Background Clinical trials and practice guidelines have identifiedclinical criteria for the use of coronary angiography and revascularizationprocedures after thrombolysis for acute myocardial infarction.The effect of these criteria on clinical practice has not beenextensively evaluated.
Methods We used classification-and-regression-tree (CART) andlogistic-regression models to study the patients in the firstGlobal Utilization of Streptokinase and Tissue Plasminogen Activatorfor Occluded Coronary Arteries trial, to identify the variablesthat best predicted the use of angiography and revascularizationprocedures after thrombolysis.
Results Among the 21,772 U.S. patients in the trial, 71 percentunderwent coronary angiography before discharge from the hospital.Of these, 58 percent underwent revascularization (73 percentreceiving angioplasty). The CART model for the use of angiographyshowed that age was the variable most predictive of angiography;only 53 percent of patients at least 73 years of age underwentangiography, as compared with 76 percent of those under 73.Among the older patients, age was again the most predictivefactor; among the younger patients, the availability of angioplastywas a more important predictor (67 percent of patients in hospitalswithout angioplasty facilities underwent angiography, as comparedwith 83 percent in hospitals with such facilities). The nextmost important variable was recurrent ischemia, which was morepredictive at hospitals without angioplasty facilities thanat those with them. Both statistical models identified coronaryanatomy as the most important predictor of the use and typeof revascularization.
Conclusions More patients treated with thrombolysis underwentangiography and revascularization before discharge than mightbe expected. Younger age and the availability of the proceduresappeared to be the major determinants of the use of coronaryangiography, whereas coronary anatomy largely determined theuse and type of revascularization. This process appeared toselect low-risk patients for intervention rather than thoseat higher risk, who would be the most likely to benefit.
Physicians routinely make difficult decisions about which patientswith acute myocardial infarction should undergo angiography,which to refer for revascularization, and which type of revascularizationprocedure to use.1 Although clinical trials have evaluated theeffect of different strategies on outcome, the behavior andpractice of physicians have rarely been studied in large trials.2-7In most practice guidelines, clinical risk factors are usedto identify candidates appropriate for certain treatments.8-10Whether in clinical practice these characteristics identifypatients suitable for the treatments remains unknown. The firstprospective Global Utilization of Streptokinase and Tissue PlasminogenActivator for Occluded Coronary Arteries trial (GUSTO-1) provideda large cohort of patients with acute myocardial infarction.11Because coronary angiography and revascularization were performedat the discretion of the treating physicians, this study offersan excellent opportunity to examine the use of these proceduresin current clinical practice.
Methods
Study Population
We studied the 21,772 U.S. patients enrolled in the GUSTO-1study,11 excluding 2371 patients in the angiographic substudy.12Patients with acute myocardial infarction who presented withinsix hours of the onset of pain and ST-segment elevation wereeligible. The criteria for exclusion from the study were relatedto bleeding and risks of allergy to the study treatment.
The study patients were randomly assigned to one of four thrombolyticstrategies. Because the use of cardiac procedures did not differsignificantly according to treatment assignment, the four groupswere combined for this analysis.13,14 Information on the patients'clinical characteristics, the use of angiography, and the useof revascularization procedures was collected on data forms,and information on the characteristics of the hospitals wasobtained by matching the GUSTO-1 and American Hospital Associationdata bases.15,16
Statistical Analysis
Demographic and clinical information was expressed as percentagesin the case of categorical variables and as medians in the caseof continuous variables.
Logistic-Regression Models
We used multiple logistic-regression analysis with specifiedpredictor variables to predict three binary outcomes: the in-hospitaluse of angiography, of revascularization procedures, and ofbypass surgery as compared with angioplasty. We calculated oddsratios and 95 percent confidence intervals in the case of continuousvariables by using the difference between the 25th and 75thpercentiles as the exponent of the regression coefficient.
Classification-And-Regression-Tree Models
We used the classification-and-regression-tree (CART) techniqueto construct models with the same binary outcomes as the logistic-regressionmodels. The independent predictors of angiography and revascularizationin the logistic-regression models were compared qualitativelywith those in the CART models.
CART modeling identifies variables that delineate subgroupsof patients with distinct patterns of postinfarction use ofcardiac procedures.17,18 First, the overall probability of theoutcome is estimated. Then, the overall population is dividedinto subgroups based on the categorical variables (or, in thecase of continuous variables, the categories chosen on the basisof the CART algorithm) that best reflect the probability ofthe outcome. This process continues with each subgroup untilthe added level of complexity cannot be justified by effortsat validation. Because the subgroups become smaller with eachdivision, the model becomes less stable after three or fourdivisions.
Both the logistic-regression model and the CART model are describedmore completely in material deposited with the National AuxiliaryPublications Service (*).
Independent Variables
In-hospital complications that occurred only before angiographyor revascularization were included in the model. The categoricalvariables included the Killip class (1, 2, 3, or 4) and thelocation of the infarct (inferior, anterior, or other). Continuousvariables were represented by spline functions in the regressionmodels, if appropriate.
In the models predictive of the use of revascularization andthe type of procedure used, we included angiographic informationin addition to the variables used to predict the use of angiography.A variable representing the extent of coronary disease8,9 wasadded to the CART models, but not to the regression models,to avoid collinearity with the degree of stenosis. Single-vessel(or left main) disease was considered to be present when therewas at least 70 percent stenosis of an artery; for double-vesselor triple-vessel disease to be considered present, at least50 percent stenosis of one or two additional arteries was required.
To account for missing information about non–infarct-relatedarteries, we first constructed models under the assumption thatthe missing values for all stenoses equaled zero, because clinicallysignificant stenoses were most likely to have been includedand clinically insignificant stenoses were most likely to havebeen omitted. Because the models yielded similar results whetherthey were constructed under this assumption, whether patientsfor whom there was missing information were excluded from theanalysis, whether variables (for other missing data) were imputedin the logistic regression, or whether we assigned variablesto different categories in the CART models, we have reportedmodels that are based on complete data.
Results
Overall, 15,471 U.S. patients (71 percent of the study population)underwent angiography in the hospital (Table 1). Among thesepatients, 8973 (58 percent) underwent revascularization —of this group, 2467 (27 percent) underwent bypass surgery and6506 (73 percent) underwent angioplasty.
Table 1. Base-Line Characteristics, Angiographic Characteristics, and Outcomes of the Study Patients.
Logistic-regression modeling identified several independentpredictors of the use of angiography among 18,837 patients (Figure1). The major predictors of the use of angiography (those withodds ratios of at least 1.5) were reinfarction, recurrent ischemia,acute mitral regurgitation, the availability of angiography,and the availability of bypass surgery. Cardiogenic shock, stroke,and older age predicted a decreased use of angiography (oddsratio, <0.65). Other predictors of the use of a procedurethat were included in the regression models but that did notreach this level of statistical significance were congestiveheart failure, the presence of a ventricular septal defect,diabetes, smoking, hypertension, hypercholesterolemia, familyhistory of myocardial infarction, sex, Killip class, infarctlocation, peak creatine kinase level, left ventricular ejectionfraction, time from the infarction to thrombolysis, prior infarction,prior angina, prior angioplasty, type of hospital ownership,teaching status of the hospital, hospital location, and hospitalsize. The relation between the use of angiography and age wasnonlinear, with an increasingly inverse relation in older patients(Figure 2).
Figure 2. Spline Function of the Relation between Age and the Use of Angiography.
The odds ratio represents the odds of undergoing angiography for a person of a specified age, as compared with the group median. The associations are not linear, and a variable for age that takes into account this lack of linearity was included in the regression analysis. The dashed lines indicate the 95 percent confidence interval.
The CART model indicated that age was the primary determinantof the use of angiography (Figure 3). Patients 73 years of ageor older (the age cutoff chosen by the CART algorithm) had onlya 53 percent chance of undergoing angiography, whereas thoseunder 73 had a 76 percent chance. Among the patients at least73 years old, the next division was also according to age: patients79 years old or older had only a 38 percent chance of undergoingangiography, whereas those under the age of 79 had a 60 percentchance. Among the 1003 patients who were at least 79 years old,the only additional factor that was important was again age— that is, whether they were younger than 87 years (inwhich case they had a 40 percent chance of undergoing angiography,as compared with a 14 percent chance if they were 87 or older).Among patients between the ages of 73 and 79 years, however,the next most important determinant was the presence or absencein the hospital of facilities for bypass surgery (chance ofundergoing angiography, 67 percent vs. 53 percent, respectively).
Figure 3. CART Model Showing the Variables Used to Discriminate between Subgroups According to the Likelihood of the Patients' Undergoing Angiography after Thrombolysis for Acute Myocardial Infarction.
The white area of each pie chart represents the percentage of patients in that subgroup in whom angiography was used, and the overall area of the pie chart indicates the size of the subgroup relative to the total population.
Among patients under the age of 73, the primary predictor ofthe use of angiography was the presence in the hospital of facilitiesfor angioplasty, followed by the presence of recurrent ischemia.Recurrent ischemia was more predictive of the use of angiographyat sites without facilities for angioplasty (84 percent of patientswith recurrent ischemia underwent angiography vs. 63 percentof patients without recurrent ischemia) than at sites with suchfacilities (91 percent vs. 81 percent). Among patients withrecurrent ischemia, younger age was the variable next most predictiveof angiography.
The 5668 patients without recurrent ischemia who were treatedat sites without facilities for angioplasty had a 63 percentchance of undergoing angiography. In contrast, the 7146 patientswithout recurrent ischemia who were treated at sites with angioplastyfacilities had an 81 percent chance of undergoing angiography.
Major predictors of revascularization in the logistic-regressionmodel (those with an odds ratio of at least 1.5) were reinfarction,on-site facilities for bypass surgery, emergency angiography,and the degree of stenosis, especially of the left anteriordescending artery (Figure 1). Severe stenoses of the left anteriordescending and right coronary arteries carried the highest oddsof revascularization. Patients with total occlusions were lesslikely to undergo revascularization, but the threshold for revascularizationwas lower in the case of stenoses of the left main coronaryartery. Among the predictors with odds ratios below 0.65 wereprior bypass surgery and on-site facilities for angiography.
Figure 1. Odds Ratios and 95 Percent Confidence Intervals (CI) for Factors Influencing the Use of Cardiac Procedures before Hospital Discharge in Patients with Acute Myocardial Infarction.
In the CART model, the distribution and severity of stenoses,particularly those of the infarct-related artery, were the principaldeterminants of revascularization. Only 13 percent of patientswith less than 75 percent stenosis of the infarct-related arteryunderwent revascularization, as compared with 67 percent ofpatients with stenoses of at least 75 percent. The latter groupwas subdivided further according to the degree of occlusion.In those without total occlusion (1218 patients), a greaterdegree of stenosis (>90 percent vs. <90 percent) increasedthe probability of revascularization (75 percent vs. 57 percent),but total occlusion carried a lower probability of revascularizationthan did partial occlusion (48 percent vs. 71 percent). In patientswith total occlusions, revascularization followed emergencyangiography more often than it followed elective angiography(75 percent vs. 39 percent).
In the logistic-regression model, the major independent predictorsof bypass surgery as compared with angioplasty (the variableswith odds ratios of at least 1.5) were recurrent ischemia, congestiveheart failure, and a greater degree of stenosis of the right,left circumflex, left anterior, and left main coronary arteries(Figure 1). The distribution of stenoses predictive of revascularizationgenerally predicted bypass surgery as compared with angioplasty.Predictors that had odds ratios below 0.65 were prior bypasssurgery and emergency angiography.
The classifications in the CART model that discriminated betweenpatients who underwent bypass surgery and those who underwentangioplasty mainly reflected the extent of coronary artery disease.Among patients with three-vessel or left main disease, the likelihoodof undergoing bypass surgery as compared with angioplasty was64 percent, whereas it was 16 percent among those with no morethan two diseased vessels. Patients with single-vessel diseasehad only a 5 percent chance of undergoing bypass surgery, butthe rate increased to 59 percent in those who had stenoses ofat least 46 percent in the left main coronary artery. Priorbypass surgery was the only other risk factor described in themodel; patients with previous bypass surgery were less likelythan other patients to undergo surgery.
Discussion
The majority of this large group of U.S. patients treated withthrombolysis for acute myocardial infarction underwent angiographyand revascularization before discharge, although in multipleclinical trials no advantage has been shown for an aggressiveuse of angiography (one that includes revascularization) ascompared with a conservative approach.2-7 Furthermore, youngerage and the availability of the procedure in the same hospitalwere major predictors of the use of angiography, whereas recurrentischemia, cardiogenic shock, and other high-risk characteristics— all of which are major indications for angiography inthe published guidelines — were less important. Once angiographywas performed, however, the coronary anatomy chiefly determinedthe use of revascularization and the choice of the procedure.This process appeared to decrease the chance that patients athigh risk, the group most likely to benefit from intervention,would undergo such procedures.
The large proportion of patients undergoing cardiac proceduresand the variables that predicted the use and type of the proceduresare not completely consistent with the criteria identified inpublished guidelines and clinical trials. These sources suggestthat uncomplicated infarction has an excellent prognosis regardlessof whether efforts at revascularization are made, whereas patientswith complications have the most to gain from the procedure.In our study, patients under the age of 80 who had cardiogenicshock or congestive heart failure had a likelihood of undergoingangiography similar to that of patients without these complications.This effect is tempered by a high mortality rate in this high-riskgroup. Even among survivors, however, the rate of angiographywas still higher among patients without the complications. Moreover,patients with left ventricular dysfunction and triple-vesseldisease or left main disease had a rate of revascularizationidentical to that of patients with left ventricular ejectionfractions above 50 percent, single- or double-vessel disease,and stenoses of less than 50 percent in the left anterior descendingartery.
Thus, in clinical practice, the use of angiography appears tobe influenced not only by known indicators of the need for revascularization,but also by a low-risk profile. Spertus and coworkers came toa similar conclusion in an analysis of 4823 infarct survivors.19In their study, except for recurrent ischemia, clinical variablespredictive of higher mortality were associated with a lowerlikelihood of angiography and angioplasty.
We found that age was the primary predictor of the use of cardiacprocedures in clinical practice. The relation between age andthe use of angiography was nonlinear, with a particularly steepnegative slope after the age of 73, although even among patientsless than 73 years old, younger patients were more likely toundergo angiography than older ones (Figure 2).
Age has been an independent predictor of 30-day mortality inalmost all studies of acute20,21 and chronic22,23 ischemic heartdisease.In GUSTO-1, the patients in the highest quartile ofthe age distribution (those at least 70 years old) were 3.9times more likely to die within 30 days than the patients inthe lowest quartile (those 52 years old or less).24 Furthermore,age has been shown to be directly related to the presence ofmultivessel disease. Among the U.S. patients under the age of50 in the GUSTO-1 study who underwent angiography, the incidenceof left main disease or three-vessel disease was 15 percent,as compared with 34 percent among patients between the agesof 70 and 80 years.
Recent overviews of randomized trials of bypass surgery andangioplasty show that higher-risk patients gain more from revascularizationthan lower-risk patients.25,26 Large observational studies havefound more benefit of revascularization in older patients andin those with cardiogenic shock.27,28 Since all patients inGUSTO-1 were eligible for thrombolysis, extensive coexistingconditions are not likely to explain the lower use of angiographyin the older patients.
The second determinant of the use of angiography in our studywas whether the admitting institution could perform angioplasty.Other studies have also found the presence of on-site facilitiesfor angioplasty to be an independent predictor of the use ofthat procedure.16,29,30 In this study, after adjustment forthe severity of illness, rural and teaching hospitals were lesslikely to perform angiography, whereas larger hospitals andthose owned by investors were more likely to do so. These numerouscomplex factors highlight the need for health systems that promotetreatment on the basis of the patient's needs rather than thecharacteristics of the admitting hospital.
Most guidelines for postinfarction management stress the importanceof recurrent ischemia as a criterion for the use of angiography.The emergence of recurrent ischemia as a critical factor inthis study is reassuring; however, 84 percent of patients underthe age of 73 who had recurrent ischemia underwent angiographyif the angioplasty facilities were not available in the hospital,as compared with 91 percent of such patients if the facilitieswere available on the site. Recurrent ischemia was only a minorfactor in patients over the age of 73. The results of the recentDanish Acute Myocardial Infarction Study, a multicenter, randomizedtrial of invasive as compared with conservative treatment ofpatients with inducible ischemia, suggest a better outcome amongpatients treated aggressively.31 These findings reinforce theimportance of an aggressive approach in patients with recurrentischemia.
Patients with left ventricular dysfunction derive more benefitfrom revascularization than do those with normal function.27-30,32In a systematic overview,25 patients with left ventricular dysfunctiongained 12 months of life as compared with almost 3 months forpatients with normal ventricular function. In our adjusted analysis,however, patients with heart failure, higher peak levels ofcreatine kinase, and prior infarctions were less likely to undergoangiography, whereas those with anterior infarctions were morelikely to undergo angiography.
Several groups have recommended that patients with cardiogenicshock undergo angiography,33-35 yet the presence of cardiogenicshock was associated with a substantial reduction in the useof angiography. Although this finding may be explained in partby the occurrence of early deaths before angiography could beperformed, the results are problematic; if the clinical guidelinesare to be followed, better systems are needed to ensure thatthese patients can have more rapid access to angiography facilities.
The severity and distribution of coronary-artery stenoses werethe primary determinants of the use of revascularization andthe choice of procedure. The choice of angioplasty as comparedwith bypass surgery followed the pattern of "angioplasty forone-vessel disease" and "bypass surgery for triple-vessel orleft main disease," as is consistent with current guidelines.10However, more patients underwent revascularization than wouldbe expected from the guidelines. Many patients with one- ortwo-vessel disease (without involvement of the left anteriordescending artery) underwent revascularization without havingrecurrent ischemia, even though evidence is lacking that theseprocedures affect the outcome in such patients.
These findings point to a divergence between an ideal decision-makingstrategy and the reality of clinical practice. First, practitionersappear to see younger age as a key factor, either because theymisunderstand the manner in which a likely benefit should beassessed or because they simply identify with younger patients.Second, in many cases the convenience of having facilities availableseems to outweigh the needs of the patient. Finally, practitionersappear to be selecting the patients likely to have good proceduraloutcomes rather than those who would derive the most benefitfrom a procedure. In an ideal world, clinicians would developan overall estimate of the likely outcome given one treatmentas compared with another, and would preferentially allocateresources to patients with more potential for benefit. If thishad occurred in the cases we studied, angiography would havebeen performed much more often in older patients who had evidenceof left ventricular dysfunction, and the majority of revascularizationprocedures would have been done in patients with left main orthree-vessel disease, for whom there is substantial evidenceof enhanced survival.
The two analytic methods we used provide an interesting contrastin perspectives. The CART model provides a streamlined way toidentify the major factors associated with an outcome, whereaslogistic regression allows the joint effects of multiple characteristicsto be evaluated simultaneously. In the delivery of health careand the allocation of resources, identifying large groups inwhich procedures are used in distinct patterns is a useful aidto decision making; for this purpose, the CART model providesa simple pattern that is easily grasped but that may not includeuncommon characteristics (such as cardiogenic shock and severemitral regurgitation) that are critical to individual patients.The logistic-regression model provides a more complex picturethat may be more difficult for policy makers and cliniciansto understand, because of the simultaneous interplay of multiplefactors. We believe that including both approaches providesa complementary overview of clinical practice patterns.
Many study subjects were excluded from the modeling of the useand choice of revascularization procedures because informationon their coronary anatomy was missing. Since a blank is oftenleft on the data-entry form when there is no stenosis, we repeatedour analysis using values of zero to represent the missing data;very similar results were obtained in the logistic-regressionand CART models in which patients with missing values were excluded.Furthermore, the clinical profiles and outcomes of the patientswith missing values were similar to those of the patients withoutmissing values. Results of provocative testing for ischemiawere not recorded in this trial. Despite the importance of suchtesting in the stratification of risks, preliminary findingsfrom the GUSTO-IIa study indicate that less than 30 percentof patients are tested.36 Finally, we lacked information oncoexisting conditions (other than infarct-related complicationsand cardiac risk factors), which could have affected the decisionto perform angiography or revascularization.
In conclusion, the use of cardiac procedures after thrombolysisfor acute myocardial infarction is more prevalent in U.S. clinicalpractice than would be expected from the current guidelines.Instead of the presence of high-risk clinical characteristics,which is associated with a greater benefit from revascularization,younger age appears to be the primary predictor of the use ofcoronary angiography. Aside from age, the availability of facilitiesfor cardiac procedures predominantly determines the use of angiography,although once angiography is performed, the coronary anatomylargely determines the use of revascularization and the choiceof the procedure.
We are indebted to the GUSTO-1 Steering Committee for theirreview of the manuscript, to Frank E. Harrell, Jr., Ph.D., forhelpful comments on statistical-modeling strategies, to PatWilliams for editorial comments, to Vesna Savor for technicalassistance, and to Barbara Cont for secretarial assistance.
* See NAPS document no. 05337 for 4 pages of supplementary material.To order, contact NAPS c/o Microfiche Publications, 248 HempsteadTpk., West Hempstead, NY 11552.
Source Information
From Montreal General Hospital, Montreal (L.P.); the Cleveland Clinic Foundation, Cleveland (D.P.M., J.S.R., E.J.T.); Duke University, Durham, N.C. (R.M.C.); and the University of Washington, Seattle (W.D.W.).
Address reprint requests to Dr. Topol at the Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195.
References
Topol EJ, Holmes DR, Rogers WJ. Coronary angiography after thrombolytic therapy for acute myocardial infarction. Ann Intern Med 1991;114:877-885.
Rogers WJ, Baim DS, Gore JM, et al. Comparison of immediate invasive, delayed invasive, and conservative strategies after tissue-type plasminogen activator: results of the Thrombolysis in Myocardial Infarction (TIMI) Phase II-A trial. Circulation 1990;81:1457-1476. [Free Full Text]
Barbash GI, Roth A, Hod H, et al. Randomized controlled trial of late in-hospital angiography and angioplasty versus conservative management after treatment with recombinant tissue-type plasminogen activator in acute myocardial infarction. Am J Cardiol 1990;66:538-545. [CrossRef][Medline]
The TIMI Research Group. Immediate vs delayed catheterization and angioplasty following thrombolytic therapy for acute myocardial infarction: TIMI II A results. JAMA 1988;260:2849-2858. [Abstract]
Simoons ML, Arnold AE, Betriu A, et al. Thrombolysis with tissue plasminogen activator in acute myocardial infarction: no additional benefit from immediate percutaneous coronary angioplasty. Lancet 1988;1:197-203. [Medline]
Topol EJ, Califf RM, George BS, et al. A randomized trial of immediate versus delayed elective angioplasty after intravenous tissue plasminogen activator in acute myocardial infarction. N Engl J Med 1987;317:581-588. [Abstract]
Califf RM, Topol EJ, Stack RS, et al. Evaluation of combination thrombolytic therapy and timing of cardiac catheterization in acute myocardial infarction: results of Thrombolysis and Angioplasty in Myocardial Infarction -- phase 5 randomized trial. Circulation 1991;83:1543-1556. [Free Full Text]
Gunnar RM, Bourdillon PD, Dixon DW, et al. ACC/AHA guidelines for the early management of patients with acute myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on assessment of diagnostic and therapeutic cardiovascular procedures. Circulation 1990;82:664-707. [Free Full Text]
Guidelines for coronary angiography: a report of the American College of Cardiology/American Heart Association Task Force on assessment of diagnostic and therapeutic cardiovascular procedures. Circulation 1987;76:963A-977A.
Guidelines and indications for coronary artery bypass graft surgery: a report of the American College of Cardiology/American Heart Association Task Force on assessment of diagnostic and therapeutic cardiovascular procedures. J Am Coll Cardiol 1991;17:543-589. [Medline]
The GUSTO Investigators. An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. N Engl J Med 1993;329:673-682. [Free Full Text]
The GUSTO Angiographic Investigators. The effects of tissue plasminogen activator, streptokinase, or both on coronary-artery patency, ventricular function, and survival after acute myocardial infarction. N Engl J Med 1993;329:1615-1622. [Free Full Text]
Lee KL, Califf RM, Simes J, Van de Werf F, Topol EJ. Holding GUSTO up to the light: global utilization of streptokinase and tissue plasminogen activator for occluded coronary arteries. Ann Intern Med 1994;120:876-881. [Free Full Text]
Ridker PM, O'Donnell CJ, Marder VJ, Hennekens CH. A response to "holding GUSTO up to the light." Ann Intern Med 1994;120:882-885. [Free Full Text]
American Hospital Association guide to the health care field. 1993 ed. Chicago: American Hospital Association, 1993.
Pilote L, Califf RM, Sapp S, et al. Regional variation in the United States for the management of acute myocardial infarction. N Engl J Med 1995;333:565-572. [Free Full Text]
Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and regression trees. Belmont, Calif.: Wadsworth International Group, 1984.
Chambers JM, Hastie TJ. Statistical models in S. New York: Chapman & Hall, 1993:414.
Spertus JA, Weiss NS, Every NR, Weaver WD. The influence of clinical risk factors on the use of angiography and revascularization after acute myocardial infarction. Arch Intern Med 1995;155:2309-2316. [Abstract]
Henning H, Gilpin EA, Covell JW, Swan EA, O'Rourke RA, Ross J Jr. Prognosis after acute myocardial infarction: a multivariate analysis of mortality and survival. Circulation 1979;59:1124-1136. [Free Full Text]
The International Study Group. In-hospital mortality and clinical course of 20,891 patients with suspected acute myocardial infarction randomized between alteplase and streptokinase with or without heparin. Lancet 1990;336:71-75. [CrossRef][Medline]
Emond M, Mock MB, Davis KB, et al. Long-term survival of medically treated patients in the Coronary Artery Surgery Study (CASS) Registry. Circulation 1994;90:2645-2657. [Free Full Text]
Pryor DB, Shaw LK, McCants CB, et al. Value of the history and physical in identifying patients at increased risk for coronary artery disease. Ann Intern Med 1993;118:81-90. [Free Full Text]
Lee KL, Woodlief LH, Topol EJ, et al. Predictors of 30-day mortality in the era of reperfusion for acute myocardial infarction: results from an international trial of 41,021 patients. Circulation 1995;91:1659-1668. [Free Full Text]
Yusuf S, Zucker D, Peduzzi P, et al. Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration. Lancet 1994;344:563-570. [Erratum, Lancet 1994;344:1446.] [CrossRef][Medline]
Michels KB, Yusuf S. Does PTCA in acute myocardial infarction affect mortality and reinfarction rates? A quantitative overview (meta-analysis) of the randomized clinical trials. Circulation 1995;91:476-485. [Free Full Text]
Califf RM, Harrell FE Jr, Lee KL, et al. The evolution of medical and surgical therapy for coronary artery disease: a 15-year perspective. JAMA 1989;261:2077-2086. [Abstract]
Alderman EL, Bourassa MG, Cohen LS, et al. Ten-year follow-up of survival and myocardial infarction in the randomized Coronary Artery Surgery Study. Circulation 1990;82:1629-1646. [Free Full Text]
Every NR, Larson EB, Litwin PE, et al. The association between on-site cardiac catheterization facilities and the use of coronary angiography after acute myocardial infarction. N Engl J Med 1993;329:546-551. [Free Full Text]
Blustein J. High-technology cardiac procedures: the impact of service availability on service use in New York State. JAMA 1993;270:344-349. [Abstract]
Ferguson JJ. Meeting highlights, American Heart Association 68th Scientific Sessions, Anaheim, California, November 13 to 15, 1995. Circulation 1996;93:843-846. [Free Full Text]
Detre KM, Peduzzi P, Hammermeister KE, Murphy ML, Hultgren HN, Takaro T. Five-year effect of medical and surgical therapy on resting left ventricular function in stable angina: Veterans Administration Cooperative Study. Am J Cardiol 1984;53:444-450. [CrossRef][Medline]
Holmes DR Jr, Bates ER, Kleiman NS, et al. Contemporary reperfusion therapy for cardiogenic shock: the GUSTO-I trial experience. J Am Coll Cardiol 1995;26:668-674. [Abstract]
Meyer J, Merx W, Dorr R, Lambertz H, Bethge C, Effert S. Successful treatment of acute myocardial infarction shock by combined percutaneous transluminal coronary recanalization (PTCR) and percutaneous transluminal coronary angioplasty (PTCA). Am Heart J 1982;103:132-134. [CrossRef][Medline]
Lee L, Bates ER, Pitt B, Walton JA, Laufer N, O'Neill WW. Percutaneous transluminal coronary angioplasty improves survival in acute myocardial infarction complicated by cardiogenic shock. Circulation 1988;78:1345-1351. [Free Full Text]
Granger CB, Califf RM, Armstrong PW, et al. Noninvasive testing is done only in low risk patients with unstable angina and non-Q-wave myocardial infarction (MI): results from GUSTO-IIa. J Am Coll Cardiol 1996;27:Suppl:181A-181A.abstract
Ko, D. T., Wang, Y., Alter, D. A., Curtis, J. P., Rathore, S. S., Stukel, T. A., Masoudi, F. A., Ross, J. S., Foody, J. M., Krumholz, H. M.
(2008). Regional Variation in Cardiac Catheterization Appropriateness and Baseline Risk After Acute Myocardial Infarction.. J Am Coll Cardiol
51: 716-723
[Abstract][Full Text]
Spertus, J. A., Furman, M. I.
(2007). Translating Evidence Into Practice: Are We Neglecting the Neediest?. Arch Intern Med
167: 987-988
[Full Text]
Stukel, T. A., Fisher, E. S., Wennberg, D. E., Alter, D. A., Gottlieb, D. J., Vermeulen, M. J.
(2007). Analysis of Observational Studies in the Presence of Treatment Selection Bias: Effects of Invasive Cardiac Management on AMI Survival Using Propensity Score and Instrumental Variable Methods. JAMA
297: 278-285
[Abstract][Full Text]
Stukel, T. A., Lucas, F. L., Wennberg, D. E.
(2005). Long-term Outcomes of Regional Variations in Intensity of Invasive vs Medical Management of Medicare Patients With Acute Myocardial Infarction. JAMA
293: 1329-1337
[Abstract][Full Text]
Ko, D. T., Austin, P. C., Chan, B. T. B., Tu, J. V.
(2005). Quality of Care of International and Canadian Medical Graduates in Acute Myocardial Infarction. Arch Intern Med
165: 458-463
[Abstract][Full Text]
Velazquez, E. J., Francis, G. S., Armstrong, P. W., Aylward, P. E., Diaz, R., O'Connor, C. M., White, H. D., Henis, M., Rittenhouse, L. M., Kilaru, R., Gilst, W. v., Ertl, G., Maggioni, A. P., Spac, J., Weaver, W. D., Rouleau, J.-L., McMurray, J. J.V., Pfeffer, M. A., Califf, R. M.
(2004). An international perspective on heart failure and left ventricular systolic dysfunction complicating myocardial infarction: the VALIANT registry. Eur Heart J
25: 1911-1919
[Abstract][Full Text]
Ritchie, J. L., Wolk, M. J., Hirshfeld, J. W. Jr, Messer, J. V., Peterson, E. D., Prystowsky, E. N., Gardner, T. J., Kimball, H. R., Popp, R. L., Smaha, L., Smith, S. C. Jr, Wann, L. S.
(2004). Task force 4: Appropriate clinical care and issues of "self-referral". J Am Coll Cardiol
44: 1740-1746
[Full Text]
Ritchie, J. L., Wolk, M. J., Hirshfeld, J. W. Jr, Messer, J. V., Peterson, E. D., Prystowsky, E. N., Gardner, T. J., Kimball, H. R., Popp, R. L., Smaha, L., Smith, S. C. Jr, Wann, L. S.
(2004). Task Force 4: Appropriate Clinical Care and Issues of "Self-Referral". Circulation
110: 2528-2534
[Full Text]
Brophy, J. M., Bogaty, P.
(2004). Primary Angioplasty and Thrombolysis Are Both Reasonable Options in Acute Myocardial Infarction. ANN INTERN MED
141: 292-297
[Abstract][Full Text]
Quaas, A, Curzen, N, Garratt, C
(2004). Non-clinical factors influencing the selection of patients with acute coronary syndromes for angiography. Postgrad. Med. J.
80: 411-414
[Abstract][Full Text]
Zeymer, U., Uebis, R., Vogt, A., Glunz, H.-G., Vohringer, H.-F., Harmjanz, D., Neuhaus, K.-L.
(2003). Randomized Comparison of Percutaneous Transluminal Coronary Angioplasty and Medical Therapy in Stable Survivors of Acute Myocardial Infarction With Single Vessel Disease: A Study of the Arbeitsgemeinschaft Leitende Kardiologische Krankenhausarzte. Circulation
108: 1324-1328
[Abstract][Full Text]
Petersen, L. A., Normand, S.-L. T., Leape, L. L., McNeil, B. J.
(2003). Regionalization and the Underuse of Angiography in the Veterans Affairs Health Care System as Compared with a Fee-for-Service System. NEJM
348: 2209-2217
[Abstract][Full Text]
Martin, R M, Hemingway, H, Gunnell, D, Karsch, K R, Baumbach, A, Frankel, S
(2002). Population need for coronary revascularisation: are national targets for England credible?. Heart
88: 627-633
[Abstract][Full Text]
Mehta, R. H., Criger, D. A., Granger, C. B., Pieper, K. K., Califf, R. M., Topol, E. J., Bates, E. R.
(2002). Patient outcomes after fibrinolytic therapy for acute myocardial infarction at hospitals with and without coronary revascularization capability. J Am Coll Cardiol
40: 1034-1040
[Abstract][Full Text]
Marso, S. P
(2002). Review: The pathogenesis of type 2 diabetes and cardiovascular disease. British Journal of Diabetes & Vascular Disease
2: 350-356
[Abstract]
Califf, R. M., DeMets, D. L.
(2002). Principles From Clinical Trials Relevant to Clinical Practice: Part I. Circulation
106: 1015-1021
[Full Text]
Ferreiros, E.R., Kevorkian, R., Fuselli, J.J., Guetta, J., Boissonnet, C.P., di Toro, D., Cragnolino, R., Masoli, O., Cagide, A., Krauss, J.
(2002). First national survey on management strategies in non ST-elevation acute ischaemic syndromes in Argentina. Results of the STRATEG-SIA study. Eur Heart J
23: 1021-1029
[Abstract][Full Text]
Pilote, L., Lauzon, C., Huynh, T., Dion, D., Roux, R., Racine, N., Carignan, S., Diodati, J. G., Levesque, C., Charbonneau, F., Pouliot, J., Joseph, L., Eisenberg, M. J.
(2002). Quality of Life After Acute Myocardial Infarction Among Patients Treated at Sites With and Without On-site Availability of Angiography. Arch Intern Med
162: 553-559
[Abstract][Full Text]
Llevadot, J, Giugliano, R.P, Antman, E.M, Wilcox, R.G, Gurfinkel, E.P, Henry, T, McCabe, C.H, Charlesworth, A, Thompson, S, Nicolau, J.C, Tebbe, U, Sadowski, Z, Braunwald, E, for the InTIME II Invatigators,
(2001). Availability of on-site catheterization and clinical outcomes in patients receiving fibrinolysis for ST-elevation myocardial infarction. Eur Heart J
22: 2104-2115
[Abstract]
Kennelly, C, Bowling, A
(2001). Suffering in deference: a focus group study of older cardiac patients' preferences for treatment and perceptions of risk. Qual Saf Health Care
10: i23-28
[Abstract][Full Text]
Mathew, V., Farkouh, M., Grill, D. E., Urban, L. H., Cusma, J. T., Reeder, G. S., Holmes, D. R. Jr, Gersh, B. J.
(2001). Clinical risk stratification correlates with the angiographic extent of coronary artery disease in unstable angina. J Am Coll Cardiol
37: 2053-2058
[Abstract][Full Text]
Fu, Y., Chang, W.-C., Mark, D., Califf, R. M., Mackenzie, B., Granger, C. B., Topol, E. J., Hlatky, M., Armstrong, P. W.
(2000). Canadian-American Differences in the Management of Acute Coronary Syndromes in the GUSTO IIb Trial : One-Year Follow-Up of Patients Without ST-Segment Elevation. Circulation
102: 1375-1381
[Abstract][Full Text]
Lincoff, A. M., Harrington, R. A., Califf, R. M., Hochman, J. S., Guerci, A. D., Ohman, E. M., Pepine, C. J., Kopecky, S. L., Kleiman, N. S., Pacchiana, C. M., Berdan, L. G., Kitt, M. M., Simoons, M. L., Topol, E. J.
(2000). Management of Patients With Acute Coronary Syndromes in the United States by Platelet Glycoprotein IIb/IIIa Inhibition : Insights From the Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy (PURSUIT) Trial. Circulation
102: 1093-1100
[Abstract][Full Text]
Friesinger, G. C. II, Smith, R. F.
(2000). Old age, left bundle branch block and acute myocardial infarction: a vexing and lethal combination. J Am Coll Cardiol
36: 713-716
[Full Text]
Pilote, L., Lavoie, F., Ho, V., Eisenberg, M. J.
(2000). Changes in the treatment and outcomes of acute myocardial infarction in Quebec, 1988-1995. CMAJ
163: 31-36
[Abstract][Full Text]
Dodek, A.
(2000). Acute myocardial infarction in Canada: improvement with time. CMAJ
163: 41-42
[Full Text]
Califf, R. M., Pieper, K. S., Lee, K. L., Van de Werf, F., Simes, R. J., Armstrong, P. W., Topol, E. J.
(2000). Prediction of 1-Year Survival After Thrombolysis for Acute Myocardial Infarction in the Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries Trial. Circulation
101: 2231-2238
[Abstract][Full Text]
Marrugat, J, Ferrieres, J, Masia, R, Ruidavets, J, Sala, J
(2000). Differences in use of coronary angiography and outcome of myocardial infarction in Toulouse (France) and Gerona (Spain). Eur Heart J
21: 740-746
[Abstract]
Druss, B. G., Bradford, D. W., Rosenheck, R. A., Radford, M. J., Krumholz, H. M.
(2000). Mental Disorders and Use of Cardiovascular Procedures After Myocardial Infarction. JAMA
283: 506-511
[Abstract][Full Text]
Hasdai, D., Califf, R. M., Thompson, T. D., Hochman, J. S., Ohman, E. M., Pfisterer, M., Bates, E. R., Vahanian, A., Armstrong, P. W., Criger, D. A., Topol, E. J., Holmes, D. R. Jr.
(2000). Predictors of cardiogenic shock after thrombolytic therapy for acute myocardial infarction. J Am Coll Cardiol
35: 136-143
[Abstract][Full Text]
Lupon, J., Valle, V., Marrugat, J., Elosua, R., Seres, L., Pavesi, M., Freixa, R., Sanz, G., Masia, R., Molina, L., Sala, J., Serra, J., for the R.E.S.C.A.T.E. Investigators,
(1999). Six-month outcome in unstable angina patients without previous myocardial infarction according to the use of tertiary cardiologic resources. J Am Coll Cardiol
34: 1947-1953
[Abstract][Full Text]
Bowling, A.
(1999). Ageism in cardiology. BMJ
319: 1353-1355
[Full Text]
Cohen, M. G., Granger, C. B., Ohman, E. M., Stebbins, A. L., Grinfeld, L. R., Cagide, A. M., Elizari, M. V., Betriu, A., Kong, D. F., Topol, E. J., Califf, R. M.
(1999). Outcome of hispanic patients treated with thrombolytic therapy for acute myocardial infarction: Results from the GUSTO-I and -III trials. J Am Coll Cardiol
34: 1729-1737
[Abstract][Full Text]
Gottlieb, S., Boyko, V., Harpaz, D., Hod, H., Cohen, M., Mandelzweig, L., Khoury, Z., Stern, S., Behar, S., for the Israeli Thrombolytic Survey Group,
(1999). Long-term (three-year) prognosis of patients treated with reperfusion or conservatively after acute myocardial infarction. J Am Coll Cardiol
34: 70-82
[Abstract][Full Text]
Brown, N, Melville, M, Gray, D, Young, T, Skene, A M, Wilcox, R G, Hampton, J R
(1999). Relevance of clinical trial results in myocardial infarction to medical practice: comparison of four year outcome in participants of a thrombolytic trial, patients receiving routine thrombolysis, and those deemed ineligible for thrombolysis. Heart
81: 598-602
[Abstract][Full Text]
Krumholz, H. M., Chen, J., Murillo, J. E., Cohen, D. J., Radford, M. J.
(1998). Admission to Hospitals With On-Site Cardiac Catheterization Facilities : Impact on Long-Term Costs and Outcomes. Circulation
98: 2010-2016
[Abstract][Full Text]
Dakik, H. A., Kleiman, N. S., Farmer, J. A., He, Z.-X., Wendt, J. A., Pratt, C. M., Verani, M. S., Mahmarian, J. J.
(1998). Intensive Medical Therapy Versus Coronary Angioplasty for Suppression of Myocardial Ischemia in Survivors of Acute Myocardial Infarction : A Prospective, Randomized Pilot Study. Circulation
98: 2017-2023
[Abstract][Full Text]
Ayanian, J. Z., Landrum, M. B., Normand, S.-L. T., Guadagnoli, E., McNeil, B. J.
(1998). Rating the Appropriateness of Coronary Angiography -- Do Practicing Physicians Agree with an Expert Panel and with Each Other?. NEJM
338: 1896-1904
[Abstract][Full Text]
Lange, R. A., Hillis, L. D.
(1998). Use and Overuse of Angiography and Revascularization for Acute Coronary Syndromes. NEJM
338: 1838-1839
[Full Text]
Abrams, J., Pilote, L., Califf, R. M., Topol, E. J.
(1997). Use of Coronary Angiography and Revascularization after Acute Myocardial Infarction. NEJM
336: 1024-1025
[Full Text]
Quinones, M. A.
(1997). Risk Stratification After Myocardial Infarction: Clinical Science Versus Practice Behavior. Circulation
95: 1352-1354
[Full Text]
(1996). Age Is Primary Determinant of Angiography Use. Journal Watch Cardiology
1996: 4-4
[Full Text]
Previous
