The dark blue line indicates the mortality rate of all hospital admissions within the Nationwide Inpatient Sample (NIS) database by year. The light blue line indicates the mortality rate of only hospital admissions where ablation was performed.
The dark blue line indicates the total number of admissions in the Nationwide Inpatient Sample (NIS) database by year, associated with the right y-axis. The light blue line indicates the total number of ablations in the NIS database, by year, associated with the left y-axis.
Curley M, Berger M, Roth J, Benjamin I, Rubenstein J. Predictors of Mortality and Major In-Hospital Adverse Events Associated With Electrophysiology Catheter Ablation. JAMA Intern Med. 2014;174(5):815-817. doi:10.1001/jamainternmed.2014.415
Catheter ablation has become an important strategy for the treatment of cardiac arrhythmias and is generally safe. Conflicting data have emerged on the mortality and complication rates in these procedures. This study examines the in-hospital outcomes for ablation for supraventricular and ventricular arrhythmias in the large Nationwide Inpatient Sample (NIS) Registry.
This study was granted an exemption from oversight by the Medical College of Wisconsin institutional review board. The NIS database includes 20% of hospital discharges in the United States. We queried in-hospital mortality associated with ablations performed for any indication (supraventricular and ventricular arrhythmias) between 1998 and 2009. For our analysis, we used survey commands in STATA 12.1 statistical software (StataCorp). The NIS database was queried for procedure codes that may indicate complications of ablation, including blood transfusions, cardiothoracic surgery, pericardiocentesis, and pacemaker implantation.
During the 92 848 710 hospitalizations, 115 955 catheter ablations were performed. The mean (SD) patient age was 60.6 (17.8) years, 43.8% of patients were female, and 52.3% of procedures were elective. There were 708 in-hospital deaths reported in patients who underwent ablation during the index admission (0.6%). This is comparable to the overall mortality of all admissions during the same period of 2.2% (Figure 1). There was no difference in mortality (P = .53) for patients undergoing catheter ablation based on the year of the ablation. Ablation was listed as the primary procedure in 72.5% of cases. The total number of ablations and hospital admissions over the study period within the NIS database are shown in Figure 2.
Catheter ablation may have been associated with 17 628 potential major in-hospital complications, reaching an overall risk of complications of 15.2%. There were 80 emergent cardiac surgical procedures (0.1%), 255 pericardiocenteses (0.2%), 2304 blood transfusions (2.0%), and 14 989 implantations (12.9%) of a permanent pacemaker during the index hospitalization.
Previously reported mortality rates for catheter ablation have ranged from 0.1% to 0.3%.1- 3 This study demonstrates that the in-hospital mortality on admissions where catheter ablation was performed was 0.6%. This likely overestimates the mortality of ablation because this includes all deaths during the index hospitalization, regardless of the association with the ablation. No significant trend was demonstrated in ablation-related mortality over the 11-year period and was lower than the mean mortality rate of all NIS admissions.
This study identified an overall potential rate of periprocedural complications of 15.2%. The most common complications were blood transfusions and implantation of a permanent pacemaker. It is not possible to discern from the NIS database if the ablation procedure caused these potential complications or if they arose for other unrelated reasons; 15.2% should represent the upper bound of the true complication rate.
There are several important limitations related to the use of an administrative in-hospital database. Discharge data captures only events that occur during an index hospitalization, and as such, mortality or complications that occur outside of the hospital are not counted. Mortality, as well as complications such as transfusions, are not known to be caused by the catheter ablation. The actual mortality associated with ablation should be no higher than these data demonstrate and is likely lower. Only in-hospital ablations are considered, so the actual mortality of all ablations (including lower-risk procedures performed on an outpatient basis) may be lower. The NIS database element of elective status was reported in the Results section to attempt to account for this. In addition, ablations are heterogeneous; thus the overall complication rate is sensitive to the mix of procedures.
In conclusion, there is a 0.6% risk of in-hospital mortality and as high as a 15% risk of complications associated with catheter ablation. Because the causes of mortality and potential complications are unknown, these rates represent the upper bounds of ablation-related mortality and morbidity. These results should inform discussions of potential procedural risk between an electrophysiologist and a patient considering ablation procedures.
Corresponding Author: Jason Rubenstein, MD, Division of Cardiology, Department of Medicine, Medical College of Wisconsin, Froedtert East Clinics, 9200 W Wisconsin Ave, Milwaukee, WI 53226 (firstname.lastname@example.org).
Published Online: March 31, 2014. doi:10.1001/jamainternmed.2014.415.
Author Contributions: Drs Curley and Rubenstein had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Curley, Berger, Rubenstein.
Acquisition, analysis, or interpretation of data: Curley, Roth, Benjamin, Rubenstein.
Drafting of the manuscript: Curley, Rubenstein.
Critical revision of the manuscript for important intellectual content: Berger, Roth, Benjamin, Rubenstein.
Statistical analysis: Curley, Rubenstein.
Administrative, technical, or material support: Curley, Roth, Benjamin, Rubenstein.
Study supervision: Berger, Rubenstein.
Conflict of Interest Disclosures: None reported.
Previous Presentation: These data were presented at the Wisconsin–American College of Cardiology annual meeting; November 10, 2012; Kohler, Wisconsin.