Author + information
- Received April 25, 2011
- Revision received August 22, 2011
- Accepted August 29, 2011
- Published online January 10, 2012.
- Rashmee U. Shah, MD, MS⁎,⁎ (, )
- James V. Freeman, MD, MPH⁎,
- David Shilane, PhD⁎,
- Paul J. Wang, MD⁎,
- Alan S. Go, MD†,‡ and
- Mark A. Hlatky, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Rashmee U. Shah, Cedars-Sinai Medical Center, Department of Cardiology, 8700 Beverly Boulevard, Room 5536B, Los Angeles, California 90048
Objectives The purpose of this study was to estimate rates and identify predictors of inpatient complications and 30-day readmissions, as well as repeat hospitalization rates for arrhythmia recurrence following atrial fibrillation (AF) ablation.
Background AF is the most common clinically significant arrhythmia and is associated with increased morbidity and mortality. Radiofrequency or cryotherapy ablation of AF is a relatively new treatment option, and data on post-procedural outcomes in large general populations are limited.
Methods Using data from the California State Inpatient Database, we identified all adult patients who underwent their first AF ablation from 2005 to 2008. We used multivariable logistic regression to identify predictors of complications and/or 30-day readmissions and Kaplan-Meier analyses to estimate rates of all-cause and arrhythmia readmissions.
Results Among 4,156 patients who underwent an initial AF ablation, 5% had periprocedural complications, most commonly vascular, and 9% were readmitted within 30 days. Older age, female, prior AF hospitalizations, and less hospital experience with AF ablation were associated with higher adjusted risk of complications and/or 30-day readmissions. The rate of all-cause hospitalization was 38.5% by 1 year. The rate of readmission for recurrent AF, atrial flutter, and/or repeat ablation was 21.7% by 1 year and 29.6% by 2 years.
Conclusions Periprocedural complications occurred in 1 of 20 patients undergoing AF ablation, and all-cause and arrhythmia-related rehospitalizations were common. Older age, female sex, prior AF hospitalizations, and recent hospital procedure experience were associated with a higher risk of complications and/or 30-day readmission after AF ablation.
The prevalence of atrial fibrillation (AF) among U.S. adults is expected to rise from 2.3 million to at least 5.6 million by 2050, with an increasing burden of medical costs (1,2). Catheter ablation, a newer treatment option for AF that uses radiofrequency or cryotherapy to electrically isolate the pulmonary veins and ablate arrhythmia foci, is increasing in use. Randomized studies have suggested that AF ablation may decrease arrhythmia burden and improve quality of life, and observational studies have suggested that ablation may be associated with lower ischemic stroke and death rates (3).
Procedural complications have been reported to affect between 1% and 8% of patients undergoing AF ablation at experienced academic hospitals (4,5) and 7% to 10% among Medicare beneficiaries treated in centers of varying types (6). However, relatively little is known about use of AF ablation and complication rates in contemporary patient cohorts across the age spectrum treated in community settings. Furthermore, the frequency of rehospitalizations and repeat ablation procedures among unselected patients is not well understood.
Using data from the Healthcare Utilization Project California State Inpatient Database (7), we characterized trends in AF ablation use, procedural complications, and repeat hospitalizations in a large, diverse contemporary population. We also sought to evaluate predictors of poor outcomes, including procedural complications and 30-day rehospitalizations.
Catheter ablation and arrhythmia
The California State Inpatient Database, from Healthcare Utilization Project (Agency for Healthcare Research and Quality), contains administrative data for all discharges from nonfederal hospitals in the state plus an encrypted linking variable to track hospitalizations of the same patient over time (7). We identified patients who received an initial AF ablation procedure between January 1, 2005, and November 31, 2008, by searching for discharges with a principal diagnosis of AF (International Classification of Diseases-9th Edition [ICD-9] 427.31) and a concurrent catheter ablation procedure code (ICD-9 37.34). We excluded patients with secondary diagnosis codes for atrial flutter, Wolff-Parkinson-White syndrome, nonparoxysmal atrioventricular nodal tachycardia, paroxysmal supraventricular tachycardia, paroxysmal ventricular tachycardia, and ventricular premature beats. We also excluded patients with diagnostic or procedural codes indicating implantation of a pacemaker or implantable cardioverter-defibrillator during the index admission (to exclude those undergoing atrioventricular node ablation) and patients undergoing open surgical ablation (Online Table A1). Patients were excluded from this analysis if they had received an AF ablation in 2004 to include only patients who presumably had their first AF ablation during the study period.
Comorbid conditions were identified based on secondary diagnosis codes for coronary artery disease, heart failure, cerebrovascular disease, renal disease, hypertension, diabetes, valvular heart disease, and lung disease during the index AF ablation hospitalization and a concurrent “present on admission” indicator (Online Table A2). In addition, patients were classified as having coronary artery disease, heart failure, cerebrovascular disease, or lung disease if they had been hospitalized during the year before the index procedure with a primary diagnosis of these conditions. Patients were classified as having chronic renal disease, hypertension, diabetes, valvular heart disease, lung disease, or cancer if they had been hospitalized during the year before the index procedure with the relevant ICD-9 code as either the primary or a secondary diagnosis (Online Table A2).
The database does not provide detail to characterize the type or duration of AF. Therefore, we used the number of AF hospitalizations as a proxy to account for disease severity, defined as the number of hospitalizations with AF in the primary diagnostic position during the year before index ablation.
Acute procedural complications included cardiac perforation and/or tamponade, pneumothorax, hemothorax, procedure-related stroke, transient ischemic attack, vascular access complication (consisting of hemorrhage/hematoma, vascular complication requiring surgical repair, and accidental puncture), and in-hospital death (Online Table A2). Procedural complications were identified using secondary diagnoses coded during the index AF ablation admission.
We estimated rates of all-cause and cause-specific 30-day rehospitalizations. Cause-specific reasons for 30-day readmissions included procedural complications and primary diagnoses of AF or atrial flutter (Online Table A3). Long-term outcomes of interest included hospitalization for any cause and admission for arrhythmia recurrence at any time after ablation through the end of the study period. Arrhythmia recurrence was defined as a repeat hospitalization with a primary diagnosis of AF or atrial flutter or a repeat hospitalization with a procedure code for catheter ablation.
Categorical data are presented as frequencies and percentages; continuous data are presented as mean ± SD. Univariate comparisons were performed using t tests for continuous variables and chi-square tests for categorical and binary variables. Multivariable logistic regression with generalized estimating equations (to account for within-hospital clustering) were used to identify predictors of inpatient complication and/or 30-day all-cause rehospitalization. Patient demographics (age, sex, and race/ethnicity), primary payer, comorbid conditions, number of AF hospitalizations during the prior year, and recent hospital procedure experience were included in the multivariable model. Recent procedure experience was defined as the number of cases completed at a given center during the 12 months before each index AF ablation divided into quartiles for comparison.
We used Kaplan-Meier analyses to estimate all-cause readmission and arrhythmia rehospitalization rates. For the readmission analysis, patients were censored at the end of the study period (December 31, 2008). For the arrhythmia rehospitalization analysis, patients were censored if they died during a hospitalization or at the end of the study period. Ascertainment of death was limited to those that occurred within a California hospital because the database cannot be linked with the National Death Index.
A value of p < 0.05 was considered statistically significant. All analyses were performed using Stata version 11.0 (StataCorp LP, College Station, Texas).
We identified 4,156 patients who received an initial AF ablation in California between January 1, 2005, and November 31, 2008. The number of initial procedures almost doubled from 684 cases in 2005 to 1,332 cases in 2008; the mean annual volume of AF ablations per hospital was 15.4 throughout the study period. A total of 98 unique hospitals performed AF ablations during the study period; the number of hospitals increased from 60 in 2005 to 77 in 2008. In this sample, mean patient age was 61.7 years; hypertension (50.3%) and coronary artery disease (14.7%) were the most common comorbidities (Table 1). Most patients had not been hospitalized primarily for AF during the year before ablation (79.1%). Recent hospital procedural experience during the preceding 12 months varied from 7.0 ± 4.4 procedures in the lowest quartile to 136.9 ± 27.8 in the highest quartile. The observation time (from time of initial AF ablation through December 31, 2008) was a median (interquartile range) of 1.53 years (0.78 to 2.62).
Inpatient complications occurred in 5.1% (n = 211) of AF ablations and did not vary according to study year. Vascular complications, primarily bleeding-related events, accounted for more than half of all complications (Table 2). One patient died during the index admission. Length of stay was 1.46 ± 1.96 days for those without complications and 3.42 ± 3.44 days for those with complications (p < 0.001). Within 30 days of index hospitalization, 9.4% (n = 390) of patients discharged alive were rehospitalized. AF or atrial flutter accounted for 26.9% of readmissions (n = 105), and procedural complications accounted for another 19.5% of readmissions (n = 76) (Table 2).
Age, sex, primary payer, heart failure, hypertension, chronic renal disease, lung disease, number of AF hospitalizations during the prior year, and procedure volume quartile were univariate predictors of inpatient complications and/or 30-day rehospitalization (p < 0.05 for each) (Table 1). In the multivariable analyses, female sex (odds ratio [OR]: 1.38 compared with males; 95% confidence intervals [CI]: 1.13 to 1.69) and number of AF hospitalizations during prior year (OR: 1.19 for each additional admission; 95% CI: 1.05 to 1.35) were significant predictors of the outcome (Table 3). The odds of inpatient complication and/or 30-day readmission increased in a stepwise fashion with increasing patient age, up to age 85 years (Table 3). Less hospital experience with AF ablation was also a significant predictor of the outcome, with a 57% increase in the odds of the outcome in the lowest compared with the highest quartile (Table 3). Every 5-U increase in the number of procedures completed during the prior 12 months decreased the odds of the outcome by 2% (OR: 0.98; 95% CI: 0.97 to 0.99).
In long-term follow up, 1,816 patients were readmitted after AF ablation for any reason. At 1 year, 61.5% of patients were free of hospital admission (95% CI: 59.9% to 63.1%) (Fig. 1). A total of 1,022 patients were rehospitalized for arrhythmia recurrence or repeat ablation. At 1 year, 78.3% (95% CI: 76.9% to 79.6%) of patients were free of readmission for recurrent arrhythmia or repeat ablation; at 2 years, 70.4% (9% CI: 68.7% to 72.0%) of patients were free of rehospitalization for recurrent arrhythmia or repeat ablation (Fig. 2). Among the 4,156 recipients of an initial AF ablation, 17.4% received additional ablation; 682 patients received 1 additional ablation, 36 received 2 additional ablations, 3 received 3 additional ablations, and 3 received 4 additional ablations during the study period.
In a large contemporary cohort of patients who underwent AF ablation in California, 5% of patients experienced a periprocedural complication and almost 10% were rehospitalized with 30 days. By 1 year, 22% of patients were rehospitalized for arrhythmia recurrence or repeat ablation and almost 40% were rehospitalized for any reason.
Periprocedural complication rates following AF ablation have ranged from 1% to 8% in smaller studies of high-volume centers (4,9), and the complication rate was 9.1% in a study of Medicare beneficiaries (6). The periprocedural complication rate of 5.1% in this study is derived from a large contemporary cohort that is more representative of the general AF population than prior estimates from selected patient cohorts (6,10).
Policymakers have begun to focus on 30-day readmission rates as a potential marker of quality of care. Prior studies of AF ablation have not reported 30-day readmission rates, which we found to be 10% overall (most commonly for AF and atrial flutter) and 12.1% among patients age 65 years and older. This readmission rate after AF ablation is lower than the 14.6% 30-day readmission rate of Medicare beneficiaries after percutaneous coronary intervention (8) and lower than the 20.1% readmission rate of Medicare beneficiaries after heart failure hospitalization (11), although still high for an elective procedure.
Hospitalization for recurrent arrhythmia or repeat ablation occurred frequently, affecting 22% of patients at 1 year and 30% of patients at 2 years (Fig. 2). Randomized controlled trials have reported AF recurrence in 13% to 44% of AF ablation recipients (12,13), and a worldwide survey of centers offering AF ablation reported recurrence in 48% to 57% of recipients at a median of 12 months' follow-up (4). The estimate in our study was at the lower end of the reported range because the database does not capture outpatient or asymptomatic recurrence. Our findings were therefore consistent with previous studies that observed that AF ablation has limited success in preventing arrhythmia recurrence over longer-term follow-up (4). We further demonstrated that these recurrences were often severe enough to result in costly hospitalizations.
Slightly <20% of patients in this cohort of unselected AF ablation recipients received multiple ablations, which is lower than the rates reported from specialized centers. The worldwide survey mentioned previously reported that more than 25% of patients received repeat ablations within a median follow-up time of 12 months (4). The difference between our findings and those from specialized centers may be due to a more aggressive approach of specialty centers in treating recurrent arrhythmias. Alternatively, specialized centers may select patients who are healthier or who have disease that is more amenable to ablation, making repeat ablation more attractive.
The increased procedural risk for women undergoing AF ablation was consistent with patterns seen in other cardiovascular procedures. Women undergoing percutaneous coronary intervention have more than a 2-fold higher risk of vascular complications than men (14–16), perhaps because of smaller body size or higher levels of anticoagulation among women (17). However, a report on risk factors for bleeding with unfractionated heparin found that female sex remained a risk factor after controlling for weight, heparin dose, and activated partial thromboplastin time, suggesting that mechanisms other than medication dosing could account for increased bleeding risk in women (18).
We found that that recent procedural experience with AF ablation was inversely related to complication rates, which was consistent with the general observation that higher hospital volume has been associated with better outcomes for various procedures (19–21) and medical conditions (22). Our measure of procedural experience differed from the methods used to estimate hospital volume in prior publications in that we assessed the number of procedures performed in the prior 12 months as the metric, rather than the average procedure volume over several years. This measure of recent experience captured the effect of the “learning curve” for new procedures such as AF ablation. Our finding of better outcomes among patients treated at centers with more procedural experience in the prior 12 months suggested that higher procedure volumes are needed to gain proficiency, maintain proficiency, or both.
The administrative data used in this analysis lack the detail that is available in trials and registries. Physician characteristics and full assessment of comorbidities (i.e., including outpatient diagnoses) were unavailable for the analyses. This limitation is counterbalanced by the larger sample size and the absence of reporting bias introduced by selective publication of results from specialized centers. Data about procedural technique, medication use, and type of AF (e.g., paroxysmal or persistent) were unavailable. We did, however, use the number of AF hospitalizations during the year before index ablation as a proxy for disease severity.
The patients in this study were representative of the general AF ablation population, but certain limitations in our cohort assembly should be noted. We identified patients who had their “first” procedure between 2005 and 2008, looking back through 2004. Some patients may have had procedures before 2004. Given the smaller number of AF ablations completed in earlier years, however, the number of misclassified patients should be small. Patients who had pacemakers placed before the index hospitalization and who may have had an atrioventricular junction ablation were not excluded.
Our study did not include a comparison of patients with AF treated with antiarrhythmic drugs. We found high rehospitalization rates following ablation, but patients with AF treated with other methods may have similar or higher rates. Also, 18% of patients without reported complications were hospitalized for 2 or more days, some of whom may have had a complication that was not coded in the hospital discharge abstract. Finally, the database could not be linked with the National Death Index; therefore, we could not accurately assess mortality late after the procedure, although we could detect all deaths during the index admission.
AF ablation is a newer treatment option that is being used with increasing frequency. In this large, unselected cohort of AF ablation recipients, 1 in 20 patients experienced a periprocedural complication and almost 1 in 10 were rehospitalized within 30 days of the procedure. Older age, female sex, prior AF hospitalizations, and recent hospital procedure experience were factors associated with a high risk of complications and/or 30-day rehospitalization after AF ablation. Future investigations should be directed at identifying ways to decrease procedural complication rates, arrhythmia recurrence, and readmissions.
For supplementary tables, please see the online version of this article.
This study was funded by a grant from the American Heart Association, Dallas, Texas, National Institutes of Health grant 5UL1RR025744, and Stanford NIH/NCRR CTSA grant KL2RR025743. Dr. Wang has stock ownership in Hansen Medical; and has received honoraria from Medtronic. All other authors have reported that they have no relationships to disclose relevant to the contents of this paper to disclose. The first two authors contributed equally to this work. Bruce D. Lindsay, MD, served as Guest Editor for this paper.
- Abbreviations and Acronyms
- atrial fibrillation
- International Classification of Diseases-9th Edition
- Received April 25, 2011.
- Revision received August 22, 2011.
- Accepted August 29, 2011.
- American College of Cardiology Foundation
- Miyasaka Y.,
- Barnes M.E.,
- Gersh B.J.,
- et al.
- Lubitz S.A.,
- Fischer A.,
- Fuster V.
- Cappato R.,
- Calkins H.,
- Chen S.-A.,
- et al.
- Weerasooriya R.,
- Khairy P.,
- Litalien J.,
- et al.
- ↵HCUP State Inpatient Databases (SID): Healthcare Cost and Utilization Project (HCUP) (Agency for Healthcare Research and Quality, Rockville, MD) http://www.hcup-us.ahrq.gov/sidoverview.jsp. Accessed July 10, 2011. 2005–2009.
- Curtis J.P.,
- Schreiner G.,
- Wang Y.,
- et al.
- Calkins H.,
- Brugada J.,
- Packer D.L.,
- et al.
- Stabile G.,
- Bertaglia E.,
- Senatore G.,
- et al.
- Ahmed B.,
- Piper W.D.,
- Malenka D.,
- et al.
- Cheng S.,
- Morrow D.A.,
- Sloan S.,
- Antman E.M.,
- Sabatine M.S.
- Freeman J.V.,
- Wang Y.,
- Curtis J.P.,
- Heidenreich P.A.,
- Hlatky M.A.