Previous studies have reported a wide range of mortality rates in patients with acute myocardial infarction and left BBB (19% to 65%) [4, 6, 8-1012-15, 19] or right BBB (11% to 77%) [3-58, 9, 11-1315-17, 19]. However, although some studies suggest that left or right BBB is an independent predictor of in-hospital death in patients with acute myocardial infarction [11, 12, 19], others have found either no effect [9] or an effect dependent on age [25] or location of infarction [16, 17]. Previous studies had small numbers of patients with BBB [1, 2, 4, 5, 9-19, 25]; were done in the prethrombolytic era [1, 2, 4-68, 10, 13-18, 25]; did not compare both left and right BBB in the same population [11, 16, 17]; or did not assess the association of BBB with adverse outcomes after adjustment for potential confounding by differences in clinical characteristics, location of infarction, and treatment [1-19, 25].

To address these problems, we studied 297 832 patients with acute myocardial infarction to estimate the prevalence of left and right BBB, compare the clinical characteristics of and treatments received by patients, and assess the association of left and right BBB with in-hospital death.

Methods

Top Methods Results Discussion Author & Article Info References

Data Source and Study Sample

The National Registry of Myocardial Infarction (NRMI) 2 is a voluntary, prospective, observational registry that includes patients with acute myocardial infarction from all 50 states who were admitted to participating U.S. hospitals. The data set for our analysis included 510 044 patients from 1571 hospitals enrolled between 1 June 1994 and 30 April 1997. A trained registry coordinator at each participating hospital retrospectively reviewed charts and recorded information on individual patients onto standardized case report forms. Completed forms were processed by an independent central data collection center. Extensive electronic data checks were done; errors in the forms were resolved before patient data were included in the database [26].

We analyzed 297 832 patients who were at least 18 years of age and were admitted to participating hospitals with confirmed acute myocardial infarction, which was defined on the basis of typical symptoms and signs accompanied by 1) a total creatine kinase level or creatine kinase-MB fraction at least twice the upper limit of normal; 2) electrocardiographic evidence of acute myocardial infarction; 3) other enzymatic, scintigraphic, echocardiographic, or autopsy evidence indicating myocardial infarction; or, if the preceding three variables were unavailable, 4) a principal discharge diagnosis of acute myocardial infarction (code 410, International Classification of Diseases, 9th Revision, Clinical Modification). The diagnosis of myocardial infarction was not independently validated. Patients who were transferred to or out of a participating hospital (n = 212 212) were excluded from the cohort to ensure that data on demographic characteristics, clinical presentation, myocardial infarction care, and in-hospital death were more complete.

Measurements

The main predictor variable was the presence of left or right BBB on the first 12-lead electrocardiogram, which was ascertained from responses on case report forms describing initial electrocardiography results. Diagnosis of left or right BBB was not independently validated. Information on the timing of onset (new or old) or persistence of BBB was not collected. Data on the presence of left anterior or posterior fascicular block were unavailable. The small proportion of patients identified with both left and right BBB (0.1%; n = 410) was categorized as having left BBB for the purpose of our analysis. Our results were not significantly affected by the inclusion or exclusion of these patients, and we therefore included them in all analyses.

Other predictor variables included age, sex, ethnicity (white, black, or other/unknown), cardiovascular history (previous myocardial infarction, angina, congestive heart failure, stroke, percutaneous transluminal coronary angioplasty [PTCA], or coronary artery bypass graft surgery [CABG]), known cardiac risk factors (diabetes mellitus, hypertension, current cigarette smoking, family history of coronary heart disease, or hypercholesterolemia), chest pain at admission, interval between symptom onset and arrival at the hospital, body weight, systolic and diastolic blood pressure at admission, heart rate at admission, Killip class, and location of infarction. If more than one infarction location was noted, patients were hierarchically assigned to the category with the worst prognosis: anterior, which could include any other specified category; right ventricular involvement, which could include any other specified category except anterior; inferior, which could include any other specified category except anterior and right ventricular involvement; and other/unknown, which included all remaining categories other than anterior, right ventricular involvement, and inferior and also included patients with unspecified location of infarction and patients with nondiagnostic electrocardiograms.

The main outcome variable was in-hospital death. Secondary clinical outcome variables were in-hospital cardiovascular events other than death: development of congestive heart failure, hypotension requiring treatment, cardiogenic shock, recurrent ischemia or angina, second- or third-degree heart block, and cardiac arrest. Process-measure outcome variables included use of intravenous thrombolytic therapy, location at which thrombolytic therapy was initiated (emergency department, intensive care or cardiac care unit, or catheterization laboratory/other), time from symptom onset to initiation of thrombolytic therapy, time from arrival at the hospital to initiation of thrombolytic therapy, reason for not using thrombolytic therapy (advanced age, nondiagnostic electrocardiogram, duration of symptoms, other contraindications, or other/reason not specified), use of primary PTCA, use of initial reperfusion strategies other than PTCA (immediate CABG or intracoronary thrombolytic therapy), medical therapies given within the first 24 hours (aspirin, ß-blockers, heparin, nitroglycerin, angiotensin-converting enzyme [ACE] inhibitors, or calcium-channel blockers), elective PTCA, elective CABG, and pacemaker insertion.

Statistical Analysis

Demographic and clinical characteristics of, treatments received by, and outcomes for patients with left, right, or no BBB were compared by using the Student t-test or the Wilcoxon rank-sum test for continuous variables and the chi-square test for proportions. All pairwise comparisons were made. All analyses were planned a priori and were treated with equal importance. Only P values of 0.001 or less were considered statistically significant, and any significant differences were assessed for clinical importance. No additional adjustments were made for multiple comparisons [27].

We used logistic regression models to evaluate the independent association of left BBB and right BBB with in-hospital death; patients with no BBB and no ST-segment elevation served as the reference group. In model 1, we adjusted for 1) differences in baseline demographic and clinical variables previously shown to affect risk for in-hospital death in patients with acute myocardial infarction and 2) any other variables associated with in-hospital death on bivariate analyses with a P value less than 0.01. In model 2, we included all of the variables from model 1 and added treatment variables previously shown to be related to in-hospital death, including time-dependent use of intravenous thrombolytic therapy or primary PTCA (<6 hours, 6 to 12 hours, or >12 hours from symptom onset); use of aspirin, ß-blockers, or ACE inhibitors within the first 24 hours; and in-hospital CABG. On the basis of an a priori hypothesis, we also tested for an interaction between left or right BBB and the presence of chest pain at presentation. Only a minimal interaction was found between left BBB and chest pain on stratified analyses; therefore, results of models without interaction terms are presented. All analyses were done by using SAS statistical software, version 6.12 (SAS Institute, Inc., Cary, North Carolina).

Role of the Funding Source

Two of the authors are employees of Genentech, Inc. (South San Francisco, California). One of the authors is a consultant for Genentech, Inc., and is on the NRMI Advisory Board. The funding source sponsored the development of NRMI 2 and provided funding for data analysis. The funding source works in conjunction with an independent NRMI Advisory Board made up of nationally recognized experts in cardiology and emergency medicine. The final version of the manuscript was reviewed by a member of the NRMI Advisory Board.

Results

Top Methods Results Discussion Author & Article Info References

Prevalence of Bundle-Branch Block and Patient Characteristics

The prevalence of right BBB on the first 12-lead electrocardiogram was 6.2% (95% CI, 6.1% to 6.3%); the prevalence of left BBB was 6.7% (CI, 6.6% to 6.8%) (Table 1).

Compared with patients with no BBB, patients with right or left BBB were more likely to be older; to be white; and to have a history of infarction, angina, congestive heart failure, CABG, stroke, diabetes mellitus, and hypertension. The proportion of female patients was greater among patients with left BBB than among those with right or no BBB (P < 0.001). Patients with right or left BBB were less likely than patients with no BBB to have had PTCA, to have a family history of coronary heart disease, to have hypercholesterolemia, and to be current cigarette smokers.

Clinical Presentation

Chest pain at presentation was more common in patients with no BBB (69.8%) than in patients with right or left BBB (60.9% and 54.4%, respectively) (P < 0.001). No clinically important differences in mean systolic blood pressure at admission were seen, but the average heart rate at admission was significantly higher in patients with left BBB (97 beats/min) than in patients with right BBB (89 beats/min) (P < 0.001) or no BBB (86 beats/min) (P < 0.001). Congestive heart failure (Killip class II to IV) at presentation was documented in 56.8% of patients with left BBB compared with 36.4% of patients with right BBB (P < 0.001) and 26.2% of patients with no BBB (P < 0.001). As expected, location of infarction was more frequently reported as "unspecified" in patients with left BBB (66.0%) or right BBB (45.7%) than in patients with no BBB (37.9%) (P < 0.001 for all comparisons).

Immediate Reperfusion Therapy

Overall, patients with no BBB were 1.7-fold more likely than patients with right BBB and 4.2-fold more likely than patients with left BBB to receive thrombolytic therapy (P < 0.001 for all comparisons) (Table 2). When analysis was restricted to patients for whom thrombolytic therapy was clearly indicated (patients < 75 years of age with chest pain of 12 hours' duration who had no recorded contraindication to this therapy [plus ST-segment elevation for patients with no BBB]), only 19.7% of those with no BBB, 21.2% of those with right BBB, and 15.1% of those with left BBB qualified for analysis. Of these patients, fewer of those with right or left BBB (32.0% and 16.6%, respectively) than those with no BBB (65.5%) received thrombolytic therapy (P < 0.001 for all comparisons), and the median time from hospital arrival to initiation of thrombolytic therapy was longer for patients with right or left BBB than for patients with no BBB. The observed underuse of thrombolytic therapy in patients with right or left BBB was not explained by an increased use of other initial reperfusion strategies. Any initial reperfusion strategy (thrombolytic therapy, primary PTCA, or immediate CABG) was used substantially less in patients with right or left BBB (44.6% and 23.6%, respectively) than in patients with no BBB (84.8%) for whom it was indicated (P < 0.001 for all comparisons).

Other Medical Therapies and Procedures

Patients with right or left BBB were also much less likely than patients with no BBB to receive aspirin, ß-blockers, heparin, and intravenous nitroglycerin within the first 24 hours (Table 2). However, patients with right or left BBB were more likely to receive ACE inhibitors, and use of calcium-channel blockers was similar in all groups. Elective PTCA (used in 11.8% of patients with no BBB, 7.4% of patients with right BBB, and 4.6% of patients with left BBB [P < 0.001]) and elective CABG (used in 7.8% of patients with no BBB, 5.6% of patients with right BBB, and 3.7% of patients with left BBB [P < 0.001]) were used less often during the index hospitalization in those with right or left BBB. Only minimal differences were seen in the use of pacemakers among all groups.

In-Hospital Death and Cardiovascular Events

In patients with right or left BBB, the unadjusted risk for in-hospital death was similar (23.0% compared with 22.6%; P > 0.2) but was almost twofold greater than that in patients with no BBB (13.1%) (P < 0.001 for both comparisons) (Table 3).

Patients with right or left BBB were also more likely than patients with no BBB to have their hospital course complicated by congestive heart failure, hypotension requiring treatment, cardiogenic shock, and cardiac arrest. Patients with left BBB were one third more likely than patients with right BBB to receive a diagnosis of congestive heart failure during hospitalization (P < 0.001). No clinically significant differences in the development of recurrent ischemia or angina were seen between patients with and patients without BBB.

The development of second- or third-degree heart block was slightly more common in patients with right BBB than in patients with left or no BBB. However, complete heart block as the cause of death was rare; it was seen in 0.3% of patients with no BBB, 0.7% of patients with right BBB, and 0.5% of patients with left BBB.

Association of Bundle-Branch Block with Death

After we controlled for differences in demographic and clinical characteristics (model 1), right BBB was associated with a 64% increase (CI, 57% to 71%) and left BBB was associated with a 33% increase (CI, 28% to 38%) in the odds of dying in the hospital; this estimate for left BBB was similar to the risk for in-hospital death seen in patients with ST-segment elevation and no BBB (35% [CI, 31% to 39%]) (Table 4). After additional adjustment for differences in use of immediate reperfusion therapy (intravenous thrombolytic therapy or primary PTCA), proven medical therapies (aspirin, ß-blockers, and ACE inhibitors) within the first 24 hours, and elective, in-hospital CABG (model 2), right BBB (odds ratio [OR], 1.64 [CI, 1.57 to 1.71]) remained the strongest predictor of death compared with ST-segment elevation with no BBB (OR, 1.53 [CI, 1.49 to 1.58]) or compared with left BBB (OR, 1.34 [CI, 1.28 to 1.39]). Right BBB was a marginally stronger predictor of death than was ST-segment elevation with no BBB (OR, 1.07 [CI, 1.01 to 1.12]), whereas left BBB was associated with a decreased risk for in-hospital death (OR, 0.87 [CI, 0.83 to 0.93]).

Discussion

Top Methods Results Discussion Author & Article Info References

In our study, right or left BBB at hospital admission was seen in one of eight patients with acute myocardial infarction and was associated with a substantially increased risk for in-hospital death. Right BBB was found to be a more important independent predictor of in-hospital mortality than left BBB. In addition, whereas patients presenting with right or left BBB had more comorbid illness and complicated infarctions, they were much less likely than patients with no BBB to receive immediate reperfusion therapy (thrombolytic therapy or primary PTCA) or other proven medical treatments.

Our study has several important strengths. The multicenter registry supplied us with a large, diverse cohort that included patients with myocardial infarction from all parts of the United States. The data reflect current patterns of use of initial reperfusion therapy and the acute management of high-risk subgroups with myocardial infarction. In addition, although mortality rates were calculated on the basis of chart review, the accuracy of coding for in-hospital death is 99.4% (based on comparisons using a subset of Medicare-eligible patients found in both NRMI 2 and the Cooperative Cardiovascular Project database sponsored by the Health Care Financing Administration) [28]. Finally, we more adequately assessed the independent effect of right or left BBB on in-hospital death in the thrombolytic era by adjusting for potential confounders, including differences in the types of therapy received, disease severity, and demographic and clinical characteristics.

Studies done in the prethrombolytic era estimated the prevalence of BBB in patients with myocardial infarction to be 2.7% to 13.4% for right BBB [1-49-1725, 29] and 0.3% to 8% for left BBB [1-49, 10, 12-1625, 29]. Four other published studies have reported data on the prevalence of BBB since the introduction of thrombolytic therapy. Sgarbossa and coworkers [19] found prevalences of 1% for right BBB and 0.5% for left BBB among 26 003 patients enrolled in the Global Utilization of Streptokinase and t-PA for Occluded Arteries (GUSTO)-1 trial. The low prevalence of BBB in the GUSTO-1 trial probably represents selection bias based on the entry criteria for and clinical presentation of included patients [19]. Newby and colleagues [12] reported prevalences of 13% for right BBB and 7% for left BBB in a subset of 681 patients enrolled simultaneously in the GUSTO-1 and Thrombolysis and Angioplasty in Myocardial Infarction (TAMI)-9 trials. Melgarejo-Moreno and associates [11] reported a prevalence of right BBB of 10.9% in 1238 patients with myocardial infarction who were admitted to three coronary care units, and the Fibrinolytic Therapy Trialists' (FTT) Collaborative Group [30] noted a combined prevalence of 4% for BBB (type unspecified) at presentation in pooled data on patients suspected of having myocardial infarction from three randomized trials of thrombolytic therapy. Using estimates based on more than 18 000 patients in each BBB group, we found prevalences of 6.2% for right BBB and 6.7% for left BBB.

In our study, the risk for in-hospital death was equally high for patients with myocardial infarction and right BBB at presentation (23.0%) or myocardial infarction and left BBB at presentation (22.6%). These risks are similar to the 23.6% risk for death at 35 days seen in patients with BBB who did not receive thrombolytic therapy in the FTT pooled analysis [28], and they are consistent with the reported 95% CIs for risk for death at 30 days in patients with new, persistent, left or right BBB who received thrombolytic therapy [12, 19].

Although right and left BBB were both independent predictors of increased in-hospital mortality, our most striking finding was that after adjustment for measured confounders, right BBB was a much stronger independent predictor of early death than left BBB. Compared with no BBB with no ST-segment elevation, right BBB was associated with a 64% increase in the odds of death whereas left BBB was associated with a 34% increase in the odds of death. In addition, the contribution of left BBB was less than that of ST-segment elevation with no BBB after we controlled for differences in patient characteristics and treatments received. However, right BBB remained a marginally stronger predictor than ST-segment elevation with no BBB (adjusted OR, 1.07 [CI, 1.01 to 1.12]). This finding extends the work of other researchers by showing that the effect of right BBB persists even after adjustment for differences in the time-dependent use of immediate reperfusion therapy (thrombolytic therapy or primary PTCA); other medical treatments received; and clinical and demographic features, including reported location of infarction [9, 11, 16, 17, 19, 25]. Furthermore, our results suggest that right BBB may contribute more independent information than left BBB in the short-term prognosis of patients with myocardial infarction.

Although the presence of right or left BBB at presentation gives the same overall prognostic information with respect to in-hospital death, the mechanisms by which right and left BBB increase risk for in-hospital death may differ substantially. The effect of left BBB seems to be partly explained by the fact that patients with left BBB are older, are more likely to be women, have a higher prevalence of preexisting congestive heart failure and other comorbid conditions, and receive substantially less reperfusion therapy and other medical therapies than those with right BBB or no BBB. On the other hand, despite adjustment for differences in comorbid conditions and therapy, right BBB remained a stronger predictor of death and is probably an independent marker of a larger, anterior infarction. Because of its dual blood supply [31], the right bundle branch may be more resistant to ischemia; thus, the development of right BBB complicating acute myocardial infarction may indicate a more extensive anteroseptal infarction compared with new left BBB [8, 17, 32, 33]. Right BBB does not seem to lead directly to an increased risk for death primarily through progressive conduction defects and complete heart block: Patients with right BBB had a relatively low incidence of second- or third-degree heart block overall, and complete heart block was the identified cause of death only rarely. This supports previous work showing that complete heart block is an uncommon cause of death without the presence of severe left ventricular dysfunction [4, 8, 34]. Another explanation for the potential association of right BBB with increased severity of infarction is that in our study, location of infarction was unspecified or unknown in almost half of patients with right BBB and two thirds of patients with left BBB, so residual confounding may exist as a result of our inability to more precisely determine the location of infarction. Finally, we could not determine timing of onset for right or left BBB, and the incidence of new right BBB may have been greater than the incidence of new left BBB in our cohort. Because the rate of early death associated with new BBB is higher than that associated with indeterminate or old BBB [3, 8, 17], a higher incidence of new right BBB may partly explain why right BBB was found to be a stronger risk factor than left BBB.

We also found that among patients for whom thrombolytic therapy was clearly indicated [35], the use of this therapy was low in those with right or left BBB compared with those with ST-segment elevation and no BBB. This supports previous findings of the underuse of thrombolytic therapy in patients with myocardial infarction and right [11] or left BBB [36, 37]. Furthermore, in patients who received thrombolytic therapy, delays were seen in the median time from hospital arrival to initiation of thrombolysis, especially in those with left BBB, suggesting that either the diagnosis of acute infarction or the benefit of reperfusion therapy was unclear to the treating physician. Knowledge of the efficacy of thrombolysis in patients with myocardial infarction and BBB is primarily based on 2146 patients with acute myocardial infarction and BBB at presentation from three randomized trials [30]. In these patients, no distinction was made between right and left BBB (Baigent C. Personal communication), but the pooled analysis showed an absolute risk reduction of 4.9% (relative risk reduction, 21%) in death at 35 days with thrombolytic therapy [30]. Our data reveal that despite the generally greater importance placed on left BBB as a poor prognostic factor in myocardial infarction [35], right BBB should receive as least as much attention in the risk assessment of patients with myocardial infarction. Furthermore, although evaluation of ST-segment elevation and pathologic Q-waves should be unaffected by complete right BBB [38], fewer than one third of the most "ideal" patients with right BBB received thrombolytic therapy and fewer than half received any initial reperfusion strategy (thrombolytic therapy, primary PTCA, or immediate CABG). Aspirin and ß-blockers were also underused in patients with right and left BBB; this suggests a need for improvement in the management of these high-risk subgroups.

Our study has several limitations. Because we were unable to determine timing of onset or persistence of BBB, we could not specifically determine the incidence of new BBB [12] or confirm the importance of the timing [12] or persistence [12, 19] of BBB. However, in actual clinical practice, providers often face the problem of not knowing whether BBB is new or old, so our study provides a generalizable estimate of the overall association between right or left BBB at presentation and in-hospital death. In addition, NRMI 2 hospitals are not necessarily representative of all U.S. hospitals: Participating facilities tend to be larger, tend to be more procedure-oriented, and may use thrombolytic therapy more often than nonparticipating hospitals [26]. Patient selection into the registry is neither all-inclusive nor random; this may introduce bias into patient sampling but with no clear effect on estimates of the association between BBB and death. In addition, because only nontransferred patients-who tended to be sicker and to have a higher prevalence of BBB-were included, care must be taken to avoid overgeneralizing these results to all patients with acute myocardial infarction. Finally, data on contraindications to therapies other than thrombolytic therapy were unavailable; thus, data on medication use must be prudently interpreted because the proportion of eligible patients may differ substantially between subgroups. Nevertheless, the overall use of proven therapies remains low in patients with myocardial infarction and right or left BBB.

In conclusion, right or left BBB at presentation in patients with acute myocardial infarction is associated with an equally high unadjusted risk for in-hospital death, but right BBB is a stronger independent predictor than left BBB for in-hospital death. In addition, patients with myocardial infarction who present with right or left BBB are much less likely to receive immediate reperfusion and other proven medical therapies. While we await better tests for the rapid diagnosis of myocardial infarction in patients with BBB, the noted substantial undertreatment of these high-risk patients strongly supports the most recent American College of Cardiology/American Heart Association Task Force recommendations: that reperfusion therapy (thrombolytic therapy or primary PTCA) be initiated immediately in patients with ischemic chest pain suggestive of acute myocardial infarction and BBB obscuring ST-segment analysis on electrocardiography at presentation [35].

From Veterans Affairs Medical Center and University of California, San Francisco, San Francisco, California; Kaiser Permanente Medical Care Program, Oakland, California; Genentech, Inc., South San Francisco, California; and Virginia Commonwealth University and Medical College of Virginia, Richmond, Virginia.

Dr. Barron and Ms. Rundle: Medical Affairs, Genentech, Inc., 1 DNA Way, Mail Stop 59, South San Francisco, CA 94080-4990.

Dr. Ornato: Department of Emergency Medicine, Main Hospital, G-503, 401 North 12th Street, PO Box 980401, Richmond, VA 23298-0401.

Dr. Avins: General Internal Medicine Section (111A-1), Department of Veterans Affairs Medical Center, 4150 Clement Street, San Francisco, CA 94121.