Cost Comparison of Radiofrequency Modification and Ablation of the Atrioventricular Junction in Patients With Chronic Atrial Fibrillation
Background Because it is not clear which technique is less expensive, the purpose of this study was to compare the cost of radiofrequency modification and ablation of the atrioventricular (AV) node in drug-refractory patients with atrial fibrillation and an uncontrolled ventricular rate.
Methods and Results The initial nominal charges for a successful procedure were compared in 10 patients with chronic atrial fibrillation who underwent modification of the AV node ($13 109±2002) and 14 similar patients who underwent ablation and pacemaker implantation ($28 302±2023, P<.001). On the basis of the long-term follow-up of patients who underwent each procedure, it was assumed that 31% of patients selected for the modification procedure would require a permanent pacemaker for inadvertent AV block or because of AV nodal ablation after a failed modification procedure and that the recurrence rate after AV node ablation would be 2%. The annual charges during follow-up were predicted and adjusted for recurrences and the need for additional procedures. The adjusted total charges at 1 year of follow-up were significantly lower for the modification procedure ($19 389±2002) than for the ablation procedure ($28 485±2023, P<.001). After 10 years of follow-up, the cumulative, adjusted charges for modification were $20 016 (42%) less than for ablation.
Conclusions The initial charges generated by AV node modification are significantly lower than for AV node ablation in patients with chronic atrial fibrillation. Even when adjusted for higher failure and recurrence rates, the modification procedure retains a major cost advantage over ablation during long-term follow-up.
Radiofrequency ablation of the atrioventricular (AV) junction has been shown to result in significant improvement in symptoms, functional capacity, and left ventricular function in patients with drug-refractory atrial fibrillation and an uncontrolled ventricular rate.1 2 3 Similar beneficial effects have been demonstrated with radiofrequency modification of the AV node.4 5 Compared with complete ablation, radiofrequency modification of the AV node has the advantage of avoiding the need for a permanent pacemaker, thereby eliminating the short- and long-term expenses associated with pacemaker therapy. Conversely, modification of the AV node has a lower success rate and a higher recurrence rate than the ablation procedure, which could negate the initial cost savings stemming from avoidance of a pacemaker. Because it is not clear which technique is less expensive overall when applied to a population of patients with atrial fibrillation and an uncontrolled ventricular rate, the purpose of this study was to compare the charges generated by radiofrequency modification and ablation of the AV node in patients with atrial fibrillation.
The initial charges generated by radiofrequency modification and ablation of the AV node were tabulated by review of the billing records of the initial hospitalization for 10 patients who underwent successful modification of the AV node and 14 patients who underwent successful ablation of the AV node during the same time period. This allowed the determination of the initial mean nominal charges for the two procedures. Each patient had chronic atrial fibrillation, and all permanent pacemakers were single-lead ventricular pacemakers.
The mean nominal charges were then adjusted to take into account the additional procedures needed because of inefficacy of the initial procedure or because of a recurrence of rapid AV conduction during follow-up. The experience with these procedures at the University of Michigan was used to create a model to determine the adjusted charges.6 By use of the follow-up data over a mean of 30 months in 62 patients who underwent the modification procedure, it was determined that 19% of patients who undergo AV node modification require a permanent pacemaker during the initial hospitalization because of unintentional high-degree AV block or because the modification procedure is ineffective and complete ablation is performed at the time of the initial procedure. The recurrence rate of rapid AV conduction after modification of the AV node was determined to be 11%, with one half of these patients undergoing a second modification procedure and the other half undergoing radiofrequency ablation of the AV node and pacemaker implantation. It was estimated that 3% of patients would require AV node ablation and pacemaker implantation because of recurrent atrial flutter with rapid AV conduction. All recurrences of rapid AV conduction were assumed to occur within 3 months after the initial procedure. Another 3% of patients undergoing radiofrequency modification of the AV node were determined to need a permanent pacemaker because of bradycardia-dependent polymorphic ventricular tachycardia after the modification procedure. Therefore, it was assumed that a total of 31% of patients who undergo the modification approach would undergo pacemaker implantation within 3 months of the procedure.
With regard to radiofrequency ablation of the AV node, it was assumed that the initial success rate of the procedure was 100% and that 2% of patients would require a second ablation procedure because of a recurrence of rapid AV conduction within 3 months of follow-up. These assumptions were based on the results of ablation of the AV node in 109 consecutive patients at the University of Michigan (unpublished data).
To determine the charges during follow-up, a standardized follow-up schedule was assumed for both groups of patients. In the patients who underwent radiofrequency modification of the AV node, this consisted of an outpatient clinic visit at 3 months and annually thereafter, with a resting ECG at the time of each visit and a 24-hour Holter monitor recording at the 3-month and 1-year visits. In the patients who underwent radiofrequency ablation of the AV node, the standardized pacemaker follow-up schedule was derived from Medicare reimbursement guidelines.7 This consisted of a 1-week wound check, 3-month chronic threshold determination, and follow-up threshold determinations once a year for the first 5 years, then every 6 months thereafter. In addition, transtelephonic telemetric monitoring was performed every 3 months for the first 4 years, every 2 months for the next 2 years, and monthly thereafter. The time to replacement of the pacemaker generator replacement was assumed to be 8 years, and the charges for generator replacement were determined from the total charges for replacement on an outpatient basis. The incidence of premature battery depletion and generator or lead malfunction was assumed to be zero. The charges for each follow-up visit or test were derived from the combined facility and professional fees at our institution. The charges for an outpatient generator change were based on the actual hospital bills of typical patients undergoing this procedure in 1996.
Characteristics of Patients
The patients whose billing records were reviewed in this study consisted of 10 patients who underwent radiofrequency modification of the AV node and 14 patients who underwent radiofrequency ablation of the AV node at the University of Michigan between March 1993 and June 1996. These patients were selected from a pool of 62 patients who underwent the modification procedure and 109 patients who underwent the ablation procedure during this time period. The selection criteria consisted of chronic atrial fibrillation, a successful long-term outcome, and the availability of complete billing records. Exclusion criteria consisted of a permanent pacemaker that had been implanted previously, the need for hospitalization immediately before or after the modification or ablation procedure for some reason other than the procedure, and a prosthetic heart valve requiring hospitalization before the procedure for heparinization during washout of a warfarin effect. The patients in each group represent all patients who met the entry criteria.
The demographic and clinical characteristics of the patients in the two treatment groups are described in Table 1⇓. The only significant difference between the two groups of patients was a lower left ventricular ejection fraction in the patients who underwent modification compared with ablation of the AV node (0.42±0.12 versus 0.54±0.10, P=.02).
Modification and Ablation Protocols
The AV nodal modification and ablation procedures were performed on the day of admission or within 24 hours of transfer from another hospital. The techniques used for both types of procedures have been described previously.4 8 9
The patients in whom the modification procedure was uncomplicated underwent continuous ECG monitoring by telemetry for 2 days; if transient AV block had occurred during the modification procedure, monitoring was performed for 4 days. Before discharge from the hospital, a 24-hour Holter monitor recording and exercise treadmill test were performed.
AV node ablation was done with a right-sided approach in 7 patients and a left ventricular approach in the remaining 7. A rate-responsive ventricular pacemaker was implanted immediately before or after the ablation procedure. A chest roentgenogram and ECG were performed after permanent pacemaker implantation, and the patients underwent 2 days of inpatient telemetric monitoring.
Analysis of Charges
In this study, the charges for the modification and ablation procedures were determined for each patient. The charges were determined by review of the computerized billing system of the Patient Accounts Department and included the professional fees, laboratory fees, and hospital fees. All charges were adjusted for inflation and are expressed in fiscal year 1996 dollars.
The nominal initial charge for the modification or ablation procedure was defined as the sum of all charges generated during the procedure and initial hospitalization in patients with a successful outcome and no complications. The adjusted charges were defined as the mean charges per patient in a hypothetical group of 100 patients undergoing the modification or ablation procedure, taking into account the need for repeat procedures and complications. The adjusted charges for the two procedures were calculated for yearly intervals over a follow-up of 10 years.
To determine the difference in actual cost of the two procedures and to validate the use of billing records as a surrogate for true cost, the actual costs of the initial hospitalization were obtained in a subgroup of patients from the Budget Office of the Department of Clinical Affairs at the University of Michigan Medical Center. Cost data were available for patients who underwent procedures in 1995 through 1996 and included 11 of the 14 patients in the ablation group and 5 of the 10 patients in the modification group. The actual total costs represent the sum of the direct costs, divided into variable direct (such as nursing salaries and surgical supply expenses) and fixed direct (such as administrative salaries and depreciation), and the indirect costs (such as overhead and malpractice insurance). The cost data do not include professional fees.
Continuous variables are expressed as the mean±SD. Student’s t test was used to compare continuous variables, and nominal variables were compared by χ2 analysis. ANOVA with repeated measures was used to compare the annual, cumulative adjusted costs of the modification and ablation procedures. A value of P<.05 was considered statistically significant.
Nominal Charges and Costs During First Year
The mean initial nominal charges for modification of the AV node, $13 109±2001, were 54% lower than the mean initial nominal charges for ablation of the AV node, $28 302±2023 (P<.001) (Table 2⇓). The mean number of days of hospitalization for the patients undergoing the modification and ablation procedures did not differ significantly (2.5±1.0 versus 2.0±0.6 days, respectively, P=.1).
The mean actual hospital cost for the modification procedure, excluding professional fees, was $4726±1639, which was 75% lower than the mean actual hospital cost for ablation, $18 809±1572 (P<.001). The difference in mean hospital cost of the two approaches, $14 083, was similar to the difference in mean initial nominal total charges, $15 193. The lower total cost of the modification procedure was attributable to a $7840 decrease in variable costs (P<.001), a $6016 decrease in indirect costs (P<.001), and a $227 decrease in fixed direct costs (P=.01) (Table 2⇑).
The total nominal charges for the first year after the procedure were $14 369±2002 in the modification group, compared with $28 935±2023 in the ablation group (P<.001).
In the ablation group, the mean initial nominal charges in patients who underwent a left-sided approach did not differ from those who underwent a right-sided approach, $27 946±2427 versus $28 657±1638 (P=.5).
Adjusted Charges for First Year
Taking into account the need for additional procedures, the mean adjusted initial charges for the modification procedure were $19 389±2002. This was significantly lower than the mean charges for the ablation procedure when adjusted for the need for a second ablation procedure in 2% of patients, $28 485±2023 (P<.001) (Table 2⇑). The adjusted total charges during the first year after the procedure also were significantly lower in the modification group than in the ablation group, $20 453±2002 versus $29 118±2023, respectively (P<.001).
Adjusted Charges During 10 Years
The predicted, cumulative, annual charges over 10 years of follow-up after radiofrequency modification and ablation of the AV node are shown in the Figure⇓. Between years 2 and 8 after the procedure, the annual charges varied from $285 to $572 in the modification group and from $477 to $1410 in the ablation group, depending on the frequency of transtelephonic monitoring. The charges for outpatient replacement of the pacemaker generator between years 8 and 9 of follow-up were calculated to be $10 525. At the end of 10 years of follow-up, the cumulative adjusted charges per patient were $20 016 (42%) less in the modification group than in the ablation group ($27 331±2002 versus $47 347±2023; P<.001).
Modification Versus Ablation of the AV Node
In properly selected patients who have drug-refractory atrial fibrillation with an uncontrolled ventricular rate, radiofrequency modification of the AV node has been demonstrated to result in a degree of improvement in symptoms, functional capacity, and left ventricular function that is comparable to the improvement realized after radiofrequency ablation of the AV node.6 Both the modification and ablation procedures may increase the risk of sudden death,6 10 although for both procedures, the sudden deaths during follow-up may be attributable to the underlying heart disease instead of the procedures themselves. Therefore, with regard to clinical response, neither technique has a large advantage over the other. However, ablation of the AV node has a higher success rate and lower recurrence rate than the modification procedure. The ablation procedure also has the advantage of being more generally applicable among patients with atrial fibrillation and an uncontrolled ventricular rate. For example, patients with atrial fibrillation and an uncontrolled ventricular rate who have symptoms caused by irregularity of the rhythm even when the rate is controlled, have periods of excessive bradycardia, or have demonstrated a predilection for torsade de pointes are better suited for the ablation procedure than the modification procedure. Therefore, for the modification procedure to have a legitimate role in the management of patients with atrial fibrillation, it would have to have a major cost advantage over ablation of the AV node.
The results of this study demonstrate that the short-term charges generated by successful radiofrequency modification of the AV node are ≈50% less than the short-term charges generated by a successful ablation procedure. However, because the recurrence rate is higher after modification than after ablation of the AV node and because ≈30% of patients selected for the modification procedure nevertheless may require a permanent pacemaker, a realistic cost comparison between the two techniques must take these factors into account. When the average charges generated by the modification and ablation procedures are adjusted to take these factors into account, the modification procedure retains a major advantage, with the adjusted total charges during the first year after the procedure being ≈33% lower than for the ablation procedure. Moreover, this cost advantage is maintained and amplified during additional years of follow-up. Because of the expenses associated with pacemaker surveillance and eventual replacement, the cumulative adjusted charges in the first decade after the procedure are 42% less after the modification procedure than after the ablation procedure. Therefore, although the modification procedure may have no advantage over the ablation procedure from the clinical standpoint, a major cost savings may provide justification for its use in patients who are suitable candidates for either the modification or ablation procedure.
Compared with medical therapy, radiofrequency catheter ablation of the AV junction for atrial fibrillation has been demonstrated to reduce the need for outpatient visits, hospital admissions, and antiarrhythmic drug therapy.1 2 3 A study conducted in Sweden found a 75% reduction in the use of antiarrhythmic drugs and a reduction in the number of days of hospitalization from 17±16 to 7±10 days per year.3 It was estimated that the cost of the ablation procedure was $10 400 and that the procedure resulted in an annual savings of $4056. Therefore, the break-even point in costs was calculated to be 2.6 years. Because of differences in healthcare costs between Sweden and the United States, it is difficult to relate these results to the results of the present study. Nevertheless, the results of the present study indicate that the modification procedure may further augment the savings realized by ablation therapy in patients who do not respond to pharmacological therapy.
Costs Associated With Pacemaker Implantation
Very few studies have quantified the long-term costs of pacemaker therapy. A previous study compared the long-term costs of dual-chamber and single-chamber pacemakers and emphasized the substantial costs incurred during follow-up.11 On the basis of a review of the literature at the time, the authors estimated that the incidence of early reoperation for lead fracture, dislodgment, and pocket infection or hemorrhage ranged from 1% to 8%. Because the study was designed to compare the costs of dual-chamber and single-chamber pacing, the determination of total cost excluded costs common to both types of pacemakers, such as the cost of the initial hospitalization. When corrected for inflation and expressed in 1996 dollars, the average cumulative cost of single-chamber pacing from after implantation through 12 years of follow-up was $6183, compared with $11 107 for dual-chamber pacing. The annual cost of follow-up in 1996 dollars in their study was $518, which is comparable to the annual charges of $477 incurred during the first 5 years of follow-up in the present study. Because the present study predicted charges on the basis of Medicare guidelines, which call for intensified follow-up as battery depletion approaches, the annual charges for follow-up in the present study increased to $1410 by year 8 after pacemaker implantation.
For the sake of homogeneity, the present study was limited to patients with chronic atrial fibrillation who received a rate-responsive ventricular pacemaker. Assuming that a patient with paroxysmal atrial fibrillation would usually be an appropriate candidate for a dual-chamber pacemaker, the results of the present study underestimate the potential cost advantage of the modification procedure compared with ablation therapy in patients who have paroxysmal atrial fibrillation. In addition, in the present study, the prediction of the expenses associated with pacemaker therapy did not take into account the possibility of lead malfunction or premature battery depletion. Because of unforeseeable problems such as atrial J-lead retention wire fractures,12 the actual long-term expense of pacemaker therapy may be higher than predicted in this study.
Failure Rate of the Modification Procedure
The difference in cost between the modification and ablation procedures is largely a function of the proportion of patients selected to undergo modification who eventually require a permanent pacemaker. To apply the results of this study to other institutions, the failure rate of the modification procedure would have to be comparable to the failure rate of ≈30% used to predict total charges in the present study. However, not all techniques for modification of the AV node have been equally successful. For example, a study that described a stepwise technique for ablation of the “slow” and “fast” AV nodal pathways reported a success rate of only 15%.13 To take this variability in success rate into account, the cost data in this study were reanalyzed to determine the minimum success rate needed for the modification procedure to retain a cost advantage over ablation.
To depict a “worst-case” scenario, it was assumed that one half of modification failures consist of patients who require a permanent pacemaker because of AV block at the time of the initial procedure and that the other half initially have a successful outcome but then have a recurrence of rapid conduction and undergo complete ablation and pacemaker implantation. It was determined that even if 60% of patients (or approximately twice as many as in the model used in this study) undergoing the modification procedure had a failed outcome, the adjusted initial charges for the modification procedure, $26 158±2002, still would be significantly less than for the ablation procedure, $28 301±2023 (P<.05).
This study has several limitations. First, the data were derived from patients who were referred to a single university hospital, and the charges incurred for the same procedures at other institutions may vary. Second, although billing data were available for every patient in this study, data on actual costs were not. However, the comparison of charges was found to provide an accurate estimate of the absolute differences in actual costs. Third, patients were not randomly assigned to undergo the modification or ablation procedures; because the mean left ventricular ejection fraction was lower in the patients who underwent the modification procedure, there may have been a bias toward higher costs in this group, and the cost advantage of the modification procedure may have been underestimated. Fourth, a large proportion of the patients in the ablation group underwent a left ventricular approach to ablation; however, the mean charges were no different between the patients who underwent a right-sided approach and a left ventricular approach, and there is no evidence that the use of the left ventricular approach affected the results of the study. Last, follow-up charges were predicted from a hypothetical model instead of actual billing records. This allowed a comparison of charges over 10 years of follow-up.
In conclusion, in patients with atrial fibrillation and an uncontrolled ventricular rate who are not successfully managed with pharmacological therapy and who are candidates for either modification or ablation of the AV node, the significantly lower short- and long-term charges for the modification procedure, even when adjusted for the need for a permanent pacemaker in approximately one third of patients, provide a rationale for attempting the modification procedure before ablation.
- Received January 16, 1997.
- Revision received March 26, 1997.
- Accepted April 4, 1997.
- Copyright © 1997 by American Heart Association
Fitzpatrick AP, Kourouyan HD, Siu A, Lee RJ, Lesh MD, Epstein LM, Griffin JC, Scheinman MM. Quality of life and outcomes after radiofrequency His-bundle catheter ablation and permanent pacemaker implantation: impact of treatment in paroxysmal and established atrial fibrillation. Am Heart J. 1996;131:499-507.
Brignole M, Gianfranchi L, Menozzi C, Bottoni N, Bollini R, Lolli G, Oddone D, Gaggioli G. Influence of atrioventricular junction radiofrequency ablation in patients with chronic atrial fibrillation and flutter on quality of life and cardiac performance. Am J Cardiol. 1994;74:242-246.
Jensen SM, Bergfeldt L, Rosenqvist M. Long-term follow-up of patients treated by radiofrequency ablation of the atrioventricular junction. Pacing Clin Electrophysiol. 1995;9:1609-1614.
Feld GK, Fleck RP, Fujimura O, Prothro DL, Bahnson TD, Ibarra M. Control of rapid ventricular response by radiofrequency catheter modification of the atrioventricular node in patients with medically refractory atrial fibrillation. Circulation. 1994;90:2299-2307.
Morady F, Hasse C, Strickberger SA, Man KC, Daoud E, Bogun F, Goyal R, Harvey M, Knight BP, Weiss R, Bahu M. Long-term follow-up after radiofrequency modification of the atrioventricular node in patients with atrial fibrillation. J Am Coll Cardiol. 1997;27:113-121.
HCFA Publication 6: Medicare Coverages Issues Manual. Section 50-1, transmittal number 1051, transmittal date October 1984.
Sousa J, El-Atassi R, Rosenheck S, Calkins H, Langberg J, Morady F. Radiofrequency catheter ablation of the atrioventricular junction from the left ventricle. Circulation. 1991;84:567-571.
Eagle KA, Mulley AG, Singer DE, Shoenfeld D, Harthorne JW, Thibault GE. Single-chamber and dual-chamber pacemaker: a formal cost comparison. Ann Intern Med. 1986;105:264-271.