Background Patients who come to the emergency department withchest pain are a heterogeneous group. Some have ischemic heartdisease that may lead to serious complications, whereas othershave minor disorders. We performed a study to identify clinicalfactors that predict which patients will have complicationsrequiring intensive care.
Methods We first studied 10,682 patients with acute chest painat seven hospitals between 1984 and 1986 (derivation set) toidentify potential clinical predictors of the development ofmajor complications. We then validated these predictors in aseparate set of 4676 patients at one hospital between 1990 and1994 (validation set).
Results In the derivation set of patients, we identified thefollowing clinical features, which, if present in the emergencydepartment, were associated with an increased risk of complications:ST-segment elevation or Q waves on the electrocardiogram thoughtto indicate acute myocardial infarction, other electrocardiographicchanges indicating myocardial ischemia, low systolic blood pressure,pulmonary rales above the bases, or an exacerbation of knownischemic heart disease. On the basis of these criteria, thepatients in the validation set were stratified into four groups,with the risk of major complications in the first 12 hours rangingfrom 0.15 to 8 percent. After 12 hours, the probability of amajor complication could be updated on the basis of whetherthe patient had already had a complication of major severity,a complication of intermediate severity, or a myocardial infarction(independent relative risks, 18.9, 7.7, and 4.0, respectively,as compared with patients without prior complications or myocardialinfarction).
Conclusions The risk of major complications in patients withacute chest pain can be estimated on the basis of the clinicalpresentation and new clinical observations made during the hospitalcourse. These estimates of risk help in making rational decisionsabout the appropriate level of medical care for patients withacute chest pain.
Clinical factors used to predict whether patients who come toan emergency department with acute chest pain are having anacute myocardial infarction or have acute ischemic heart disease1,2,3,4,5,6,7serve as surrogate criteria for another important issue: decidingwhether a patient will benefit from hospitalization and, ifso, determining the appropriate level of care. The findingson the initial electrocardiogram recorded in the emergency departmentpredict not only the presence of acute myocardial infarction1,2,3,4,5,6but also the subsequent development of life-threatening complications.7,8,9Acute myocardial infarction develops in 75 to 86 percent ofpatients with classic electrocardiographic abnormalities,4,10but only about 50 to 65 percent of patients with acute myocardialinfarction present with these electrocardiographic findings.4
It may be appropriate to observe certain patients for 6 to 12hours without admitting them to the hospital, and certain patientswho are admitted may be moved safely from an intensive or intermediatecare unit to a general unit or even sent home within 12 to 24hours.11,12,13,14,15 Some patients with acute myocardial infarctionmay even be ready for discharge by the fourth hospital day.16
In this report from the Multicenter Chest Pain Study, we describethe derivation and validation of a straightforward clinicalapproach to assessing the risk of major complications in patientswho present to the emergency department with chest pain andupdating the risk periodically during the first 72 hours ofhospitalization.
Methods
The Multicenter Chest Pain Study2,4,12,13,17,18 assessed therisk of complications in patients who came to any of seven emergencydepartments with a primary symptom of chest pain unexplainedby obvious local trauma or abnormalities on the chest radiograph.The three university hospitals and four community hospitalsparticipating in the study and the enrollment periods were asfollows: Brigham and Women's Hospital, Boston, January 1984to November 1986; Yale–New Haven Hospital, New Haven,Connecticut, February 1984 to June 1986; University of CincinnatiHospital, Cincinnati, July 1984 to October 1986; Milford Hospital,Milford, Connecticut, June 1984 to September 1986; Danbury Hospital,Danbury, Connecticut, January 1984 to August 1986; St. Mary'sHospital, Waterbury, Connecticut, April 1984 to April 1985;and William Beaumont Hospital, Royal Oak, Michigan, June 1985to August 1985. A total of 10,682 patients were enrolled (thederivation set). An additional 4676 patients at Brigham andWomen's Hospital who met the same entry criteria were enrolledbetween July 2, 1990, and February 18, 1994 (the validationset), to test the accuracy of risk estimates based on data fromthe earlier group. The protocol was approved by the institutionalreview board at each hospital.
Admitted patients underwent serial electrocardiography and cardiac-enzymemeasurements at the discretion of their physicians. The physicianswho examined the patient in the emergency department and theresearch nurse or other personnel who occasionally transcribedpertinent data onto our standard study form did so at a timewhen they had no knowledge of the patient's subsequent course.
A patient could be included in the study for each separate visitto the emergency department, up to a total of three visits.Patients who had a cardiac arrest in the emergency departmentor who left against medical advice were excluded.
Patients were enrolled in the study only if they were admittedor if they consented to return for follow-up within 24 to 72hours, if requested. In the derivation set, which included 14,832potentially eligible patients, 2140 did not consent to participate,and 1576 had incomplete study forms or insufficient clinicaldata for a meaningful assessment of the risk of major complications(primarily because of missing or insufficiently explicit interpretationof the electrocardiogram). For 434 patients, follow-up datawere inadequate to allow the investigators, who were blindedto the initial data, to determine whether a myocardial infarctionhad occurred. The remaining 10,682 patients were included inall subsequent analyses. Of the 4637 patients (43 percent) whowere not initially admitted to the hospital, 65 percent underwentsubsequent serum cardiac-enzyme measurements or examinationsby the investigators or other physicians. In the validationset, 292 (17 percent) of the discharged patients saw a physicianwithin one month after discharge. Twenty-nine patients in thevalidation set were excluded because of insufficiently explicitinterpretation of the electrocardiogram. For all patients inboth sets, there were sufficient follow-up data from examinationsby the patients' physicians or the study investigators or fromtelephone interviews or a review of medical records by the investigatorsto determine whether complications had occurred.
Definition of Events
We recorded all major events, defined as those thought on aclinical basis to require intensive care, and all intermediateevents, defined as those that did not necessarily require intensivecare but did require careful monitoring or active treatment(Table 1). All events were defined before any analyses wereperformed with data from the validation set. Complications thatoccurred during or after cardiac surgery were excluded.
Table 1. Complications Defined as Major or Intermediate Events in Patients with Acute Chest Pain.
A myocardial infarction was diagnosed according to the study'sstandard criteria, which included an elevated serum concentrationof creatine kinase MB isoenzymes, an elevated concentrationof lactate dehydrogenase isoenzyme 1 in the absence of anotherexplanation, or new pathologic Q waves with a decrease in theamplitude of the following R wave.2,4,12,13,17 Patients withsudden cardiac arrest within 72 hours without another knowncause were considered to have a myocardial infarction. Thisdiagnosis was also made in patients who received intravenousthrombolytic agents or primary percutaneous angioplasty, ifST-segment elevation was present on the electrocardiogram andthen resolved or evolved in the next 24 hours after therapyand if the patient also had complete occlusion of the infarct-relatedartery, echocardiographic wall-motion abnormalities correspondingto the acute electrocardiographic changes, or a creatine kinaseMB level above 2.5 percent of an elevated total creatine kinaselevel that returned to a normal value. The time at which themyocardial infarction occurred was considered to be the timeat which the patient met the enzymatic criteria (in over 98percent of the patients) or other criteria (in less than 2 percent).Nine percent of the patients with myocardial infarction in thederivation set and 45 percent of those in the validation setunderwent thrombolytic therapy or primary percutaneous angioplasty.
Statistical Analysis
All analyses of data from the derivation set were performedbefore the analyses of data from the validation set. The associationbetween clinical data at the time of presentation and the riskof major events was analyzed by the chi-square test, with theappropriate degrees of freedom, for categorical variables, byStudent's t-test for continuous variables, and by a univariateCox proportional-hazards test for life-table analyses.19
Recursive-partitioning analysis was used to assess the abilityof about 50 potential predictive variables from the history,physical examination, and electrocardiogram to discriminatebetween patients with major events within 24 hours and thosewithout such events.2,4,20 For continuous variables, any cutoffpoint could theoretically be chosen, but we selected only thosecutoff points that would be easy to apply clinically.
We used the Mantel–Haenszel test to determine whetherany variables added statistically significant information acrossall strata of patients.21 On the basis of clinical judgment,we hypothesized that patients with a myocardial infarction ora prior intermediate or major event would be at increased riskfor a subsequent major event. Because of the size of the database and the large number of statistical tests performed, onlytwo-tailed P values of less than 0.01 were considered to indicatestatistical significance, except where otherwise noted.
We compared key results in the derivation set with those inthe validation set by the chi-square test for association andthe Mantel–Haenszel stratified test.21 The area underthe receiver-operating-characteristic curve, which is a measureof overall discrimination, was calculated22 and compared23 forvarious predictive models.
Results
Timing and Type of Events
Of the 10,682 patients in the derivation set, 462 had a totalof 947 major events within 72 hours after they had presentedto the emergency department (Table 2). Except for cardiac catheterizationfollowed by coronary-artery bypass surgery in patients withrecurrent ischemic pain, the likelihood of each major eventdeclined as the elapsed time after presentation to the emergencydepartment increased, with 384 major events (41 percent) occurringwithin the first 12 hours and 585 (62 percent) occurring within24 hours. Of the 462 patients in the derivation set who hadmajor events within 72 hours, 190 (41 percent) had the firstevent within 12 hours, and 287 (62 percent) within 24 hours.
Table 2. Major Events after Arrival at the Emergency Department.
Among the patients in the validation set, major events werealso more common in the first 24 hours than in any subsequent24-hour period, but the overall event rate in the first 24 hourswas lower than in the derivation set (1.8 percent vs. 2.8 percent,P<0.01), and most of the individual events were also lessfrequent. Patients in the validation set were significantlymore likely than those in the derivation set to undergo a revascularizationprocedure that met the study criteria for a major event in thefirst 24 hours and in the interval between 24 and 72 hours.
Risk of a Major Event
In the derivation set, the factors associated with the developmentof major events within 72 hours were older age, male sex, descriptionof pain as the same as that during a prior myocardial infarctionor worse than prior angina, systolic blood pressure below 110mm Hg, rales above the bases on the initial physical examination,and initial electrocardiographic changes suggestive of acutemyocardial infarction or ischemia (Table 3).
Table 3. Relative Risk of a Major Event within 72 Hours in the Derivation and Validation Sets, According to Data on Presentation, the Ultimate Diagnosis, and Discharge or Admission.
Event rates within 72 hours were much higher among the patientsin the derivation set who met the diagnostic criteria for anacute myocardial infarction (23 percent) than among those withunstable angina (5 percent) or other cardiac diagnoses (2 percent).Of the 82 patients with unstable angina in whom major eventsdeveloped, 43 (52 percent) met our criteria because they underwentrevascularization after the development of recurrent ischemicpain but did not have another major event before the procedurewas performed. Overall, 445 of the 483 patients (92 percent)with events within 72 hours were admitted by their physiciansto intensive care units.
The factors that were correlated with the development of majorevents in the derivation set were predictive in the validationset. Although the overall complication rate was lower in thevalidation set than in the derivation set, the complicationrates among the patients with myocardial infarction were similarin the two groups (22 percent in the validation set and 23 percentin the derivation set, P = 0.67) and higher among the patientswith unstable angina in the validation set (8 percent, vs. 5percent in the derivation set; P<0.001), indicating thatthe patients in the validation set were at lower risk becausefewer of them met these diagnostic criteria. Of the events inthe group of patients with unstable angina, 59 (74 percent)were procedures that followed recurrent ischemic pain.
As compared with patients in the derivation set, those in thevalidation set were significantly more likely to be admittedto an intermediate care unit (52 percent vs. 10 percent, P<0.001)and less likely to be admitted to an intensive care unit (10percent vs. 43 percent, P<0.001). The proportion of patientswith myocardial infarctions who were initially admitted to anintermediate care unit increased from 4 percent in the derivationset to 31 percent in the validation set. Along with the increaseduse of intermediate care, the frequency of documented myocardialinfarction among patients admitted to an intensive care unitrose from 31 percent in the derivation set to 50 percent inthe validation set (P<0.001), and the proportion of patientsadmitted to an intensive care unit without myocardial infarctionor unstable angina fell from 41 percent to 22 percent (P<0.001).The rate of major complications remained low (3.3 percent) amongpatients in the validation set who were admitted to an intermediatecare unit, though it was significantly higher than the ratein the derivation set (1.6 percent, P<0.01).
Multivariate Correlates of the Risk of an Event within 24 Hours
Patients were assigned to one of four groups according to therisk of an event within 24 hours, from very low (0.3 percent)to high (16 percent) (Figure 1). Five factors were used to assignpatients to a risk group: electrocardiographic evidence of anST-segment elevation or Q waves, not known to be old, in twoor more leads; ST-segment or T-wave changes, not known to beold, indicative of myocardial ischemia; pain worse than priorangina or the same as the pain associated with a prior myocardialinfarction; systolic blood pressure of less than 110 mm Hg;or rales above the bases bilaterally.
Figure 1. Derivation of the Four Initial Risk Groups on the Basis of Data Available at the Time of Presentation in the Emergency Department.
Myocardial infarction (MI) was suspected if the electrocardiogram (ECG) showed ST-segment elevation of 1 mm or more or pathologic Q waves in two or more leads, and these findings were not known to be old. Ischemia was suspected if the ECG showed ST-segment depression of 1 mm or more or T-wave inversion in two or more leads, and these findings were not known to be old. Risk factors included systolic blood pressure below 110 mm Hg, rales heard above the bases bilaterally on physical examination, and known unstable ischemic heart disease, defined as a worsening of previously stable angina, the new onset of postinfarction angina or angina after a coronary-revascularization procedure, or pain that was the same as that associated with a prior myocardial infarction. The difference between each adjacent pair of risk groups was significant (P<0.001).
The risk of complications associated with each of the four groupsidentified in the derivation set was remarkably consistent whenapplied to the validation set (Table 4). With the use of a systolicblood pressure of less than 100 mm Hg, which appeared to bea better univariate threshold for predicting complications inthe validation set, the results were similar both clinicallyand statistically.
Table 4. Rate of First Major Event According to the Level of Risk Identified in the Emergency Department.
Updated Risk of a Major Event after the Initial Presentation
Among the patients in the derivation group, the relative riskof a major event more than 12 hours after presentation was associatedwith prior events (Table 5). Each prior event remained a significantpredictor of a subsequent major event after adjustment for thepatient's characteristics at presentation. After adjustmentfor the initial risk group, their constituent factors, and aprior major or intermediate event or myocardial infarction,no other factor in Table 2 for which data were available atthe time of presentation was a significant correlate of a subsequentmajor event in any period.
Table 5. Relative Risk of a Subsequent Major Event More Than 12 Hours after Presentation, According to the Prior Event.
The influence of a prior major or intermediate event or documentedmyocardial infarction on the risk of a subsequent major eventwas generally stronger in the validation set (Table 5), especiallyafter adjustment for the patient's characteristics at presentation.
Overall Risk Assessment Based on Data at Presentation and Subsequent Events
By considering events that occurred during intervals after presentation,in addition to the data at the time of presentation, we assignedpatients to one of four risk groups that accurately predictedthe rates of major events in the validation set (Table 6). Therisk of a major event declined as the time after presentationincreased, except among the patients with a major or intermediateevent or a myocardial infarction diagnosed in a prior period.In both the derivation and the validation sets, after an initialperiod of 12 hours, the development of an event was more importantthan the patient's original risk factors in predicting the probabilityof a subsequent event.
Table 6. Rate of New Major Events, According to the Original Risk Group and as Updated on the Basis of the Occurrence of a Myocardial Infarction or Intermediate or Major Event after Admission.
The development of a major or intermediate event or a myocardialinfarction after the first 12 hours increased the area underthe receiver-operating-characteristic curve in the validationset as follows: >12 to 24 hours, from 0.77 to 0.85; >24to 48 hours, from 0.77 to 0.91; and >48 to 72 hours, from0.81 to 0.88 (P<0.001 for all three comparisons).
Discussion
Given substantial evidence from previous studies that electrocardiographicabnormalities recorded in the emergency room predict subsequentin-hospital complications, it is not surprising that in ourstudy electrocardiographic abnormalities were the most importantpredictor of major complications within the first 24 hours afterpresentation. A systolic blood pressure below 110 mm Hg, andespecially below 100 mm Hg, also predicted complications, althoughwe were not able to determine whether patients with chronicallylow blood pressure were at lower risk. Overall, our risk factorsare similar to the original Killip criteria24 and the predictorsof mortality at 30 days among hospitalized patients with acutemyocardial infarction in the era of thrombolytic therapy.25
Updating the Risk during the Hospital Course
Previous studies have concentrated on the identification ofpatients at very low risk 12 hours13 or 24 hours11,12,14,15after admission or even patients at low risk four days afteran acute myocardial infarction.6 The large number of patientsin our series permitted us to update the risk during the first72 hours on the basis of the development of major events, intermediateevents, or evidence of an acute myocardial infarction. Afterthe first 12 hours, the early hospital course was more importantthan the original presenting characteristics in predicting thesubsequent risk of complications. These observations, as wellas the specific risk probabilities, should be helpful in decisionmaking, even in the era of thrombolytic therapy.
Rational Use of Hospital Facilities
To date, decision-making aids in the management of acute chestpain have not been widely used by clinicians,26 probably becausethese aids have focused on diagnoses rather than on the needfor hospitalization or intensive care. Although there is interestin finding a way to predict the hospital course on the basisof the initial emergency department data alone,27 our findingthat the early in-hospital course is also a strong predictorof later in-hospital events confirms similar findings in otherstudies.11,12,13,16,28
There is no universal consensus on the criteria for admittingpatients with acute chest pain or for determining the optimallevel of care for those who are admitted. Several analyses suggestthat admission is appropriate for patients with a probabilityof acute myocardial infarction as low as 5 to 7 percent,29 althoughthe threshold for admission to an intensive care unit shouldprobably be higher (e.g., closer to 20 percent) among patientswho do not have other problems that require intensive care.30When beds are limited, physicians seem remarkably able to shiftpatients from intensive care units to less intensive facilitieswithout any loss of safety.31,32 Since low-risk patients ininitially stable condition who meet the diagnostic criteriafor acute myocardial infarction seem to do well in intermediatecare units, low-intensity care with continuous electrocardiographicmonitoring has been developed as a cost-effective alternativeto intensive care for low-risk patients with acute chest painwho have no other indications for admission.33
Interventions relying on conferences, discharge-planning rounds,or utilization review have reduced the length of stay for patientswith cardiac conditions in intensive care units or other settingswith monitoring,14,15,34 whereas the simple feedback of informationhas not.35 The events and complications that formed the basisfor the consensus recommendations in these intervention studieswere generally consistent with those that were empirically identifiedin the current research.14,15
There is no precise risk threshold that can uniformly dictatewhich patients should be admitted from the emergency departmentto a specific hospital unit or how long they should remain hospitalized.However, our prospectively derived and validated risk probabilitiesshould aid in decision making. There seems little doubt thatadmission to an intensive care unit is appropriate for patientsin the emergency department who are considered to be at highrisk according to our protocol and that if there is any evidenceof major or intermediate complications or an acute myocardialinfarction, such patients should probably remain in the intensivecare unit for at least two or three days. Admission to an intensivecare unit also appears to be appropriate for patients at moderaterisk, whether the assessment is based on electrocardiographiccriteria30 or on the combination of unstable angina and abnormalpump function, as manifested by relative hypotension or substantialheart failure. Any patient in whom a major event develops isat high risk for a subsequent major event, whereas a patientwith an intermediate event or evidence of an acute myocardialinfarction has about a 3.5 to 7.5 percent risk of a major eventper 24-hour period. Patients without initial complications whoare at too high a risk to be sent home but who have a low probabilityof an acute myocardial infarction or a major event can be monitoredin an intermediate care unit or in an area adjacent to or inthe emergency department.18,30,33,36 Patients who have no complicationsand no recurrent chest pain at 24 hours are also at very lowrisk. An observation period of 12 hours or less seems sufficientfor many low-risk patients.11,12,13,36,37
In our approach to risk stratification, we did not considerother information, such as echocardiographic findings or theresults of other diagnostic tests.28,37,38,39,40 Studies areneeded to determine the extent to which such additional informationcan result in a cost-effective improvement in predictions basedon simple clinical characteristics alone.
Supported in part by grants from the John A. Hartford Foundation(83102-2H), the Robert Wood Johnson Foundation (12543), andthe Agency for Health Care Policy and Research (RO1-HS06452).Dr. Lee is the recipient of an Established Investigator Award(900119) from the American Heart Association. Dr. Rouan wasa Teaching and Research Scholar of the American College of Physiciansduring the study. Dr. Johnson is the recipient of a MinorityFaculty Development Award from the Robert Wood Johnson Foundation.
We are indebted to the investigators in the Multicenter ChestPain Study, who gathered the data on the patients in the derivationset.
Source Information
From the Department of Medicine, University of California, San Francisco, School of Medicine, San Francisco (L.G.); the Division of Clinical Epidemiology (E.F.C., P.A.J., T.H.L.), Cardiovascular Division (P.A.J., T.H.L.), and Division of General Medicine (E.F.C., T.H.L.), Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston; the Department of Epidemiology, Harvard School of Public Health, Boston (E.F.C.); the Primary Care Research Unit and Department of Medicine, New York Medical College, Valhalla (D.A.B.), and the Division of General Internal Medicine, University of Cincinnati Medical Center, Cincinnati (G.W.R.).
Address reprint requests to Dr. Goldman at the Department of Medicine, University of California, San Francisco, School of Medicine, 505 Parnassus Ave., San Francisco, CA 94143.
References
Pozen MW, D'Agostino RB, Mitchell JB, et al. The usefulness of a predictive instrument to reduce inappropriate admissions to the coronary care unit. Ann Intern Med 1980;92:238-242.
Goldman L, Weinberg M, Weisberg M, et al. A computer-derived protocol to aid in the diagnosis of emergency room patients with acute chest pain. N Engl J Med 1982;307:588-596. [Abstract]
Pozen MW, D'Agostino RB, Selker HP, Sytkowski PA, Hood WB Jr. A predictive instrument to improve coronary-care-unit admission practices in acute ischemic heart disease: a prospective multicenter clinical trial. N Engl J Med 1984;310:1273-1278. [Abstract]
Goldman L, Cook EF, Brand DA, et al. A computer protocol to predict myocardial infarction in emergency department patients with chest pain. N Engl J Med 1988;318:797-803. [Abstract]
Baxt WG. Use of an artificial neural network for the diagnosis of myocardial infarction. Ann Intern Med 1991;115:843-848. [Erratum, Ann Intern Med 1992;116:94.]
Baxt WG, Skora J. Prospective validation of artificial neural network trained to identify acute myocardial infarction. Lancet 1996;347:12-15. [CrossRef][Medline]
Brush JE Jr, Brand DA, Acampora D, Chalmer B, Wackers FJ. Use of the initial electrocardiogram to predict in-hospital complications of acute myocardial infarction. N Engl J Med 1985;312:1137-1141. [Abstract]
Stark ME, Vacek JL. The initial electrocardiogram during admission for myocardial infarction: use as a predictor of clinical course and facility utilization. Arch Intern Med 1987;147:843-846. [Abstract]
Yusuf S, Pearson M, Sterry H, et al. The entry ECG in the early diagnosis and prognostic stratification of patients with suspected acute myocardial infarction. Eur Heart J 1984;5:690-696. [Free Full Text]
Rude RE, Poole WK, Muller JE, et al. Electrocardiographic and clinical criteria for recognition of acute myocardial infarction based on analysis of 3,697 patients. Am J Cardiol 1983;52:936-942. [CrossRef][Medline]
Mulley AG, Thibault GE, Hughes RA, Barnett GO, Reder VA, Sherman EL. The course of patients with suspected myocardial infarction: the identification of low-risk patients for early transfer from intensive care. N Engl J Med 1980;302:943-948. [Abstract]
Lee TH, Rouan GW, Weisberg MC, et al. Sensitivity of routine clinical criteria for diagnosing myocardial infarction within 24 hours of hospitalization. Ann Intern Med 1987;106:181-186.
Lee TH, Juarez G, Cook EF, et al. Ruling out acute myocardial infarction: a prospective multicenter validation of a 12-hour strategy for patients at low risk. N Engl J Med 1991;324:1239-1246. [Abstract]
Weingarten S, Ermann B, Bolus R, et al. Early "step-down" transfer of low-risk patients with chest pain: a controlled intervention trial. Ann Intern Med 1990;113:283-289.
Eagle KA, Mulley AG, Skates SJ, et al. Length of stay in the intensive care unit: effects of practice guidelines and feedback. JAMA 1990;264:992-997. [Abstract]
Mark DB, Sigmon K, Topol EJ, et al. Identification of acute myocardial infarction patients suitable for early hospital discharge after aggressive interventional therapy: results from the Thrombolysis and Angioplasty in Acute Myocardial Infarction Registry. Circulation 1991;83:1186-1193. [Free Full Text]
Lee TH, Rouan GW, Weisberg MC, et al. Clinical characteristics and natural history of patients with acute myocardial infarction sent home from the emergency room. Am J Cardiol 1987;60:219-224. [CrossRef][Medline]
Fiebach NH, Cook EF, Lee TH, et al. Outcomes in patients with myocardial infarction who are initially admitted to stepdown units: data from the Multicenter Chest Pain Study. Am J Med 1990;89:15-20. [CrossRef][Medline]
Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: John Wiley, 1980.
Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and regression trees. Belmont, Calif.: Wadsworth International Group, 1984.
Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:719-748.
Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29-36. [Free Full Text]
Hosmer DW Jr, Lemeshow S. Applied logistic regression. New York: John Wiley, 1989.
Killip T III, Kimball JT. Treatment of myocardial infarction in a coronary care unit: a two year experience with 250 patients. Am J Cardiol 1967;20:457-464. [CrossRef][Medline]
Newby LK, Califf RM, Guerci A, et al. Early discharge in the thrombolytic era: an analysis of criteria for uncomplicated infarction from the Global Utilization of Streptokinase and t-PA for Occluded Coronary Arteries (GUSTO) Trial. J Am Coll Cardiol 1996;27:625-632. [Abstract]
Davison G, Suchman AL, Goldstein BJ. Reducing unnecessary coronary care unit admissions: a comparison of three decision aids. J Gen Intern Med 1990;5:474-479. [Medline]
Sekler HP, Griffith JL, D'Agostino RB. A time-insensitive predictive instrument for acute myocardial infarction mortality: a multicenter study. Med Care 1991;29:1196-1211. [Medline]
Hamm CW, Ravkilde J, Gerhardt W, et al. The prognostic value of serum troponin T in unstable angina. N Engl J Med 1992;327:146-150. [Abstract]
Fineberg HV, Scadden D, Goldman L. Care of patients with a low probability of acute myocardial infarction: cost effectiveness of alternatives to coronary-care-unit admission. N Engl J Med 1984;310:1301-1307. [Abstract]
Tosteson ANA, Goldman L, Udvarhelyi IS, Lee TH. Cost-effectiveness of a coronary care unit versus a stepdown unit for emergency department patients with acute chest pain. Circulation (in press).
Singer DE, Carr PL, Mulley AG, Thibault GE. Rationing intensive care -- physician responses to a resource shortage. N Engl J Med 1983;309:1155-1160. [Abstract]
Selker HP, Griffith JL, Dorey FJ, D'Agostino RB. How do physicians adapt when the coronary care unit is full? A prospective multicenter study. JAMA 1987;257:1181-1185. [Abstract]
Gaspoz JM, Lee TH, Weinstein MC, et al. Cost-effectiveness of a new short-stay unit to "rule-out" acute myocardial infarction in low risk patients. J Am Coll Cardiol 1994;24:1249-1259. [Abstract]
Weingarten SR, Riedinger MS, Conner L, et al. Practice guidelines and reminders to reduce duration of hospital stay for patients with chest pain: an interventional trial. Ann Intern Med 1994;120:257-263. [Free Full Text]
Lee TH, Pearson SD, Johnson PA, et al. Failure of information as an intervention to modify clinical management: a time-series trial in patients with acute chest pain. Ann Intern Med 1995;122:434-437. [Free Full Text]
Gibler WB, Young GP, Hedges JR, et al. Acute myocardial infarction in chest pain patients with nondiagnostic ECGs: serial CK-MB sampling in the emergency department: the Emergency Medicine Cardiac Research Group. Ann Emerg Med 1992;331:504-512.
Puleo PR, Meyer D, Wathen C, et al. Use of a rapid assay of subforms of creatine kinase MB to diagnose or rule out acute myocardial infarction. N Engl J Med 1994;331:561-566. [Free Full Text]
Ravkilde J, Nissen H, Horder M, Thygesen K. Independent prognostic value of serum creatine kinase isoenzyme MB mass, cardiac troponin T and myosin light chain levels in suspected acute myocardial infarction: analysis of 28 months of follow-up in 196 patients. J Am Coll Cardiol 1995;25:574-581. [Abstract]
Hilton TC, Thompson RC, Williams HJ, Saylors R, Folmer H, Stowers SA. Technetium-99m sestamibi myocardial perfusion imaging in the emergency room evaluation of chest pain. J Am Coll Cardiol 1994;23:1016-1022. [Abstract]
Berning J, Launbjerg J, Appleyard M. Echocardiographic algorithms for admission and predischarge prediction of mortality in acute myocardial infarction. Am J Cardiol 1992;69:1538-1544. [Medline]
Gonnah, R, Hegazi, M O, Hmdy, I, Shenoda, M M M
(2008). Can a change in policy reduce emergency hospital admissions? Effect of admission avoidance team, guideline implementation and maximising the observation unit. Emerg. Med. J.
25: 575-578
[Abstract][Full Text]
Dodd, J. D., Kalva, S., Pena, A., Bamberg, F., Shapiro, M. D., Abbara, S., Cury, R. C., Brady, T. J., Hoffmann, U.
(2008). Emergency Cardiac CT for Suspected Acute Coronary Syndrome: Qualitative and Quantitative Assessment of Coronary, Pulmonary, and Aortic Image Quality. Am. J. Roentgenol.
191: 870-877
[Abstract][Full Text]
Hagberg, S. M., Woitalla, F., Crawford, P.
(2008). 2002 ACC/AHA Guideline Versus Clinician Judgment as Diagnostic Tests for Chest Pain. J Am Board Fam Med
21: 101-107
[Abstract][Full Text]
Ekelund, U, Forberg, J L
(2008). New methods for improved evaluation of patients with suspected acute coronary syndrome in the emergency department. Postgrad. Med. J.
84: 83-86
[Abstract][Full Text]
Ekelund, U, Forberg, J L
(2007). New methods for improved evaluation of patients with suspected acute coronary syndrome in the emergency department. Emerg. Med. J.
24: 811-814
[Abstract][Full Text]
Benezet-Mazuecos, J., Ibanez, B., Rubio, J. M., Navarro, F., Martin, E., Romero, J., Farre, J.
(2007). Utility of in-hospital cardiac remote telemetry in patients with unexplained syncope. Europace
9: 1196-1201
[Abstract][Full Text]
de Feyter, P. J., van Pelt, N.
(2007). Spiral Computed Tomography Coronary Angiography: A New Diagnostic Tool Developing Its Role in Clinical Cardiology. J Am Coll Cardiol
49: 872-874
[Full Text]
Goldstein, J. A., Gallagher, M. J., O'Neill, W. W., Ross, M. A., O'Neil, B. J., Raff, G. L.
(2007). A Randomized Controlled Trial of Multi-Slice Coronary Computed Tomography for Evaluation of Acute Chest Pain. J Am Coll Cardiol
49: 863-871
[Abstract][Full Text]
Hoffmann, U., Nagurney, J. T., Moselewski, F., Pena, A., Ferencik, M., Chae, C. U., Cury, R. C., Butler, J., Abbara, S., Brown, D. F., Manini, A., Nichols, J. H., Achenbach, S., Brady, T. J.
(2006). Coronary Multidetector Computed Tomography in the Assessment of Patients With Acute Chest Pain. Circulation
114: 2251-2260
[Abstract][Full Text]
Hoffmann, U., Pena, A. J., Moselewski, F., Ferencik, M., Abbara, S., Cury, R. C., Chae, C. U., Nagurney, J. T.
(2006). MDCT in early triage of patients with acute chest pain.. Am. J. Roentgenol.
187: 1240-1247
[Abstract][Full Text]
Hoffmann, U., Pena, A. J., Cury, R. C., Abbara, S., Ferencik, M., Moselewski, F., Siebert, U., Brady, T. J., Nagurney, J. T.
(2006). Cardiac CT in Emergency Department Patients with Acute Chest Pain.. RadioGraphics
26: 963-978
[Abstract][Full Text]
Reilly, B. M., Evans, A. T.
(2006). Translating Clinical Research into Clinical Practice: Impact of Using Prediction Rules To Make Decisions. ANN INTERN MED
144: 201-209
[Abstract][Full Text]
Kennedy, R L, Harrison, R F
(2006). Identification of patients with evolving coronary syndromes by using statistical models with data from the time of presentation. Heart
92: 183-189
[Abstract][Full Text]
Swap, C. J., Nagurney, J. T.
(2005). Value and Limitations of Chest Pain History in the Evaluation of Patients With Suspected Acute Coronary Syndromes. JAMA
294: 2623-2629
[Abstract][Full Text]
Drew, B. J., Califf, R. M., Funk, M., Kaufman, E. S., Krucoff, M. W., Laks, M. M., Macfarlane, P. W., Sommargren, C., Swiryn, S., Van Hare, G. F.
(2004). Practice Standards for Electrocardiographic Monitoring in Hospital Settings: An American Heart Association Scientific Statement From the Councils on Cardiovascular Nursing, Clinical Cardiology, and Cardiovascular Disease in the Young: Endorsed by the International Society of Computerized Electrocardiology and the American Association of Critical-Care Nurses. Circulation
110: 2721-2746
[Abstract][Full Text]
Reilly, B. M., Evans, A. T.
(2004). Assessment of Patients with Chest Pain. ANN INTERN MED
141: 325-325
[Full Text]
Crausman, R. S., Glod, D. J.
(2004). Perioperative Medical Assessment of the Podiatric Surgical Patient. J. Am. Podiatr. Med. Assoc.
94: 86-89
[Abstract][Full Text]
Goldman, L., Kirtane, A. J.
(2003). Triage of Patients with Acute Chest Pain and Possible Cardiac Ischemia: The Elusive Search for Diagnostic Perfection. ANN INTERN MED
139: 987-995
[Abstract][Full Text]
Sharma, S., Riviello, J. J., Harper, M. B., Baskin, M. N.
(2003). The Role of Emergent Neuroimaging in Children With New-Onset Afebrile Seizures. Pediatrics
111: 1-5
[Abstract][Full Text]
Amsterdam, E. A., Kirk, J. D., Diercks, D. B., Lewis, W. R., Turnipseed, S. D.
(2002). Immediate exercise testing to evaluate low-risk patients presenting to the emergency department with chest pain. J Am Coll Cardiol
40: 251-256
[Abstract][Full Text]
Reilly, B. M., Evans, A. T., Schaider, J. J., Das, K., Calvin, J. E., Moran, L. A., Roberts, R. R., Martinez, E.
(2002). Impact of a Clinical Decision Rule on Hospital Triage of Patients With Suspected Acute Cardiac Ischemia in the Emergency Department. JAMA
288: 342-350
[Abstract][Full Text]
Welch, R. D., Zalenski, R. J., Frederick, P. D., Malmgren, J. A., Compton, S., Grzybowski, M., Thomas, S., Kowalenko, T., Every, N. R., for the National Registry of Myocardial Infarction,
(2001). Prognostic Value of a Normal or Nonspecific Initial Electrocardiogram in Acute Myocardial Infarction. JAMA
286: 1977-1984
[Abstract][Full Text]
Bholasingh, R, de Winter, R J, Fischer, J C, Koster, R W, Peters, R J G, Sanders, G T
(2001). Safe discharge from the cardiac emergency room with a rapid rule-out myocardial infarction protocol using serial CK-MBmass. Heart
85: 143-148
[Abstract][Full Text]
Herren, K R, Mackway-Jones, K
(2001). Emergency management of cardiac chest pain: a review. Emerg. Med. J.
18: 6-10
[Full Text]
Katz, D. A.
(2000). Risk stratification in unstable angina: the role of clinical prediction models. J Am Coll Cardiol
36: 1809-1811
[Full Text]
Heeschen, C., Goldmann, B.U., Terres, W., Hamm, C.W.
(2000). Cardiovascular risk and therapeutic benefit of coronary interventions for patients with unstable angina according to the troponin T status. Eur Heart J
21: 1159-1166
[Abstract]
Lee, T. H., Goldman, L.
(2000). Evaluation of the Patient with Acute Chest Pain. NEJM
342: 1187-1195
[Full Text]
Ewy, G. A., Ornato, J. P.
(2000). Emergency cardiac care: introduction. J Am Coll Cardiol
35: 825-880
[Full Text]
Bachur, R., Harper, M. B.
(2000). Reevaluation of Outpatients With Streptococcus pneumoniae Bacteremia. Pediatrics
105: 502-509
[Abstract][Full Text]
Littmann, L.
(1999). Large T wave inversion and QT prolongation associated with pulmonary edema: A report of nine cases. J Am Coll Cardiol
34: 1106-1110
[Abstract][Full Text]
Savonitto, S., Ardissino, D., Granger, C. B., Morando, G., Prando, M. D., Mafrici, A., Cavallini, C., Melandri, G., Thompson, T. D., Vahanian, A., Ohman, E. M., Califf, R. M., Van de Werf, F., Topol, E. J.
(1999). Prognostic Value of the Admission Electrocardiogram in Acute Coronary Syndromes. JAMA
281: 707-713
[Abstract][Full Text]
Farkouh, M. E., Smars, P. A., Reeder, G. S., Zinsmeister, A. R., Evans, R. W., Meloy, T. D., Kopecky, S. L., Allen, M., Allison, T. G., Gibbons, R. J., Gabriel, S. E., The Chest Pain Evaluation in the Emergency Room (C,
(1998). A Clinical Trial of a Chest-Pain Observation Unit for Patients with Unstable Angina. NEJM
339: 1882-1888
[Abstract][Full Text]
Newby, L. K., Christenson, R. H., Ohman, E. M., Armstrong, P. W., Thompson, T. D., Lee, K. L., Hamm, C. W., Katus, H. A., Cianciolo, C., Granger, C. B., Topol, E. J., Califf, R. M., Investigators, f. t. G.-I.
(1998). Value of Serial Troponin T Measures for Early and Late Risk Stratification in Patients With Acute Coronary Syndromes. Circulation
98: 1853-1859
[Abstract][Full Text]
Iskander, S., Iskandrian, A. E.
(1998). Risk assessment using single-photon emission computed tomographic technetium-99m sestamibi imaging. J Am Coll Cardiol
32: 57-62
[Abstract][Full Text]
Rosenthal, G. E., Sirio, C. A., Shepardson, L. B., Harper, D. L., Rotondi, A. J., Cooper, G. S.
(1998). Use of Intensive Care Units for Patients With Low Severity of Illness. Arch Intern Med
158: 1144-1151
[Abstract][Full Text]
Hamm, C. W., Goldmann, B. U., Heeschen, C., Kreymann, G., Berger, J., Meinertz, T.
(1997). Emergency Room Triage of Patients with Acute Chest Pain by Means of Rapid Testing for Cardiac Troponin T or Troponin I. NEJM
337: 1648-1653
[Abstract][Full Text]
Previous
