Cost‐utility analysis of radiofrequency ablation versus optimal medical therapy in managing supraventricular tachycardia among Filipinos

Abstract

Background

Radiofrequency ablation (RFA) is the standard of care in the management of supraventricular tachycardia (SVT). Its cost‐effectiveness in an emerging Asian country has not been studied.

Objectives

A cost‐utility analysis of RFA versus optimal medical therapy (OMT) among Filipinos with SVT was conducted using the public healthcare provider's perspective.

Methods

A simulation cohort using a lifetime Markov model was constructed using patient interviews, a review of literature, and expert consensus. Three basic health states were defined: stable, SVT recurrence, and death. The incremental cost per quality‐adjusted life year (ICER) was determined for both arms. Utilities for the entry states were derived from patient interviews using the EQ5D‐5L tool; utilities for other health states were taken from publications. Costs were assessed from the healthcare payer perspective. A sensitivity analysis was done.

Results

Base case analysis showed that RFA versus OMT is both highly cost‐effective at 5 years and over a lifetime. RFA at 5 years costs about PhP276,913.58 (USD5,446) versus OMT of PhP151,550.95 (USD2,981) per patient. Discounted lifetime costs were PhP280,770.32 (USD5,522) for RFA, versus PhP259,549.74 (USD5,105) for OMT. There was improved quality of life with RFA (8.1 vs. 5.7 QALYs per patient). The 5‐year and lifetime incremental cost‐effectiveness ratios were PhP148,741.40 (USD2,926) and Php15,000 (USD295), respectively. Sensitivity analysis showed 56.7% of simulations for RFA fell below a GDP‐benchmarked willingness‐to‐pay (WTP) threshold.

Conclusion

Despite the initial higher cost, RFA versus OMT for SVT is highly cost‐effective from the Philippine public health payer's perspective.

Radiofrequency ablation for Supraventricular tachycardia has been found to be highly cost‐effective, costing less than one GDP, based on a Markov model of patients from an emerging economy. This finding is robust across a wide range of inputs in the sensitivity analysis. This may be explained by the curative nature of the procedure.

1. INTRODUCTION

Supraventricular tachycardia (SVT) refers to rapid arrhythmias (>100/min), originating in tissue within or above the His bundle. Hemodynamic compromise and cerebral hypoperfusion may occur. ¹ Treatment options include drug therapy or ablation, the choice of which depends on the severity of SVT. ² Catheter ablation is first‐line therapy for symptomatic, recurrent SVT (Class I), especially among higher‐risk individuals, such as those with preexcited atrial fibrillation or high‐risk occupations. It is also an alternative to drugs for patients with symptomatic focal AT (Class I). Drug therapy with oral Verapamil, Diltiazem, and beta‐blockers may be considered (Class I for AVNRT or AVRT; Class IIa for focal AT). In patients with AVRT and/or preexcited AF, oral Flecainide or Propafenone (Class IIa) is considered reasonable. ² , ³

Access to RFA in the Philippines is limited. Asia‐Pacific Heart Rhythm Society (APHRS) data show a mere 117 procedures annually from 2014 to 2017. ⁴ More recently, 1121 and 1149 government insurance claims for SVT in 2017 and 2018, respectively, were documented. ⁵ The procedure is costly and paid for out of pocket; hence, the need to increase government subsidies, which would improve compliance with treatment guidelines. ⁶

The Universal Health Care Act, recently signed into law in February 2019, emphasizes the role of the State “to protect and promote the right to health of every Filipino.” In determining which health services should be covered by the government, such services should be proven to be cost‐effective, as well as socially, organizationally, and ethically acceptable. ⁷ This study aimed to conduct a comparative analysis of both the costs and benefits (in quality‐adjusted life years, or QALYs) and cost‐effectiveness (reflected by incremental cost‐effectiveness ratio [ICER]) of RFA versus OMT among patients with SVT.

2. METHODS

2.1. Patient population and interventions

A hypothetical population of adult Filipino patients with recurrent, symptomatic SVT was created. Exclusion criteria were adapted from ACC/AHA clinical practice guidelines for the use of RFA in the management of adult patients with SVT. ² Comparative Costs, QALYs, and the cost‐effectiveness (ICER) of RFA relative to OMT were evaluated. OMT included medications, treatment for emergent arrhythmia‐associated events, and consultations.

2.2. Model structure

The model considered two treatment arms: status quo or OMT and RFA. Both models simulated the progression of a treated, stable SVT patient through different health states (Figure 1).

FIGURE 1.

Markov model used for cost‐utility analysis comparing RFA with OMT. ADR, adverse drug reaction; OPT, optimal medical therapy; SVT, supraventricular tachycardia.

The OMT or “status quo” model considered five health states: treated/stable, adverse drug reactions (ADR), SVT recurrence, dropouts, and death. The ADR and SVT recurrence states are considered transient states; that is, cohort members entering these states remain within this state for only one model cycle and transition immediately to the next cycle. In case of ADR, patients immediately either drop out of the treatment regimen or recover and return to a stable state. For SVT recurrence, patients recover and return to a stable state or transition to death. The OMT arm assumes a passive or reactive approach.

The RFA scenario, on other hand, assumes more aggressive management where the procedure is performed within a year of diagnosis. There is a small probability of treatment failure. These are accounted for in the model as transitional states. Mortality rates for dropouts and SVT recurrence were considered similar, and it was assumed that SVT recurrence portends the same mortality risks regardless of treatment arm. Post‐RFA complications were assumed to occur only within the first year.

2.3. Model inputs

2.3.1. Clinical inputs

A focused literature review was done to synthesize published values and outcomes. PubMed was searched using the terms “ablation,” “standard of care,” “supraventricular tachycardia,” and variations thereof. Articles that compared the effectiveness of antiarrhythmic drugs versus catheter ablation for SVT were reviewed. Studies involving patients with Implantable Cardioverter Defibrillator or Atrial Fibrillation were excluded. For OMT, there is a high variation reported for failure. Studies with the lowest failure rate, for conservative estimates, were selected. ⁸ , ⁹ Meanwhile, the acute success rates were consistently high in all studies for ablation success. To note, a large multicenter study reported major complications as 3%. ¹⁰ Model input parameters are in Table 1.

TABLE 1.

Input parameters used in the model.

Parameter	Description	Value	References
Comparator: Optical Medical Therapy (OMT)
tpEsvt	Annual progression risk from stable to recurrent SVT, on OMT	0.40	[8, 10, 12]
tpEadr	Annual progression risk from stable to ADR	0.09	[11]
tpRadr	Annual probability of recovering from ADR	0.80	[8, 13]
tpRsvt	Annual probability of recovering from recurrent SVT	0.95	[8, 13]
tpDsvt	Annual mortality risk due to recurrent SVT	0.05	[15]
tpDall	Annual all‐cause mortality	0.074	[16]
Treatment: Radiofrequency ablation (RFA)
tpEsvtrf	Annual progression risk from stable to recurrent SVT, post‐RFA	0.032	[10]
tpRFArep	Annual probability of repeat ablation during first year post‐RFA	0.083	[10, 14]
tpRFArep_rec	Annual probability of recurrent SVT after repeat ablation	0.014	[10, 12]
tpRFA_comp	Annual probability of post‐RFA complications	0.049	[10, 17]
tpDsvtrf	Annual mortality risk due to recurrent SVT	0.05	[15]
tpDall	Annual all‐cause mortality	0.074	[16]
Health utilities
qSomt	Utility weight for stable SVT, on OMT	0.83 0.69	[9] Patient survey
qSrfa	Utility weight for stable SVT, post‐RFA	0.98 0.96	[9] Patient survey
qcomp	Utility weight for post‐RFA complications: complete heart block, hemodynamic instability, major bleeding, vascular injury	0.854	[10]

Note: These are the values entered in the Excel Markov model and their sources (from literature or from Patient surveys).

2.3.2. Quality of life inputs

To assess the changes in QALY before and after RFA—corresponding to the utility values in each treatment arms' entry state—the EuroQol Valuation Technology (EQVT) was utilized. ¹⁸ The EQVT uses a valuation tool, EQ‐5D‐5L, a multi‐attribute utility instrument (MAUI) used to measure QALYs. The tool is a self‐administered or facilitated questionnaire containing (a) a descriptive system where respondents score each of the five domains: mobility, self‐care, usual activities, pain/discomfort, and anxiety/depression and (b) a visual analog scale (VAS). Permission was secured for use of this valuation tool.

A retrospective observational cohort was purposively sampled from three referral hospitals in Metro Manila to determine QALY using the EQ‐5D‐5L tool. Ethical approval was secured from the Philippines' Department of Health Single Joint Research Ethics Board (DOH‐SJREB) and each hospital's ethical research board. Informed consent was taken prior to each enrollment.

In the descriptive sections, respondents were asked to describe their situation in each domain. The VAS allowed respondents to rate and mark their health status quantified on a 20 cm vertical visual analog scale.

The scores from the descriptive sections were assessed against a published tariff to obtain a utility value, with 1.0 as the maximum and −1.0 as the minimum. Death corresponds to a utility score of 0; negative values may imply a health state worse than death. Utility values multiplied with the life years for a defined period yielded QALY. Each of the 3125 possible health states has been valued for the Philippines by the National Institutes of Health—Institute for Health Policy and Development Studies (NIH‐IHPDS), and the tool is available for use in the Philippines using a self‐administered or facilitated questionnaire. To our knowledge, this study is one of the first studies in the Philippines that utilizes the EQ‐5D‐5L after a clinical intervention for the purpose of economic evaluation. EQ‐5D‐5L is known to be a valid and reliable tool to measure the health‐related quality of life for cardiovascular diseases. ¹⁹ , ²⁰

2.3.3. Cost inputs

Resource use was estimated from a combination of published literature and expert panel validation. Drug prices were derived from the latest Philippine Drug Price Reference Index (DPRI) and the DOH‐sourced Maximum Drug Retail Price (MDRP). Costs for complications were from published prices from authoritative sources. Inputs and their values are summarized in Table 2.

TABLE 2.

Cost parameters used in the model.

Parameter	Percentage use	Unit cost [range] PhP	References
Comparator: Optical Medical Therapy (OMT) a
Flecainide 100 mg tab	52.1%	72.60 [MDRP] (USD1.42)	[21, 22]
Metoprolol 100 mg tab	10.0%	1.55 [0.83–4.85] (USD0.03)
Procainamide	4.4%	18 [5.30–27.00] (USD 0.35)
Verapamil 80 mg tab	47.9%	100.25 [18.75–46.50] (USD0.22)
Digoxin 250 mcg tab	28.5%	3.97 [3.00–5.00]
Follow‐up consultations	3.7 times b	1000 (USD19.67)	[21]
Treatment: Radiofrequency ablation (RFA)
Procedure cost		250,000.00 [50,000.00‐450,000.00] (USD4,917)	[24]
Procedural complications c		50,000 (USD983)	[25]
Maintenance	33.33% d	1200 (USD23.‐60)	[25]
Regular routine follow‐up	1.2 times b	1000 (USD19.67)	23
SVT recurrence (ER visits) e		20,000.00 [6200.00–12,200.00] (USD393.40)	[26]

^a Maintenance drugs computed annually; MDRP‐ maximum drug retail price.

^b Average annual clinic visits per patient.

^c Mean is from 20% of the Ablation cost. The reason is that complications end up with additional 1–2 ICU days hence the assumption.

^d Maintenance of drugs is assumed one‐third of the cost compared with the standard of care since one study reported that one‐third of patients continued to receive antiarrhythmic drug therapy at the end of the follow‐up period 28.

^e Mean cost is from PHIC case rates of any unspecified cardiac arrhythmia. We assumed identical costs of ER visits for both RFA and OMT.

2.4. Base case analysis

With expert consensus and inputs from a review of literature, a Markov model in MS Excel with annual cycles was developed which followed the natural disease progression of a patient with SVT treated with either OMT or RFA. Calculations of costs were constructed from the Philippine healthcare payer's perspective over a 5‐year lifetime horizon. The primary cost outcome was total lifetime direct medical costs, while the health effect outcome was measured in QALYs. The incremental cost‐effectiveness ratio (ICER) for OMT versus RFA was then calculated, followed by calculating the net monetary benefit (NMB) of each intervention. All outcomes and costs were discounted at 3.5% for the base case.

2.5. Sensitivity analysis

A probabilistic sensitivity analysis (PSA) using 10,000 Monte‐Carlo simulations to account for uncertainties surrounding the input parameters. The results were plotted on a cost‐effectiveness plane. The net monetary benefits were calculated, plotting the results in a cost‐effectiveness acceptability curve (CEAC) representing the GoP's willingness‐to‐pay (WTP) for RFA instead of OMT. One‐way sensitivity analysis (OWSA) was likewise performed, where specific parameters were varied between low and high values to determine the main drivers of the ICER. As such, QALY inputs of RFA and OMT, discount rate, cost of ablation, and complication cost to OWSA were analyzed and summarized in a Tornado plot.

3. RESULTS

The characteristics of the interview cohort are summarized in Table 3.

TABLE 3.

Characteristics of the EQ‐5D‐5L interview respondents.

Characteristics	RFA (n = 21)
Age
Range	24 to 85 years
Average	55 years
% Male	38.1%
Average number of years from diagnosis to ablation	5.86 years
Average number of years since ablation	2.29 years
Prominent symptoms, No. (%)
Fainting	2 (9.5%)
Dizziness	8 (38.1%)
Choking sensation/coughing	2 (9.5%)
Dyspnea	10 (47.6%)
Fatigue	3 (14.3%)
Palpitations	6 (28.6%)
Preprocedure drugs used
Antiarrhythmics	20 (95.2%)
Beta‐blockers (Metoprolol, etc.)	6 (28.6%)
Calcium blockers (Verapamil, etc.)	6 (28.6%)
Sodium blockers (Flecainide)	3 (14.3%)
Potassium blockers (Amiodarone)	5 (23.8%)
Other drugs	12 (57.1%)
Anticoagulants	6 (28.6%)
Statins	2 (9.5%)
Angiotensin II receptor blockers	4 (19.0%)
None stated	3 (14.3%)
Could not recall at least one drug taken	6 (28.6%)

Note: Respondents were all adult post‐RFA patients who were at least 10 weeks postprocedure at the time of the interview.

Quality of life measurements (QALYs) before and after RFA measured using the validated EQ‐5D are shown in Table 4.

TABLE 4.

QALYs before and after RFA.

	Pre‐RFA	Post‐RFA
Mobility	1.86	0.01
Self‐care	1.38	0.01
Usual activity	1.9	0.01
Pain/discomfort	2.43	0.01
Anxiety/Depression	2.62	0.00
VAS	69.76	92.48
Adjusted utility weight	0.69	0.96

Abbreviation: VAS, visual analog scale.

3.1. Base case

Base case results are in Table 5. Using the discounted case, within 5 years, RFA was both more expensive and more beneficial in terms of QALYs gained than OMT. The greatest cost was incurred during the first 2 years; RFA cost was about PHP277,000.00 (USD5,448)/capita, more expensive than OMT by PHP125,000.00 (USD2,458), yet resulting in a health benefit of 3.2 QALYs/patient (RFA) versus 2.3 QALY/patient (OMT). The 5‐year ICER of RFA relative to OMT was thus positive, indicative of the higher costs but also higher health benefits, at PHP148,741.40 (USD2,925) per QALY gained.

TABLE 5.

Results of base case analysis of radiofrequency ablation (RFA) vs optimal medical therapy (OMT), over 5 years and over a lifetime.

Treatment arm	Total cost per capita (PhP)	Total QALY gain per capita	Incremental cost (PhP)/incremental effectiveness (QALYs)	ICER (PhP/QALY gained)
5‐year horizon
Discounted
RFA	276,913.58	3.180	125,362.63/0.843	148,741.40
OMT	151,550.95	2.338
Undiscounted
RFA	279,271.47	3.842	103,741.45/1.016	102,132.90
OMT	175,530.03	2.826
Lifetime horizon
Discounted
RFA	280,770.32	6.213	21,220.58/1.420	14,940.10
OMT	259,549.74	4.792
Undiscounted
RFA	289,586.82	11.951	(172,267.66)/2.241	(76,878.03)
OMT	461,854.48	9.711

Abbreviations: OMT, optimal medical therapy; RFA‐radiofrequency ablation.

Over a lifetime, the discounted ICER improved to PHP15,000.00 (USD 295) per QALY gained, attributable to the drop in expenses of care in the years following RFA. In contrast, OMT patients continue to require medications and ER visits. The lifetime health benefit was 6.2 QALYs/capita (RFA) and 4.8 QALY/capita (OMT).

3.2. Sensitivity analysis

The scatterplot (Figure 2) shows the simulated ICERs in all four quadrants of the cost‐effectiveness plane. The joint density appears concentrated around the x‐axis in quadrants I and II. This indicates relative dominance in the lifetime benefit of RFA versus OMT, reflected by more dense points to the right of the y‐axis. This is further supported by the average incremental per capita benefit of about 0.9 more QALYs gained. Similar to other new technologies, RFA was also slightly more expensive, although highly more effective over a lifetime, with an average incremental cost of PHP31,000.00 (USD609)/capita. The average ICER over all 10,000 simulations amounted to about PHP54,000.00 (USD1,062) per QALY gained.

FIGURE 2.

Scatterplot of Monte‐Carlo simulation results on the cost‐effectiveness plane. x‐axis: QALY per 1000 persons; y‐axis: Cost in Philippine pesos.

The net monetary benefits were plotted in the cost‐effectiveness acceptability curve (Figure 3). A majority (56.73%) of lifetime simulations fall below a willingness‐to‐pay (WTP) threshold of PhP490,000.00 (USD9,638) or three times the 2021 per capita GDP. RFA was more cost‐effective than OMT through a broad range of simulated thresholds, beginning at a lower value of PhP40,000.00 (USD786)/QALY.

FIGURE 3.

Cost‐effectiveness Acceptability Pane. In red‐RFA (radio frequency ablation); in blue‐ OMT(optimal medical therapy); x‐axis‐ Cost‐effectiveness threshold in Philippine Pesos; y‐axis‐ proportion of simulations.

A deterministic one‐way sensitivity analysis of ICERs is shown in Figure 4. Both 5‐year and lifetime ICERs were significantly influenced by RFA cost; 5‐year ICERs were more sensitive to procedure costs than lifetime ICERs. The 5‐year ICERs were about equally sensitive to variations in discount and utility weights. In contrast, lifetime ICERs were moderately sensitive to discount and less sensitive to utility weights: changes in discount rates created ICER variations of 20–30% and utility weights caused variations of <2%. Both 5‐year and lifetime ICERs were relatively insensitive to peri‐procedural complications cost.

FIGURE 4.

Tornado plot for results of the deterministic sensitivity analyses for the ICER of RFA vs OMT. OMT for SVT, within (A) a 5‐year time horizon and (B) a lifetime horizon.

4. DISCUSSION

Securing government subsidy for radiofrequency ablation of SVT in the Philippines remains a challenge. In this CUA, the Philippines age‐adjusted mortality rate and local costs were incorporated into a Markov model constructed by content and method experts to come up with the cost per quality adjusted life year of RFA versus optimal medical therapy for the management of SVT. The results showed that the discounted incremental cost per quality adjusted life year using the present unit price of RFA and optimal medical therapy was quite low, at PhP 148,741.40 (USD2,925) per QALY gained within the first 5 years, and even lower over a lifetime, at PhP 14,940.10 (USD293) per QALY gained. This signifies that even within the first 5 years post‐RFA, where most of the large medical costs associated with the procedure are incurred, RFA is highly cost‐effective, with the base case ICER amounting to <1 GDP per capita.

The cost‐effectiveness of RFA is traceable in the model to the curative nature of the procedure, and low rates of expensive complications and indications for repeat procedures. On the other hand, recurrent ER admissions are expected for OMT due to breakthrough episodes, in spite of regular intake of medications. In calculating costs, we included downstream costs, as recommended by health economics experts, such as the cost of complications and the cost of ER visits. ²⁷ Not included is the cost of inpatient hospitalizations, which although not recommended in international guidelines, is often required of SVT patients locally, and loss of patient and family member productivity during hospitalizations.

A similar cost‐effectiveness study of RFA versus medical therapy was conducted in Guatemala showing that the cost of the RFA procedure itself was USD5,411. RFA gains 1.46 QALYs and saves USD7,993 compared with OMT for patients with SVT. RFA in that study was considered a safe procedure. No deaths were documented, even during redo ablation. ²⁸ This study suggested that RFA, especially in low‐income countries, should be the most appropriate therapeutic choice, especially among the young who would otherwise be on maintenance OMT. RFA is either cost‐saving or has a cost‐effectiveness ratio compared with many commonly accepted therapies for other diseases.

The WHO Choosing Interventions that are Cost‐effective project (WHO‐CHOICE) considers an intervention to be cost‐effective if the cost of preventing one "disability‐adjusted life‐year (DALY)" is less than three times its annual GDP per capita, while it is very cost‐effective if it prevents one DALY at a cost of less than one GDP per capita. ²⁹ This has empirically been used especially among countries where no specific cutoff has been identified. Local policymakers have selected a more stringent cutoff of 1GDP per capita in efforts to prioritize the highly cost‐effective interventions. RFA achieves this cutoff within 5 years, with cost‐effectiveness further increasing over a lifetime. These results may be explained by the high success rate of RFA, and the potentially “curative” nature of RFA, creating marked savings as far as ER visits and even hospital admissions. The figures estimated by our model may in fact underestimate the true cost savings because in the Philippines many SVT patients brought to the ER are not discharged directly from the ER but are admitted for two nights to the hospital.

Sensitivity analysis proves the robustness of our findings. The cost‐effectiveness acceptability curve shows that RFA is uniformly cost‐effective compared with OMT over a broad range of costs. Our sensitivity analyses indicate that RFA is likely to be cost‐saving or cost‐effective under a wide range of assumptions. We made several assumptions to produce a tractable model. We used a long‐term perspective to consider in more detail the cost and risk associated with the strategies. Prolonged drug therapy leads to both increased expenditures and an increased risk of side effects. We also explicitly modeled crossovers between therapies.

Aside from the GDP threshold, some authors recommend the use of other approaches in addition to a threshold ICER in coming up with a decision to adopt or provide coverage for a new intervention. ²⁷ Two approaches that may be used are the Willingness to Pay approach and the Budget Impact Analysis. Budget impact analysis showed the cost of RFA at 250,000 at government hospitals, which is below the WTP of Php450,000 (USD8,851). ³⁰

There are three strengths of this study. First, the utilities used to derive the QALYs were directly measured from a local cohort and not derived from previous trials, providing the advantage of reflecting real‐world experience. Second is that the instrument used to measure utilities, the EuroQol, is a validated and standardized tool. Third, it uses a broad perspective, that of the government payer. This increases the reliability and applicability of our results to other emerging economies. ²⁷ , ²⁹

The analysis is based on a simplified disease and treatment pathway that considers only AVNRT and AVRT as indications for ablation. We also assumed that post‐procedural complications as well as events necessitating re‐ablation occur only within the year following the procedure. We also assume that the entire population of SVT‐diagnosed individuals undergoes either RFA or the designated comparator treatment. While we do not foresee any significant effect of including other indications for ablation here, ³¹ we do not discount that there may be subpopulations of the diagnosed group for whom ablation might not be appropriate, and hence recommend that further subgroup analyses be done in future studies.

FUNDING INFORMATION

This work was supported by a research grant from the Philippine Heart Rhythm Society.

CONFLICT OF INTEREST STATEMENT

None of the authors has any personal or commercial interest in any CIED company or distributor.

APPROVAL OF THE RESEARCH PROTOCOL BY AN INSTITUTIONAL REVIEWER BOARD

UPMREB 2020–318‐01‐SJREB, 21 September 2020.

INFORMED CONSENT

Taken for each patient.

REGISTRY AND THE REGISTRATION NO. OF THE STUDY/TRIAL

RGAO‐2019‐0499.

Gervacio GG, Kimwell MJM, Fadreguilan EC, De Guzman DC, Gabriel EA, Tolentino CS, et al. Cost‐utility analysis of radiofrequency ablation versus optimal medical therapy in managing supraventricular tachycardia among Filipinos. J Arrhythmia. 2023;39:175–184. 10.1002/joa3.12833