Implantable cardioverter defibrillator therapy is cost effective for primary prevention patients in Taiwan: An analysis from the Improve SCA trial.

OBJECTIVE: Implantable cardiac defibrillators (ICDs) for primary prevention (PP) of sudden cardiac arrest (SCA) are well-established but underutilized globally. The Improve SCA study has identified a cohort of patients called 1.5 primary prevention (1.5PP) based on PP patients with the presence of certain risk factors. We evaluated the cost-effectiveness of ICD therapy compared to no ICD among the PP population and the subset of 1.5PP patients in Taiwan.
METHODS: A Markov model was run over a lifetime time horizon from the Taiwan payer perspective. Mortality and utility estimates were obtained from the literature (PP) and the IMPROVE SCA trial (1.5PP). Cost inputs were obtained from the Taiwan National Health Insurance Administration (NHIA), Ministry of Health and Welfare. We used a willingness-to-pay (WTP) threshold of NT$2,100,000, as established through standard WTP research methods and in alignment with World Health Organization recommendations.
RESULTS: The total discounted costs for ICD therapy and no ICD therapy were NT$1,664,259 and NT$646,396 respectively for PP, while they were NT$2,410,603 and NT$905,881 respectively for 1.5PP. Total discounted QALYs for ICD therapy and no ICD therapy were 6.48 and 4.98 respectively for PP, while they were 10.78 and 7.71 respectively for 1.5PP. The incremental cost effectiveness ratio was NT$708,711 for PP and NT$441,153 for 1.5PP, therefore ICD therapy should be considered cost effective for PP and highly cost effective for 1.5PP.
CONCLUSIONS: ICD therapy compared to no ICD therapy is cost-effective in the whole PP population and highly cost-effective in the subset 1.5PP population in Taiwan.

Introduction

Evidence for the use of implantable cardioverter defibrillators (ICDs) for primary prevention of sudden cardiac arrest (SCA) in patients with moderately symptomatic heart failure and reduced systolic function is well-established through multiple randomized clinical trials [1, 2] and confirmed in real-world observational evidence [3, 4]. This evidence has led to strong recommendations for ICD use in society guidelines [5, 6] and has been leveraged to establish the cost-effectiveness of ICD therapy in multiple healthcare systems [7, 8]. Despite this strong evidence base, ICD therapy remains underutilized globally, due at least in part to cost considerations and lack of reimbursement [9]. ‎

The Improve SCA study has identified a high-risk subset of primary prevention patients called 1.5 primary prevention based on the presence of at least one of the following documented risk factors: non-sustained ventricular tachycardia (NSVT), frequent premature ventricular contractions (PVCs) >10/h, left ventricular ejection fraction (LVEF) <25%, pre-syncope or syncope [10]. Improve SCA patients with 1.5 primary prevention characteristics were found to have a higher rate of treatment with appropriate therapy than primary prevention patients, and when treated with an ICD, 1.5 primary prevention patients experienced a 49% relative risk reduction in all-cause mortality.

While the cost-effectiveness of ICD therapy for primary prevention patients has been established in western countries, it has not been previously established for the healthcare system in Taiwan. Furthermore, the cost-effectiveness of ICD therapy for 1.5 primary prevention patients is not well known. The 1.5 primary prevention cohort could be used to prioritize health care resources in geographies where such resources are insufficient to cover the full primary prevention population. The aim of the present study was to critically evaluate the cost-effectiveness models of ICD therapy for both the superset of primary prevention and the subset 1.5 of primary prevention patients with heart failure in the Taiwan healthcare system, and to identify the main factors influencing the cost-effectiveness of ICD therapy [11].

Methods

We used an existing Markov decision model [7] to estimate the lifetime cost, quality of life, survival, and incremental cost-effectiveness of ICD therapy versus no ICD therapy for a Taiwanese population at risk for SCA (both primary prevention and 1.5 primary prevention). The Improve SCA study [11] protocol was approved by the Institutional Review Board or Medical Ethics Committee of each respective study center. This analysis is a modeling exercise based on previously published data from the Improve SCA study [11] and does not involve any additional human research. No ICD therapy was selected as the control instead of pharmacologic therapy based SCD-HeFT study findings that indicated no significant difference in the risk of death between treatment with amiodarone and treatment with a placebo [1]. Model inputs are shown in Table 1 and described in detail, below. The model was implemented in Microsoft Excel, as described previously [7].

Table 1

Model input parameters.

Model Parameters	Base Case Value	Standard Error	Distribution	Reference
Monthly Risk of Mortality (ICD Therapy 1.5PP)
Sudden cardiac death	0.0007	0.0003	Beta	[11]
Non-sudden cardiac death	0.0014	0.0004	Beta
Non-cardiac death	0.0005	0.0003	Beta
Unknown death	0.0013	0.0003	Beta
Monthly Risk of Mortality (No ICD Therapy 1.5PP)
Sudden cardiac death	0.0028	0.0005	Beta	[11]
Non-sudden cardiac death	0.0021	0.0004	Beta
Non-cardiac death	0.0010	0.0004	Beta
Unknown death	0.0014	0.0004	Beta
Monthly Risk of Mortality, ICD Therapy (Primary Prevention)
Sudden cardiac death	0.0015	0.0001	Beta
Heart failure death	0.0029	0.0002	Beta
Other cardiac death	0.0004	0.00002	Beta
Non-cardiac death	0.0024	0.0002	Beta
Monthly Risk of Mortality, No ICD Therapy (Primary Prevention)
Sudden cardiac death	0.0042	0.0004	Beta	[7]
Heart failure death	0.0029	0.0003	Beta
Other cardiac death	0.0002	0.00002	Beta
Non-cardiac death	0.0031	0.0003	Beta
ICD-Related Probabilities
Initial operative death	0.0002	0.00002	Beta	[12]
Continue ICD therapy after shock	0.0034	0.0002	Beta	[13–17]
Discontinue ICD therapy after shock	0.0001	0.00007	Beta
Lead replacement (initial implant)	0.0004	0.0005	Beta	[18, 19]
Lead replacement (replacement implant)	0.0008	0.0009	Beta	[20]
Lead dislodgement (initial implant)	0.018	0.0012	Beta
Lead dislodgement (replacement implant)	0.005	0.0009	Beta	[13]
ICD infection (initial implant)	0.0244	0.0049	Beta	[21]
ICD infection (replacement implant)	0.0432	0.0064	Beta	[22]
Costs, 2017 New Taiwan Dollar (NT$)
ICD implant procedure (initial)	NT$633,678			[23]
ICD implant procedure (replacement)	NT$315,513
Lead replacement	NT$62,757
ICD infection	NT$765,213
ICD lead dislodgement	NT$61,566
ICD generator removal	NT$64,290
ICD inappropriate shocks	NT$670			[24]
Monthly inpatient cost	NT$6,828			[25]
Monthly outpatient cost	NT$858
Discount rate	1.375%			[26]
Utility Primary Prevention
Annual utility	0.7315	0.0126	Beta	[14]
ICD complication state	0.6474	0.112	Beta
Utility 1.5 Prevention
Annual utility	0.8683	0.0360	Beta	[27]
ICD complication state	0.7685	0.0360	Beta

Abbreviations: ICD, Implantable Cardioverter-Defibrillator

Model structure

The model is structured as a decision tree with two treatment arms, ICD therapy or no ICD therapy, followed by consecutive Markov models (Fig 1). The model was run in two separate scenarios, following a simulated cohort of 1,000 patients with a standard indication for primary prevention ICD therapy, and for patients with a standard indication for primary prevention ICD therapy with one or more 1.5 primary prevention risk factors. Patients who enter the model in the ICD arm are at an initial risk of operative death or survival. Patients who survive the ICD surgery enter the Markov model in the well state. From the well state, ICD patients stay well or progress to ICD complications, sudden cardiac death, non-sudden cardiac death, non-cardiac death, unknown death. Patients remain in the same state or progress to a different state at the beginning of each cycle, except for the complication state. Patients who experience an ICD complication remain in the complication state for only one cycle, then progress to continued ICD therapy or discontinued ICD therapy. In the event of therapy discontinuation, ICD patients stay well without ICD treatment or progress to sudden cardiac death, non-sudden cardiac death, non-cardiac death, unknown death. Patients in the no ICD arm enter the model in the well state and remain well or progress to sudden cardiac death, non-sudden cardiac death, non-cardiac death, unknown death.

Fig 1

Model schematic.

Patients incur costs and effects by progressing through the model in monthly increments over a lifetime (420 months); a lifetime perspective allows the model to account for all costs incurred by patients that survive without a sudden cardiac arrest event. Patients in both treatment arms incur monthly inpatient and outpatient costs. In the ICD therapy arm, patients also incur the cost of the device and ICD implant procedure. ICD patients who remain alive long enough to require a device replacement incur additional device and procedure costs at the time of replacement. ICD patients may receive an inappropriate shock or other ICD-related complication that incurs a cost and affects treatment adherence. After experiencing an inappropriate shock or other ICD-related complication, patients remain in the ICD therapy arm with ICD treatment or progress to discontinued use of ICD therapy. We assumed ICD patients who discontinue use of ICD therapy have the same mortality risk as patients in the no ICD arm.

Clinical data

Mortality inputs to the model for primary prevention patients were based on a previously published meta-analysis and for 1.5 primary prevention patients were based on Improve SCA clinical study results. Primary prevention patients had a mean age of 61.1 years and were 76.3% male, while 1.5 primary prevention patients had a mean age of 61.1 years and were 79.5% male; other characteristics for each population are included in the original study publications [7, 11]. Other clinical inputs to the model for both ICD cohorts were based on the United States (US) National ICD registry, literature, and administrative claims-based analyses. The probability of implant-related operative death (0.0002) was based on the US National ICD Registry and applied only to the ICD treatment arms [12].‎ Inappropriate shock probability was derived from a weighted average based on the MADIT RIT, ADVANCE III, PROVIDE, and PainFree SST clinical trials that demonstrated a reduction in inappropriate shock rates due to device programming [13-16]. Probabilities of lead failure or dislodgement after initial implant were based on studies of annual incidence of lead failure and ICD lead dislodgement at one year after implant, 0.45% and 1.8% respectively [18, 19]. ‎ Probability of lead dislodgement or replacement after ICD replacement was based on data from the REPLACE registry that reported a 1% combined dislodgement and replacement rate [20]. We assumed half of the combined rate reported in the REPLACE registry could be attributed to lead failure (0.5%) and half could be attributed to lead dislodgement (0.5%). We estimated the one-year probability of lead infection after initial implant (1.22%) and after device replacement (2.16%) with a retrospective data analysis based on administrative claims from a large US insurer [21]. The lifetime risk of lead infection after the first year of an initial or replacement implant was double the value of the one-year claims-based probability [22].

Economic data

Device related costs and long-term health care utilization costs associated with heart disease were modeled over a lifetime. Costs for individual events were assumed to be the same regardless of the indication (primary prevention or 1.5 primary prevention) for the ICD. The procedural costs of an initial ICD implant, subsequent revision or replacement, and ICD-related complications (infection and dislodgement) which including cost of devices, admission fee, drug fee, examination etc. were derived from the Taiwan Diagnosis-Related Group (Tw-DRG), edition 3.4 of the National Health Insurance Administration (NHIA), Ministry of Health and Welfare [23]. The procedural cost of inappropriate shocks and monthly long-term inpatient and outpatient costs were derived for the base case by evaluating the practice and calculated the cost base on "Fee Schedule for Medical Services of National Health Insurance" of the National Health Insurance Administration (NHIA), Ministry of Health and Welfare [24, 25]. Monthly inpatient and outpatient costs related to heart failure were estimated from the NHIA inpatient expenditures by admissions (DD) database and NHIA outpatient expenditures (OD) database, respectively.

Health-related quality of life

Quality of life was based on an analysis of EQ-5D data collected in the PainFree SST clinical trial [28]. Taiwan-specific utilities were derived by mapping each patient’s EQ-5D state using country specific societal preferences (S2 Table) [29]. We assumed the baseline utility for ICD patients and no ICD patients was the same [30]. Patients who experienced an ICD-related complication received a short-term utility decrement of 0.096 that is equivalent to 3.5 days [31].

Construction of the ICER (w/WTP) and sensitivity analysis

Total lifetime costs and quality-adjusted life years (QALYs) between ICD therapy and no ICD therapy were simulated to calculate the incremental cost effectiveness ratio (ICER). Both undiscounted and discounted results were calculated to best represent the time value of costs and outcomes. We conducted one-way sensitivity and probabilistic sensitivity analyses to assess the impact of model inputs and parameter uncertainty. Beta distributions were used for probabilities and utilities in the probabilistic sensitivity analysis. We used a willingness-to-pay (WTP) threshold value of NT$2.1 million for this analysis [32].

Results

Base case scenario for primary prevention

Table 2 shows the results of the base-case scenario for primary prevention.

Table 2

Base case scenario results (primary prevention and 1.5 primary prevention).

PP Base Case Scenario Results		ICD therapy	No ICD Therapy
Undiscounted	Aggregated costs	NT$1,785,966	NT$646,396
	Differential cost	NT$1,139,570
	Effectiveness (life-years saved)	9.82	7.36
	Effectiveness (QALY saved)	7.16	5.39
	Differential effectiveness (QALY)	1.77
	ICER (costs per QALY saved)	NT$642,272
Discounted	Aggregated costs	NT$1,664,259	NT$597,087
	Differential cost	NT$1,067,172
	Effectiveness (life-years saved)	8.88	6.80
	Effectiveness (QALY saved)	6.48	4.98
	Differential effectiveness (QALY)	1.51
	ICER (Costs per QALY saved)	NT$708,711
1.5PP Base Case Scenario Results		ICD therapy	No ICD Therapy
Undiscounted	Aggregated costs	NT$2,410,603	NT$905,881
	Differential cost	NT$1,504,722
	Effectiveness (life-years saved)	14.17	9.82
	Effectiveness (QALY saved)	12.27	8.53
	Differential effectiveness (QALY)	3.75
	ICER (costs per QALY saved)	NT$401,722
Discounted	Aggregated costs	NT$2,175,478	NT$818,782
	Differential cost	NT$1,356,695
	Effectiveness (life-years saved)	12.45	8.88
	Effectiveness (QALY saved)	10.78	7.71
	Differential effectiveness (QALY)	3.08
	ICER (Costs per QALY saved)	NT$441,153

Abbreviations: PP, Primary Prevention; ICD, Implantable Cardioverter-Defibrillator; QALY, quality-adjusted life year; ICER incremental cost-effectiveness ratio.

ICD therapy for primary prevention resulted in a benefit of 8.88 (discounted) and 9.82 (undiscounted) life-years saved, while no ICD therapy resulted in a benefit of 6.80 and 7.36 life-years saved, respectively. Measured in QALYs, the discounted benefit from ICD therapy is 6.48 and 4.98 from no ICD therapy, resulting in an incremental effectiveness of 1.51 QALYs. Discounted costs from ICD therapy and no ICD therapy account for NT$1,664,259 and NT$597,087, respectively. The ICER for ICD therapy is NT$708,711 per QALY; ICD therapy for primary prevention is cost-effective at NT$2.1 million.

Base case scenario for 1.5 primary prevention

Table 2 (above) also shows the results of the base-case scenario for 1.5 primary prevention. ICD therapy for 1.5 primary prevention resulted in a benefit of 12.45 (discounted) and 14.17 (undiscounted) life-years saved, while no ICD therapy resulted in a benefit of 8.88 and 9.82 life-years saved, respectively. Measured in QALYs, the discounted benefit from ICD therapy is 10.78 and 7.71 from no ICD therapy, resulting in an incremental effectiveness of 3.08 QALYs. Discounted costs from ICD therapy and no ICD therapy account for NT$2,175,478 and NT$818,782, respectively. The ICER for ICD therapy is NT$441,153 per QALY; ICD therapy for 1.5 prevention is below one third of the WTP and is highly cost-effective at NT$2.1 million.

Sensitivity analyses

Results of the sensitivity analyses are presented for the 1.5 primary prevention indication only (results for the primary prevention indication are in the S1 Table). Results of the one-way sensitivity analyses show that costs per QALY are more responsive to the age at implant, conventional mortality, replacement period and quality of life (Fig 2, Panel A). The incremental costs per QALY remained below the WTP threshold for all values of the one-way sensitivity analysis.

Fig 2

Fig 2, Panel B shows the simulated costs per QALY of the probabilistic sensitivity analysis. Results for 1.5 primary prevention show a mean cost per QALY of NT$434,053 (median cost per QALY of NT$435,301, 95-percent Credible Interval [NT$291,938 –NT$1,115,321] per QALY) after 1,000 iterations. For 1.5 primary prevention, 99.5 percent of simulations result in costs per QALY below the WTP threshold, and 90 percent of simulations result in costs per QALY below one third of the WTP threshold, indicating ICD therapy is highly cost effective for this population.

Discussion

Our results indicate that ICD therapy is cost effective for the whole primary prevention patient population and highly cost effective for the subset of 1.5 primary prevention patients in the Taiwan healthcare system. The primary prevention and 1.5 primary prevention are at ICERs of NT$708,711 per QALY and NT$441,153 per QALY respectively; while 1.5 primary prevention was more cost effective, both are less than the WTP value of NT$2.1 million. This finding is robust, with sensitivity analyses indicating that the cost effectiveness is preserved in nearly all reasonable variations of model inputs. To our knowledge, this is the first evaluation of the cost-effectiveness of ICD therapy compared to no ICD therapy among the whole primary prevention patient population and the subset of 1.5 primary prevention patients from the perspective of the public healthcare system in Taiwan.

Prior estimates of the cost effectiveness of ICD therapy have been performed in the primary prevention population. Mark et al [8]‎ performed an analysis of the randomized SCD-HeFT trial and found ICD therapy to be economically attractive at $41,530/QALY (at a WTP of $100,000) in the US healthcare system. An analysis in the healthcare system of a European country using a meta-analysis of six randomized primary prevention trials and the same model used in this study showed similar results [7].‎ The cost-effectiveness of ICD therapy has also been confirmed in a real world setting outside of clinical trials [33].‎ Our study shows that ICD therapy for primary prevention patients cost less than 1 GDP per capita per QALY and appear even more cost-effective in the Taiwan healthcare system compared to previous reports in other countries (S1 Table).

Despite convincing evidence from multiple randomized clinical trials [1-3], and strong recommendations in international society guidelines [5, 6], ICD therapy remains underutilized in Asian countries [34, 35]. In particular, a report by Chia et al in 2017 [9] found that among guideline recommended ICD-eligible patients in Taiwan, only 7.7% had received ICD therapy, and by comparison, in Japan 52.5% of ICD-eligible patients had received ICD therapy [14]. National health expenditure (NHE) in Taiwan was 6.1% of GDP in 2017 [36] approximately one-third of the US (17.1%) and 69% of the average for OECD (Organization for Economic Cooperation and Development) countries (8.8%). Health spending per capita in 2017 in Taiwan was PPPUS$3,047 [36] less than one-third (30%) of the US total (PPPUS$10,209) and 76% of the average for OECD countries (PPPUS$3,992). To the extent that economic factors play a role, this study provides information for decision makers to direct scarce resources first toward those who can benefit the most. While it remains cost effective to treat the entire primary prevention population with ICD therapy, from an economic standpoint a priority could be placed on treating patients with a 1.5 primary prevention indication.

It is important to acknowledge the limitations of this analysis. The Improve SCA trial was not randomized, however the mortality benefit from the trial remained significant after adjusting for baseline characteristics likely to have an impact on mortality, and the effectiveness of ICD therapy for the primary prevention population has been shown to be replicated in both randomized and non-randomized observational trials. Costs and benefits were modeled beyond the timeline of direct observation in the Improve SCA trial, however this is a standard approach in economic modeling and necessary for the proper perspective for decision makers. For the full primary prevention analysis ICD effectiveness data were taken from a global meta-analysis that did not include Taiwan, however such data from Taiwan were not available. For the 1.5 primary prevention analysis patients in the Improve SCA trial were not all from Taiwan, yet the majority were from Asia and ICD therapy application is well developed and largely standardized around the world. Conclusions from this report may not be generalizable beyond the Taiwan healthcare system.

Conclusion

ICD therapy is cost effective for primary prevention patients in the Taiwan healthcare system, and highly cost effective for 1.5 prevention patients. These data provide guidance as to an efficient way to address underutilization of ICD therapy in indicated patients in Taiwan.

(TIF)

Click here for additional data file.

(TIF)

Click here for additional data file.

Characteristics and result of economic evaluations of ICD for primary prevention.

(DOCX)

Click here for additional data file.

Taiwan utility analysis results.

(DOCX)

Click here for additional data file.

We have included a trace of the base case calculations for each patient population as a supplementary files, which give transparent and granular information on the calculations performed to produce the results.

This, together with the information in the manuscript (the specific model inputs and sources, and the structure of the model) is meant to allow for sufficient transparency to understand how the results were produced. S1 File, PP patients. S2 File, 1.5PP patients.

(XLS)

Click here for additional data file.

This, together with the information in the manuscript (the specific model inputs and sources, and the structure of the model) is meant to allow for sufficient transparency to understand how the results were produced. S1 File, PP patients. S2 File, 1.5PP patients.

(XLS)

Click here for additional data file.

19 Aug 2020

PONE-D-20-18597

Implantable cardioverter defibrillator therapy is cost effective for primary prevention patients in Taiwan: an analysis from the Improve SCA trial

PLOS ONE

Dear Dr. Yen Bin Liu

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 02 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Giuseppe Coppola

Academic Editor

PLOS ONE

Additional Editor Comments:

Kind authors, your paper underwent 3 different reviewers. Please read carefully comments and criticisms by statistic’s reviewer

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

2. Thank you for stating the following in the Financial Disclosure section:

'The overall work was funded by Medtronic, plc. Mr. Holbrook, Mr. Higuera, Ms. Wherry, and Mr. Phay are Medtronic, plc. employees and stockholders. Drs. Hsieh, Lin, and I are consultants for Medtronic, Inc. and have not received compensation for the participation in this work.'

We note that one or more of the authors have an affiliation to the commercial funders of this research study: Medtronic.

2.1. Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

2.2. Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc.

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If this adherence statement is not accurate and  there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

4. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files.

5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The paper shows a very interesting work that confirm the cost-effectiveness of ICD therapy. It's particularly interesting the sub-categorization of primary prevention patients. This sub-category includes patients at higher risk and therefore more likely to benefit from ICD implantation. In conclusion, a well-designed and interesting work although limited to the health reality of Taiwan

Reviewer #2: Excellent work, well written and structured, which further confirms the importance of ICD implantation in primary prevention of sudden cardiac death, also proving the cost-effectivenes of this practice in a particular subcategory of Taiwan population.

Reviewer #3: PONE-D-20-18597: statistical review

SUMMARY. This paper introduces a Bayesian network (called a Markov decision model by the authors) to evaluate the cost-effectiveness of Implantable cardiac defibrillators (ICD) compared to no ICD therapy for a

Taiwanese population at risk for sudden cardiac arrest (SCA). The network is initialized using input parameters obtained by multiple sources (Table 1) and exploited to estimate costs and life years saved (Table 2). Although the choice of a Bayesian network is potentially a correct method to evaluate cost-effectiveness, the paper lacks detailed methods (see major issues 1-5). Without these details, it is unclear whether the paper represents a technically sound and reproducible piece of scientific research.

MAJOR ISSUES

1. Study population. The study population of the primary prevention patients is not defined. A table should be provided with the main characteristics of the subjects under analysis.

2. Input parameters. Parameter values are obtained from multiple sources with different degrees of uncertainty (see the standard errors of Table 1). How were these standard errors integrated in the analysis?

3. Estimation. Nothing is said about the estimation procedure exploited to obtain the output of Table 2. How were these estimates obtained? In addition, standard errors should be provided along with the estimates. The output of a Bayesian network is typically the probability distribution of the output variables, given the inputs. What do the outputs of Table exactly represent? The expectation of the probability distribution? Or its mode?

4. The excel sheet that has been exploited to obtain the results should be provided as a supplementary information file, to guarantee reproducibility.

5. Details about the sensitivity analysis should be provided. How many times was the network simulated? Which values of the variables of Figure 2, panel A, were exploited?

SPECIFIC COMMENTS

1. Table 1 includes a strange statement "Error! Reference source not found". Please check. In addition: what is the "beta distribution" mentioned in the table?

2. Figure 1. Transition probabilities should be attached to each arc of the graph.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Gianfranco Ciaramitaro, M.D. Ph.D

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

31 Aug 2020

Dear Reviewers,

Thank you for your thoughtful review and questions. Please refer to the "Reviewer Response Letter" for the following responses to your questions.

1. Study population. The study population of the primary prevention patients is not defined. A table should be provided with the main characteristics of the subjects under analysis.

The study populations (both primary prevention and 1.5 primary prevention) are not new cohorts, they have been previously published. We have added general characteristics to this manuscript and given a clear reference to the publications where more detailed characteristics can be found. The following sentence was added to the methods section of the manuscript in the subheading “Clinical Data”: Primary prevention patients had a mean age of 61.1 years and were 76.3% male, while 1.5 primary prevention patients had a mean age of 61.1 years and were 79.5% male; other characteristics for each population are included in the original study publications.

2. Input parameters. Parameter values are obtained from multiple sources with different degrees of uncertainty (see the standard errors of Table 1). How were these standard errors integrated in the analysis?

The analysis was done in two parts, following standard cost-effectiveness analysis methodology (Ramsey SD, Willke RJ, Glick HA, et al. Cost-effectiveness analysis alongside clinical trials II: an ISPOR Good Research Practices Task Force Report. Value Health. 2015;18(2):161-172).

The first part is a deterministic base case analysis, in which the base case inputs from Table 1 are entered into the model as point estimates and the ICER is calculated. The result of the base case analysis, which does not use the standard error values, is documented in Table 2. This ICER is deterministic and does not account for any variability and results in a single numeric point estimate for the ICER. The following sentence in the methods section was modified to make this clearer:

Total lifetime costs and quality-adjusted life years (QALYs) between ICD therapy and no ICD therapy were simulated using base case model inputs to calculate the deterministic incremental cost effectiveness ratio (ICER).

After completion of the deterministic base case analysis, we performed 2 different types of sensitivity analyses. The first is the one-way sensitivity analysis, which is also deterministic, but varies a single input to its high and low values while holding all other inputs constant resulting in a set of numeric point estimate ICER values which are represented in Figure 2, panel A. We updated the following sentence in the methods section to more clearly indicate that the one-way sensitivity analysis is deterministic (the probabilistic sensitivity analysis is already aptly named):

We conducted deterministic one-way sensitivity and probabilistic sensitivity analyses to assess the impact of model inputs and parameter uncertainty.

The second sensitivity analysis was a probabilistic sensitivity analysis, where input parameters were all varied simultaneously by selecting random values from a probability distribution based on the mean and standard error values of the parameters. This analysis produced the findings in the final paragraph of the results section:

Results for 1.5 primary prevention show a mean cost per QALY of NT$434,053 (median cost per QALY of NT$435,301, 95-percent Credible Interval [NT$291,938 – NT$1,115,321] per QALY) after 1,000 iterations. For 1.5 primary prevention, 99.5 percent of simulations result in costs per QALY below the WTP threshold, and 90 percent of simulations result in costs per QALY below one third of the WTP threshold, indicating ICD therapy is highly cost effective for this population.

3. Estimation. Nothing is said about the estimation procedure exploited to obtain the output of Table 2. How were these estimates obtained? In addition, standard errors should be provided along with the estimates. The output of a Bayesian network is typically the probability distribution of the output variables, given the inputs. What do the outputs of Table exactly represent? The expectation of the probability distribution? Or its mode?

As stated in the response to question #2 above, the results in Table 2 represent a single point estimate and are considered the base case results per standard cost effectiveness analysis methodology. Uncertainty and variability of the model are addressed with the sensitivity analyses which are also described in the response to question #2 above.

4. The excel sheet that has been exploited to obtain the results should be provided as a supplementary information file, to guarantee reproducibility.

We have included a trace of the base case calculations for each population as supplementary files, which gives transparent and granular information on the calculations performed to produce the results. This, together with the information in the manuscript (the specific model inputs and sources, and the structure of the model) is meant to allow for sufficient transparency to understand how the results were produced.

5. Details about the sensitivity analysis should be provided. How many times was the network simulated? Which values of the variables of Figure 2, panel A, were exploited?

Some of these details, including how many times the network was simulated, are included in the manuscript:

Results for 1.5 primary prevention show a mean cost per QALY of NT$434,053 (median cost per QALY of NT$435,301, 95-percent Credible Interval [NT$291,938 – NT$1,115,321] per QALY) after 1,000 iterations. For 1.5 primary prevention, 99.5 percent of simulations result in costs per QALY below the WTP threshold, and 90 percent of simulations result in costs per QALY below one third of the WTP threshold, indicating ICD therapy is highly cost effective for this population.

The variables that could change in the probabilistic sensitivity analysis were the variables that included standard error values in Table 1. The following sentence was added to the last paragraph in the methods section to make that clearer for the reader:

For the probabilistic sensitivity analysis, the inputs that were varied were the ones with standard errors reported in Table 1, and for each run of the model those inputs were randomly varied according to a beta probability distribution based on the base case and standard error values of the corresponding input.

SPECIFIC COMMENTS

1. Table 1 includes a strange statement "Error! Reference source not found". Please check. In addition: what is the "beta distribution" mentioned in the table?

Thanks for pointing out the missing reference error message, it is now corrected. Regarding the beta distribution, see the answer to comment #5 above.

2. Figure 1. Transition probabilities should be attached to each arc of the graph.

It is not common for publications on cost effectiveness of medical therapies to include transition probabilities in the figure on the model structure (see references included below). There are many probabilities, and they are quite precise, so inclusion on the graph directly would cloud understanding of the general state transitions and pathways inherent to the model structure, which is the purpose of the figure. There is full transparency in the manuscript, as the transition probabilities are all included in Table 1 in the description of model inputs.

Representative publications on the cost effectiveness of medical therapies that include figures on model structure while reporting transition probabilities separately:

• Sanders GD et al. Cost-Effectiveness of Implantable Cardioverter–Defibrillators. N Engl J Med. 2005 Oct 6;353(14):1471-80.

• Gada H et al. Markov Model for Selection of Aortic Valve Replacement Versus Transcatheter Aortic Valve Implantation (Without Replacement) in High-Risk Patients. Am J Cardiol. 2012 May 1;109(9):1326-33.

• Canestaro WJ et al. Cost-Effectiveness of Oral Anticoagulants for Treatment of Atrial Fibrillation. Circ Cardiovasc Qual Outcomes. 2013 Nov;6(6):724-31

Dear Reviewers, the cover letter, ethics statement, manuscript and reviewer response letter has been updated to address you questions included in your 29 August 2020 email. Please review these revisions. Thank you, Dr Liu

Submitted filename: Response to Reviewers 31 Aug 2020. .docx

Click here for additional data file.

20 Oct 2020

Implantable cardioverter defibrillator therapy is cost effective for primary prevention patients in Taiwan: an analysis from the Improve SCA trial

PONE-D-20-18597R1

Dear Dr. Liu,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Andrea Ballotta

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Congratulations nulla obsta for the publication of the above mentioned manuscript.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: No

30 Oct 2020

PONE-D-20-18597R1

Implantable cardioverter defibrillator therapy is cost effective for primary prevention patients in Taiwan: an analysis from the Improve SCA trial

Dear Dr. Liu:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Andrea Ballotta

Academic Editor

PLOS ONE