fourteen-year (1987 to 2000) trends in the attack rates of, therapy for, and mortality from...

FcmAarhoBi

BHc

Ma

HAsH5H

0d

Fourteen-Year (1987 to 2000) Trends in the Attack Rates of,Therapy for, and Mortality from Non–ST-Elevation AcuteCoronary Syndromes in Four United States Communities

Stanley Watkins, MD, MHSa, David Thiemann, MDa,*, Josef Coresh, MD, PhDa,Neil Powe, MD, PhD, MBAa, Aaron R. Folsom, MDb, and Wayne Rosamond, PhDc

During the past 2 decades, randomized trials have proved the efficacy of severaltreatments for non–ST-elevation acute coronary syndromes (NSTE-ACSs), includingaspirin, � blockers, and coronary revascularization. However, the cumulative effec-tiveness of these evolving therapies in actual clinical practice remains unknown. TheAtherosclerosis Risk In Communities (ARIC) surveillance study uses rigorous pro-spective community surveillance to monitor the epidemiology of coronary heartdisease among subjects who are 35 to 74 years of age and reside in 4 United Statescommunities, with a population totaling 370,000 subjects. We identified 6,379 ARICsurveillance patients who were hospitalized with NSTE-ACS (defined as cardiacchest pain and ST depression or T-wave inversion on the presenting electrocardio-gram) between 1987 and 2000 and then analyzed 30-day and 1-year mortalities bycalendar year of admission. Using logistic regression, 30-day mortality was modeledfirst using predictor variables of the calendar year, ARIC community, and indicatorsof severity and co-morbidity and then by adding variables for treatment with aspirin,� blockers, and coronary revascularization to this model. Crude 30-day mortalitydecreased from 8.6% in 1988 to 3.6% in 2000 (p for trend <0.001), a trend thatremained significant (p � 0.006) after adjustment for case severity and co-morbidity.The trend became nonsignificant after adjustment for treatment variables, suggestingthat newer treatments may explain the improved survival. In conclusion, 30-daymortality from NSTE-ACS has decreased as treatment has improved. © 2005
Elsevier Inc. All rights reserved. (Am J Cardiol 2005;96:1349–1355)
fNattecpfcpss

M

BieUlsM

rom a clinical standpoint, acute coronary syndromes (ACSs)an be roughly divided into 2 syndromes: ST-elevationyocardial infarction and non–ST-elevation ACS (NSTE-CS), a term that subsumes NSTE myocardial infarction

nd unstable angina. Although NSTE-ACS comprisesoughly 50% of ACS incidence and mortality,1 surveillanceistorically has concentrated on ST-elevation infarction orn the general category of acute coronary artery disease.ecause NSTE-ACS differs from ST-elevation syndromes

n pathophysiology, therapy, and outcome, there is a need

The aJohns Hopkins University Bloomberg School of Public Health,altimore, Maryland; the bUniversity of Minnesota, School of Publicealth, Division of Epidemiology, Minneapolis, Minnesota; and the

University of North Carolina, Chapel Hill, Chapel Hill, North Carolina.anuscript received April 13, 2005; revised manuscript received and

ccepted July 1, 2005.Dr. Watkins was funded by Grant T32 HL07024 from the National

eart, Lung, and Blood Institute (NHLBI), Bethesda, Maryland. Thetherosclerosis Risk in Communities (ARIC) study is a collaborative

tudy supported by the NHLBI contract numbers: N01-HC-55015, N01-C-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-5021, and N01-HC-55022. Data collection was funded by NHLBI N01-C-55020.

*Corresponding author: Tel: 410-955-6558; fax: 410-955-3478.
iE-mail address: [email protected] (D. Thiemann).
002-9149/05/$ – see front matter © 2005 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2005.07.037

or descriptive epidemiology that focuses specifically onSTE-ACS.2–5 In randomized trials, aspirin, � blockers,

nd invasive treatment strategy have shown benefit for in-ermediate end points, such as recurrent myocardial infarc-ion, angina, and repeat procedures, but the cumulativeffect of these therapies on mortality due to NSTE-ACS inommunity practice remains unknown. We examined tem-oral trends in the attack rates, treatment, and mortalityrom NSTE-ACS from 1987 to 2000 in the 4 United Statesommunities of the ongoing community surveillance com-onent of the Atherosclerosis Risk In Communities (ARIC)tudy, which provided continuous and standardized hospitalurveillance throughout the study period.

ethods

etween 1987 and 2000 the ARIC study maintained ongo-ng surveillance of hospitalizations for coronary heart dis-ase in 4 demographically diverse, geographically definednited States communities (Forsyth County, North Caro-

ina; Washington County, Maryland; regions within Jack-on, Mississippi; and regions within suburban Minneapolis,innesota), with a total population of approximately 370,000

n 2000. All hospital admissions for community residents

www.AJConline.org

woCacacohwfifa

etmwvotsepb

tNwoi4e

wmr

aTutsfibvgfP

dwrhpriPvpbsiMh

in an A

1350 The American Journal of Cardiology (www.AJConline.org)

ho are 35 to 74 years of age are screened for a broad rangef hospital-determined discharge diagnosis (Internationallassification of Diseases, Ninth Revision) codes indicatingcardiac diagnosis (code 402, hypertensive heart disease;

ode 410, acute myocardial infarction; code 411, othercute/subacute coronary disease; code 412, previous myo-ardial infarction; code 413, angina pectoris; code 414,ther chronic ischemic heart disease; code 428, congestiveeart failure) and selected other codes. Different samplingeights were employed for different International Classi-cation of Diseases, Ninth Revision code groups, rangingrom 1.0 for myocardial infarction to 0.1 for heart failurend pulmonary edema.6

Trained abstractors reviewed medical records for eachligible discharge, recording demographics, history, symp-oms, cardiac enzymes, diagnostic procedures, inpatient phar-acologic treatment (excluding lipid-lowering therapy, whichas not abstracted until 1998), and revascularization. Re-ascularization procedures (coronary artery bypass graftingr percutaneous coronary intervention) were recorded ifhey were performed during the initial hospitalization orubsequent transfer. If available, 3 electrocardiograms fromach admission (initial and final electrocardiograms for hos-italization and 1 from hospital day 3) were blindly interpretedy a core laboratory according to the Minnesota Code.7

To maintain a consistent surveillance definition in an erahat spanned the introduction of troponin assays, we definedSTE-ACS as an admission for a chest pain syndrome thatas determined after inpatient evaluation to be cardiac inrigin and was accompanied by ST depression or T-wavenversion on the initial electrocardiogram (Minnesota codes-1-1 and 4-1-2 and codes 5-1 and 5-2, respectively). For

Figure 1. Age-adjusted attack rates of NSTE-ACS

ach patient, only the initial ARIC NSTE-ACS encounter t

as included. The main outcome measurement was 30-dayortality as confirmed by medical record review, state death

ecords, or the National Death Index.To adjust for co-morbidity and disease severity, we used

modified Predicting Risk of Death in Cardiac Diseaseool (PREDICT) score,8 a validated metric for ACS thatses presenting clinical variables (cardiogenic shock, his-ory of myocardial infarction or cardiac procedures, age,everity of electrocardiographic changes, congestive heartailure, and Charlson Comorbidity Index) to predict mortal-ty. We omitted renal failure from the PREDICT calculationecause serum creatinine level is not available in ARIC sur-eillance data. The linear validity of the PREDICT score wasraphically confirmed, with 30-day mortality that rangedrom 0% with a PREDICT score of 0 to 32% with aREDICT score of 17.

We hypothesized that mortality after NSTE-ACS wouldecrease with advancing calendar time and that this decreaseould coincide with the adoption of ACS therapies. Attack

ates were calculated by dividing the number of NSTE-ACSospitalizations per year by census-determined communityopulation and using direct age adjustment.9 Simple linearegression was used to test the significance of temporal trendsn attack rates, mortality, and severity (measured by modifiedREDICT score). After exploratory analysis of independentariables with correlation coefficients, box plots, and scatterlots, logistic regression was used to examine the relationetween calendar year and 30-day mortality. Logistic regres-ion was also used to model 30-day mortality using the mod-fied PREDICT score and ARIC catchment area (model 1).

odel 2 added to these covariates separate variables for in-ospital treatment with aspirin, � blockers, and revasculariza-

RIC surveillance population from 1987 to 2000.

ion (with percutaneous coronary intervention or coronary ar-

twav1wti

ew

R

O

patient

FT the Na

1351Coronary Artery Disease/Outcomes from Acute Coronary Syndromes, 1987 to 2000

ery bypass surgery). Because mortality was not purely linearith passing calendar year, the 14-year course of the study was

lso divided into 5 periods that were analyzed as indicatorariables (1987 to 1988, 1989 to 1991, 1992 to 1994, 1995 to997, and 1998 to 2000). In a third analysis each patient-yearas modeled as an indicator variable with 1994 (the year with

he lowest mortality) as the reference year. There were no

Figure 2. Inpatient therapies administered to

igure 3. Crude mortality rates for patients who had NSTE-ACS from 198hirty-day and 1-year mortalities were defined by state death records and

nteractions between independent variables in the logistic mod- N

ls. All analyses were weighted according to original samplingeights using STATA 8.10

esults

f 50,352 abstracted events from 1987 to 2000, 6,379 were

s who had NSTE-ACS from 1987 to 2000.

00. In-hospital mortality was defined by abstractions of medical records.tional Death Index. Error bars represent 95% confidence intervals.

7 to 20

STE-ACSs according to our study definition. The crude at-

twi(Ndwfozd

eoscAmuc3apir

c

a1bhqsimulmP

m(Pptfttyy0iibtdititw

atlmi

D

Tmb3tcmt1wA

Fs2


ack rate of NSTE-ACS in the ARIC communities (Figure 1)as 29.8 of 10,000 subjects who were 35 to 74 years of age

n 1987, with a gradual decrease to 22.9 of 10,000 in 2000p for trend � 0.003). Mean age of patients who hadSTE-ACS was 61.7 years and 58% were men, values thatid not change with time; the proportion of patients whoere of African-American ethnicity gradually increased

rom 13.6% in 1987 to 26.5% in 2000. Thirty-three percentf patients who had NSTE-ACS had abnormal cardiac en-ymes during their hospitalization, 21% had equivocal car-iac enzymes, and 46% had normal cardiac enzymes.

Treatment during hospitalization for NSTE-ACSvolved markedly during the study (Figure 2). Inpatient usef aspirin increased from 48% in 1987 to 87% in 1994 andubsequently remained at �90%. Beta-blocker use in-reased consistently from 43% in 1987 to 78% in 2000.lthough the use of telemetry or coronary care units re-ained relatively constant at 96% in 1987 and 99% in 2000,

se of coronary angiography and revascularization in-reased markedly, from 38% and 24% in 1987 to 54% and7% in 2000, respectively. Among patients who underwentngiography after 1996, there was an increase in the pro-ortion of patients who received percutaneous revascular-zation and a decrease in the proportion of patients whoeceived surgical revascularization.

Crude in-hospital, 30-day, and 1-year mortality de-

igure 4. Mean (A) peak creatine kinase (CK) level and (B) PREDICTcore among patients who were hospitalized for NSTE-ACS from 1987 to000. Error bars represent 95% confidence intervals.

reased from the late 1980s to the early 1990s (Figure 3) d

nd then increased somewhat after 1994, most notably in-year mortality. This trend persisted in stratified analysisy age quartiles, gender, race, center, presence of coronaryeart failure, history of myocardial infarction, PREDICTuartiles, and peak creatine kinase quartiles, all of whichhowed an initial decrease in mortality followed by a smallncrease in the late 1990s. Myocardial infarction size aseasured by peak creatine kinase decreased gradually (Fig-

re 4) from 600 IU/L in the late 1980s to �500 IU/L in theate 1990s (p for trend � 0.001), whereas combined co-orbidity and illness severity as measured by the modifiedREDICT score (Figure 4) remained essentially unchanged.

Multiple logistic regression (Table 1) showed that 30-dayortality decreased significantly with passing calendar time

odds ratio 0.75 per 5-year period, p � 0.004). ModifiedREDICT score and ARIC community also were independentredictors of 30-day mortality. The Hosmer-Lemeshow test forhis model showed good fit (p � 0.673); variance inflationactors were �1.5, thus ruling out multicollinearity. Whenreatments (aspirin administration, �-blocker administra-ion, and revascularization) were added to the logistic anal-sis (model 2, Table 1), the association between calendarear and mortality was markedly attenuated (odds ratio.91, p � 0.323), suggesting an association between evolv-ng therapy and temporal changes in 30-day mortality. Sim-lar results were seen with the outcome of 1-year mortality,ut the decrease in 1-year mortality was considerably lesshan that in 30-day mortality (odds ratio 0.85, 95% confi-ence interval 0.75 to 0.99, p � 0.047). Adjustment fornfarct size (peak creatine kinase) and race did not alterhese findings. Aspirin, � blockers, and revascularization werendividually associated with lower severity-adjusted mor-ality. In multivariate analysis only aspirin and � blockersere associated with lower severity-adjusted mortality.Adjustment for case mix (model 1) did not significantly

lter the trend in crude mortality (Figure 5). Adjustment forreatment parameters significantly lowered mortality in theate 1980s, when aspirin and �-blocker use was lowest; inore recent years, crude and adjusted mortality were nearly

dentical.

iscussion

his epidemiologic study of 4 diverse United States com-unities found that the NSTE-ACS attack rates measured

y hospitalization-based surveillance for patients who were5 to 74 years of age decreased from 29.4 of 10,000 in 1987o 25.2 of 10,000 in 1994 and has remained relativelyonstant since then. Crude and case mix–adjusted 30-dayortality decreased markedly from 1987 to 1994 and pla-

eaued or increased slightly thereafter. Crude and adjusted-year mortality followed a similar pattern (Figure 4), albeitith a more marked increase in 1-year mortality since 1994.s a result, overall 1-year mortality remained fairly stable
uring the 14-year study.

NEdAu1h

piitNvs

tment p

TA1

3

1

ated ye


There are several possible explanations for decreasingSTE-ACS attack rates, none of them entirely persuasive.ffective primary and secondary prevention of coronaryisease11 in theory may decrease the incidence of NSTE-CS, although national surveys of serum cholesterol andse of lipid-lowering therapy have shown little change from988 to 2000.12 Alternatively, a national trend toward 23-

Figure 5. Crude mortality adjusted by case mix (model 1) and trea

able 1ssociation of calendar time with mortality after non–ST-elevation acute987 to 2000

Crude p Value

0-day mortalityCalendar Year (per 5 yrs)‡ 0.75 (0.63–0.91) 0.004Period§

1987–1988 1.98 (1.28–3.06) 0.0021989–1991 1.23 (0.83–1.82) 0.3051992–1994 0.73 (0.48–1.12) 0.1491995–1997 0.96 (0.62–1.47) 0.8461998–2000 1 (reference) —

-year mortalityCalendar year (per 5 yrs)‡ 0.86 (0.74–1.01) 0.065Period§

1987–1988 1.29 (0.89–1.84) 0.1701989–1991 1.08 (0.77–1.51) 0.6631992–1994 0.58 (0.40–0.85) 0.0051995–1997 0.74 (0.52–1.06) 0.1041998–2000 1 (reference) —

* Adjusted for PREDICT score and center.† Adjusted for PREDICT score, center, �-blocker use, aspirin use, and‡ Odds ratios represent the odds of death after NSTE-ACS for a 5-year§ Odds ratios represent the odds of death after NSTE-ACS for the indic

our observation admissions and emergency room chest t

ain units (neither of which is sampled by ARIC) mayncrease outpatient management of NSTE-ACS, thus caus-ng an artifactual decrease in NSTE-ACS attack rate inhe ARIC cohort. However, outpatient management ofSTE-ACS would in all likelihood increase illness se-erity for the remaining inpatients, but severity (as mea-ured by modified PREDICT score) did not change during

arameters (model 2) using the population of 1994 as the standard.

y syndromes in the Atherosclerosis Risk In Communities Study,

Odds Ratios

Model 1* p Value Model 2† p Value

.62–0.91) 0.006 0.91 (0.74–1.10) 0.323

.37–3.29) 0.001 1.29 (0.81–2.05) 0.282

.80–1.83) 0.390 0.89 (0.56–1.39) 0.603

.48–1.16) 0.205 0.65 (0.41–1.01) 0.054

.65–1.59) 0.884 1.01 (0.64–1.57) 0.976rence) — 1 (reference) —

.73–0.99) 0.047 1.06 (0.89–1.27) 0.491

.94–2.03) 0.096 0.86 (0.57–1.29) 0.466

.76–1.56) 0.638 0.79 (0.53–1.16) 0.230

.40–0.88) 0.009 0.49 (0.33–0.74) 0.001

.53–1.13) 0.188 0.74 (0.51–1.07) 0.113rence) — 1 (reference) —

larization.ersus a 5-year earlier time.ars compared with 1998 to 2000.

coronar

0.75 (0

2.13 (11.21 (00.76 (01.01 (01 (refe

0.85 (0

1.38 (01.09 (00.59 (00.77 (01 (refe

revasculater v

he study.

oOrmair

criiws

iNohsfao

tAptpmtdmacN

tm1maaftoti

sucpf1

msN7Awcnmtmpehsdbiimbii

Ap


Trends in pharmacologic therapies that were observed inur ARIC data are similar to previously reported trends.ther studies have found similar increases in use of aspi-

in13,14 and � blockers.15,16 It is likely that many inpatientedicines were used before and/or after admission, so the

ssociation between improved outcome and medical therapyn our study may well reflect outpatient medical treatmentather than isolated inpatient efficacy.

Use of inpatient cardiac catheterization in our ARICohort increased almost 50% from 1987 to 2000, as didevascularization. Mode of revascularization changed abruptlyn the late 1990s with a decrease in bypass surgery and anncrease in percutaneous procedures, a trend that coincidesith the introduction of second-generation intracoronary

tents and matches national revascularization data.17

We found a statistically significant 47% relative decreasen 30-day mortality among patients in ARIC who hadSTE-ACS from 1987 to 2000, an annual relative decreasef 5%. This finding correlates with several other studies thatave reported relative 1.4% to 5% annual decreases inhort-term mortality for patients who had myocardial in-arction during the previous 2 decades.1,14,18–20 These par-llel trends suggest that similar factors may have improvedutcomes from ST-elevation infarction and NSTE-ACS.

Although randomized trials have found that no singleherapy has consistently decreased mortality due to NSTE-CS, many studies have shown benefit for surrogate endoints such as recurrent myocardial infarction and refrac-ory ischemia,21–23 and the cumulative effect of variousroved therapies would also be expected to affect 30-dayortality. Our analysis showed no confounding by temporal

rends in illness severity or co-morbidity, so the observedecrease in mortality is most likely due to improving treat-ent rather than to changes in case mix. However, this

nalysis includes only subjects who were hospitalized andannot be generalized to encompass overall survival fromSTE-ACS.The initial decrease in mortality from NSTE-ACS from

he late 1980s through the mid-1990s was followed byodest increases in crude and adjusted mortality in the late

990s for reasons that are unclear. This late increase inortality was seen in all centers, in both men and women,

nd across different age strata. Measured illness severitynd co-morbidity did not change with time, and adjustmentor patient variables did not alter the mortality trend. Al-hough ARIC data do not include postdischarge medicationsr compliance, newer treatments such as statins and angio-ensin-converting enzyme inhibitors would be expected toncrease long-term survival.24,25

One possible explanation for the mortality plateau is aaturation effect, if current treatments for NSTE-ACS aresed so widely that further decreases in mortality are diffi-ult to achieve. However, the most likely explanation for thelateau and for the small, unexpected increase in mortalityrom NSTE-ACS in the ARIC data in the late 1990s is the
997 replacement of creatine phosphokinase-MB as the
ain biomarker for myocyte necrosis with more sensitiveerum troponin assays. In 1996 14% of patients who hadSTE-ACS had a troponin measurement compared with6% in 1997. The effect of troponin screening on NSTE-CS epidemiology is conjectural and complicated. Patientsho have an abnormal troponin level and a normal serum

reatine phosphokinase level reportedly have a worse prog-osis than patients who have abnormal levels for thesearkers,26 although this may reflect treatment bias rather

han pathophysiology. In contrast, negative troponin assaysay allow discharge from the emergency department of

atients who do not have ACS but do have chest pain andlectrocardiographic abnormalities and might otherwiseave been included in our study, thus skewing the ARICample toward higher risk patients. This bias might not beetected by PREDICT score or other severity measurementsecause the troponin assay could be the only marker forncreased risk. In a similar fashion, increasing use of screen-ng tests, such as nuclear stress tests in emergency depart-ents, over the time course of the ARIC study may have

etter identified patients who had cardiac pain, thus increas-ng the risk burden of the study population and artifactuallyncreasing late mortality.

cknowledgment: The investigators thank the staff andarticipants of the ARIC study for important contributions.

1. Furman MI, Dauerman HL, Goldberg RJ, Yarzebski J, Lessard D,Gore JM. Twenty-two year (1975 to 1997) trends in the incidence,in-hospital and long-term case fatality rates from initial Q-wave andnon–Q-wave myocardial infarction: a multi-hospital, community-wideperspective. J Am Coll Cardiol 2001;37:1571–1580.

2. Savonitto S, Ardissino D, Granger CB, Morando G, Prando MD,Mafrici A, Cavallini C, Melandri G, Thompson TD, Vahanian A, et al.Prognostic value of the admission electrocardiogram in acute coronarysyndromes. JAMA 1999;281:707–713.

3. Roe MT, Ohman EM, Pollack CV Jr, Peterson ED, Brindis RG,Harrington RA, Christenson RH, Smith SC Jr, Califf RM, Gibler WB.Changing the model of care for patients with acute coronary syn-dromes. Am Heart J 2003;146:605–612.

4. ISIS-2 (Second International Study of Infarct Survival) CollaborativeGroup. Randomised trial of intravenous streptokinase, oral aspirin,both, or neither among 17,187 cases of suspected acute myocardialinfarction: ISIS-2. Lancet 1988;2:349–360.

5. Grines CL, Browne KF, Marco J, Rothbaum D, Stone GW, O’Keefe J,Overlie P, Donohue B, Chelliah N, Timmis GC, et al. A comparison ofimmediate angioplasty with thrombolytic therapy for acute myocardialinfarction. The Primary Angioplasty in Myocardial Infarction StudyGroup. N Engl J Med 1993;328:673–679.

6. White AD, Folsom AR, Chambless LE, Sharret AR, Yang K, ConwillD, Higgins M, Williams OD, Tyroler HA. Community surveillance ofcoronary heart disease in the Atherosclerosis Risk in Communities(ARIC) Study: methods and initial two years’ experience. J ClinEpidemiol 1996;49:223–233.

7. Prineas RJ, Crow RS, Blackburn HW. The Minnesota Code Manual ofElectrocardiographic Findings: Standards and Procedures for Measure-ment and Classification. Boston: J. Wright, 1982.

8. Jacobs DR Jr, Kroenke C, Crow R, Deshpande M, Gu DF, GatewoodL, Blackburn H. PREDICT: a simple risk score for clinical severity
and long-term prognosis after hospitalization for acute myocardial

11

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2


infarction or unstable angina: the Minnesota heart survey. Circulation1999;100:599–607.

9. Gordis L. Epidemiology. Philadelphia: W.B. Saunders, 2000.0. STATA Version 8. College Station, TX: STATA, 2002.1. Reikvam A, Kvan E, Aursnes I. Use of cardiovascular drugs after

acute myocardial infarction: a marked shift towards evidence-baseddrug therapy. Cardiovasc Drugs Ther 2002;16:451–456.

2. Ford ES, Mokdad AH, Giles WH, Mensah GA. Serum total cholesterolconcentrations and awareness, treatment, and control of hypercholes-terolemia among US adults: findings from the National Health andNutrition Examination Survey, 1999 to 2000. Circulation 2003;107:2185–2189.

3. Jackson EA, Sivasubramian R, Spencer FA, Yarzebski J, Lessard D,Gore JM, Goldberg RJ. Changes over time in the use of aspirin inpatients hospitalized with acute myocardial infarction (1975 to 1997):a population-based perspective. Am Heart J 2002;144:259–268.

4. McGovern PG, Jacobs DR Jr, Shahar E, Arnett DK, Folsom AR,Blackburn H, Luepker RV. Trends in acute coronary heart diseasemortality, morbidity, and medical care from 1985 through 1997: theMinnesota heart survey. Circulation 2001;104:19–24.

5. Burwen DR, Galusha DH, Lewis JM, Bedinger MR, Radford MJ,Krumholz HM, Foody JM. National and state trends in quality of carefor acute myocardial infarction between 1994–1995 and 1998–1999:the medicare health care quality improvement program. Arch InternMed 2003;163:1430–1439.

6. Pashos CL, Normand SL, Garfinkle JB, Newhouse JP, Epstein AM,McNeil BJ. Trends in the use of drug therapies in patients with acutemyocardial infarction: 1988 to 1992. J Am Coll Cardiol 1994;23:1023–1030.

7. Frye RL, Brooks MM, Nesto RW. Gap between clinical trials andclinical practice: lessons from the Bypass Angioplasty Revasculariza-tion Investigation (BARI). Circulation 2003;107:1837–1839.

8. Tunstall-Pedoe H, Kuulasmaa K, Mahonen M, Tolonen H, Ruokoko-ski E, Amouyel P. Contribution of trends in survival and coronary-
event rates to changes in coronary heart disease mortality: 10-year
results from 37 WHO MONICA project populations. Monitoringtrends and determinants in cardiovascular disease. Lancet 1999;353:1547–1557.

9. McGovern PG, Pankow JS, Shahar E, Doliszny KM, Folsom AR,Blackburn H, Luepker RV. Recent trends in acute coronary heart dis-ease—mortality, morbidity, medical care, and risk factors. The MinnesotaHeart Survey Investigators. N Engl J Med 1996;334:884–890.

0. Rosamond WD, Chambless LE, Folsom AR, Cooper LS, Conwill DE,Clegg L, Wang CH, Heiss G. Trends in the incidence of myocardialinfarction and in mortality due to coronary heart disease, 1987 to 1994.N Engl J Med 1998;339:861–867.

1. The PURSUIT Trial Investigators. Inhibition of platelet glycoproteinIIb/IIIa with eptifibatide in patients with acute coronary syndromes.Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppres-sion Using Integrilin Therapy. N Engl J Med 1998;339:436–443.

2. Theroux P, Ouimet H, McCans J, Latour JG, Joly P, Levy G, PelletierE, Juneau M, Stasiak J, deGuise P, et al. Aspirin, heparin, or both totreat acute unstable angina. N Engl J Med 1988;319:1105–1111.

3. Colivicchi F, Guido V, Tubaro M, Ammirati F, Montefoschi N,Varveri A, Santini M. Effects of atorvastatin 80 mg daily early afteronset of unstable angina pectoris or non–Q-wave myocardial infarc-tion. Am J Cardiol 2002;90:872–874.

4. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effectsof an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovas-cular events in high-risk patients. The Heart Outcomes PreventionEvaluation Study Investigators. N Engl J Med 2000;342:145–153.

5. Heart Protection Collaborative Study Group. MRC/BHF Heart Protec-tion Study of cholesterol lowering with simvastatin in 20,536 high-riskindividuals: a randomised placebo-controlled trial. Lancet 2002;360:7–22.

6. Meier MA, Al-Badr WH, Cooper JV, Kline-Rogers EM, Smith DE,Eagle KA, Mehta RH. The new definition of myocardial infarction:diagnostic and prognostic implications in patients with acute coronary
syndromes. Arch Intern Med 2002;162:1585–1589.

fourteen-year (1987 to 2000) trends in the attack rates of, therapy for, and mortality from...

Documents