.001

Valvular heart disease, n (%) 6446 (1) 26 (2) 0.359

Renal disease, n (%) 13 196 (3) 66 (4) 0.001

Previous PCI, n (%) 113 876 (23) 435 (26) 0.001

Previous CABG, n (%) 31 575 (8) 176 (14) <0.001

LV function category, n (%) <0.001

>50% 185 637 (73) 691 (66)

30%–49% 54 125 (21) 266 (25)

<30% 15 249 (6) 88 (8)

Cardiogenic shock, n (%) 10 555 (2) 48 (3) 0.020

Circulatory support, n (%) 10 067 (2) 144 (9) <0.001

Mechanical ventilation, n (%) 6695 (1) 18 (1) 0.421

Antiplatelet therapy, n (%) <0.001

Clopidogrel 419 250 (93) 1459 (94)

Prasugrel 18 186 (4) 38 (2)

Ticagrelor 12 114 (2) 57 (4)

Warfarin, n (%) 5142 (1) 24 (1) 0.095

Stent diameter, mm 3.3±0.6 3.4±0.7 0.002

Stent length, mm 24±13 33±22 <0.001

Year, n (%) 0.359

2006 47 006 (9) 137 (8)

2007 55 308 (11) 167 (9)

2008 62 859 (12) 202 (11)

2009 65 883 (13) 223 (13)

2010 68 688 (13) 245 (14)

2011 72 024 (14) 250 (14)

2012 76 089 (14) 275 (16)

2013 77 502 (15) 263 (15)

Indication for PCI, n (%) <0.001

Stable angina 218 824 (42) 843 (48)

NSTEMI 192 051 (37) 618 (35)

STEMI 108 509 (21) 285 (16)

BMI indicates body mass index; CABG, coronary artery bypass graft; CAD, coronary artery disease; MI, myocardial infarction; LV, left ventricular; NSTEMI, non–ST-segment–elevation myocardial infarction; PCI, percutaneous coronary intervention; and STEMI, ST-segment–elevation myocardial infarction.

Table 1. Continued

VariableNo Perforation (n=525 359)

Perforation (n=1762) P Value

(Continued )

5 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

has limited the robustness of any conclusions and in particular multiple logistic regression analyses of the predictors of CP and predictors of mortality in those who experience perforation have not been performed before because of small sample sizes.

The incidence of CP during PCI in the current study is consistent with that reported in the previous smaller stud-ies and is midway between the lowest reported frequency of 0.12% and the upper reported frequency of 0.82%.10,11 Two summary studies have reported on the occurrence of CP. The 16 published CP studies involving 197 061 PCIs were sum-marized recently. In this largely descriptive article, the perfo-ration frequency was 0.43% although no additional analyses were presented.1 In the second summary analysis, Patel et al25 reported 419 perforations from 65 studies of 18 061 patients undergoing CTO PCI. In this series, the reported frequency of CP was 2.9% with a tamponade frequency of 0.2%. Although

this is a large series of CP, no further data were presented on the perforation subset of the main study.

In the current study, a subanalysis of the annualized data demonstrates not only a year-on-year increase in the crude numbers of perforation as the number of UK PCIs increase but also a nonsignificant trend for the percentage frequency to increase. Given the independent predictors of perforation identified in this study, it is perhaps not surprising in view of the aging population with increasingly complex non-CTO and CTO procedures that the frequency of perforation is unlikely to drop.26–28 Based on the current data, it would seem reasonable to predict that the frequency of CP might actually increase in coming years.

Several previous studies have examined the baseline and procedural factors associated with perforation although none have had the sample size to perform multiple logistic regres-sion analyses.4–20 The findings of these smaller studies have reported somewhat divergent findings although increasing age, use of rotational atherectomy, complex lesions, and CTO inter-vention all feature more commonly. The current study largely confirms the tentative findings of previous studies confirming women and the elderly to be at risk of CP and those cases in which atherectomy devices are utilized. Novel predictors of CP

Table 2. Procedural Variables by Coronary Perforation Status

VariableNo Perforation (n=525 359)

Perforation (n=1762) P Value

Bivalirudin, n (%) 16 562 (3) 47 (3) 0.267

Glycoprotein IIb/IIIa inhibitor, n (%)

116 206 (24) 289 (18) <0.001

Vessel attempted, n (%)

Vein graft 21 779 (4) 88 (5) 0.107

Left main 17 474 (3) 135 (8) <0.001

LAD 240 583 (48) 794 (46) 0.182

Circumflex 123 108 (24) 373 (22) 0.008

Right coronary 185 106 (37) 737 (43) <0.001

Postprocedural TIMI flow, n (%)

<0.001

0 16 325 (7) 127 (17)

1 2542 (1) 24 (3)

2 7883 (3) 42 (6)

3 222 526 (89) 545 (74)

Radial access, n (%) 243 622 (48) 811 (47) 0.502

Stent implanted, n (%) <0.001

No stent 49 965 (10) 403 (23)

Bare-metal stent 118 475 (23) 284 (17)

Drug-eluting stent 347 231 (67) 1031 (60)

Rotational atherectomy, n (%)

8047 (2) 103 (6) <0.001

Laser angioplasty, n (%) 890 (0.2) 18 (1) <0.001

Cutting balloon, n (%) 16 990 (3) 59 (4) 0.584

Chronic occlusion, n (%) 25 195 (5) 363 (21) <0.001

Surgical cover, n (%) 0.148

None or off-site 177 921 (35) 589 (37)

On-site 324 701 (65%) 997 (63)

LAD indicates left anterior descending artery; and TIMI, Thrombolysis in Myocardial Infarction.

Table 3. Significant Associations Between Covariates and Coronary Perforation

Variable Odds Ratio (95% CI) P Value

Age per y 1.03 (1.02–1.03) <0.001

Male sex 0.76 (0.67–0.87) <0.001

Hypercholesterolaemia 1.16 (1.01–1.33) 0.035

Diabetes mellitus 0.84 (0.71–0.98) 0.028

Previous CABG 1.44 (1.17–1.77) <0.001

Shock 0.60 (0.38–0.92) 0.021

Circulatory support 4.22 (3.22–5.54) <0.001

Left main stem PCI 1.54 (1.21–1.96) 0.001

Chronic occlusions 3.96 (3.28–4.78) <0.001

Stent

Bare-metal 0.37 (0.30–0.46) <0.001

Drug eluting 0.43 (0.37–0.51) <0.001

Rotational atherectomy 2.37 (1.80–3.11) <0.001

Cutting balloon 0.62 (0.43–0.89) 0.010

Side branch occlusion 4.07 (2.93–5.67) <0.001

Coronary dissection 3.31 (2.78–3.94) <0.001

NSTEMI indication 1.26 (1.07–1.47) 0.004

Potential predictors in the model: age, sex, smoking status, body mass index, family history of coronary artery disease, hypercholesterolaemia, hypertension, diabetes mellitus, previous myocardial infarction, previous stroke, peripheral vascular disease, valvular heart disease, renal disease, previous PCI, previous CABG, left ventricular function, cardiogenic shock, ventilator use, circulatory support, radial access, surgical cover, year, vessel of PCI, chronic occlusion, stent type used, rotational atherectomy, laser angioplasty, cutting balloon, side-branch occlusion, coronary dissection and diagnosis. CABG indicates coronary artery bypass graft; CI, confidence interval; NSTEMI, non–ST-segment–elevation myocardial infarction; and PCI, percutaneous coronary intervention.

6 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

in the current study include patients with previous CABG, left main stem intervention, and a reduced frequency in diabetics and in cases with cutting balloon use. The observation that both BMS and DES use were associated with a lower likelihood of perforation may be explained by several factors including guidewire perforation, and the fact that the use of covered stent is not recorded as a separate field in the database. In addition, the occurrence of a perforation with predilation may lead oper-ators to terminate the procedure after control of the perforation, without continuing to stent implantation. The observation that the use of a cutting balloon seemed to be protective against perforation is a novel finding and may support the hypothesis

that careful lesion preparation thus necessitating less aggres-sive postdilatation could reduce the incidence of perforation. The multicenter nature of the database also allowed an analysis of the influence of on-site versus off-site cardiac surgical sup-port, and no relationship was observed.

The incidence of major adverse events in patients with CP was high with a 13-fold increase in in-hospital MACE and a 5-fold increase in 30-day mortality. These data are a stark reminder that while CP during PCI is a relatively rare event, when it does occur the outcome remains extremely poor. In addition, we observed a significant increase in mortality at 30 days and 1 year in patients who had CP and patients who did not have CP. This may reflect an increase in undertaking PCI on higher risk patients with time. However, the magnitude of increase in mortality was higher in the CP group than in the non-perforation group that may also reflect the increased complexity of patients experiencing a perforation. Tamponade occurred in 14% of patients with CP, a finding that is testament to the skill of the operators necessitating rapid resuscitation, balloon occlu-sion, and placement of a pericardial drain to avoid its occur-rence. The rate of surgical repair was also extremely low (3%) in the current series and likely reflects the development of inter-ventional tools such as covered stents and embolization coils, as well as a wider appreciation of techniques such as distal fat and thrombus embolization to treat distal wire tip perforations.2,3 The size of the current analysis also enabled a multiple logistic regression analysis of the predictors of an adverse outcome in those patients who experienced a CP. The results are largely intuitive with increasing age, previous myocardial infarction, and renal disease, all strongly predictive of increased mortality. The observed increased mortality associated with stent use may be explained by the lack of a dedicated covered stent field in the database, ie, the use of a covered stent may be recorded as no stent used. This analysis also supports the anecdotal experience that if a CP does occur, concomitant periprocedural use of a gly-coprotein IIb/IIIa inhibitor is associated with worse outcomes, with an almost doubling of 30-day mortality.

Figure 3. Thirty-day mortality by perfora-tion status. PCI indicates percutaneous coronary intervention.

Table 4. Outcomes by Coronary Perforation Status

VariableNo Perforation

(n=525 359), n (%)Perforation

(n=1762) , n (%) P Value

In-hospital MACE 10 705 (2) 406 (26) <0.001

In-hospital mortality

5490 (1) 145 (8) <0.001

Mortality at 30 d 10 586 (2) 185 (11) <0.001

Mortality at 1 y 24 485 (5) 244 (15) <0.001

Mortality at 5 y 57 897 (25) 402 (47) <0.001

In-hospital bleed 3171 (0.6) 246 (14.0) <0.001

Reinfarction 3032 (0.6) 87 (5.6) <0.001

Emergency CABG 903 (0.2) 42 (2.7) <0.001

Stroke 833 (0.2) 20 (1.3) <0.001

Cardiac tamponade

248 (0.05) 222 (14.18) <0.001

Side branch occlusion

3658 (0.7) 51 (2.9) <0.001

Coronary dissection

19 001 (3.6) 225 (12.8) <0.001

CABG indicates coronary artery bypass graft; and MACE, major adverse cardiac events.

7 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

This analysis has several strengths. The BCIS data set includes >98% of all PCI procedures performed in the UK, which therefore reflects a national, real-world experience that includes high-risk patients encountered in daily interventional practice (who are often excluded from randomized controlled trials). Therefore, for the first time, the occurrence of perfora-tion is reported on a national basis. Our analysis of over ½ million PCI procedures represents that largest analysis to date by several orders of magnitude, analyzing temporal trends, predictors and outcomes for the first time derived from >1700 perforations. Such large national registry data with unselected enrollment are important for evaluation of low-frequency com-plications such as CP, particularly given that such low event rates would mean that single-center registries/RCTs would be grossly underpowered.

LimitationsIn addition, although the BCIS database does not specifically define perforation (leaving recording of this complication at the discretion of the operator), the observed incidence of perforation in our large series implies that reasonable clini-cal judgment was applied by UK operators. In addition, the BCIS database does not differentiate between CPs resulting from guidewire and perforations because of vessel rupture by balloon or stent inflation or provide data on perforation sever-ity. Therefore, we are unable to provide separate analyses on the predictors and outcomes of these subgroups of CP or a substratification by perforation severity is not possible within the constraints of the current data set. The independent asso-ciations between covariates and a perforation complication that we have reported cannot infer causality, but may relate to changes in the procedure/clinical status occurring as a con-sequence of the perforation. For example, the use of circula-tory support is associated with increased odds of perforation 4-fold, but this may not be a causal but merely reflect that post perforation, patients are more likely to require circulatory support because of significant hemodynamic compromise.

Similarly, side-branch occlusion may be associated with the use of a covered stent. Also, as the BCIS database only records emergency CABG as an outcome after CP. As a result, we are unable to provide data on the use of covered stents, pericardial drains, or embolization coils. Finally, census data were only available at specific time points with time to census data not available in the data extract that was used for this analysis. It has, however, been shown that for relatively short follow-up

Table 6. Predictors of 30-Day Mortality in Patients Who Experienced a Coronary Perforation

Variable Adjusted Odds Ratio (95% CI) P Value

Age per y 1.05 (1.03–1.08) <0.001

Diabetes mellitus 1.78 (1.02–3.11) 0.043

Previous myocardial infarction

2.08 (1.25–3.46) 0.005

Renal disease 4.03 (1.76–9.26) 0.001

Ventilatory support 8.83 (1.69–46.15) 0.010

Circulatory support 6.58 (3.72–11.63) <0.001

Glycoprotein IIb/IIIa inhibitor use

2.02 (1.19–3.44) 0.010

Postprocedure TIMI flow 3 0.41 (0.23–0.75) 0.004

Stent type

Bare metal 3.70 (1.74–7.88) 0.001

Drug eluting 2.27 (1.14–4.51) 0.020

Potential predictors in the model: age, sex, smoking status, body mass index, family history of coronary artery disease, hypercholesterolaemia, hypertension, diabetes mellitus, previous myocardial infarction, previous stroke, peripheral vascular disease, valvular heart disease, renal disease, previous PCI, previous coronary artery bypass graft, left ventricular function, cardiogenic shock, ventilator use, circulatory support, antiplatelet therapy, warfarin, glycoprotein IIb/IIIa inhibitor, bivalirudin use, radial access, surgical cover, year, vessel of PCI, chronic occlusion, postprocedure TIMI flow, stent type used, rotational atherectomy, laser angioplasty, cutting balloon, side-branch occlusion, and coronary dissection and diagnosis. CI indicates confidence interval; PCI, percutaneous coronary intervention; and TIMI, Thrombolysis in Myocardial Infarction.

Table 5. Adjusted Odds of Adverse Outcomes by Coronary Perforation Status

Adverse OutcomeMultiple Logistic Regression Adjusted

Odds Ratio (95% CI) P ValueInverse Probability Weighting by

Propensity Score Odds Ratio (95% CI) P Value

In-hospital MACE 13.20 (11.13–15.65) <0.001 20.22 (15.43–26.49) <0.001

30-d mortality 4.86 (3.84–6.15) <0.001 6.54 (4.40–9.71) <0.001

1-y mortality 2.54 (2.08–3.09) <0.001 4.01 (2.87–5.61) <0.001

5-y mortality 1.37 (1.04–1.80) 0.025 2.93 (2.12–4.06) <0.001

In-hospital reinfarction 4.61 (3.46–6.14) <0.001 11.41 (7.16–18.19) <0.001

In-hospital emergency CABG 5.34 (3.35–8.51) <0.001 11.65 (5.24–25.91) <0.001

In-hospital cardiac tamponade 220.06 (169.24–286.14) <0.001 513.14 (342.83–768.05) <0.001

In-hospital stroke 4.30 (2.42–7.64) <0.001 6.75 (2.48–18.34) <0.001

In-hospital bleed 20.86 (17.21–25.28) <0.001 30.87 (22.81–41.77) <0.001

Adjusted for age, sex, smoking status, body mass index, family history of coronary artery disease, hypercholesterolaemia, hypertension, diabetes mellitus, previous myocardial infarction, previous stroke, peripheral vascular disease, valvular heart disease, renal disease, previous PCI, previous CABG, left ventricular function, cardiogenic shock, ventilator use, circulatory support, antiplatelet therapy, warfarin, glycoprotein IIb/IIIa inhibitor, bivalirudin use, radial access, surgical cover, year, vessel of PCI, chronic occlusion, postprocedure TIMI (Thrombolysis in Myocardial Infarction) flow, stent type used, rotational atherectomy, laser angioplasty, cutting balloon, side-branch occlusion, coronary dissection, and diagnosis. CABG indicates coronary artery bypass graft; CI, confidence interval; MACE, major adverse cardiac events; and PCI, percutaneous coronary intervention.

8 Kinnaird et al Coronary Perforation From an Analysis of 527 121 Cases

times (up to 5 years), the performance of logistic regression models is similar to survival analysis models.29

ConclusionsUsing data derived from a national PCI database, CP occurred as a PCI procedural complication in 0.33% of cases. Indepen-dent predictors of CP included age, female sex, CTO interven-tion, and atherectomy. CP was strongly associated with poor outcomes.

DisclosuresNone.

References 1. Shimony A, Joseph L, Mottillo S, Eisenberg MJ. Coronary artery per-

foration during percutaneous coronary intervention: a systematic review and meta-analysis. Can J Cardiol. 2011;27:843–850. doi: 10.1016/j.cjca.2011.04.014.

2. Tarar MN, Christakopoulos GE, Brilakis ES. Successful management of a distal vessel perforation through a single 8-French guide catheter: combining balloon inflation for bleeding control with coil embolization. J Invasive Cardiol. 2013;25:486–491.

3. Chen S, Lotan C, Jaffe R, Rubinshtein R, Ben-Assa E, Roguin A, Varshitzsky B, Danenberg HD. Pericardial covered stent for coronary perforations. Catheter Cardiovasc Interv. 2015;86:400–404. doi: 10.1002/ccd.26011.

4. Hendry C, Fraser D, Eichhofer J, Mamas MA, Fath-Ordoubadi F, El-Omar M, Williams P. Coronary perforation in the drug-eluting stent era: incidence, risk factors, management and outcome: the UK experience. EuroIntervention. 2012;8:79–86. doi: 10.4244/EIJV8I1A13.

5. Al-Lamee R, Ielasi A, Latib A, Godino C, Ferraro M, Mussardo M, Arioli F, Carlino M, Montorfano M, Chieffo A, Colombo A. Incidence, predic-tors, management, immediate and long-term outcomes following grade III coronary perforation. JACC Cardiovasc Interv. 2011;4:87–95. doi: 10.1016/j.jcin.2010.08.026.

6. Romaguera R, Sardi G, Laynez-Carnicero A, Ben-Dor I, Maluenda G, Wakabayashi K, Hauville C, Torguson R, Xue Z, Kent KM, Satler LF, Suddath WO, Pichard AD, Lindsay J, Waksman R. Outcomes of coro-nary arterial perforations during percutaneous coronary intervention with bivalirudin anticoagulation. Am J Cardiol. 2011;108:932–935. doi: 10.1016/j.amjcard.2011.05.025.

7. Kiernan TJ, Yan BP, Ruggiero N, Eisenberg JD, Bernal J, Cubeddu RJ, Witzke C, Don C, Cruz-Gonzalez I, Rosenfield K, Pomersantev E, Palacios I. Coronary artery perforations in the contemporary interventional era. J Interv Cardiol. 2009;22:350–353. doi: 10.1111/j.1540-8183.2009.00469.x.

8. Shimony A, Zahger D, Van Straten M, Shalev A, Gilutz H, Ilia R, Cafri C. Incidence, risk factors, management and outcomes of coronary artery perforation during percutaneous coronary intervention. Am J Cardiol. 2009;104:1674–1677. doi: 10.1016/j.amjcard.2009.07.048.

9. Fasseas P, Orford JL, Panetta CJ, Bell MR, Denktas AE, Lennon RJ, Holmes DR, Berger PB. Incidence, correlates, management, and clini-cal outcome of coronary perforation: analysis of 16,298 procedures. Am Heart J. 2004;147:140–145.

10. Gunning MG, Williams IL, Jewitt DE, Shah AM, Wainwright RJ, Thomas MR. Coronary artery perforation during percutaneous intervention: inci-dence and outcome. Heart. 2002;88:495–498.

11. Fejka M, Dixon SR, Safian RD, O’Neill WW, Grines CL, Finta B, Marcovitz PA, Kahn JK. Diagnosis, management, and clinical outcome of cardiac tamponade complicating percutaneous coronary intervention. Am J Cardiol. 2002;90:1183–1186.

12. Ellis SG, Ajluni S, Arnold AZ, Popma JJ, Bittl JA, Eigler NL, Cowley MJ, Raymond RE, Safian RD, Whitlow PL. Increased coronary perforation in the new device era. Incidence, classification, management, and outcome. Circulation. 1994;90:2725–2730.

13. Stathopoulos I, Kossidas K, Panagopoulos G, Garratt K. Cardiac tampon-ade complicating coronary perforation during angioplasty: short-term out-comes and long-term survival. J Invasive Cardiol. 2013;25:486–491.

14. Ajluni SC, Glazier S, Blankenship L, O’Neill WW, Safian RD. Perforations after percutaneous coronary interventions: clinical, angiographic, and therapeutic observations. Cathet Cardiovasc Diagn. 1994;32:206–212.

15. Dippel EJ, Kereiakes DJ, Tramuta DA, Broderick TM, Shimshak TM, Roth EM, Hattemer CR, Runyon JP, Whang DD, Schneider JF, Abbottsmith

CW. Coronary perforation during percutaneous coronary intervention in the era of abciximab platelet glycoprotein IIb/IIIa blockade: an algorithm for percutaneous management. Catheter Cardiovasc Interv. 2001;52:279–286. doi: 10.1002/ccd.1065.

16. Friedrich SP, Berman AD, Baim DS, Diver DJ. Myocardial perforation in the cardiac catheterization laboratory: incidence, presentation, diagnosis, and management. Cathet Cardiovasc Diagn. 1994;32:99–107.

17. Javaid A, Buch AN, Satler LF, Kent KM, Suddath WO, Lindsay J Jr, Pichard AD, Waksman R. Management and outcomes of coronary artery perforation during percutaneous coronary intervention. Am J Cardiol. 2006;98:911–914. doi: 10.1016/j.amjcard.2006.04.032.

18. Gruberg L, Pinnow E, Flood R, Bonnet Y, Tebeica M, Waksman R, Satler LF, Pichard AD, Kent KM, Leon MB, Lindsay J Jr. Incidence, manage-ment, and outcome of coronary artery perforation during percutaneous coronary intervention. Am J Cardiol. 2000;86:680–682, A8.

19. Ben-Gal Y, Weisz G, Collins MB, Genereux P, Dangas GD, Teirstein PS, Singh VP, Rabbani LE, Kodali SK, Sherman W, Leon MB, Moses JW. Dual catheter technique for the treatment of severe coronary artery per-forations. Catheter Cardiovasc Interv. 2010;75:708–712. doi: 10.1002/ccd.22331.

20. Bauer T, Boeder N, Nef HM, Möllmann H, Hochadel M, Marco J, Weidinger F, Zeymer U, Gitt AK, Hamm CW. Fate of patients with coro-nary perforation complicating percutaneous coronary intervention (from the Euro Heart Survey Percutaneous Coronary Intervention Registry). Am J Cardiol. 2015;116:1363–1367. doi: 10.1016/j.amjcard.2015.07.056.

21. Ludman PF; British Cardiovascular Intervention Society. British Cardiovascular Intervention Society Registry for audit and quality as-sessment of percutaneous coronary interventions in the United Kingdom. Heart. 2011;97:1293–1297. doi: 10.1136/heartjnl-2011-300299.

22. Mamas MA, Anderson SG, Carr M, Ratib K, Buchan I, Sirker A, Fraser DG, Hildick-Smith D, de Belder M, Ludman PF, Nolan J; British Cardiovascular Intervention Society; National Institute for Cardiovascular Outcomes Research. Baseline bleeding risk and arterial access site prac-tice in relation to procedural outcomes after percutaneous coronary in-tervention. J Am Coll Cardiol. 2014;64:1554–1564. doi: 10.1016/j.jacc.2014.05.075.

23. Mamas MA, Anderson SG, O’Kane PD, Keavney B, Nolan J, Oldroyd KG, Perera D, Redwood S, Zaman A, Ludman PF, de Belder MA; British Cardiovascular Intervention Society and the National Institute for Cardiovascular Outcomes Research. Impact of left ventricular function in relation to procedural outcomes following percutaneous coronary in-tervention: insights from the British Cardiovascular Intervention Society. Eur Heart J. 2014;35:3004–3012a. doi: 10.1093/eurheartj/ehu303.

24. Ratib K, Mamas MA, Routledge HC, Ludman PF, Fraser D, Nolan J. Influence of access site choice on incidence of neurologic complications after percutaneous coronary intervention. Am Heart J. 2013;165:317–324. doi: 10.1016/j.ahj.2012.10.015.

25. Patel VG, Brayton KM, Tamayo A, Mogabgab O, Michael TT, Lo N, Alomar M, Shorrock D, Cipher D, Abdullah S, Banerjee S, Brilakis ES. Angiographic success and procedural complications in patients undergo-ing percutaneous coronary chronic total occlusion interventions: a weight-ed meta-analysis of 18,061 patients from 65 studies. JACC Cardiovasc Interv. 2013;6:128–136. doi: 10.1016/j.jcin.2012.10.011.

26. Kragholm K, Halim SA, Yang Q, Schulte PJ, Hochman JS, Melloni C, Mahaffey KW, Moliterno DJ, Harrington RA, White HD, Armstrong PW, Ohman EM, Van de Werf F, Tricoci P, Alexander JH, Giugliano RP, Newby LK. Sex-stratified trends in enrollment, patient charac-teristics, treatment, and outcomes among non-ST-segment elevation acute coronary syndrome patients: insights from clinical trials over 17 years. Circ Cardiovasc Qual Outcomes. 2015;8:357–367. doi: 10.1161/CIRCOUTCOMES.114.001615.

27. Rosengren A, Wallentin L, Simoons M, Gitt AK, Behar S, Battler A, Hasdai D. Age, clinical presentation, and outcome of acute coronary syn-dromes in the Euroheart acute coronary syndrome survey. Eur Heart J. 2006;27:789–795. doi: 10.1093/eurheartj/ehi774.

28. George S, Cockburn J, Clayton TC, Ludman P, Cotton J, Spratt J, Redwood S, de Belder M, de Belder A, Hill J, Hoye A, Palmer N, Rathore S, Gershlick A, Di Mario C, Hildick-Smith D; British Cardiovascular Intervention Society; National Institute for Cardiovascular Outcomes Research. Long-term follow-up of elective chronic total coronary occlu-sion angioplasty: analysis from the UK Central Cardiac Audit Database. J Am Coll Cardiol. 2014;64:235–243. doi: 10.1016/j.jacc.2014.04.040.

29. Annesi I, Moreau T, Lellouch J. Efficiency of the logistic regression and Cox proportional hazards models in longitudinal studies. Stat Med. 1989;8:1515–1521.

Cardiovascular Outcomes Researchon behalf of the British Cardiovascular Intervention Society and the National Institute for

Ludman, Mark deBelder, Richard Anderson and Mamas A. MamasTim Kinnaird, Chun Shing Kwok, Evangelos Kontopantelis, Nicholas Ossei-Gerning, Peter

121 Cases From the British Cardiovascular Intervention Society Database527Coronary Intervention in the United Kingdom Between 2006 and 2013: An Analysis of

Incidence, Determinants, and Outcomes of Coronary Perforation During Percutaneous

Print ISSN: 1941-7640. Online ISSN: 1941-7632 Copyright © 2016 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleCirculation: Cardiovascular Interventions

doi: 10.1161/CIRCINTERVENTIONS.115.0034492016;9:Circ Cardiovasc Interv. 

http://circinterventions.ahajournals.org/content/9/8/e003449World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circinterventions.ahajournals.org/content/suppl/2016/08/02/CIRCINTERVENTIONS.115.003449.DC1.htmlData Supplement (unedited) at:

  http://circinterventions.ahajournals.org//subscriptions/

is online at: Circulation: Cardiovascular Interventions Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Answer

Permissions and Rights Question andunder Services. Further information about this process is available in thepermission is being requested is located, click Request Permissions in the middle column of the Web pageClearance Center, not the Editorial Office. Once the online version of the published article for which

can be obtained via RightsLink, a service of the CopyrightCirculation: Cardiovascular Interventionsin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

Supplemental Material

Supplemental Table 1: 30 day and 1-year mortality over the study period in those patients with and without coronary perforation.

Variable 2006 2007 2008 2009 2010 2011 2012 2013 P-value

30-day mortality for patients with perforation

12/136 (8.82, 4.05-13.59%)

19/165 (11.52, 6.65-16.39%)

16/201 (7.96, 4.22-11.70%)

15/218 (6.88, 3.52-10.24%)

28/244 (11.48, 7.48-15.48%)

21/243 (8.64, 5.11-12.17%)

42/271 (15.5, 11.19-19.81%)

32/258 (12.4, 8.38-16.42%)

0.049

30-day mortality for patients without perforation

663/46083 (1.44, 1.33-1.55%)

829/54362 (1.52, 1.42-1.62%)

1015/61881 (1.64, 1.54-1.74%)

1230/64630 (1.90, 1.79-2.01%)

1395/67835 (2.06, 1.95-2.17%)

1676/70699 (2.37, 2.26-2.48%)

1901/74655 (2.55, 2.44-2.66%)

1877/74158 (2.53, 2.42-2.64%)

<0.001

1-year mortality for patients with perforation

16/136 (11.76, 6.35-17.17%)

26/164 (15.85, 10.26-21.44%)

19/200 (9.50, 5.44-13.56%)

25/218 (11.47, 7.24-15.70%)

41/244 (16.80, 12.11-21.49%)

24/243 (9.88, 6.13-13.63%)

52/269 (19.33, 14.61-24.05%)

41/148 (27.70, 20.49-34.91%)

<0.001

1-year mortality for patients without perforation

1787/46070 (3.88, 3.70-4.06%)

2124/54257 (3.91, 3.75-4.07%)

2574/61695 (4.17, 4.01-4.33%)

3022/64586 (4.68, 4.52-4.84%)

3176/67732 (4.69, 4.53-4.85%)

3766/70619 (5.33, 5.16-5.50%)

4,112/74114 (5.55, 5.39-5.71%)

3,924/43181 (9.09, 8.82-9.36%)

<0.001

2

Supplementary Table 2: Exploration of how significant predictors of coronary perforation complication changes over time

Variable 2006 2007 2008 2009 2010 2011 2012 2013 P-value

Age (95%CI) 64.08 (63.98-64.18)

64.19 (64.09-64.28)

64.47 (64.38-64.56)

64.92 (64.84-65.01)

64.91 (64.82-65.00)

65.24 (65.16-65.33)

65.07 (64.98-65.16)

65.03 (64.94-65.11)

<0.001

Male gender (95%CI)

34842/47143 (73.91, 73.51-74.31%)

41062/55475 (74.02, 73.66-74.38%)

46589/63061 (73.88, 73.54-74.22%)

48937/66106 (74.03, 73.70-74.36%)

51015/68933 (74.01, 73.68-74.34%)

53443/72274 (73.94, 73.62-74.26%)

56374/76364 (73.82, 73.51-74.13%)

57615/77765 (74.09, 73.78-74.40%)

0.958

Hypercholestolaemia (95%CI)

20046/42883 (46.75, 46.28-47.22%)

29965/53512 (56.00, 55.58-56.42%)

33090/61168 (54.10, 53.71-54.49%)

38345/64879 (59.10, 58.72-59.48%)

38946/67631 (57.59, 57.22-57.96%)

38533/70281 (54.83, 54.46-55.20%)

40085/74395 (53.88, 53.52-54.24%)

40467/74925 (54.01, 53.65-54.37%)

<0.001

Diabetes (95%CI) 6910/39744 (17.39, 17.02-17.76%)

9619/52904 (18.18, 17.85-18.51%)

11310/60954 (18.55, 18.24-18.86%)

12001/64727 (18.54, 17.24-18.84%)

12815/66889 (19.16, 18.86-19.46%)

13664/70078 (19.50, 19.21-19.79%)

15013/72533 (20.70, 20.41-20.99%)

15919/74203 (21.45, 21.15-21.75%)

<0.001

Previous CABG (95%CI)

2633/28191 (9.34, 9.00-9.68%)

3406/38336 (8.88, 8.60-9.16%)

3904/46983 (8.31, 8.06-8.56%)

4053/48308 (8.39, 8.14-8.64%)

4316/53251 (8.11, 7.88-8.34%)

4349/56122 (7.75, 7.53-7.97%)

4578/57875 (7.91, 7.69-8.13%)

4512/57437 (7.86, 7.64-8.08)

<0.001

Shock (95%CI) 468/40046 (1.17, 1.06-1.28%)

748/50998 (1.47, 1.37-1.57%)

903/58847 (1.53, 1,43-1.63%)

1149/61719 (1.86, 1.75-1.97%)

1453/64722 (2.24, 2.13-2.35%)

1722/67941 (2.53, 2.41-2.65%)

1850/72490 (2.55, 2.44-2.66)

2310/73761 (3.13, 3.00-3.26%)

<0.001

Circulatory support (95%CI)

634/39744 (1.60, 1.48-1.72%)

955/50927 (1.88, 1.76-2.00%)

1080/59227 (1.82, 1.71-1.93%)

1259/62393 (2.02, 1.91-2.13%)

1435/65585 (2.19, 2.08-2.30%)

1609/70244 (2.29, 2.18-2.40%)

1618/74424 (2.17, 2.07-2.27%)

1621/74893 (2.16, 2.06-2.26%)

<0.001

Left main stem PCI (95%CI)

1221/42268 (2.89, 2.73-3.05%)

1452/53043 (2.74, 2.60-2.88%)

1681/60781 (2.77, 2.64-2.90%)

2053/64405 (3.19, 3.05-3.33%)

2124/66448 (3.20, 3.07-3.33%)

2508/70537 (3.56, 3.42-3.70%)

2863/74548 (3.84, 3.70-3.98%)

3707/75027 (4.94, 4.78-5.10%)

<0.001

Chronic occlusions (95%CI)

2421/42026 (5.76, 5.54-5.98%)

3044/53949 (5.64, 5.45-5.83%)

3439/61732 (5.57, 5.39-5.75%)

3461/63941 (5.41, 5.23-5.59%)

3306/66588 (4.96, 4.80-5.12%)

3353/69965 (4.79, 4.63-4.95%)

3309/74161 (4.46, 4.31-4.61%)

3225/75637 (4.26, 4.12-4.40%)

<0.001

Rotational atherectomy (95%CI)

326/42424 (0.77, 0.69-0.85%)

474/52657 (0.90, 0.82-0.98%)

780/60031 (1.30, 1.21-1.39%)

879/63352 (1.39, 1.30-1.48%)

1055/65529 (1.61, 1.51-1.71%)

1275/69350 (1.84, 1.74-1.94%)

1621/73280 (2.21, 2.10-2.32%)

1740/74529 (2.33, 2.22-2.44%)

<0.001

Cutting balloon (95%CI)

338/42424 (0.80, 0.72-0.88%)

443/52657 (0.84, 0.76-0.92%)

2591/60031 (4.32, 4.16-4.48%)

2695/63352 (4.25, 4.09-4.41%)

2489/65529 (3.80, 3.65-3.95%)

2619/69350 (3.78, 3.64-3.92%)

2803/73280 (3.83, 3.69-3.97%)

3071/74529 (4.12, 3.98-4.26%)

<0.001

NSTEMI indication (95%CI)

16141/45528 (35.45, 35.01-35.89%)

21275/54427 (39.09, 38.68-39.50%)

22784/62134 (36.67, 36.29-37.05%)

24595/65661 (37.46, 37.09-37.83%)

25424/68486 (37.12, 36.76-37.48%)

26243/71837 (36.53, 36.18-36.88%)

27365/75838 (36.08, 35.74-36.42%)

28842/77219 (37.35, 37.01-37.69%)

<0.001

95% Confidence intervals calculated from http:www.mccallum-layton.co.uk/tools/statistic-calculators/confidence-interval-for-proportions-calculator/

3

Supplemental Table 3: Propensity score matching analysis on 10 imputed datasets, reporting average treatment effects (ATE).

Analysis of propensity score matching with ATE

N Coefficient 95% CI p-value

In-hospital MACE 341,143 0.2054 0.1507 0.2601 <0.001

30 day mortality 346,808 0.0654 0.0330 0.0979 <0.001

1 year mortality 294,424 0.0596 0.0202 0.0990 0.003

5 year mortality 97,742 0.0503 -0.0320 0.1326 0.229

In-hospital re-infarction 352,822 0.0484 0.0247 0.0721 <0.001

In-hospital emergency CABG 352,822 0.0085 -0.0014 0.0185 0.091

In-hospital cardiac tamponade 352,822 0.1248 0.0910 0.1587 <0.001

In-hospital stroke 341,143 0.0056 -0.0039 0.0150 0.245

In-hospital bleed 352,822 0.1240 0.0894 0.1584 <0.001

4

SupplementaryTable4:Balancediagnosticsforpropensitymodel

Propensitymodel

Group Mean(SD) Median(Q1,Q3)

Perforation vs. noperforation

Case 0.9969(0.0037)

0.9978(0.9966,0.9985)

Control 0.9969(0.0037)

0.9978(0.9966,0.9985)

Abs(Case-Control) 9.1*10-6(0.00028)

2.4*10-6(9.5*10-7,5.5*10-6)

5

Supplementary Table 5: Adjusted odds of adverse outcomes by coronary perforation status with and without adjustments for center volume

Adverse outcome Multiple logistic regression adjusted odds ratio without center volume variable (95% CI)

p-value Multiple logistic regression adjusted odds ratio with center volume variable (95% CI)

p-value

In-hospital MACCE 16.14 (13.73-18.98) <0.001 16.49 (14.02-19.39) <0.001 30-day mortality 5.54 (4.40-6.97) <0.001 5.57 (4.42-7.03) <0.001 1-year mortality 2.72 (2.24-3.30) <0.001 2.73 (2.24-3.31) <0.001 5-year mortality 1.42 (1.08-1.86) 0.011 1.42 (1.09-1.86) 0.010 In-hospital re-infarction

6.84 (5.26-8.89) <0.001 7.30 (5.61-9.50) <0.001

In-hospital emergency CABG

6.87 (4.38-10.77) <0.001 6.91 (4.42-10.81) <0.001

In-hospital cardiac tamponade

234.13 (181.36-302.27) <0.001 239.90 (186.14-309.19) <0.001

In-hospital stroke 5.32 (3.01-9.38) <0.001 5.48 (3.10-9.67) <0.001 In-hospital bleed 22.98 (19.04-27.73) <0.001 24.59 (20.34-29.73) <0.001 Adjusted for age, gender, smoking status, body mass index, family history of coronary artery disease, hypercholesterolaemia, hypertension, diabetes, previous myocardial infarction, previous stroke, peripheral vascular disease, valvular heart disease, renal disease, previous PCI, previous CABG, left ventricular function, cardiogenic shock, ventilator use, circulatory support, antiplatelet therapy, warfarin, glycoprotein IIb/IIIa inhibitor, bivalirudin use, radial access, surgical cover, year, vessel of PCI, chronic occlusion, rotational atherectomy, laser angioplasty, cutting balloon and diagnosis.

6