defining the optimal degree of heparin anticoagulation

1941-7632American Heart Association. All rights reserved. Print ISSN: 1941-7640. Online ISSN: 2010 Copyright ©

Avenue, Dallas, TX 72514Circulation: Cardiovascular Interventions is published by the American Heart Association. 7272 Greenville

DOI: 10.1161/CIRCINTERVENTIONS.110.957381 2010;3;593-601; originally published online November 9, 2010;Circ Cardiovasc Interv

Share and P. Michael GrossmanSmith, Herbert D. Aronow, Elias H. Kassab, Michael F. Knox, Mauro Moscucci, David

Christos Kasapis, Hitinder S. Gurm, Stanley J. Chetcuti, Khan Munir, Ann Luciano, DeanInterventions : Insight From a Large, Regional, Multicenter Registry

Defining the Optimal Degree of Heparin Anticoagulation for Peripheral Vascular

http://circinterventions.ahajournals.org/content/3/6/593.full

on the World Wide Web at: The online version of this article, along with updated information and services, is located

.110.957381.DC1.htmlhttp://circinterventions.ahajournals.org/content/suppl/2010/12/15/CIRCINTERVENTIONS

Data Supplement (unedited) at:

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

[email protected]. E-mail:Health, 351 West Camden Street, Baltimore, MD 21201-2436. Phone: 410-528-4050. Fax: Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters Kluwer

http://circinterventions.ahajournals.org/site/subscriptions/Subscriptions: Information about subscribing to Circulation: Cardiovascular Interventions is online at

by guest on January 20, 2011circinterventions.ahajournals.orgDownloaded from

http://circinterventions.ahajournals.org/content/3/6/593.full

http://circinterventions.ahajournals.org/content/suppl/2010/12/15/CIRCINTERVENTIONS.110.957381.DC1.html

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

mailto:[email protected]

http://www.lww.com/reprints

http://circinterventions.ahajournals.org/

Defining the Optimal Degree of Heparin Anticoagulation forPeripheral Vascular Interventions

Insight From a Large, Regional, Multicenter Registry

Christos Kasapis, MD; Hitinder S. Gurm, MD, FACC; Stanley J. Chetcuti, MD, FACC;Khan Munir, PhD, MBA; Ann Luciano, RN, ACNP-BC; Dean Smith, PhD;

Herbert D. Aronow, MD, MPH, FACC; Elias H. Kassab, MD, FACC; Michael F. Knox, MD, FACR;Mauro Moscucci, MD, MBA, FACC; David Share, MD, MPH; P. Michael Grossman, MD, FACC

Background—The optimal degree of heparin anticoagulation for peripheral vascular interventions (PVIs) has not beendefined. We sought to correlate total heparin dose and peak procedural activated clotting time (ACT) withpostprocedural outcomes in patients undergoing PVI.

Methods and Results—We studied 4743 patients who received heparin during PVIs in a regional, multicenter registry.From those, 1246 had recorded peak procedural ACT with the same point-of-care device. Periprocedural and in-hospitaloutcomes were compared between patients who received a total heparin dose �60 U/kg (n�2161) and �60 U/kg(n�2582). Similarly, outcomes were evaluated between groups with a peak procedural ACT �250 seconds (n�855)and �250 seconds (n�391). Technical and procedural success as well as intraprocedural thrombotic events did notdiffer between groups. Patients with heparin dose �60 U/kg had a higher rate of postprocedural hemoglobin drop �3g/dL (7.09% versus 5.09%, respectively, P�0.004) and a higher transfusion rate compared with those with heparin dose�60 U/kg (4.92% versus 3.15%, respectively, P�0.002). In multivariate analysis, independent predictors of bleedingrequiring transfusion were total heparin dose �60 U/kg, ACT �250 seconds, female sex, age �70 years, prior anemia,prior heart failure, low creatinine clearance, hybrid vascular surgery, rest pain, and below-knee intervention. Inpropensity-matched, risk-adjusted models and after hierarchical modeling, total heparin dose �60 U/kg and ACT �250seconds remained strong predictors of post-PVI drop in hemoglobin �3 g/dL or transfusion.

Conclusions—During PVI, higher total heparin dose (�60 U/kg) and peak ACT �250 seconds were predictors ofpostprocedural transfusion. The high technical and procedural success in all groups suggests that use of weight-basedheparin dosing with a target ACT �250 seconds in PVI may minimize the bleeding risk without compromisingprocedural success or increasing thromboembolic complications. (Circ Cardiovasc Interv. 2010;3:593-601.)

Key Words: peripheral vascular disease � heparin � blood coagulation tests � peripheral interventions

Unfractionated heparin (UFH) is the most commonly usedantithrombotic agent in percutaneous peripheral vascu-

lar interventions (PVIs). The availability of a rapid “point ofcare” assay (the activated clotting time [ACT]) led to theassumption that an optimal level of anticoagulation could beidentified in patients undergoing percutaneous coronary in-terventions (PCIs).1 However, the optimal heparin anticoag-ulation in PVI is unknown, with current recommendationsempirically based on data from the coronary literature and theTASC II guidelines recommending heparinization to achievea target ACT between 200 and 250 seconds for PVI.2 Even in

the case of coronary interventions, there is a wide range of“optimal” ACT levels from as low as 200 to 250 seconds toas high as 350 to 375 seconds suggested by differentstudies,3–6 whereas more contemporary data favor a down-ward shift to the target ACT.4–6

Clinical Perspective on p 601

Adequate dosing with UFH is essential to suppress throm-bin generation associated with balloon-induced vascular in-jury.7 In contrast, higher doses of heparin are associated withbleeding complications. It is well recognized that bleeding

Received April 21, 2010; accepted September 9, 2010.From the Division of Cardiovascular Medicine (C.K., H.S.G., S.J.C., K.M., A.L., D.S., P.M.G.), University of Michigan Health System, Ann Arbor,

Mich; the Department of Cardiology (H.S.G., S.J.C., P.M.G.), Veterans Administration Health System Ann Arbor, Ann Arbor, Mich; Michigan Heart andVascular Institute (H.D.A.), St Joseph Mercy Hospital, Ann Arbor, Mich; Dearborn Cardiology Associates (E.H.K.), Oakwood Hospital and MedicalCenter, Dearborn, Mich; the Department of Radiology (M.F.K.), Spectrum Health Hospitals, Grand Rapids, Mich; the Department of CardiovascularMedicine (M.M.), University of Miami, Miami, Fla; and Healthcare Quality (D.S.), Blue Cross Blue Shield of Michigan, Ann Arbor, MI.

The online-only Data Supplement is available at http://circinterventions.ahajournals.org/cgi/content/full/CIRCINTERVENTIONS.110.957381/DC1.Correspondence to P. Michael Grossman, MD, University of Michigan Health System Cardiovascular Center, 1500 E Medical Center Dr, SPC 5869,

Ann Arbor, MI 48109-5869. E-mail [email protected]© 2010 American Heart Association, Inc.

Circ Cardiovasc Interv is available at http://circinterventions.ahajournals.org DOI: 10.1161/CIRCINTERVENTIONS.110.957381

593 by guest on January 20, 2011circinterventions.ahajournals.orgDownloaded from

http://circinterventions.ahajournals.org/cgi/content/full/CIRCINTERVENTIONS.110.957381/DC1


complications, especially those requiring transfusion, are animportant predictor of adverse outcome in patients withcoronary artery disease who undergo surgical or percutaneousrevascularization.8–13 The purpose of this study was toevaluate the relationship between the degree of heparinanticoagulation and clinical efficacy as well as bleeding riskin patients undergoing PVI in a regional, multicenter, multi-disciplinary registry. We also investigated independent riskfactors of periprocedural transfusion because knowledge ofthese factors would potentially guide a more “conservative”anticoagulation target or an alternative anticoagulant strategyin the subgroups of patients with an exaggerated bleedinghazard.

MethodsBlue Cross Blue Shield of Michigan PVI RegistryDetails of the construct, data collection, and data quality assurancefor the Blue Cross Blue Shield of Michigan Cardiovascular Consor-tium Peripheral Vascular Intervention Quality Improvement Initia-tive (BMC2 PVI), have been described previously.14 In brief, theBMC2 PVI registry is a prospective, multicenter registry of patientsundergoing PVI at the participating hospitals (supplemental mate-rial). Endovascular carotid interventions and aortic endografts arenot included in the registry because procedural and clinical param-eters differ in these patients. A data form is compiled for eachprocedure. Data quality and the inclusion of consecutive proceduresare contractually required and assured by ad hoc queries, randomchart review, and a series of diagnostic routines. The data formcomprises demographic, clinical and procedural variables, baselineand adjunctive pharmacotherapies, and in-hospital outcome data. Alist of standard definitions using the American College of CardiologyData Standards Committee has been used as reference.

Data Quality ControlAll data undergo a 3-step validation process, including manualreview for completeness and validity, review of rejected data formsduring the import process, and review of forms that fail diagnosticinquiries. All sites are audited twice yearly by a nurse investigatorfrom the coordinating center. In addition to examining cases inwhich an adverse event was recorded, a random 5% sample of casesis audited for completeness and accuracy. The compliance andaccuracy of data in the audits is �95%. Data that are found to be

inaccurate or incomplete are corrected and feedback is provided tothe responsible centers.

Study Patients and DefinitionsThe study cohort for this analysis consisted of 6020 patientsundergoing PVI between January 2001 and December 2007. Patientswho received periprocedural GP IIb/IIIa antagonists, thrombolyticagents, or who were already on heparin before the procedure, wereexcluded from the analysis. We also excluded patients whose valuesfor heparin dose or peak ACT were missing because measurement ofpeak intraprocedural ACT was not standard practice in a majority oflabs during the study period. In addition, we excluded patients withrecorded ACT �200 seconds because they would not representclinically acceptable targets, based on the current practice supportedby the coronary literature and the TASC II guidelines, and thesepatients were more likely to be rebolused with additional heparinafter the initial ACT was checked. Because of reported variability inACT measurements with different point-of-care assays, we restrictedour ACT analysis to those using the same ACT device (Hemochron,ITC, Edison, NJ).15,16 Thus, 4743 cases were considered for theheparin analysis and 1246 cases for the ACT analysis. Of those,42.8% had femoropopliteal interventions, 21.4% had iliac interven-tions, 17.2% had renal interventions, 14.1% had below-knee inter-ventions, and 4.6% had upper extremity interventions. All patientshad confirmed peripheral arterial disease, based on 1 or more of thefollowing: abnormal ankle-brachial index, noninvasive imagingstudies (ultrasound, computed tomographic angiography, or mag-netic resonance angiography) or invasive angiography. Initial anal-yses were done on 4 clinically relevant categories of heparin doserange and 3 categories of peak ACT range, respectively, suggestingthat an increment in the risk of post-PVI drop in hemoglobin �3g/dL and/or transfusion was associated with a total heparin dose �60U/kg and a peak ACT �250 seconds (Figures 1 and 2). Postproce-dural and in-hospital outcomes were compared post hoc betweenpatients who received a total heparin dose �60 U/kg and �60 U/kg.Similarly, outcomes were evaluated between groups with a peakprocedural ACT �250 and �250 seconds.

The primary end point for this study was post-PVI bleeding,defined as a drop in postprocedural hemoglobin �3g/dL from thepreprocedural baseline value or need for transfusion. Preproceduraland postprocedural hemoglobin was drawn within 24 hours beforeand after the procedure, respectively. Other secondary end pointsincluded technical success (defined as successful access, deploymentof the device, and �30% diameter residual stenosis assessed byangiography after revascularization),procedural success (defined astechnical success without periprocedural complications),17 thrombo-

Figure 1. Rates of transfusion and drop inhemoglobin (Hgb) �3 g/dL in different hepa-rin dose groups. There is an increment inthe risk of bleeding and transfusion whenheparin dose exceeds 60 U/kg.

594 Circ Cardiovasc Interv December 2010



embolic complications (defined as atherosclerotic debris or bloodclot formed during the procedure with evidence of decreased bloodflow or occlusion), and major adverse cardiac events (MACE,defined as a composite of death, myocardial infarction, and transientischemic attack[TIA]/stroke). Vascular complications were definedas a composite of retroperitoneal hematoma, pseudoaneurysm, he-matoma requiring transfusion or associated with a decrease inhemoglobin �3 g/dL, arteriovenous fistula demonstrated by arte-riography or ultrasound, and acute thrombosis or need for surgicalrepair of the access site.

Statistical AnalysisStatistical analysis was performed using the SAS 9.1.3 software(SAS Institute, Cary, NC). Descriptive statistics were used to reportbaseline patient characteristics. All tests used P�0.05 as the criticalvalue of statistical significance. The �2 and Fisher exact tests wereused to analyze categorical variables. Continuous variables wereanalyzed using the Student t test and Wilcoxon rank-sum tests asneeded. The associations of ACT and weight-adjusted heparin dosewith a drop in postprocedural hemoglobin �3 g/dL or bleedingrequiring transfusion were tested by using multivariate logisticregression models to investigate independent predictors of bleeding.The clinical characteristics included in the model were sex, age (�70years), low body mass index (BMI) (�18.5 kg/m2), history ofmyocardial infarction or coronary artery bypass grafting, hyperten-sion, low creatinine clearance (�60 mL/min per 1.73 m2), hyperlip-idemia, diabetes, history of heart failure or TIA/stroke, hybridvascular surgery and rest pain, preprocedural antiplatelet therapy,treatment with statin, ACT (�250 seconds), and heparin dose (�60U/kg). We also used intervention site variables (renal, iliac, femo-ropopliteal, below-knee, and upper extremity), device variables(balloon, stent, atherectomy, intravascular ultrasound, thrombolysis,cryoballoon, and cutting balloon) and vascular access closure de-vices (manual or mechanical) in the multivariate regression models.All variables were assessed using forward selection method with aselection criterion of P�0.05.

To further adjust for the nonrandomized use of heparin dose �60U/kg and for possible selection bias, a predictive model that adjustedfor the propensity to receive heparin �60 U/kg was also devel-oped.18 The propensity score of receiving heparin �60 U/kg wascalculated using a nonparsimonious logistic regression model. Thevariables included in the model were age (�70 years), sex, low BMI(�18.5 kg/m2), history of myocardial infarction, hypertension,coronary artery bypass grafting, diabetes, history of heart failure,TIA/stroke, low creatinine clearance (�60 mL/min per 1.73m2,hyperlipidemia, hybrid vascular surgery, rest pain, prior antiplatelettherapy, and intervention site variables. The propensity score was

then included as an additional explanatory variable in the finalmodels. Furthermore, we used the Greedy matching technique usinga macro to select patients treated with heparin �60 U/kg ascounterparts to patients treated with heparin �60 U/kg.19 Morespecifically, we sought to match each patient with �60 U/kg heparinto a patient with �60 U/kg, who had a propensity score that wasidentical to 5 digits. If this could not be done, we then performed a4-, 3-, 2-, or 1-digit match. Once this threshold was exceeded, thepatient with �60 U/kg heparin was excluded. In this way, we wereable to match 1927 cases with heparin dose �60 U/kg to 1927 caseswith heparin dose �60 U/kg. Outcomes were then compared withinthis propensity-matched cohort. A similar propensity-matched cohortwas developed for peak ACT values �250 (379 cases) and �250seconds (379 cases), and outcomes were compared. We estimated theunivariate statistical significance of the effect of higher dose ofheparin and higher peak ACT, respectively, on adverse outcomesusing generalized estimating equation cluster analysis.20,21

Given the multidisciplinary origin of data, to adjust for hospital-level and physician-level clustering, we compared hierarchical re-gression modeling with the conventional logistic models to deter-mine the statistical significance of the effect of higher dose ofheparin as well as higher peak ACT values on transfusion.22–24 Thiswas performed with a random-effects modeling technique.20,21

ResultsAmong a total of 4743 patients, 2161 patients receivedheparin �60 U/kg and 2582 patients received �60 U/kg. Ofthose, 1246 patients had peak procedural ACT recorded andmeasured with the Hemochron device (Hemochron, ITC,Edison, NJ); 855 of them had ACT �250 seconds and 391had ACT �250 seconds. The baseline demographic andclinical characteristics are presented in Table 1 for the heparindose and ACT groups, respectively. In patients receivinghigher dose of heparin (�60 U/kg), there were more women,current smokers, and higher prevalence of low BMI and lowcreatinine clearance. Furthermore, in patients receiving thelower dose of heparin (�60 U/kg), there was a higherprevalence of coronary artery disease, diabetes, hypertension,hyperlipidemia, congestive heart failure, chronic obstructivelung disease, and preprocedural use of antiplatelet therapywith more than 1 agent.

Table 2 summarizes the unadjusted rate of postproceduralor in-hospital adverse events for the heparin dose and ACT

Figure 2. Rates of transfusion and drop inhemoglobin (Hgb) �3 g/dL in differentgroups of ACT. An increment in the risk ofbleeding and transfusion was observed withACT exceeding 250 seconds.

Kasapis et al Heparin Dose in Peripheral Vascular Interventions 595



groups. Technical and procedural success was similar in allgroups. Higher weight-adjusted heparin doses were associ-ated with significantly more bleeding complications. Patientswith heparin dose �60 U/kg had a higher rate of postproce-dural hemoglobin drop �3 g/dL (7.09% versus 5.09%,respectively, P�0.004) and a higher transfusion rate com-pared with those with heparin dose �60 U/kg (4.92% versus3.15%, respectively, P�0.002). There was no difference intechnical and procedural success, MACE, and thromboem-bolic complications in both groups.

Similarly, a higher ACT was associated with more bleed-ing and need for transfusion. Patients with a procedural ACT�250 seconds had a higher rate of postprocedural drop inhemoglobin �3 g/dL (9.7% versus 4.8%, respectively,P�0.0009) and a higher incidence of transfusions comparedwith the group with ACT �250 seconds (5.4% versus 2.8%,respectively, P�0.02). Technical and procedural success,

MACE, and thromboembolic complications were similar inthe two ACT groups.

There were more femoropopliteal PVIs than any othervascular bed treated (Table 3). A significantly higher numberof femoropopliteal and below-knee interventions were per-formed by using �60 U/kg of heparin. The majority of thecases were performed with retrograde arterial access withonly 12.8% of cases performed with antegrade access. Theeffect of higher heparin dose and ACT on higher bleedingrates persisted even when cases with retrograde access wereanalyzed separately.

Although it is difficult to objectively standardize thecomplexity of the procedure in a multicenter data base, westudied in a separate analysis the heparin dose, bleedingcomplications, and procedural success stratified according tothe location of the procedure—for example, renal, iliac,femoropopliteal, below-knee, and upper extremity—as well

Table 1. Baseline Demographic and Clinical Characteristics in the Heparin Dose and ACT Groups

Characteristics

Heparin Data Set (n�4743) ACT Data Set (n�1246)

Heparin�60 U/kg

Heparin�60 U/kg P Value

ACT�250 Seconds

ACT�250 Seconds P Value

Cases, N 2161 2582 855 391

Average age, y (�SD) 68.2 (�11.3) 68.7 (�11.2) 0.18 68.32 (�11.0) 67.98 (�12.4) 0.6

Female, N (%) 869 (40.2) 1158 (44.8) 0.0013 334 (39.1) 169 (43.2) 0.16

Current smoker, N (%) 580 (26.8) 804 (31.1) 0.001 234 (27.4) 101 (25.8) 0.57

Prior CAD, N (%) 1670 (77.3) 1794 (69.5) �0.0001 684 (80.0) 316 (80.8) 0.7

Prior diabetes, N (%) 997 (46.1) 1063 (41.2) 0.0006 374 (43.7) 177 (45.3) 0.6

Prior HTN, N (%) 1993 (92.2) 2329 (90.2) 0.014 801 (93.7) 369 (94.4) 0.6

Prior HL, N (%) 1766 (81.7) 1991 (77.1) �0.0001 656 (76.3) 278 (71.1) 0.03

Prior CHF, N (%) 530 (24.5) 468 (18.1) �0.0001 218 (25.5) 109 (27.9) 0.3

Prior COPD, N (%) 629 (29.1) 616 (23.8) �0.0001 246 (28.8) 106 (27.1) 0.5

Prior TIA/stroke, N (%) 646 (29.9) 729 (28.2) 0.2 254 (29.7) 113 (28.9) 0.7

Low BMI,* �18.5 kg/m2, N (%) 38 (1.8) 92 (3.6) 0.0002 22 (2.6) 8 (2.1) 0.6

Low creatinine clearance,†�60 mL/min, N (%)

870 (40.9) 1170 (46.7) �0.0001 376 (44.7) 173 (45.4) 0.8

Preprocedural mono-antiplatelettherapy,‡ N (%)

783 (36.2) 992 (38.4) 0.12 277 (32.4) 143 (36.57) 0.14

Preprocedural antiplatelettherapy,�1, N (%)

1103 (51.0) 1206 (46.7) 0.0029 531 (62.1) 222 (56.8) 0.07

Aspirin alone 652 (30.2%) 830 (32.2%) 0.14 82 (5.4%) 56 (7.5%) 0.047

Clopidogrel alone 116 (5.4%) 141 (5.5%) 0.8 14 (0.9%) 4 (0.5%) 0.3

Aspirin and clopidogrel 1019 (47.2%) 1115 (43.2%) 0.006 786 (51.9%) 354 (47.6%) 0.06

Presenting symptoms

Asymptomatic, N (%) 94 (4.3) 91 (3.5) 0.1 49 (5.7) 21 (5.4) 0.8

Claudication, N (%) 2363 (74) 3860 (77.2) 0.001 620 (72.5) 304 (77.7) 0.05

Limb salvage, N (%) 294 (9.2) 546 (10.9) 0.01 54 (6.3) 39 (10.0) 0.02

Rest pain, N (%) 647 (20.3) 1224 (24.5) �0.0001 131 (15.3) 87 (22.2) 0.003

Renal salvage, N (%) 49 (1.5) 57 (1.1) 0.1 24 (2.8) 12 (3.1) 0.8

CAD indicates coronary artery disease; HTN, hypertension; HL, hyperlipidemia; CHF, congestive heart failure; and COPD, chronicobstructive pulmonary disease.

*The BMI is the weight in kilograms divided by the square of height in meters. Low BMI is defined as �18.5 according to theCenters for Disease Control and Prevention definition.

†Low creatinine clearance is defined as �60 mL/min, which is equivalent to stage 3 or more chronic kidney disease accordingto the National Kidney Foundation definition.

‡Antiplatelets: Aspirin, clopidogrel, ticlopodine, or cilostazol.




as the procedure duration—for example, �57 minutes, 58 to90 minutes, and �90 minutes. Interventions to the renalarteries had lower heparin doses and bleeding rates as well ashigher procedural success, femoropopliteal and iliac interven-tions had intermediate heparin doses, bleeding rates andprocedural success, and below-knee interventions had higherheparin doses and bleeding rates as well as lower proceduralsuccess. Similarly, longer procedure duration was intuitivelyassociated with use of higher heparin dose, higher bleedingcomplications, and lower procedural success. The effect ofhigher heparin dose on higher bleeding rates persisted evenwhen patients were analyzed separately in short versus longprocedure duration.

In a multivariate analysis, higher heparin dose (�60 U/kg)was found to be an independent predictor of bleeding requir-ing transfusion (odds ratio [OR], 1.5; 95% confidence inter-val [CI], 1.11 to 2.0; P�0.01) (Table 4); other independentpredictors of bleeding were female sex, age �70 years, prioranemia, prior heart failure, hybrid vascular surgery, rest pain,and below-knee interventions. The c-statistic for this model

was 0.799. A separate multivariate logistic model was devel-oped to avoid collinearity between heparin and ACT and alsoidentified ACT �250 seconds as an independent predictor oftransfusion (OR, 1.9; 95% CI, 1.06 to 3.6; P�0.03). For thismodel, the c-statistic was 0.812. A significant interaction wasobserved between heparin dose and ACT levels (OR, 1.29;95% CI, 1.03 to 1.6; P�0.001).

Manual hemostasis was achieved in 69.7% of the patients.There was no statistical difference in the rate of bleedingbetween manual hemostasis and use of vascular closuredevices. In addition, when we stratified the patients accordingto sheath size, there was no significant correlation betweensheath size and increased bleeding events.

Furthermore, in a propensity-matched cohort (Table 5) inwhich each patient treated with higher dose of heparin (�60U/kg) was matched to a similar patient treated with lowerheparin dose (�60 U/kg), the higher total heparin dose wasassociated with significantly higher rates of postproceduraldrop in hemoglobin �3 g/dL (7.01% versus 5.14%, respec-tively, P�0.01) and transfusion (4.67% versus 3.17%, re-

Table 2. Postprocedural and In-Hospital Adverse Events in the Heparin Dose and ACT Groups

Heparin Data Set (n�4743) ACT Data Set (n�1246)

Heparin�60 U/kg

Heparin�60 U/kg P Value

ACT�250 Seconds

ACT�250 Seconds P Value

Cases, n 2161 2582 855 391

Mean heparin dose, U/kg �SD, or meanACT, seconds �SD

47.9�9.4 82.5�22.0 �0.0001 221.3�14.1 286�41.7 �0.0001

Drop in Hgb �3 g/dL, n (%) 110 (5.09) 183 (7.09) 0.004 41 (4.8) 38 (9.7) 0.0009

Transfusion rate, n (%) 68 (3.15) 127 (4.92) 0.002 24 (2.8) 21 (5.4) 0.02

Transfusion or drop in Hgb �3 g/dL, n (%) 142 (6.57) 255 (9.88) �0.0001 54 (6.3) 44 (11.2) 0.003

MACE: Death or MI or TIA/stroke, n (%) 15 (0.69) 16 (0.62) 0.7 6 (0.7) 3 (0.8) 1

Embolic complications, n (%) 3 (0.14) 11 (0.43) 0.07 3 (0.3) 1 (0.3) 0.8

Thrombotic complications, n (%) 7 (0.32) 16 (0.62) 0.14 2 (0.2) 2 (0.5) 0.4

Vascular complications, n (%) 57 (2.64 98 (3.8) 0.02* 28 (3.3) 16 (4.1) 0.5

Not crossed with wire or device, n (%) 140 (6.5) 161 (6.2) 0.73 37 (4.3) 17 (4.3) 0.98

Technical success, n (%) 1869 (86.49) 2223 (86.1) 0.6 744 (87.0) 339 (86.7) 0.9

Procedural success, n (%) 1765 (81.7) 2223 (86.1) 0.7 703 (82.2) 315 (80.56) 0.4

Mean procedure duration, min �SD) 69.33�42.2 92.54�49.65 �0.0001 75.1�48.2 86.9�53.0 0.0002

Hgb indicates hemoglobin; MI, myocardial infarction.*This difference is driven by higher rate of bleeding complications, for example, hematomas or retroperitoneal bleeding in the high-dose heparin group, whereas

other vascular complications were similar in both groups.

Table 3. Frequency of Interventions at Different Vascular Beds Stratified by Heparin Dose and ACT Levels

Heparin Data Set ACT Data Set

n (%) �60 U/kg �60 U/kg P Value n (%) �250 Seconds �250 Seconds P Value

Renal, n (%) 866 (18.3) 438 (20.27) 428 (16.58) 0.001 290 (23.3) 212 (24.8) 78 (19.95) 0.06

Iliac, n (%) 1347 (28.4) 654 (30.26) 693 (26.84) 0.009 351 (28.2) 240 (28.07) 111 (28.39) 0.9

Femoropopliteal, n (%) 2403 (50.6) 991 (45.86) 1412 (54.69) �0.0001 589 (47.3) 395 (46.2) 194 (49.6) 0.2

Below-knee, n (%) 650 (13.7) 224 (10.37) 426 (16.5) �0.0001 139 (11.1) 89 (10.41) 50 (12.79) 0.2

Upper extremity, n (%) 211 (4.45) 102 (4.72) 109 (4.22) 0.4 59 (4.7) 38 (4.44) 21 (5.37) 0.47

Other, n (%) 124 (2.6) 65 (3.01) 59 (2.29) 0.1 39 (3.1) 26 (1.72) 13 (1.75) 0.9

Below-knee indicates popliteal, peroneal, tibial, tibioperoneal trunk, or dorsalis pedis; upper extremity, radial, brachial, subclavian, or axillary; and other, celiac,mesenteric, distal aorta, pulmonary, or dialysis fistula.




spectively, P�0.016). Similarly, in a propensity-matched,risk-adjusted model for the ACT groups (�250 seconds and�250 seconds), patients with a higher ACT (�250 seconds)had significantly higher rates of postprocedural drop inhemoglobin (10.03% versus 5.8%, P�0.03) and a trendtoward more transfusion (5.54% versus 2.9%, P�0.07) (Ta-ble 5). Importantly, there were no differences with respect totechnical and procedural success, MACE, and thromboem-bolic complications in all groups. A separate propensity-scorematching analysis by adding intervention sites (renal, iliac,femoropopliteal, below-knee, and upper extremity) to themodels demonstrated similar findings.

The heparin patient cohort (n�4743) data were obtainedfrom 7 different hospitals involving 75 physicians, whereasthe ACT cohort (n�1246) data derived from 7 hospitalsinvolving 59 physicians. To adjust for hospital-level and

physician-level variations, a hierarchical regression modelwas performed to determine the relationships betweenphysician- and hospital-level characteristics for each of theoutcomes. The physician- and hospital-level factors did nothave any impact on the effect of higher dosage of heparinleading to transfusion as determined by similar ORs derivedfrom the hierarchical (OR, 1.4; 95% CI, 1.14 to 1.82;P�0.002) and logistical (OR, 1.5; 95% CI, 1.1 to 2.0;P�0.01) models. Similarly, the effect of higher ACT re-mained significant in the hierarchical (OR, 2.2; 95% CI, 1.3to 3.7; P�0.0003) as well as in the logistic model (OR, 1.9;95% CI, 1.06 to 3.6; P�0.03).

DiscussionThis analysis of prospectively collected individual patientdata represents the largest study correlating the degree of

Table 4. Independent Predictors of Transfusion and Drop in Hgb >3 g/dL Using LogisticRegression Models

Variable

Predictors of Drop in Hgb �3 g/dL Predictors of Transfusion

OR 95% CI P Value OR 95% CI P Value

Female sex 1.98 1.5–2.5 �0.0001 2.3 1.7–3.2 �0.0001

Age �70 y 1.3 0.98–1.7 0.06 1.4 1.05–2.0 0.02

Prior anemia 1.3 0.9–1.2 0.065 3.36 2.4–4.7 �0.0001

Prior CHF NS 1.74 1.2–2.4 0.0009

Low creatinine clearance,�60 mL/min/1.73 m2

3.01 1.6–5.4 0.0002 2.47 1.15–5.2 0.019

Hybrid vascular surgery 7.6 4.4–13.1 �0.0001 6.1 3.4–11.2 �0.0001

Rest pain 2.2 1.66–2.9 �0.0001 3.08 2.2–4.2 �0.0001

Below-knee 3.5 1.8–6.2 �0.0001 3.47 1.01–3.5 0.0007

Heparin, �60 U/kg 1.4 1.1–1.8 0.01 1.5 1.11–2.0 0.01

ACT, �250 seconds* 2.0 1.2–3.2 0.001 1.9 1.06–3.6 0.03

Interaction term forheparin and ACT†

1.3 1.06–1.6 0.01 1.29 1.03–1.6 0.001

Hgb indicates hemoglobin; CHF, congestive heart failure; and NS, nonsignificant.*The ACT was tested using a separate multivariate logistic regression analysis, excluding the heparin dose, to avoid

significant collinearity between the 2 variables.†An interaction term was created for heparin and ACT and added to the multivariate models to determine

interactions between heparin dose and ACT.

Table 5. Outcomes of the Propensity-Matched Cohorts Categorized by Heparin Dose and ACT Level

Heparin Data Set ACT Data Set

�60 U/kg �60 U/kg P Value �250 Seconds �250 Seconds P Value

Cases, N 1927 1927 379 379

Transfusion, N (%) 61 (3.17%) 90 (4.67%) 0.016 11 (2.9%) 21 (5.54%) 0.07

Drop in hemoglobin, �3 g/dL, N (%) 99 (5.14%) 135 (7.01%) 0.01 22 (5.8%) 38 (10.03%) 0.03

MACE, N (%) 15 (0.78%) 12 (0.62%) 0.5 3 (0.79%) 3 (0.79%) 1

Embolus, N (%) 2 (0.1%) 8 (0.4%) 0.06 1 (0.26%) 1 (0.26%) 1

Thrombus, N (%) 7 (0.38%) 13 (0.67%) 0.17 1 (0.26%) 2 (0.53%) 0.5

Vascular access complication, N (%) 50 (2.6%) 73 (3.8%) 0.03* 12 (3.17%) 16 (4.22%) 0.4

Not crossed with wire or device, N (%) 126 (6.5%) 103 (5.3%) 0.1 11 (2.9%) 17 (4.49%) 0.2

Procedural success, N (%) 1563 (81.1%) 1572 (81.58%) 0.7 314 (82.8%) 304 (80.2%) 0.3

Hgb indicates hemoglobin; MACE are defined as a composite of death or myocardial infarction or TIA/stroke.*This difference is driven by higher rate of bleeding complications, for example, hematomas or retroperitoneal bleeding in the

high-dose heparin group, whereas other vascular complications were similar in both groups.




anticoagulation with heparin and early postprocedural out-comes in PVI. The key findings of our study were that duringPVI, a higher total heparin dose (�60 U/kg) and a peakprocedural ACT �250 seconds were predictors of postpro-cedural bleeding events. Moreover, the technical and proce-dural success was high and the rate of thromboemboliccomplications was low and did not differ between groups.These findings suggest that use of weight-based heparindosing (initially, up to 60 U/kg) with a target ACT �250seconds may improve outcomes in PVI.

A previous single-center, retrospective study evaluatedin-hospital complications of peripheral interventions in whichUFH was the primary anticoagulant in 131 patients (withACT recorded in 114) and reported a nonsignificant trendtoward higher adverse events with increased weight-baseddoses of heparin and no significant relationship between ACTand adverse events, although there was a relationship betweenheparin dosage and ACT level.25 The rate of major bleedingin this study increased from 2.1% with heparin dose 30 to54.9 U/kg to 5.4% with heparin dose 55 to 99.9 U/kg. This isin agreement with our findings, in which patients withheparin dose �60 U/kg had higher rates of drop in hemoglo-bin of �3 g/dL (7.09%) and transfusion (4.92%) comparedwith those with heparin dose �60 U/kg (5.09% and 3.15%,respectively).

Most of the data related to the optimal degree of antico-agulation with heparin for arterial percutaneous interven-tional procedures are extrapolated from the coronary litera-ture. A previous meta-analysis of 6 randomized, controlledtrials of novel adjunctive antithrombotic regimens for PCI, inwhich UFH constituted the control arm, demonstrated an“optimal” target of ACT in the range of 350 to 375 secondsthat provided the lowest composite ischemic event rate.3

However, this degree of anticoagulation was associated witha higher incidence of bleeding, increasing from 8.6% withinan ACT range of 325 to 350 seconds to 12.4% at ACT of 350to 375 seconds.

Subsequent and more contemporary studies favored adownward shift in the target ACT levels in PCI.4–6 Accord-ingly, a post hoc analysis of the population enrolled in theESPRIT trial (Enhanced Suppression of the Platelet GPIIb/IIIa Receptor with Integrilin Therapy), comparing the useof eptifibatide versus placebo in patients undergoing PCI withstent, demonstrated that the incidence of ischemic events didnot increase as ACT decreased at least to a level of 200seconds, whereas bleeding events did increase with increas-ing ACT levels.4 Notably, the recommended initial bolus ofheparin dose in this trial was 60 U/kg, which is the cutoff thatwe used in our study. The ESPRIT investigators concludedthat an ACT level of 200 to 250 seconds is reasonable interms of efficacy and safety with the use of contemporarytechnology and pharmacotherapy. Within the same concep-tual framework, a large pooled analysis of the UFH-treatedpatients enrolled in 4 recent large, randomized trials using amedian ACT of 276 seconds demonstrated that in patientsundergoing PCI with frequent stent and potent platelet inhi-bition use, ACT does not correlate with ischemic complica-tions and has a modest linear association with bleedingcomplications, suggesting that lower ACT values do not

appear to compromise efficacy while increasing safety.6

Notably, these trials assessed the concurrent use of UFH andGPIIb/IIIa antagonists, whereas in our study the use of UFHalone was evaluated. Therefore, even in the absence of the useof GP IIb/IIIa antagonists, a lower ACT or heparin dose inPVI does not result in more thromboembolic complications.

Although it appears simple to extrapolate the evidencefrom coronary interventions and apply it to the peripheralarteries, the relationship between heparin dosing–ACT levelsand bleeding complications, as well as procedural outcomes,may be less evident in straightforward cases of coronarystenting than in more prolonged and complex peripheralarterial procedures. Therefore, the results of our analysisprovide further insight in defining the optimal degree ofheparin anticoagulation during PVI. In addition, it is welldocumented that bleeding complications, especially thoserequiring transfusion, remain an independent, strong predic-tor of adverse outcomes in patients undergoing PCI.8–11,13,26

Even though there is no current evidence on the prognosticimplications of bleeding complications in PVI, patients withperipheral arterial disease represent an equivalent high-riskcohort, often with significant comorbidities, such as coronaryartery disease, diabetes, and renal dysfunction. Therefore, it isessential to minimize bleeding risk while maintaining a safeanticoagulation level during PVI.

Although our registry analysis establishes the evidencebase for a higher safety threshold for a target ACT in PVI, thequestion of a lower safe ACT threshold remains to beanswered. It is possible that lower levels of anticoagulationmay be sufficient and safe in PVI, given the large vessel size,especially in those patients already on dual antiplatelettherapy. Within this concept, a randomized, double-blind,prospective study demonstrated that in the treatment ofuncomplicated coronary lesions and in the presence of dualantiplatelet therapy with aspirin and clopidogrel, elective PCIcould be safely performed without systemic anticoagulationand was associated with a reduced incidence of bleedingcomplications.27 However, this interesting hypothesis has notbeen studied yet in PVI, it cannot be answered within thesetting of a “real-life” registry using the current guideline-driven standards of care, and it will require a future random-ized, prospective clinical study to address this question.

Beyond the heparin dose and the ACT level, in themultivariate analysis we found that female sex, creatinineclearance �60 mL/min per 1.73m2 age �70 years, prepro-cedural anemia, history of heart failure, hybrid vascularsurgery, rest pain, and below-knee interventions were inde-pendent predictors of higher postprocedural bleeding risk.These findings are consistent with those from 2 large retro-spective databases of patients undergoing PCI13,28 and aregistry of 24 045 patients with acute coronary syndromesfrom the Global Registry of Acute Coronary Events(GRACE).29 Less aggressive anticoagulation, or use of analternative anticoagulant strategy such as a direct thrombininhibitor, may decrease the incidence of bleeding complica-tions in these subgroups of high-risk patients.30–33

LimitationsOur findings are based on observational, prospectively col-lected data, and, inevitably, there were significant differences




in the demographic and clinical characteristics. Although weadjusted using multivariate logistic regression and propensityanalysis, it is possible that other confounding factors mayhave influenced outcomes. In addition, this database spans aperiod of 7 years and encompasses a heterogeneous group ofendovascular interventions and advancing interventionaltechniques as well as reflecting multiple institutions andoperators. Even though we have tried to adjust for hospital-and operator-specific clustering by using hierarchical regres-sion techniques, operator-specific variations in proceduraloutcomes as well as in heparin dosing and ACT target are alsoinherent limitations of our study. However, this represents a“real-life” data base, and these procedures will continue to beperformed by multiple operators from different specialties,including interventional cardiologists, interventional radiolo-gists, and vascular surgeons as well as in both academic andprivate institutions. Because of inherent differences in themeasurement of ACT by different ACT devices15,16 and to beconsistent, we have restricted our ACT data to those mea-sured by the Hemochron device only, which was the mostcommon device used in our registry. This obviously limitsour ability to generalize the findings to other devices. Itshould be also noted that the study cohort for this analysisconsisted of patients undergoing PVI using the old formula-tion of heparin available in the United States. However,effective as of October 2009, a new formulation of heparin isavailable in the United States, and, according to the FDApublic health alert, there is an approximate 10% decrease inthe potency of the “new heparin” compared with the “oldheparin.”34 This dose adjustment should be taken into con-sideration when the “new heparin” is used. Furthermore, thesuggested weight-based heparin dosing (initially up to 60U/kg) with a target ACT of �250 seconds, as opposed to“empirical” dosing, should serve as a general guide forheparin anticoagulation in PVI and should not substitute forthe individualization of dose in specific subgroups of patientswho are at higher bleeding risk, necessitating a more conser-vative dose, or, conversely, are resistant to heparin requiringhigher doses.

ConclusionNotwithstanding these limitations, this is the first large-scalestudy to evaluate the optimal degree of heparin anticoagula-tion in PVI. This study suggests that use of weight-basedheparin dosing (initially up to 60 U/kg) with a target ACT�250 seconds in PVI may minimize the bleeding risk withoutcompromising procedural success or increasing thromboem-bolic complications. Future prospective, randomized studiesare warranted to further validate these findings and poten-tially establish a lower safe ACT threshold for peripheralvascular interventions.

Sources of FundingThis work was funded by Blue Cross Blue Shield of Michigan, aUniversity of Michigan Cardiovascular Center McKay ResearchGrant, and an unrestricted educational grant from Bristol-MyersSquibb, Incorporated. The funding organizations had no role in thedesign and conduct of the study; collection, analysis, or preparationof the data; or preparation, review, or approval of the manuscript.

DisclosuresDr Gurm received research support from Blue Cross Blue Shield ofMichigan and the National Institutes of Health. Dr Kassab is aspeaker/consultant for Abbott Vascular/Guidant Corporation, BostonScientific Corporation, Cardiovascular Systems Inc, ev3/FoxHollowCorporation, Spectranetics Corporation, and Medtronic VascularCorporation, holding stocks in Cardiovascular Systems, Inc, and isfounder/co-owner of Kassab, Kughn, Endovascular, LLC (KKED,LLC). Dr Grossman received research support from Blue Cross BlueShield of Michigan.

References1. Ogilby JD, Kopelman HA, Klein LW, Agarwal JB. Adequate hepa-

rinization during PTCA: assessment using activated clotting times. CathetCardiovasc Diagn. 1989;18:206–209.

2. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, FowkesFG. Inter-Society Consensus for the Management of Peripheral ArterialDisease (TASC II). J Vasc Surg. 2007;45(Suppl S):S5–S67.

3. Chew DP, Bhatt DL, Lincoff AM, Moliterno DJ, Brener SJ, Wolski KE,Topol EJ. Defining the optimal activated clotting time during percutane-ous coronary intervention: aggregate results from 6 randomized, con-trolled trials. Circulation. 2001;103:961–966.

4. Tolleson TR, O’Shea JC, Bittl JA, Hillegass WB, Williams KA, LevineG, Harrington RA, Tcheng JE. Relationship between heparin anticoagu-lation and clinical outcomes in coronary stent intervention: observationsfrom the ESPRIT trial. J Am Coll Cardiol. 2003;41:386–393.

5. Brener SJ, Bhatt DL, Moliterno DJ, Schneider JP, Ellis SG, Topol EJ.Revisiting optimal anticoagulation with unfractionated heparin duringcoronary stent implantation. Am J Cardiol. 2003;92:1468–1471.

6. Brener SJ, Moliterno DJ, Lincoff AM, Steinhubl SR, Wolski KE, TopolEJ. Relationship between activated clotting time and ischemic or hemor-rhagic complications: analysis of 4 recent randomized clinical trials ofpercutaneous coronary intervention. Circulation. 2004;110:994–998.

7. Ragosta M, Karve M, Brezynski D, Humphries J, Sanders JM, SarembockIJ, Gimple LW, Powers ER. Effectiveness of heparin in preventingthrombin generation and thrombin activity in patients undergoingcoronary intervention. Am Heart J. 1999;137:250–257.

8. Rao SV, Jollis JG, Harrington RA, Granger CB, Newby LK, ArmstrongPW, Moliterno DJ, Lindblad L, Pieper K, Topol EJ, Stamler JS, CaliffRM. Relationship of blood transfusion and clinical outcomes in patientswith acute coronary syndromes. JAMA. 2004;292:1555–1562.

9. Yatskar L, Selzer F, Feit F, Cohen HA, Jacobs AK, Williams DO, SlaterJ. Access site hematoma requiring blood transfusion predicts mortality inpatients undergoing percutaneous coronary intervention: data from theNational Heart, Lung, and Blood Institute Dynamic Registry. CatheterCardiovasc Interv. 2007;69:961–966.

10. Rao SV, O’Grady K, Pieper KS, Granger CB, Newby LK, Van de WerfF, Mahaffey KW, Califf RM, Harrington RA. Impact of bleeding severityon clinical outcomes among patients with acute coronary syndromes.Am J Cardiol. 2005;96:1200–1206.

11. Yang X, Alexander KP, Chen AY, Roe MT, Brindis RG, Rao SV, GiblerWB, Ohman EM, Peterson ED. The implications of blood transfusions forpatients with non-ST-segment elevation acute coronary syndromes:results from the CRUSADE National Quality Improvement Initiative.J Am Coll Cardiol. 2005;46:1490–1495.

12. Koch CG, Li L, Duncan AI, Mihaljevic T, Loop FD, Starr NJ, BlackstoneEH. Transfusion in coronary artery bypass grafting is associated withreduced long-term survival. Ann Thorac Surg. 2006;81:1650–1657.

13. Kinnaird TD, Stabile E, Mintz GS, Lee CW, Canos DA, Gevorkian N,Pinnow EE, Kent KM, Pichard AD, Satler LF, Weissman NJ, Lindsay J,Fuchs S. Incidence, predictors, and prognostic implications of bleedingand blood transfusion following percutaneous coronary interventions.Am J Cardiol. 2003;92:930–935.

14. Mukherjee D, Munir K, Hirsch AT, Chetcuti S, Grossman PM,Rajagopalan S, Nallamothu BK, Moscucci M, Henke P, Kassab E, SohalC, Riba A, Person D, Luciano AE, DeGregorio M, Patel K, RutkowskiKC, Eagle KA. Development of a multicenter peripheral arterial inter-ventional database: the PVD-QI2. Am Heart J. 2005;149:1003–1008.

15. Avendano A, Ferguson JJ. Comparison of Hemochron and HemoTecactivated coagulation time target values during percutaneous transluminalcoronary angioplasty. J Am Coll Cardiol. 1994;23:907–910.

16. Doherty TM, Shavelle RM, French WJ. Reproducibility and variability ofactivated clotting time measurements in the cardiac catheterization labo-ratory. Catheter Cardiovasc Interv. 2005;65:330–337.




17. Diehm N, Baumgartner I, Jaff M, Do DD, Minar E, Schmidli J, Diehm C,Biamino G, Vermassen F, Scheinert D, van Sambeek MR, Schillinger M.A call for uniform reporting standards in studies assessing endovasculartreatment for chronic ischaemia of lower limb arteries. Eur Heart J.2007;28:798–805.

18. D’Agostino RB Jr. Propensity score methods for bias reduction in thecomparison of a treatment to a non-randomized control group. Stat Med.1998;17:2265–2281.

19. Parsons L. Reducing bias in a propensity score matched-pair sample usingGreedy matching techniques. Proceedings of the 26th Annual SAS UsersGroup International Conference, Cary, NC: SAS Institute; 2001:214–226.

20. Allison P. Logistic Regression Using the SAS System: Theory and Appli-cation. Cary, North Carolina: SAS Institute Inc; 1999:197–200.

21. Austin PC. A critical appraisal of propensity-score matching in themedical literature between 1996 and 2003. Stat Med. 2008;27:2037–2049.

22. Austin PC, Tu JV, Alter DA. Comparing hierarchical modeling withtraditional logistic regression analysis among patients hospitalized withacute myocardial infarction: should we be analyzing cardiovascularoutcomes data differently? Am Heart J. 2003;145:27–35.

23. DeLong E. Hierarchical modeling: its time has come. Am Heart J.2003;145:16–18.

24. Austin PC, Goel V, van Walraven C. An introduction to multilevelregression models. Can J Public Health. 2001;92:150–154.

25. Shammas NW, Lemke JH, Dippel EJ, McKinney DE, Takes VS,Youngblut M, Harris M, Harb C, Kapalis MJ, Holden J. In-hospitalcomplications of peripheral vascular interventions using unfractionatedheparin as the primary anticoagulant. J Invasive Cardiol. 2003;15:242–246.

26. Jani SM, Smith DE, Share D, Kline-Rogers E, Khanal S, O’Donnell MJ,Gardin J, Moscucci M. Blood transfusion and in-hospital outcomes inanemic patients with myocardial infarction undergoing percutaneouscoronary intervention. Clin Cardiol. 2007;30:II49–II56.

27. Stabile E, Nammas W, Salemme L, Sorropago G, Cioppa A, Tesorio T,Ambrosini V, Campopiano E, Popusoi G, Biondi Zoccai G, Rubino P.The CIAO (Coronary Interventions Antiplatelet-based Only) Study: a

randomized study comparing standard anticoagulation regimen toabsence of anticoagulation for elective percutaneous coronary inter-vention. J Am Coll Cardiol. 2008;52:1293–1298.

28. Chiu JH, Bhatt DL, Ziada KM, Chew DP, Whitlow PL, Lincoff AM, EllisSG, Topol EJ. Impact of female sex on outcome after percutaneouscoronary intervention. Am Heart J. 2004;148:998–1002.

29. Moscucci M, Fox KA, Cannon CP, Klein W, Lopez-Sendon J, Mon-talescot G, White K, Goldberg RJ. Predictors of major bleeding in acutecoronary syndromes: the Global Registry of Acute Coronary Events(GRACE). Eur Heart J. 2003;24:1815–1823.

30. Allie DE, Hall P, Shammas NW, Safian R, Laird JR, Young JJ, VirmaniA. The Angiomax Peripheral Procedure Registry of Vascular Events Trial(APPROVE): in-hospital and 30-day results. J Invasive Cardiol. 2004;16:651–656.

31. Shammas NW, Allie D, Hall P, Young J, Laird J, Safian R, Virmani A.Predictors of in-hospital and 30-day complications of peripheral vascularinterventions using bivalirudin as the primary anticoagulant: results fromthe APPROVE Registry. J Invasive Cardiol. 2005;17:356–359.

32. Lincoff AM, Bittl JA, Harrington RA, Feit F, Kleiman NS, Jackman JD,Sarembock IJ, Cohen DJ, Spriggs D, Ebrahimi R, Keren G, Carr J, CohenEA, Betriu A, Desmet W, Kereiakes DJ, Rutsch W, Wilcox RG, de FeyterPJ, Vahanian A, Topol EJ. Bivalirudin and provisional glycoproteinIIb/IIIa blockade compared with heparin and planned glycoproteinIIb/IIIa blockade during percutaneous coronary intervention:REPLACE-2 randomized trial. JAMA. 2003;289:853–863.

33. Stone GW, McLaurin BT, Cox DA, Bertrand ME, Lincoff AM, MosesJW, White HD, Pocock SJ, Ware JH, Feit F, Colombo A, Aylward PE,Cequier AR, Darius H, Desmet W, Ebrahimi R, Hamon M, RasmussenLH, Rupprecht HJ, Hoekstra J, Mehran R, Ohman EM. Bivalirudin forpatients with acute coronary syndromes. N Engl J Med. 2006;355:2203–2216.

34. US Food and Drug Administration. FDA Drug Safety Communication:Update: Follow up to the Public Health Alert about Changes to the HeparinSodium USP Monograph. April 7th 2010. Available at: http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm207506.htm. Accessed July 2010.

CLINICAL PERSPECTIVEThe optimal degree of heparin anticoagulation for peripheral vascular interventions (PVI) has not been defined. Currentrecommendations have been empirically based on data from the coronary literature. We sought to correlate heparin doseand peak procedural activated clotting time with postprocedural outcomes in patients undergoing PVI in a regional,multicenter registry. Total heparin dose �60 U/kg and peak activated clotting time �250 seconds were associated withsignificantly higher rates of post-PVI drop in hemoglobin �3 g/dL and/or transfusion, with no differences in technical orprocedural success or thromboembolic complications. These findings suggest that weight-based heparin dosing (initiallyup to 60 U/kg) with a target activated clotting time of 250 seconds or less may improve outcomes in PVI, with the caveatthat an acceptable lower activated clotting time threshold is unknown. Our study is important because it establishes anevidence base for the optimal degree of heparin anticoagulation for PVI.



http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm207506.htm




SUPPLEMENTAL MATERIAL

Participating hospitals in the BMC2 PVI registry:

University of Michigan Health System, Ann Arbor, MI; Veterans Administration Ann Arbor

Medical Center, Ann Arbor, MI; St. Joseph Mercy Hospital, Ann Arbor, MI; Sparrow Hospital,

Lansing, MI; Oakwood Hospital, Dearborn, MI; Meijer Heart Center, Grand Rapids, MI;

Holland Hospital, Holland, MI



defining the optimal degree of heparin anticoagulation

Documents