Atrial fibrillation (AF) is a common complication of acute myocardial infarction (AMI) and contributes to high rates of in-hospital adverse events. However, there are few contemporary studies examining rates of AF in the contemporary era of AMI or the impact of new-onset AF on key in-hospital and postdischarge outcomes. We examined trends in AF in 6,384 residents of Worcester, Massachusetts, who were hospitalized with confirmed AMI during 7 biennial periods between 1999 and 2011. Multivariate logistic regression analysis was used to examine associations between occurrence of AF and various in-hospital and postdischarge complications. The overall incidence of AF complicating AMI was 10.8%. Rates of new-onset AF increased from 1999 to 2003 (9.8% to 13.2%), and decreased thereafter. In multivariable adjusted models, patients developing new-onset AF after AMI were at a higher risk for in-hospital stroke (odds ratio [OR] 2.5, 95% confidence interval [CI] 1.6 to 4.1), heart failure (OR 2.0, 95% CI 1.7 to 2.4), cardiogenic shock (OR 3.7, 95% CI 2.8 to 4.9), and death (OR 2.3, 95% CI 1.9 to 3.0) than patients without AF. Development of AF during hospitalization for AMI was associated with higher rates of readmission within 30 days after discharge (21.7% vs 16.0%), but no significant difference was noted in early postdischarge 30-day all-cause mortality rates (8.3% vs 5.1%). In conclusion, new-onset AF after AMI is strongly related to in-hospital complications of AMI and higher short-term readmission rates.
Atrial fibrillation (AF) is the most common dysrhythmia, and the global prevalence of AF continues to increase with the aging of the population in many countries. AF is a frequent complication of acute myocardial infarction (AMI), and in previous studies of patients with AMI, has been linked to increased rates of heart failure, stroke, and death. Over the last few decades, dramatic changes have occurred in how patients with AMI are diagnosed and treated. More sensitive troponin assays, early percutaneous revascularization, and newer therapeutic options for medical management have favorably reshaped the prognosis of patients with AMI. Moreover, economic factors, including pay-for-performance and public reporting of adverse outcomes, and quality improvement programs have helped achieve better patient-related outcomes. Although these changes have had a favorable impact on overall in-hospital mortality in patients with AMI, studies suggest that the incidence of AF complicating AMI remains as common today as it was 20 years ago. However, limited studies have examined recent trends in AF or its impact on traditional in-hospital and postdischarge short-term outcomes, especially from a community-wide perspective. With the goal of filling knowledge gaps related to the descriptive epidemiology of AF in the contemporary era of AMI, we analyzed data from the population-based Worcester Heart Attack Study (WHAS).
Methods
The WHAS is an ongoing, population-based, observational investigation examining long-term trends in the incidence, morbidity, in-hospital complications, and short- and long-term mortality of patients hospitalized with AMI at all greater Worcester medical campuses. Our study population comprises of 6,384 residents of the Worcester metropolitan area in central Massachusetts who were hospitalized with a discharge diagnosis of AMI at all Worcester Standard Metropolitan Statistical Area hospitals during any of 7 biennial years from 1999 to 2011. There were originally 16 health care facilities that were included in the study. Recently, fewer hospitals (n = 11) have been providing care to residents of central Massachusetts because of hospital closures, mergers, or conversion to long-term care facilities. Among the present 11 hospitals, 3 were tertiary care/university-based medical centers where almost 85%-90% of patients with confirmed AMI had been hospitalized during the years under study with little variation in this proportion observed over time. Patients with a known history of AF, based on the review of information contained in medical records, were considered to have “Prevalent AF” (n = 493), and those who developed AF during their hospitalization for AMI were defined as having “Incident AF” (n = 693).
Trained physicians and nurses abstracted data of eligible patients with confirmed AMI from hospital medical records. Information was collected about patients’ age, gender, previous co-morbidities, type of AMI (ST-segment elevation myocardial infarction [STEMI] or non-ST-segment elevation myocardial infarction [NSTEMI], Q wave vs Non–Q wave), AMI order (initial vs previous), physiologic parameters on admission (blood pressure, respiratory rate, and so on), in-hospital medications, and in-hospital procedures. Data were also collected about occurrence of in-hospital complications such as stroke, heart failure, acute renal failure, cardiogenic shock, and death. Postdischarge readmission rates were tabulated, and postdischarge mortality was evaluated by review of medical records for subsequent hospitalizations and a nationwide search of death certificates for residents of the Worcester metropolitan area. Medical records of residents of the Worcester metropolitan area admitted for possible AMI at all Worcester Standard Metropolitan Statistical Area medical centers were individually reviewed and validated. The diagnosis of AMI was confirmed using preestablished diagnostic criteria.
The occurrence and timing of AF was determined for WHAS participants based on manual abstraction of clinical information from ambulance transport records, emergency admission notes and logs, progress notes, and all in-hospital 12-lead electrocardiograms (ECGs). Interpretation of ECG findings is not done in isolation as a computerized interpretation of each recorded ECG is performed at all participating greater Worcester hospitals in addition to an over-read by a board-certified clinical cardiologist. Prevalent AF was considered present if a history of AF was noted in the admission note or any progress note. The criteria used to define incident AF included no documentation of history of AF and either (1) AF deemed present by an interpreting cardiologist on any 12-lead ECG obtained during the index hospitalization or (2) new-onset AF documented in any clinical note during the index hospitalization. Patients who underwent coronary artery bypass grafting during hospitalization for AMI were excluded from the analysis because postoperative AF is common in patients who undergo coronary artery bypass grafting and is caused by different etiologic factors and pathophysiologic mechanisms.
Differences in characteristics of those with incident AF, prevalent AF, and no AF were examined through the use of chi-square test and analysis of variance for discrete and continuous variables, respectively. Similar methods were used to examine differences in hospital and postdischarge outcomes according to hospital development of incident AF. Short-term prognosis in each of the periods under study was examined by calculating in-hospital complication and case-fatality rates. Multivariate logistic models were constructed, with accompanying odds ratios (OR) and 95% confidence intervals (CI), to examine trends in occurrence of incident AF over the study period. The odds of developing AF were adjusted for several potentially confounding demographic and clinical factors, including age, gender, race, history of angina, hypertension, diabetes, stroke, heart failure, hyperlipidemia, chronic obstructive pulmonary disease, renal failure, and AMI-associated characteristics. To examine the potential impact of infarct type (STEMI vs NSTEMI), chronic kidney disease (CKD), and receipt of percutaneous coronary intervention (PCI) on the odds of developing AF, we conducted stratified analyses evaluating the odds of developing AF in the aforementioned groups. An additional series of multivariate logistic regression models was used to examine the association between occurrence of AF and in-hospital stroke, renal failure, heart failure, 30-day postdischarge mortality, and 30-day readmission rates while adjusting for the same set of covariates included in our previous analyses. We did not control for receipt of cardiac medications or coronary reperfusion strategies in the analyses because of the potential for confounding by treatment indication and lack of data on timing of administration of these therapies. All analyses were performed using SAS version 9.3 (SAS Institute Inc., Cary, NC).
Results
The baseline characteristics of the 6,384 study patients are listed in Table 1 . Four hundred and ninety-three patients had AF before their AMI (7.7%), whereas 693 patients (10.8%) developed new-onset AF during their index AMI-related hospitalization.
| Characteristics | Incident AF (n=693) | Prevalent AF (n=493) | No AF (n=5,198) | P value in relation to No AF |
|---|---|---|---|---|
| Age -Mean [SD] | 76.2 [11.4] | 79.8 [8.9] | 68.5 [14.5] | <0.001 |
| Age < 65 years | 15.7 % | 5.9 % | 38.1 % | <0.001 |
| Age 65-84 years | 55.8 % | 54.4 % | 44.7 % | <0.001 |
| Age >= 85 years | 28.4 % | 39.8 % | 17.3 % | <0.001 |
| Sex, % Men | 50.8 % | 54.4 % | 56.6% | 0.01 |
| White Race | 91.5 % | 94.5 % | 89.1 % | <0.001 |
| Angina Pectoris | 13.7 % | 18.3 % | 15 % | 0.08 |
| Chronic Kidney Disease | 21.9 % | 31.4 % | 17.7 % | <0.001 |
| COPD | 20.2 % | 26.4 % | 16.9 % | <0.001 |
| Diabetes | 33.2 % | 36.3 % | 33.8 % | 0.48 |
| Heart failure | 24.4 % | 48.9 % | 22.5 % | <0.001 |
| Hyperlipidemia | 52.4 % | 50.9 % | 56.6 % | <0.001 |
| Hypertension | 74.5 % | 81.3 % | 71.1 % | <0.001 |
| Stroke | 12.5 % | 20.1 % | 11.1 % | <0.001 |
| AMI Type | ||||
| ST-segment elevation myocardial infarction | 38.1 % | 20.5 % | 35.4 % | <0.001 |
| Non ST-segment elevation Myocardial infarction | 61.9 % | 79.5 % | 64.6 % | <0.001 |
| Initial AMI | 70.9 % | 53.5 % | 64.6 % | <0.001 |
| Q Wave | 26.3% | 10.5 % | 21.9 % | <0.001 |
| Non Q Wave | 73.7 % | 89.5 % | 78.1% | <0.001 |
| Troponin I Peak in ng/mL – Mean [SD] | 28.1 [83.3] | 17.9 [44.4] | 18.1 [66.9] | <0.001 |
| Creatinine in mg/dl – Mean [SD] | 1.5 [1.1] | 1.6 [1] | 1.4 [1.1] | <0.001 |
| Factors at Admission | ||||
| Initial Heart Rate (bpm)-Mean [SD] | 91.6 [27.3] | 91.6 [24.2] | 84.9 [21.1] | <0.001 |
| Systolic Blood Pressure (mmHg) -Mean [SD] | 133.5 [31.9] | 134.7 [31.5] | 142.5 [31.3] | <0.001 |
| Diastolic Blood Pressure (mmHg)-Mean [SD] | 73.7 [19.4] | 73[18.4] | 77.8 [19.5] | <0.001 |
| Respiratory Rate (per minute)- Mean [SD] | 21.9 [6.3] | 22.1 [5.9] | 21 [5.5] | <0.001 |
| In-Hospital Medications | ||||
| Beta-blocker | 85.6 % | 86.8 % | 88.1 % | 0.13 |
| Aspirin | 90.0 % | 85.2 % | 91.9 % | <0.001 |
| Statin/ Lipid Lowering Agents | 63.5 % | 55.8 % | 69 % | <0.001 |
| ACE-I/ARB | 62.5 % | 59.6 % | 62.6 % | 0.4 |
| Clopidogrel | 49.6 % | 37.3 % | 58.9 % | <0.001 |
| Calcium Channel Blockers | 33.2 % | 34.9 % | 19.1 % | <0.001 |
| Diuretic | 67.1 % | 75.7 % | 46.4 % | <0.001 |
| Digoxin | 34.1 % | 44.2 % | 10.9 % | <0.001 |
| Thrombolytics | 4.4 % | 2.0% | 6.3 % | <0.001 |
| Enoxaparin/Heparin | 27.1 % | 18.3 % | 19.8 % | <0.001 |
| Warfarin | 19.9 % | 40.8 % | 8.4 % | <0.001 |
| In-Hospital Procedures | ||||
| Percutaneous coronary Intervention | 37.2 % | 21.5 % | 44.8 % | <0.001 |
| Cardiac Catheterization | 48.1 % | 34.5 % | 59.5 % | <0.001 |
Rates of new-onset AF initially increased from 9.8% in 1999 to 13.2% in 2003, after which rates decreased to a nadir of 6.4% in 2009, before returning to near-baseline levels in the most recent study year 2011 ( Figure 1 ). After controlling for several demographic factors, co-morbid conditions, AMI characteristics, and in-hospital complications, the odds of developing AF were the highest in 2003 and the lowest in 2009 but remained relatively stable throughout the other years under study ( Table 2 ). The odds of developing AF over time did not vary significantly in groups stratified by PCI status, AMI type, or presence of CKD ( Supplemental Table 1 ).
| Period | Patients | Crude OR with 95 % CI | Adjusted OR ∗ with 95 % CI |
|---|---|---|---|
| 199 9 | 920 | 1.0 (Referent study year) | 1.0 |
| 2001 | 1065 | 1.4 (1.1-1.9) | 1.4 (1.1-1.8) |
| 2003 | 998 | 1.5 (1.1-1.9) | 1.5 (1.1-2) |
| 2005 | 788 | 1.4 (1-1.9) | 1.3 (1-1.8) |
| 2007 | 755 | 1.1 (0.8-1.5) | 1.1 (0.8-1.5) |
| 2009 | 680 | 0.6 (0.5-0.9) | 0.6 (0.4-0.9) |
| 2011 | 685 | 0.9 (0.6-1.2) | 0.8 (0.6-1.2) |
∗ Adjusted for age, gender, race, history of angina, hypertension, diabetes, stroke, heart failure, hyperlipidemia, COPD, renal failure and AMI-associated characteristics.
Patients who developed AF were older, more likely to be women, and have a history of hypertension, heart failure, stroke, chronic obstructive pulmonary disease, and CKD than those who did not develop AF ( Table 1 ). Although patients with incident AF had lower blood pressures on admission and a higher heart rate, they were less likely to have undergone in-hospital procedures such as PCI and cardiac catheterization. Patients with incident AF were less likely to have been treated with aspirin, clopidogrel, β blockers, statins, or thrombolytics. In contrast, a greater percentage of patients with AMI with new-onset AF received digoxin, diuretics, calcium channel blockers, and anticoagulants (heparin and warfarin) ( Table 1 ).
Patients developing AF were significantly more likely to have a hospital course characterized by the development of several complications such as stroke, heart failure, acute renal failure, and cardiogenic shock ( Table 3 ). After adjusting for several demographic factors, co-morbid conditions, previous medical history, and AMI-associated characteristics, patients with incident AF remained at a significantly higher risk for developing stroke (adjusted OR 2.5, 95% CI 1.6 to 4.1), heart failure (adjusted OR 2.0, 95% CI 1.7 to 12.4), and cardiogenic shock (adjusted OR 3.7, 95% CI 2.8 to 4.9) than were patients who remained free from AF during their hospitalization (p <0.001 for all) ( Figure 2 ). Although the study was limited by relatively small numbers of patients in each group, outcomes did not vary significantly in patients with incident AF further stratified by receipt of PCI, AMI type, or presence of CKD ( Supplemental Table 2 ).
