Antiplatelet Therapy Use After Discharge Among Acute Myocardial Infarction

Patients With In-Hospital Bleeding

Tracy Y. Wang, Lan Xiao, Karen P. Alexander, Sunil V. Rao, Mikhail N. Kosiborod,
John S. Rumsfeld, John A. Spertus and Eric D. Peterson
Circulation 2008;118;2139-2145; originally published online Nov 3, 2008;
DOI: 10.1161/CIRCULATIONAHA.108.787143
Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX
72514
Copyright © 2008 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online
ISSN: 1524-4539

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/cgi/content/full/118/21/2139

Subscriptions: Information about subscribing to Circulation is online at

http://circ.ahajournals.org/subscriptions/

Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
410-528-8550. E-mail:
journalpermissions@lww.com

Reprints: Information about reprints can be found online at

http://www.lww.com/reprints

Downloaded from circ.ahajournals.org by on May 5, 2009

 Health Services and Outcomes Research

Antiplatelet Therapy Use After Discharge Among Acute

Myocardial Infarction Patients With In-Hospital Bleeding
Tracy Y. Wang, MD, MHS; Lan Xiao, PhD; Karen P. Alexander, MD; Sunil V. Rao, MD;
Mikhail N. Kosiborod, MD; John S. Rumsfeld, MD, PhD;
John A. Spertus, MD, MPH; Eric D. Peterson, MD, MPH

Background—Bleeding among patients with acute myocardial infarction (AMI) is associated with worse long-term
outcomes. Although the mechanism underlying this association is unclear, a potential explanation is that withholding
antiplatelet therapies long beyond resolution of the bleeding event may contribute to recurrent events.
Methods and Results—We examined medication use at discharge, 1, 6, and 12 months after AMI among 2498 patients in
the Prospective Registry Evaluating Myocardial Infarction: Events and Recovery (PREMIER) registry. Bleeding was
defined as non– coronary artery bypass graft–related Thrombolysis of Myocardial Infarction major/minor bleeding or
transfusion among patients with baseline hematocrit ⱖ28%. Logistic regression was used to evaluate the association
between bleeding during the index AMI hospitalization and medication use. In-hospital bleeding occurred in 301
patients (12%) with AMI. Patients with in-hospital bleeding were less likely to be discharged on aspirin or
thienopyridine (adjusted odds ratio⫽0.45; 95% CI, 0.31 to 0.64; and odds ratio⫽0.62; 95% CI, 0.42 to 0.91,
respectively). At 1 month after discharge, although patients with in-hospital bleeding remained significantly less likely
to receive aspirin (odds ratio⫽0.68; 95% CI, 0.50 to 0.92), use of thienopyridines in the 2 groups started to become
similar. By 1 year, antiplatelet therapy use was similar among patients with and without bleeding. Postdischarge
cardiology follow-up was associated with greater antiplatelet therapy use than either primary care or no clinical
follow-up.
Conclusions—Patients whose index AMI is complicated by bleeding are less likely to be treated with antiplatelet therapies
during the first 6 months after discharge. Early reassessment of antiplatelet eligibility may represent an opportunity to
reduce the long-term risk of adverse outcomes associated with bleeding. (Circulation. 2008;118:2139-2145.)
Key Words: anticoagulants 䡲 antiplatelet therapy 䡲 bleeding 䡲 myocardial infarction

A ntiplatelet and antithrombotic medications and invasive

vascular procedures are important strategies for the
management of acute myocardial infarction (AMI).1,2 Al-
Cardiology/American Heart Association guidelines.8 Al-
though discontinuing these medications in the setting of
active bleeding is logical, failure to reintroduce these second-
though these therapies have decreased the rates of mortality ary prevention medications after the bleeding event has
and subsequent cardiac events,3 they have also increased the resolved may predispose such patients to future cardiac
risk for bleeding complications.4,5 Prior studies have estab- events. To date, little is known about the rates of antiplatelet
lished an association between bleeding during AMI and agent use among AMI patients with an in-hospital bleeding
worse short- and long-term outcomes.5–7 One potential expla- event.
nation for this association may be that a bleeding event during
AMI hospitalization alters the patient’s subsequent ability Clinical Perspective p 2145
and/or likelihood of receiving secondary prevention therapies To investigate this association, we used data from the
such as antiplatelet medications at hospital discharge or multicenter Prospective Registry Evaluating Myocardial In-
during follow-up after AMI hospitalization. Discharge use of farction: Events and Recovery (PREMIER) study to compare
aspirin and clopidogrel has been demonstrated in trials to patterns of antiplatelet medication use after hospital discharge
reduce death and recurrent myocardial infarction (MI) after among AMI patients whose hospitalizations were and were
AMI and is recommended by the American College of not complicated by bleeding. We hypothesized that patients

Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
Received April 18, 2008; accepted September 12, 2008.
From the Duke Clinical Research Institute and Duke University Medical Center, Durham, NC (T.Y.W., K.P.A., S.V.R., E.D.P.); Mid America Heart
Institute and University of Missouri–Kansas City, Kansas City, Mo (L.X., M.N.K., J.A.S.); and the Denver Veterans Affairs Medical Center, Denver, Colo
(J.S.R.).
Correspondence to Tracy Y. Wang, MD, MHS, Duke Clinical Research Institute, 2400 Pratt St, Room 0311, Terrace Level, Durham, NC 27705. E-mail
wang0085@mc.duke.edu
© 2008 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.108.787143

2139
Downloaded from circ.ahajournals.org by on May 5, 2009
2140 Circulation November 18, 2008

who experienced bleeding during their index AMI hospital- The association between in-hospital bleeding and medication use
ization would be significantly less likely to be discharged on at discharge, 1 month, 6 months, and 1 year was examined separately
for aspirin, thienopyridines, ␤-blockers, and statins. For each medi-
antiplatelet therapy and would remain significantly less likely
cation, a marginal repeated-measures logistic regression model (SAS
to receive these medications over 1 year of follow-up after PROC GLIMMIX) was constructed, including age, sex, hyperten-
their bleeding complication had resolved. In contrast, we sion, diabetes, heart failure, presenting heart rate and systolic blood
hypothesized that other guidelines-recommended secondary pressure, admission creatinine clearance and hematocrit, in-hospital
prevention medications (␤-blockers and statins) would be percutaneous coronary intervention (PCI) or CABG, and insurance
status as time-dependent fixed effects and an unstructured correlation
used similarly among patients with and without a bleeding structure among the 4 time points. The effect of bleeding on
complication. medication use was estimated at each time point. In the analysis of
thienopyridine use, stent use during PCI was also added as a
Methods covariate. Variables used for adjustment were selected on the basis
of previously published factors associated with bleeding in the AMI
Study Population population,7 as well as other variables likely to factor into the
Between January 1, 2003, and June 28, 2004, 2498 patients with decision for medication use. For postdischarge medication use,
AMI from 19 US hospitals were prospectively screened and enrolled rehospitalization was added as a covariate given concerns that
into the PREMIER study. The details of the PREMIER study have subsequent cardiac or bleeding events might also influence medica-
been reported previously.9 Patients were eligible if they were aged tion use; a dummy variable for rehospitalization using a single value
ⱖ18 years and had evidence supporting the diagnosis of AMI, was used as a place holder in the longitudinal model for the discharge
including prolonged (⬎20 minutes) ischemic signs/symptoms or time point. Follow-up provider specialty (cardiologist/cardiac sur-
ECG ST-segment changes. Patients who were transferred into an geon versus internist/family practitioner versus none) was also
enrolling site ⬎24 hours from symptom onset or who developed investigated as a covariate for medication use after discharge. To
cardiac marker elevation as a complication of elective coronary account for multiple comparisons among bivariate analyses of the 4
revascularization were excluded. Institutional review board approval different medications, we adjusted the significance level from 0.05 to
at each participating hospital and individual patient informed consent 0.0197 on the basis of a Bonferroni-type correction adjusted for
were obtained. correlations among the medications.11 A ␹2 test was used to assess
the significance of the interaction between bleeding and clinical
Data Collection follow-up specialty on postdischarge medication use.
The PREMIER data collection form prospectively captured demo- To further assess the robustness of our findings, we repeated the
graphic and clinical characteristics at enrollment, in-hospital use of analysis after (1) limiting it to only patients who underwent PCI, (2)
coronary angiography and revascularization procedures, and dis- limiting it to only patients who were initially medically managed,
charge medications. Telephone interviews conducted at 1, 6, and 12 and (3) excluding all patients with an interval hospitalization as a
months collected information on current medications and any inter- surrogate for clinical stability and presumed safety of continuing or
val hospitalizations. A final query of the Social Security Adminis- reinitiating antiplatelet medications.
A probability value of ⬍0.05 was considered statistically signif-
tration Death Masterfile was conducted in August 2007 to determine
icant for all tests. All analyses were performed with the use of SAS
survival through 3 years. Of the 2498 enrolled patients, 2481 patients
software, version 9.1 (SAS Institute, Cary, NC).
were discharged alive, and follow-up was completed for 2096
The authors had full access to and take full responsibility for the
patients (83.9%) at 1 month, 2064 patients (82.6%) at 6 months, and
integrity of the data. All authors have read and agree to the
1960 patients (78.5%) at 1 year.
manuscript as written.
Data Definitions
Data for the PREMIER registry were collected with the use of a
Results
standardized set of data elements and definitions (www.cvoutcomes. Baseline Characteristics
org/groups/PREMIER%20QI%20Registry/). In-hospital bleeding Among the 2498 AMI patients enrolled in PREMIER, 301
was defined as Thrombolysis in Myocardial Infarction major or
patients (12%) experienced an in-hospital bleeding event.
minor bleeding or receipt of ⱖ1 U of allogeneic red blood cell
transfusion. Thrombolysis in Myocardial Infarction bleeding has Bleeding source was gastrointestinal in 69 patients (23%) and
been defined previously as intracranial hemorrhage, observed blood related to the cardiac catheterization site in 64 patients (21%).
loss with hemoglobin drop ⱖ3 g/dL, and unobserved blood loss with Red blood cell transfusions were given to 198 patients, with
hemoglobin drop ⱖ4 g/dL.10 To increase the specificity of this a median of 2 U transfused (interquartile range, 2 to 4 U). As
definition, patients who had a transfusion alone were not considered
shown in Table 1, patients with a bleeding event were older,
to have had a bleeding event if they had anemia at the time of
admission (hematocrit ⬍28%) or underwent coronary artery bypass more likely to be female, and more likely to have comorbidi-
grafting (CABG). Hypertension was defined as blood pressure ties such as hypertension, prior heart failure, and chronic
⬎140/90 mm Hg on at least 2 occasions or currently receiving renal insufficiency. Patients with and without bleeding events
antihypertensive pharmacological therapy. Prior heart failure refers had similar rates of in-hospital coronary revascularization
to evidence of dyspnea, fluid retention, low cardiac output secondary with either PCI or CABG (67.8% versus 71.4%; P⫽0.19).
to cardiac dysfunction, rales, jugular venous distention, or pulmo-
nary edema before the current admission. Renal insufficiency was
defined as a documented history of renal failure with baseline Use of Medications at Hospital Discharge Among
creatinine ⬎1.5 mg/dL. Patients With and Without Bleeding
As shown in Table 2, AMI patients with in-hospital bleeding
Statistical Analyses were significantly less likely to be prescribed aspirin (82.8%
The study population was stratified by the occurrence of an in- versus 93.3%; P⬍0.001), thienopyridine (58.8% versus
hospital bleeding event. We first performed a univariate comparison 67.6%; P⫽0.002), ␤-blocker (86.5% versus 91.1%; P⫽0.01),
of baseline demographic and clinical characteristics, in-hospital
procedures, and postdischarge medication use between the 2 groups. or statins (78.1% versus 83.6%; P⫽0.02) at hospital dis-
Continuous variables were compared with the use of Student t tests, charge. The difference in aspirin and thienopyridine use
and categorical variables were compared with the use of ␹2 tests. persisted after multivariable adjustment with adjusted odds
Downloaded from circ.ahajournals.org by on May 5, 2009
 Wang et al Bleeding and Antiplatelet Therapy After AMI 2141

Table 1. Baseline Demographic and Clinical Characteristics significant differences in the use of aspirin, thienopyridine,
Among Patients With and Without In-Hospital Bleeding ␤-blocker, and statin remained among patients with or with-
Patients Without Patients With out in-hospital bleeding events (Table 2). A similar pattern of
Bleeding Event Bleeding Event medication use was observed among the subset of patients
(n⫽2197) (n⫽301) P (n⫽1432) treated with percutaneous coronary stents, as well
Demographics as among patients who were initially medically managed
Age, y* 60.3⫾12.8 65.1⫾13.7 ⬍0.001
(n⫽725).
In a secondary analysis that excluded all patients with
Female sex, % 31.1 43.2 ⬍0.001
interval rehospitalization, similar trends were observed in the
Body mass index, kg/m2* 29.2⫾6.5 28.6⫾7.2 0.31
use of postdischarge antiplatelet and other secondary preven-
White race, % 74.6 70.2 0.24 tion medications. At 6 months after discharge, adjusted ORs
Medical history, % of aspirin, thienopyridine, ␤-blocker, and statin use were 0.54
Hypertension 62.7 70.4 0.01 (95% CI, 0.35 to 0.82), 1.04 (95% CI, 0.70 to 1.55), 0.94
Diabetes mellitus 28.7 29.6 0.76 (95% CI, 0.60 to 1.48), and 0.88 (95% CI, 0.58 to 1.35),
Prior MI 21.4 22.3 0.73 respectively. At 1 year, there were no significant differences
Prior PCI 17.8 18.9 0.62
in medication use among patients with and without in-
hospital bleeding.
Prior CABG 13.0 12.6 0.87
Among patients who experienced a bleeding event, those
Prior heart failure 11.2 18.3 ⬍0.001
seen in follow-up by either a cardiologist or a cardiac surgeon
Renal insufficiency† 14.9 26.6 ⬍0.001 had a higher rate of antiplatelet therapy use than those seen in
Presenting features follow-up by an internist or family practitioner (Figure 2).
Heart rate ⬎100/min, % 13.8 18.3 0.04 Patients without any clinical follow-up had the lowest rates of
Systolic blood pressure 6.5 8.6 0.17 antiplatelet therapy use regardless of bleeding status (rates of
⬍90 mm Hg, % aspirin and thienopyridine use at 1 month after discharge
ST-elevation MI, % 43.0 41.5 0.43 among patients with cardiology versus internist versus no
Initial hematocrit, %* 40.5⫾6.0 37.1⫾6.6 ⬍0.001 clinical follow-up were 76% versus 65% versus 58% [P for
Baseline creatinine 74.4⫾29.8 65.6⫾40.4 ⬍0.001
trend⫽0.04] and 66% versus 56% versus 42% [P⫽0.006],
clearance, mL/min* respectively).
In-hospital procedures, %
Diagnostic coronary 87.2 85.4 0.39
Discussion
angiography
Although bleeding in the setting of AMI has been linked with
worse long-term mortality, the explanation for this associa-
PCI 61.3 57.8 0.24
tion remains uncertain. Our study is one of the first to
PCI with stent use 57.9 53.2 0.12
examine both discharge and posthospitalization medication
CABG 11.4 12.0 0.79 regimens of patients who experienced bleeding complications
Values are expressed as percentages except where indicated. during the course of their AMI. We found that patients with
*Expressed as median⫾SD. in-hospital bleeding were less likely to be discharged on
†Renal insufficiency defined as documented history of renal failure with several evidence-based therapies, including antiplatelet ther-
baseline creatinine ⬎1.5 mg/dL.
apy. The differences in antiplatelet use persisted to 6 months
and were more apparent among those patients not receiving
ratios (OR) (95% CI) of 0.45 (0.31 to 0.64) and 0.62 (0.42 to follow-up care by a cardiology subspecialist.
0.91), respectively (Figure 1). The combination of antithrombotic medications and coro-
nary revascularization procedures has been shown to lower
Postdischarge Therapies morbidity and mortality among patients with AMI.12–16 How-
At 1 month after discharge, patients with in-hospital bleeding ever, use of these class IA recommended therapies2 comes at
remained less likely to be treated with aspirin (66.1% versus the price of increased bleeding risk. Bleeding occurs at a rate
76.8%; P⬍0.001; adjusted OR⫽0.68 [95% CI, 0.50 to 0.92]), between 0.4% and 10% among hospitalized patients with
thienopyridine (54.1% versus 60.6%; P⫽0.05; adjusted AMI depending on the population studied and the bleeding
OR⫽0.83 [95% CI, 0.59 to 1.17]), or statin (60.4% versus definition used,7,17 and transfusion use has been reported in
72.9%; P⬍0.001; adjusted OR⫽0.65 [95% CI, 0.48 to 0.87]) up to 15% of patients with acute coronary syndromes.18 Our
compared with patients who did not have a bleeding event study is consistent with these results, with ⬎12% of patients
(Table 2, Figure 1). ␤-Blocker use was similar at 1 month experiencing either in-hospital bleeding or requiring non–
between those with and without bleeding (adjusted OR⫽1.05 CABG-related transfusion.
[95% CI, 0.76 to 1.44]). Bleeding as a complication of MI care is associated with
At 6 months after discharge, aspirin use was still lower poor long-term outcomes; higher rates of mortality, recurrent
among patients who had bled previously (65.4% versus infarction, and stroke have been reported up to 6 months after
77.0%; P⬍0.001; adjusted OR⫽0.63 [95% CI, 0.46 to 0.87]). the initial bleeding event.5,7 Although it is intuitive why
However, thienopyridine, ␤-blocker, and statin use were severe bleeding, such as intracranial hemorrhage, is associ-
similar between the 2 groups (Figure 1). By 1 year, no ated with increased mortality, it is less clear why milder
Downloaded from circ.ahajournals.org by on May 5, 2009
2142 Circulation November 18, 2008

Table 2. Unadjusted Rates of Medication Use After Discharge Among Patients With and Without In-Hospital Bleeding
Overall Population Patients With Coronary Stent Patients Initially Medical Managed

Patients Without Patients With Patients Without Patients With Patients Without Patients With
Bleeding Bleeding Bleeding Bleeding Bleeding Bleeding
(n⫽2197) (n⫽301) P* (n⫽1272) (n⫽160) P* (n⫽628) (n⫽97) P*
Discharge
Aspirin 93.3% 82.8% ⬍0.001 96.5% 91.1% 0.001 85.5% 69.1% ⬍0.001
Thienopyridine 67.6% 58.8% 0.002 95.6% 90.0% 0.002 30.6% 24.7% 0.24
␤-Blocker 91.1% 86.5% 0.01 95.2% 88.4% ⬍0.001 77.1% 74.2% 0.54
Statin 83.6% 78.1% 0.02 90.6% 85.6% 0.05 69.7% 59.8% 0.05
1 Month
Aspirin 76.8% 66.1% ⬍0.001 82.0% 73.8% 0.02 63.8% 51.5% 0.05
Thienopyridine 60.6% 54.1% 0.05 82.5% 79% 0.31 27.9% 22.1% 0.31
␤-Blocker 76.1% 73.7% 0.42 81.8% 79.7% 0.55 59.1% 60.3% 0.86
Statin 72.9% 60.4% ⬍0.001 81.5% 70.3% 0.002 54.3% 35.3% 0.003
6 Months
Aspirin 77.0% 65.4% ⬍0.001 83.3% 72.9% 0.002 61.0% 51.5% 0.14
Thienopyridine 41.3% 39.0% 0.49 54.3% 56.3% 0.65 23.9% 16.2% 0.16
␤-Blocker 73.8% 72.2% 0.58 80.4% 79.6% 0.82 56.3% 57.4% 0.86
Statin 69.7% 62.9% 0.04 77.1% 69.6% 0.06 51.3% 50.0% 0.84
1 Year
Aspirin 76.1% 71.9% 0.17 80.6% 76.7% 0.28 61.8% 62.5% 0.92
Thienopyridine 33.2% 33.5% 0.93 42.3% 45.9% 0.46 20.1% 12.5% 0.18
␤-Blocker 71.7% 68.8% 0.37 77.3% 75.4% 0.63 54.2% 53.6% 0.93
Statin 70.1% 65.0% 0.12 76.3% 71.3% 0.21 52.9% 48.2% 0.51
*A P value ⬍0.0197 is considered significant on the basis of a Bonferroni adjustment accounting for correlation between the 4 medications.11

degrees of hemorrhage (so-called nuisance bleeding) would ing period and set the substrate for future thrombotic
also be independently associated with long-term adverse events.18 –23
outcomes. Consistent with previously published studies,7 we Although the aforementioned mechanisms may indeed
show that patients who bled were older and more likely to play a role, a more direct explanation might be that patients
have comorbidities such as renal insufficiency, hypertension, with bleeding are less aggressively treated with guidelines-
or heart failure, which can also contribute to their worse recommended AMI therapies. Randomized trial data to guide
long-term outcomes. Other studies have suggested that proin- treatment of AMI patients with recent bleeding are lacking.
flammatory and prothrombotic effects as a result of bleeding There is legitimate concern that patients may rebleed if
or transfusion may persist well beyond the immediate bleed- exposed to antiplatelet agents before healing from the bleed-

Figure 1. Adjusted ORs (95% CIs) for aspirin, thienopyridine, ␤-blocker, and statin use from hospital discharge through 1 year. ORs are
adjusted for age, sex, hypertension, diabetes, heart failure, presenting heart rate and systolic blood pressure, admission creatinine
clearance and hematocrit, in-hospital PCI or CABG, and insurance status. For thienopyridine use, stent use was added as a covariate.
Interval rehospitalization and follow-up provider specialty were added as covariates when medication use at 1, 6, and 12 months after
discharge was examined.

Downloaded from circ.ahajournals.org by on May 5, 2009

 Wang et al Bleeding and Antiplatelet Therapy After AMI 2143

Figure 2. Unadjusted rates of aspirin

and thienopyridine use at 1 month, 6
months, and 1 year after discharge
among bleeding and nonbleeding
patients stratified by clinical follow-up.

ing event is complete. Moreover, other secondary medica- follow-up to reinitiate medications perceived to have adverse
tions such as ␤-blockers or statins may be reasonably consequences, as well as the specialists’ greater awareness of
withheld in the setting of hemodynamic uncertainty sur- the need for therapy intensification to achieve the goals of
rounding a bleeding event. As such, it is understandable cardiovascular risk reduction.24 This highlights the impor-
that a transient bleeding episode would prevent clinicians tance of close postdischarge follow-up among patients with
from discharging patients on appropriate secondary pre- bleeding, as patients without follow-up miss the opportunity
vention medications. to be evaluated for possible reinitiation of therapy.25,26
Clinicians may also defer reinitiation of antiplatelet ther- Randomized controlled trial data on patients with recent
apy until the patient is deemed “safe” from further bleeding. bleeding are lacking, and thus the timing of reinstituting
Our data demonstrate that the treatment gap persists for quite antiplatelet therapies is often based on clinical intuition. The
some time after the initial in-hospital event: beyond 6 months, results of this study argue for further investigation addressing
for example, for aspirin. Although we were unable to deter- the safety and potential benefit of early antiplatelet therapy
mine whether patients could have been reinitiated on thera- reinitiation.12,27
pies earlier, it is interesting that patients with a bleed who
were seen in follow-up by a cardiologist were more likely to Study Limitations
be treated with antiplatelet agents than those seen in Several potential issues should be considered in the interpre-
follow-up by a primary care practitioner or those receiving tation of these results. The bleeding population included
no follow-up. This may reflect a certain inertia in clinical patients who received ⱖ1 U of red blood cell transfusion.
Downloaded from circ.ahajournals.org by on May 5, 2009
2144 Circulation November 18, 2008

Although most patients received between 2 and 4 U of red Ornato JP, Pearle DL, Sloan MA, Smith SC Jr, Alpert JS, Anderson JL,
blood cells, the relationship between the number of units Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka
LF, Hunt SA, Jacobs AK, Ornato JP. ACC/AHA guidelines for the
transfused and posthospitalization medication use could not management of patients with ST-elevation myocardial infarction: a report
be studied with this small population. Given the limited of the American College of Cardiology/American Heart Association Task
sample size and number of bleeding events, we also did not Force on Practice Guidelines (Committee to Revise the 1999 Guidelines
for the Management of Patients With Acute Myocardial Infarction). J Am
have enough power to assess how the timing of antiplatelet Coll Cardiol. 2004;44:E1–E211.
medication resumption influenced long-term outcomes. The 2. Anderson JL, Adams CD, Antman EM, Bridges CR, Califf RM, Casey
PREMIER study did not capture the detailed clinical rationale DE Jr, Chavey WE II, Fesmire FM, Hochman JS, Levin TN, Lincoff AM,
behind adjustments made to the medication regimen after Peterson ED, Theroux P, Wenger NK, Wright RS, Smith SC Jr, Jacobs
AK, Halperin JL, Hunt SA, Krumholz HM, Kushner FG, Lytle BW,
discharge, nor did it capture bleeding complications after Nishimura R, Ornato JP, Page RL, Riegel B; American College of
hospital discharge that may have occurred as a result of Cardiology; American Heart Association Task Force on Practice
resuming antiplatelet agents. Outpatient follow-up, including Guidelines (Writing Committee to Revise the 2002 Guidelines for the
Management of Patients With Unstable Angina/Non ST-Elevation Myo-
intensity of follow-up, was not prespecified; as such, conclu-
cardial Infarction); American College of Emergency Physicians; Society
sions about how follow-up influences medication selection for Cardiovascular Angiography and Interventions; Society of Thoracic
cannot be drawn. As with any other observational analysis, Surgeons; American Association of Cardiovascular and Pulmonary Reha-
the results are subject to unmeasured confounders despite bilitation; Society for Academic Emergency Medicine. ACC/AHA 2007
guidelines for the management of patients with unstable angina/non–ST-
multivariable adjustment. elevation myocardial infarction: executive summary: a report of the
Conclusions American College of Cardiology/American Heart Association Task Force
on Practice Guidelines (Writing Committee to Revise the 2002
We found that patients who bleed during their index AMI Guidelines for the Management of Patients With Unstable Angina/Non
hospitalization are less likely to be treated with antiplatelet ST-Elevation Myocardial Infarction). Circulation. 2007;116:e148 – e304.
therapies for up to 6 months after their AMI event, which may 3. Fox KA, Steg PG, Eagle KA, Goodman SG, Anderson FA Jr, Granger
be a potential contributor to the higher long-term mortality CB, Flather MD, Budaj A, Quill A, Gore JM; GRACE Investigators.
Decline in rates of death and heart failure in acute coronary syndromes,
associated with bleeding in patients with AMI. Although the 1999 –2006. JAMA. 2007;297:1892–1900.
decision to treat AMI patients with antiplatelet medications 4. Rao SV, O’Grady K, Pieper KS, Granger CB, Newby LK, Van de Werf
after bleeding is based largely on clinical intuition, continuity F, Mahaffey KW, Califf RM, Harrington RA. Impact of bleeding severity
on clinical outcomes among patients with acute coronary syndromes.
of care is critical because patients without postdischarge
Am J Cardiol. 2005;96:1200 –1206.
follow-up miss the opportunity to be evaluated for possible 5. Eikelboom JW, Mehta SR, Anand SS, Xie C, Fox KA, Yusuf S. Adverse
reinitiation of medications and may be at higher risk for impact of bleeding on prognosis in patients with acute coronary syn-
downstream events. These results should drive clinicians to dromes. Circulation. 2006;114:774 –782.
6. Manoukian SV, Feit F, Mehran R, Voeltz MD, Ebrahimi R, Hamon M,
continually assess for the opportunity to safely reinitiate these Dangas GD, Lincoff AM, White HD, Moses JW, King SB 3rd, Ohman
effective secondary prevention medications after resolution EM, Stone GW. Impact of major bleeding on 30-day mortality and
of the bleeding event. clinical outcomes in patients with acute coronary syndromes: an analysis
from the ACUITY Trial. J Am Coll Cardiol. 2007;49:1362–1368.
7. Moscucci M, Fox KA, Cannon CP, Klein W, López-Sendón J, Mon-
Sources of Funding talescot G, White K, Goldberg RJ. Predictors of major bleeding in acute
The PREMIER Quality Improvement Registry was organized by coronary syndromes: the Global Registry of Acute Coronary Events
Cardiovascular Outcomes, Inc, a nonprofit 501(c)(3) organization (GRACE). Eur Heart J. 2003;24:1815–1823.
dedicated to the advancement of outcomes research and improve- 8. Collaborative meta-analysis of randomised trials of antiplatelet therapy
ment of cardiovascular healthcare quality, and is funded by Cardio- for prevention of death, myocardial infarction, and stroke in high risk
vascular Therapeutics, Inc, Palo Alto, Calif. More information on the patients. BMJ. 2002;324:71– 86.
PREMIER Quality Improvement Registry can be found at http:// 9. Spertus J, Peterson ED, Rumsfeld JS, Jones P, Decker C, Krumholz HM.
www.cvoutcomes.org. Funding and support for this publication were The PREMIER (Prospective Registry Evaluating Myocardial Infarction:
provided by the Duke Clinical Research Institute, Durham, NC. Events and Recovery) Registry: evaluating the impact of myocardial
Funding for statistical analysis was provided by CV Therapeutics, infarction on patient outcomes. Am Heart J. 2006;151:589 –597.
Inc, Palo Alto, Calif. 10. Chesebro JH, Knatterud G, Roberts R, Borer J, Cohen LS, Dalen J, Dodge
HT, Francis CK, Hillis D, Ludbrook P. Thrombolysis in Myocardial
Infarction (TIMI) trial, phase I: a comparison between intravenous tissue
Disclosures plasminogen activator and intravenous streptokinase: clinical findings
Tracy Wang, MD, MHS reports having received research grants through hospital discharge. Circulation. 1987;76:142–154.
from The Medicines Company and Heartscape, Inc; Lan Xiao, MS: 11. Sankoh AJ, Huque MF, Dubey SD. Some comments on frequently used
no relationships to disclose; Karen Alexander, MD: speakers’ bu- multiple endpoint adjustments methods in clinical trials. Stat Med. 1997;
reaus for Amgen, Pfizer, and Schering-Plough; Sunil Rao, MD: no 16:2529 –2542.
relationships to disclose; Mikhail Kosiborod, MD: no relationships 12. Yusuf S, Zhao F, Mehta SR, Chrolavicius S, Tognoni G, Fox KK. Effects
to disclose; John Rumsfeld, MD, PhD: serves on the Scientific of clopidogrel in addition to aspirin in patients with acute coronary
Advisory Board for United Healthcare and is Chief Science Officer syndromes without ST-segment elevation. N Engl J Med. 2001;345:
for the American College of Cardiology National Cardiovascular 494 –502.
Data Registries; John Spertus, MD, MPH: no relationships to 13. Boersma E, Akkerhuis KM, Theroux P, Califf RM, Topol EJ, Simoons
disclose; Eric Peterson, MD, MPH: research grants from Bristol- ML. Platelet glycoprotein IIb/IIIa receptor inhibition in non–ST-elevation
Myers Squibb, Bristol-Myers Squibb/Sanofi-Aventis Pharmaceuti- acute coronary syndromes: early benefit during medical treatment only,
with additional protection during percutaneous coronary intervention.
cals Partnership, Bristol-Myers Squibb/Merck, and Schering-Plough
Circulation. 1999;100:2045–2048.
Corporation.
14. Kastrati A, Mehilli J, Neumann FJ, Dotzer F, ten Berg J, Bollwein H,
Graf I, Ibrahim M, Pache J, Seyfarth M, Schühlen H, Dirschinger J,
References Berger PB, Schömig A; Intracoronary Stenting and Antithrombotic:
1. Antman EM, Anbe DT, Armstrong PW, Bates ER, Green LA, Hand M, Regimen Rapid Early Action for Coronary Treatment 2 (ISAR-REACT
Hochman JS, Krumholz HM, Kushner FG, Lamas GA, Mullany CJ, 2) Trial Investigators. Abciximab in patients with acute coronary syn-

Downloaded from circ.ahajournals.org by on May 5, 2009

 Wang et al Bleeding and Antiplatelet Therapy After AMI 2145

dromes undergoing percutaneous coronary intervention after clopidogrel 19. Twomley KM, Rao SV, Becker RC. Proinflammatory, immunomodu-
pretreatment: the ISAR-REACT 2 randomized trial. JAMA. 2006;295: lating, and prothrombotic properties of anemia and red blood cell trans-
1531–1538. fusions. J Thromb Thrombolysis. 2006;21:167–174.
15. Ferguson JJ, Califf RM, Antman EM, Cohen M, Grines CL, Goodman S, 20. Pawloski JR, Stamler JS. Nitric oxide in RBCs. Transfusion. 2002;42:
Kereiakes DJ, Langer A, Mahaffey KW, Nessel CC, Armstrong PW, 1603–1609.
Avezum A, Aylward P, Becker RC, Biasucci L, Borzak S, Col J, Frey MJ, 21. Payne DA, Jones CI, Hayes PD, Webster SE, Ross Naylor A, Goodall
AH. Platelet inhibition by aspirin is diminished in patients during carotid
Fry E, Gulba DC, Guneri S, Gurfinkel E, Harrington R, Hochman JS,
surgery: a form of transient aspirin resistance? Thromb Haemost. 2004;
Kleiman NS, Leon MB, Lopez-Sendon JL, Pepine CJ, Ruzyllo W,
92:89 –96.
Steinhubl SR, Teirstein PS, Toro-Figueroa L, White H; SYNERGY Trial 22. Rao SV, Jollis JG, Harrington RA, Granger CB, Newby LK, Armstrong
Investigators. Enoxaparin vs unfractionated heparin in high-risk patients PW, Moliterno DJ, Lindblad L, Pieper K, Topol EJ, Stamler JS, Califf
with non–ST-segment elevation acute coronary syndromes managed with RM. Relationship of blood transfusion and clinical outcomes in patients
an intended early invasive strategy: primary results of the SYNERGY with acute coronary syndromes. JAMA. 2004;292:1555–1562.
randomized trial. JAMA. 2004;292:45–54. 23. Reininger CB, Reininger AJ, Steckmeier B, Greinacher A, Lasser R,
16. Mehta SR, Cannon CP, Fox KA, Wallentin L, Boden WE, Spacek R, Schweiberer L. Platelet response to vascular surgery: a preliminary study
Widimsky P, McCullough PA, Hunt D, Braunwald E, Yusuf S. Routine on the effect of aspirin and heparin. Thromb Res. 1994;76:79 – 87.
vs selective invasive strategies in patients with acute coronary syndromes: 24. Phillips LS, Branch WT, Cook CB, Doyle JP, El-Kebbi IM, Gallina DL,
a collaborative meta-analysis of randomized trials. JAMA. 2005;293: Miller CD, Ziemer DC, Barnes CS. Clinical inertia. Ann Intern Med.
2908 –2917. 2001;135:825– 834.
17. Kinnaird TD, Stabile E, Mintz GS, Lee CW, Canos DA, Gevorkian N, 25. Butler J, Arbogast PG, BeLue R, Daugherty J, Jain MK, Ray WA, Griffin
Pinnow EE, Kent KM, Pichard AD, Satler LF, Weissman NJ, Lindsay J, MR. Outpatient adherence to beta-blocker therapy after acute myocardial
infarction. J Am Coll Cardiol. 2002;40:1589 –1595.
Fuchs S. Incidence, predictors, and prognostic implications of bleeding
26. Simpson E, Beck C, Richard H, Eisenberg MJ, Pilote L. Drug pre-
and blood transfusion following percutaneous coronary interventions.
scriptions after acute myocardial infarction: dosage, compliance, and
Am J Cardiol. 2003;92:930 –935. persistence. Am Heart J. 2003;145:438 – 444.
18. Yang X, Alexander KP, Chen AY, Roe MT, Brindis RG, Rao SV, Gibler 27. Spertus JA, Kettelkamp R, Vance C, Decker C, Jones PG, Rumsfeld JS,
WB, Ohman EM, Peterson ED; CRUSADE Investigators. The impli- Messenger JC, Khanal S, Peterson ED, Bach RG, Krumholz HM, Cohen
cations of blood transfusions for patients with non–ST-segment elevation DJ. Prevalence, predictors, and outcomes of premature discontinuation of
acute coronary syndromes: results from the CRUSADE National Quality thienopyridine therapy after drug-eluting stent placement: results from the
Improvement Initiative. J Am Coll Cardiol. 2005;46:1490 –1495. PREMIER registry. Circulation. 2006;113:2803–2809.

CLINICAL PERSPECTIVE
Although bleeding in the setting of an acute myocardial infarction is known to be associated with worse long-term
outcomes, the mechanism is unclear. Our study suggests an explanation: Proven secondary prevention therapies, such as
antiplatelet agents, are withheld long beyond the resolution of the bleeding event. Postdischarge cardiology follow-up was
associated with greater antiplatelet therapy use than either primary care or no clinical follow-up. Thus, early reassessment
of antiplatelet eligibility may represent an opportunity to reduce the long-term risks associated with bleeding, and future
larger studies examining the impact of early antiplatelet medication resumption on ischemic and bleeding outcomes are
warranted.

Go to http://cme.ahajournals.org to take the CME quiz for this article.

Downloaded from circ.ahajournals.org by on May 5, 2009