A Cost-Effectiveness Analysis for the Combination of Universal Screening at 9-10 Years Old and Reverse Cascade Screening of Relatives for Familial Hypercholesterolemia in Japan

Keiji Matsunaga; Mariko Harada-Shiba; Shizuya Yamashita; Hayato Tada; Akihito Uda; Katsuya Mori; Mizuki Yoshimura; Sachie Inoue; Isao Kamae; Shinji Yokoyama; Tetsuo Minamino

doi:10.5551/jat.65181

Abstract

Aim: Screening for familial hypercholesterolemia (FH) is important for reducing the incidence of cardiovascular diseases (CVDs). Cost-effectiveness was evaluated using the Kagawa FH screening model, which is a combination of universal screening (US) in the universal health examination for children 9–10 years old conducted in Kagawa Prefecture, and reverse cascade screening (RCS) of the probands’ relatives.

Methods: A lifetime simulation was conducted using mathematical models (decision tree and Markov model) to determine the cost-effectiveness of introducing a series of FH screenings (US in children + RCS in adult relatives). Only screening-related costs and direct medical costs were included, using quality-adjusted life years (QALYs) as an outcome. The costs of statins were estimated using the public health insurance claims database DeSC Healthcare, Inc. The risk of each CVD event was estimated using the same claims data and adjusted for age. We hypothesized that standard statin treatment decreases CVD risk by reducing plasma low-density lipoprotein cholesterol levels.

Results: A series of FH screenings (US in children + RCS in adult relatives) was cost-effective compared to no screening, with an incremental cost-effectiveness ratio (ICER) of approximately JPY 150,000 (USD 1,042)/QALY, which was below the willingness-to-pay threshold of JPY 5,000,000 (USD 34,722)/QALY for medical technology in Japan (USD 1 = JPY 144). The ICER for the US without RCS was also acceptable at approximately JPY 2,720,000 (USD 18,889)/QALY.

Conclusion: The cost-effectiveness analysis revealed that a series of FH screenings (US in children + RCS in adult relatives) based on the Kagawa model was cost-effective.

See editorial vol. 32: 926-928

Introduction

Familial hypercholesterolemia (FH) is an autosomal dominant genetic disorder with a prevalence of approximately 1 in 300 in the general population^1-3). The diagnosis rate of FH in Japan was reported to be approximately 2.6% by Tada et al.⁴⁾. Hyper-low-density lipoprotein (LDL)-cholesterolemia is asymptomatic, but it may cause various cardiovascular diseases (CVDs). As FH is associated with hyper-LDL-emia from the early stage of life, the cumulative LDL-cholesterol (LDL-C) level reaches the coronary artery disease threshold at a relatively young age⁵⁾. Therefore, management of plasma LDL-C should be started as early as possible. An early diagnosis and strict treatment are key to preventing premature coronary disease.

Matsunaga et al. reported that, among 15,665 children who underwent universal screening (US) for FH using the plasma LDL-C level in the Health Examination for Prevention of Lifestyle Diseases for Children 9–10 years old in Kagawa, approximately 60% of those referred to a medical institution with LDL-C levels ≥ 140 mg/dL had positive genetic testing results for FH⁶⁾. Based on this prefecture-wide administrative model, the early diagnosis of FH has been promoted in Kagawa by performing reverse cascade screening (RCS) in the relatives of children diagnosed with FH in the health examination program (hereafter referred to as the Kagawa model)⁶⁾.

Cost-effectiveness had been estimated for the US of FH with and without RCS. McKay et al. evaluated US screening strategies for FH compared to no screening in children one to two years old in the UK. They concluded that a strategy involving cholesterol screening, diagnostic genetic testing, and RCS was the most cost-effective⁷⁾. Jahn et al. conducted a systematic review of studies evaluating the cost-effectiveness of FH screening and concluded that results may vary depending on the screening methods and clinical criteria combined⁸⁾. They also noted that caution is required in the interpretation of the results, as the background of the analyses can differ among countries.

Owing to the problems described above, it is important to consider the regional background of public health and medical care systems to develop an appropriate screening model and method of evaluation. In addition, no analysis has been attempted from the viewpoint of healthcare economics for FH in the US in Japan.

Aim

This study aimed to estimate the cost-effectiveness of a series of FH screenings (US in children + RCS in adult relatives) based on the Kagawa model in Japan compared to no screening. Scenario analyses were conducted to evaluate the health economic benefits for the potential improvement of the policy-making process for the healthcare system.

Methods

Overview and Model Structure

• Overview of Kagawa’s FH Screening

The flow of the Kagawa model is illustrated in Fig.1. In US, children with serum LDL-C levels ≥ 140 mg/dL in a health examination performed in elementary schools for the prevention of lifestyle diseases among children 9–10 years old are encouraged to see a doctor at a local medical facility, and a diagnosis is made at designated hospitals via genetic testing (detailed examination) as required if FH is suspected after excluding the possibility of other diseases (secondary hyper-LDL cholesterolemia, etc.). In the RCS for children with FH, genetic testing was conducted for adult relatives (assuming six second-degree relatives).

Fig.1. Universal screening and reverse cascade screening for FH

FH, familial hypercholesterolemia; LDL-C, low-density lipoprotein cholesterol; RCS, reverse cascade screening.

• Analysis Framework

The target population was children 9–10 years old, and the target strategy was a series of US+RCS (including genetic testing) via the LDL-C test (primary screening) in a child health examination, with the efficacy of this approach then compared with that of no screening at all. The analysis was conducted from the perspective of a public healthcare payer.

A combination of the decision tree model (US+RCS) and Markov model (post-US+RCS) was used, and the time horizon was the lifetime. The decision tree model was based on McKay’s cost-utility analysis⁷⁾ to reflect the Kagawa model, and the Markov model was based on the cost-effectiveness model of the National Institute for Health and Care Excellence Technology Appraisal (NICE TA) 733 ⁹⁾. The discount rate is 2% per year for both cost and effectiveness.

• Model Structure

The decision tree model is presented in Figs.2A-C. Fig.2A shows the classification of the target population, Fig.2B shows the content of the FH screening (US+RCS) for the target population (with a series of FH screenings), and Fig.2C shows the content of the comparator (without FH screening). Fig.2A shows the classification and proportions of FH patients (FH[+] or FH[-]) and the genetic information resulting in a diagnosis of FH (Pathogenic Variant[PV][+] or PV[-]), which were combined to allow classification into four populations (FH[+]&PV[+], FH[+]&PV[-], FH[-]&PV[+], and FH[-]&PV[-]). Fig.2B shows the screening flow for the four classified groups based on the diagnostic flow of the Kagawa model (presence/absence of pediatric health checkups, consultation with a doctor at a local medical facility, and genetic testing when FH is suspected). The group diagnosed as FH(+) included actual FH patients with a high CVD risk. For children diagnosed as FH(+), the standard of care (SOC) (statins) is immediately begun, or diet and exercise therapies are provided if SOC treatment is avoided. RCS is performed for second-degree adult relatives in this case. If a child is not diagnosed with FH(+), SOC treatment is not initiated, even if they have a high risk of CVD, and RCS is not performed. RCS is assumed to be performed for up to six relatives, with up to two additional relatives receiving SOC if PV(+) is detected via genetic testing ⁷⁾ . For the other three groups (FH[+] and PV[-], FH[-] and PV[+], and FH[-] and PV[-]), screening, FH treatment, and RCS are performed. The FH(-) group included patients with a low risk of CVD, but FH treatment and/or RCS were performed depending on the diagnosis. However, treatment effects and/or detection of adult relative FH(+) using RCS were not obtained.

Fig.2 .Model structure of decision tree model

(A) Overall (child health examination + family physician consultation + genetic testing + RCS). (B) FH(+) & PV(+). (C) No screening. ^aWith RCS: It is assumed that two new PV(+) can be found in adult relatives via genetic testing conducted on six second-degree relatives per 1 FH child. ^bA group not diagnosed with FH with an LDL-C level ≥ 140 mg/dL was assumed. FH treatment other than SOC (diet and exercise therapies) was administered. CVD, cardiovascular disease; FH, familial hypercholesterolemia; LDL, low-density lipoprotein; LDL-C, low-density lipoprotein cholesterol; PV, pathogenic variant; RCS, reverse cascade screening; SOC, standard of care; US, universal screening; NA, not available.

The decision-tree model of the comparator (no screening) is shown in Fig.2C. The clinical validity of the flow and each value setting in Japan (Supplementary Table 1) was confirmed by several medical experts using the Kagawa model.

Supplementary Table 1.Model parameter

	Value	Settings in DSA			Settings in PSA			Ref
	Value	Lower limit	Upper limit	Range	Distribution	Parameters^a		Ref
Demographics
Analysis start age (year)
US	10	-	-	-	-	-	-	1)
RCS	42	-	-	-	-	-	-	2, 3)
Percentage of female
US	49%	-	-	-	-	-	-	1)
RCS	50%	-	-	-	-	-	-	Experts opinion
DM%
US	0.02%	-	-	-	-	-	-	4)
RCS	10%	-	-	-	-	-	-	Experts opinion
Baseline LDL-C (mg/dL)
US
FH(+) & SOC treatment	184	-	-	-	-	-	-	1)
FH(+) & diet and exercise therapies	152	-	-	-	-	-	-
FH(-)	93	-	-	-	-	-	-
RCS
FH(+)	248	-	-	-	-	-	-	5, 6)
LDL-C reduction effect by treatment
US
FH(+) & SOC treatment	32.7%	27.8%	37.6%	±15%	Beta	114.58	235.81	7)
FH(+) & diet and exercise therapies	8%	6.8%	9.2%	±15%	Beta	156.99	1805.42	8, 9)
RCS
FH(+)	32.7%	27.8%	37.6%	±15%	Beta	114.58	235.81	7)
FH epidemiological information
P(FH+)^b	0.396%	-	-	-	-	-	-	10, 11)
P(FH-)	99.604%	-	-	-	-	-	-	10)
P(PV+\|FH+)	61.2%	49.5%	72.9%	95%CI	Beta	40.39	25.61	1)
P(PV-\|FH+)	38.8%	-	-	-	-	-	-
P(PV+\|FH-)	0.015%	0.013%	0.018%	±15%	Beta	170.71	1108903.99	10)
P(PV-\|FH-)	99.985%	-	-	-	-	-	-
Screening
US
Child health examination rate	92.2%	78.4%	100.0%	±15%	Beta	20.86	1.76	1)
Percentage of LDL-C ≥ 140 mg/dL
FH(+)	88%	74.8%	100.0%	±15%	Beta	21.61	2.95	10)
FH(-)	0.1%	0.09%	0.12%	±15%	Beta	170.56	170389.21
Consultation rate of a doctor at a local medical facility	60%	51.0%	69.0%	±15%	Beta	67.69	45.13	Experts opinion
Genetic testing rate with suspected FH	11.6%	9.9%	13.3%	±15%	Beta	150.81	1149.28	1)
No genetic testing rate with suspected FH	50%	42.5%	57.5%	±15%	Beta	84.87	84.87	Experts opinion
Positive rate of genetic testing
FH(+)	100%^c	-	-	-	-	-	-	Assumption
FH(-)	0%^c	-	-	-	-	-	-
RCS^d
The number of adult relatives with RCS (genetic testing)	6	5.1	6.9	±15%	Gamma	170.73	0.04	Assumption
The number of patients with PV(+) detected by RCS	2	1.7	2.3	±15%	Gamma	170.73	0.01
Diagnosis rate of FH(+) in PV(+)	70%^e	-	-	-	-	-	-	10)
RR per unit change in LDL-C (1.0 mmol/L)
Revasc	0.75	0.64	0.86	±15%	Normal	0.75	0.0574	12)
UA	0.73	0.62	0.84	±15%	Normal	0.73	0.0559
MI	0.73	0.62	0.84	±15%	Normal	0.73	0.0559
Stroke	0.79	0.67	0.91	±15%	Normal	0.79	0.0605
CV death	0.84	0.71	0.97	±15%	Normal	0.84	0.0643
Event risk adjustment coefficient according to age (per year)
Other events	1.03	1.00^f	1.18	±15%	Normal	1.03	0.0788	13)
CV death	1.05	1.00^f	1.21	±15%	Normal	1.05	0.0804
Minimum age applying event risk adjustment coefficient (year)
Minimum age	20	-	-	-	-	-	-	DeSC analysis
Initiation rates of SOC treatment in the FH(+) & no SOC treatment (per year)
Initiation rate of treatment	1.04%	0.88%	1.20%	±15%	Beta	168.95	16075.81	DeSC analysis
Screening-related costs (JPY/screening) (USD 1 = JPY 144)
Child health examination^g	2,870 (USD 20)	2,440 (USD 17)	3,301 (USD 23)	±15%	Gamma	170.73	16.81	Experts opinion, 14)
A doctor at a local medical facility consultation^h	5,750 (USD 40)	4,888 (USD 34)	6,613 (USD 46)	±15%	Gamma	170.73	33.68
Genetic testing (child)ⁱ	50,000 (USD 347)	42,500 (USD 295)	57,500 (USD 399)	±15%	Gamma	170.73	292.86
Genetic testing (adult relative with RCS)^j	300,000 (USD 2,083)	-	-	-	-	-	-
FH treatment costs (JPY/year) (USD 1 = JPY 144)
With FH(+) diagnosis^k
14 years or younger	12,493 (USD 87)	10,619 (USD 74)	14,366 (USD 100)	±15%	Gamma	170.73	73.17	Experts opinion, 14, 15), DeSC analysis
15 years or older	18,759 (USD 130)	15,945 (USD 111)	21,572 (USD 150)	±15%	Gamma	170.73	109.87
Diagnosed with suspected FH(+)^l
14 years or younger	4,725 (USD 33)	4,016 (USD 28)	5,434 (USD 38)	±15%	Gamma	170.73	27.68	Experts opinion, 14)
15 years or older	4,725 (USD 33)	4,016 (USD 28)	5,434 (USD 38)	±15%	Gamma	170.73	27.68
State costs (JPY/year) (USD 1 = JPY 144)
Revasc	1,705,078 (USD 11,841)	1,449,316 (USD 10,065)	1,960,840 (USD 13,617)	±15%	Gamma	170.73	9986.90	16)
UA (acute)	2,156,290 (USD 14,974)	1,832,847 (USD 12,728)	2,479,734 (USD 17,220)	±15%	Gamma	170.73	12629.71
UA (Year1)	900,432 (USD 6,253)	765,367 (USD 5,315)	1,035,497 (USD 7,191)	±15%	Gamma	170.73	5273.97
	900,432 (USD 6,253)	765,367 (USD 5,315)	1,035,497 (USD 7,191)	±15%	Gamma	170.73	5273.97
UA (Stable)	495,600 (USD 3,442)	421,260 (USD 2,925)	569,940 (USD 3,958)	±15%	Gamma	170.73	2902.80
MI (Acute)	2,156,29 (USD 14,974)	1,832,847 (USD 12,728)	2,4797,34 (USD 17,220)	±15%	Gamma	170.73	12629.71
MI (Year1)	900,432 (USD 6,253)	765,367 (USD 5,315)	1,035,497 (USD 7,191)	±15%	Gamma	170.73	5273.97
MI (Year2)	900,432 (USD 6,253)	765,367 (USD 5,315)	1,035,497 (USD 7,191)	±15%	Gamma	170.73	5273.97
MI (Stable)	900,432 (USD 6,253)	765,367 (USD 5,315)	1,035,497 (USD 7,191)	±15%	Gamma	170.73	5273.97
Stroke (Acute)	1,699,225 (USD 11,800)	1,444,341 (USD 10,030)	1,954,109 (USD 13,570)	±15%	Gamma	170.73	9952.62
Stroke (Year1)	318,387 (USD 2,211)	270,629 (USD 1,879)	366,145 (USD 2,543)	±15%	Gamma	170.73	1864.84
Stroke (Year2)	318,387 (USD 2,211)	270,629 (USD 1,879)	366,145 (USD 2,543)	±15%	Gamma	170.73	1864.84
Stroke (Stable)	318,387 (USD 2,211)	270,629 (USD 1,879)	366,145 (USD 2,543)	±15%	Gamma	170.73	1864.84
CV death	1,955,000 (USD 13,576)	1,661,750 (USD 11,540)	2,248,250 (USD 15,613)	±15%	Gamma	170.73	11450.73
Utility score
State utility score
FH primary (No event)	1.000	-	-	-	-	-	-	17)
Post revasc	1.000	-	-	-	-	-	-
UA 0-1 year	0.765	0.728	0.802	95%CI	Normal	0.765	0.019
UA 1-2 year	0.960	0.931	0.989	95%CI	Normal	0.960	0.015
UA stable	0.960	0.931	0.989	95%CI	Normal	0.960	0.015
MI 0-1 year	0.765	0.728	0.802	95%CI	Normal	0.765	0.019
MI 1-2 year	0.906	0.867	0.945	95%CI	Normal	0.906	0.020
MI stable	0.906	0.867	0.945	95%CI	Normal	0.906	0.020
Stroke 0-1 year	0.775	0.701	0.849	95%CI	Normal	0.775	0.038
Stroke 1-2 year	0.822	0.818	0.826	95%CI	Normal	0.822	0.002
Stroke stable	0.822	0.818	0.826	95%CI	Normal	0.822	0.002
WPAI (%)
Absenteeism								DeSC analysis
Revascularisation	5.07	-	-	-	-	-	-
UA	3.01	-	-	-	-	-	-
MI	8.00	-	-	-	-	-	-
IS	10.11	-	-	-	-	-	-
HF	8.07	-	-	-	-	-	-
Presenteeism								DeSC analysis
Revascularisation	15.45	-	-	-	-	-	-
UA	15.76	-	-	-	-	-	-
MI	17.02	-	-	-	-	-	-
IS	16.82	-	-	-	-	-	-
HF	14.47	-	-	-	-	-	-

Notes:

^a Beta (α, β), gamma(α, β), and normal (mean, SE).

^b Percentage of FH(+) in overall population (i.e., potential FH patients).

^c The change was considered impossible as 100% or 0% was assumed, and sensitivity analysis was not set.

^d It was assumed that genetic testing was conducted in second-degree relatives (6 adult relatives) per child diagnosed with FH(+).

^e The change in parameter is analyzed as “the number of patients with PV(+) detected by RCS” has been changed.

^f It was adjusted to keep the minimum value above 1.00.

^g Blood collection (D400-1: 37 points) + biochemical test I fee (D007: 106 points) + assessment fee (D026-4: 144 points) were calculated.

^h First visit fee (A000: 288 points) + blood collection (D400-1: 37 points) + biochemical test I fee (D007: 106 points) + assessment fee (D026-4: 144 points) were calculated.

ⁱ Genetic testing (D006-4: 5000 points) was calculated.

^j Genetic testing (D006-4: 5000 points) * second-degree relatives (6 adult relatives) was calculated.

^k The total of yearly drug costs and outpatient management costs were set as treatment costs (per year) for patient diagnosed with FH(+). Outpatient management costs were calculated by (revisit fee [73 points] + additional outpatient management [52 points]) * 4 times/year + (blood collection [37 points] + blood chemistry test fee CPK [11 points] + assessment fee [144 points]) * 2 times/year.

^l The total of yearly outpatient management costs were set as treatment costs (per year) for patient diagnosed with FH(+). Outpatient management costs were calculated by (revisit fee [73 points] + additional outpatient management [52 points] + outpatient nutritional education fee 2 (2) (i) [190 points]) * 1.5.

USD 1 = JPY 144, as of October 1, 2024.

Abbreviations: CPK: creatine phosphokinase, CV: cardiovascular, DM: diabetes mellitus, DSA: deterministic sensitivity analysis, FH: familial hypercholesterolaemia, HF: heart failure, IS: ischemic stroke, LDL-C: low-density lipoprotein-cholesterol, MI: myocardial infarction, PSA: probabilistic sensitivity analysis, PV: pathogenic variant, RCS: reverse cascade screening, Revasc: revascularization, RR: rate ratio, SE: standard error, SOC: standard of care, UA: unstable angina, US: universal screening, WPAI: work productivity and activity impairment.

The Markov model, which analyzes the occurrence of CVD events, is shown in Fig.3. The model analyzes the states of “i. FH primary (No event)”; four events of “ii. revascularization,” “iii. unstable angina,” and iv. nonfatal myocardial infarction,” and “v. stroke”; and the post-event state. Events iii.–v. were subcategorized into three post-event states (0-1 year, 1-2 year, and stable), depending on the number of years that had lapsed since the event. Assuming that the event incidence resulted from the LDL-C level and age, the following two adjustments were made to the baseline risk to calculate the adjusted event incidence⁹⁾:

Fig.3. Model structure of Markov model

CV, cardiovascular; FH, familial hypercholesterolemia; NF, non-fatal; MI, myocardial infarction; Revasc, revascularization; UA, unstable angina.

means transition between states. means staying in the same state (no transition).

○Adjustment based on the difference between mean LDL-C level per state (i.–v.) (i.e. mean LDL-C level of the baseline risk analysis set) and LDL-C of the target population

○Adjustment based on the difference between mean age (i.–v.) (i.e. mean age of the baseline risk analysis set) and age of the target population

Ei=E0i＊αi ^A0−A1＊βi ^L0−L1

Ei: adjusted event risk (%)

E0i: baseline risk (%)

αi: Event risk adjustment coefficient according to age (per year)

A0: age of the target population

A1: mean age of the baseline risk analysis set

βi: Rate ratio per unit change in LDL-C (1.0 mmol/L)

L0: mean LDL-C level of the baseline risk analysis set

L1: LDL-C level of the target population

With regard to the baseline risk, health insurance claims data provided by DeSC were used to calculate the cumulative incidence of four events (primary and secondary or later) up to day 365 with or without diabetes mellitus using a competitive risk model (Supplementary Table 2, ⁹⁾). The DeSC database consists of data from multiple health insurers and is considered to have excellent population representativeness and patient traceability in Japan. Recurrence was considered for all events, and transition from all states was possible.

Supplementary Table 2.Summary of baseline risk analysis using insurer database

Item	Content
Data source	・Insurer database provided by DeSC (insured by National Health Insurance)
Inclusion criteria	・Insured person diagnosed with either FH, ASCVD, HL, or DM and DM drug (concurrent prescription) and with confirmation of an insured period of past 7 months including the month of issuing the relevant health insurance claim.
Group analysis
(with/without DM)	With DM	・“DM diagnosis and drug prescription” is confirmed from the first month of baseline assessment period^a to the month to which the first day of observation^b belongs.
(with/without DM)	Without DM	・No “DM diagnosis and drug prescription” is confirmed from the first month of baseline assessment period^a to the month to which the first day of observation^b belongs.
Definition of each event	Revasc	・A medical care has been performed for revasc, the diagnosis of NF-MI or UA is found in the health insurance claim ID in which a medical care for revasc is recorded, and death registry information is not recorded.
	UA	・The diagnosis of UA is found in health insurance claim of admission, and death registry information is not recorded.
	NF-MI	・The diagnosis of NF-MI is found in health insurance claim of admission, and death registry information is not recorded.
	Stroke	・The diagnosis of stroke is found in health insurance claim of admission, and death registry information is not recorded.
	Death	・Death occurs during the follow-up period.

Notes:

^a Goes back at least 7 months including the concerned month.

^b The oldest day of calculation of first or revisit fee or admission fee in a health insurance claim that satisfies inclusion criteria.

Abbreviations: ASCVD: atherosclerotic cardiovascular disease, DM: diabetes mellitus, FH: familial hypercholesterolaemia, HL: hyperlipemia, NF-MI: non-fatal myocardial infarction, Revasc: revascularization, UA: unstable angina.

Model Inputs

Model parameters are listed in Supplementary Table 1. The baseline LDL-C for each group was based on the literature^{1, 6, 10)}. The effect of treatment on LDL-C reduction was estimated based on meta-analyses^{2, 11, 12)}. The P(FH+) (i.e. potential FH patients) was estimated to be 0.396% according to McKay’s method^{5, 7)}, the P(PV+|FH+) was estimated to be 61.2%⁶⁾, and the P(PV+|FH-) was 0.015%⁷⁾. The child health examination rate was set to 92.2%⁶⁾, and the percentage of LDL-C ≥ 140 mg/dL was set to 88% for FH(+) and 0.1% for FH(-)⁷⁾ [expert opinion]. The consultation rate of a doctor at a local medical facility was set to 60% [expert opinion]; the genetic testing rate with suspected FH was 11.6%⁶⁾, and the genetic testing rate with suspected FH was 50% [expert opinion]. The positive rates of genetic testing were set to 100% for FH(+) and 0% for FH(−). The number of adult relatives who received RCS (genetic testing) was set to 6 per 1 FH child, and the number of patients with PV(+) detected by RCS was assumed to be two. The diagnosis rate of FH(+) in the PV(+) group was set to 70%⁷⁾.

The event risk per year was estimated for each event based on the insurer database provided by DeSC Healthcare, Inc.¹³⁾ (see Supplementary Tables 2 and 3).

Supplementary Table 3.Model parameter (Baseline risk)

	Value	Settings in DSA			Settings in PSA			Ref
	Value	Lower limit	Upper limit	Range	Distribution	Parameters^a		Ref
Baseline risk (per year)
With DM
To Revasc
FH primary (No event)	0.47%	0.40%	0.54%	±15%	Beta	169.92	35984.20	DeSC analysis
Post revasc	16.67%	14.17%	19.17%	±15%	Beta	142.10	710.35
UA 0-1 year	4.69%	3.99%	5.39%	±15%	Beta	162.68	3305.92
UA 1-2 year	7.57%	6.43%	8.71%	±15%	Beta	157.73	1925.91
UA stable	3.71%	3.15%	4.27%	±15%	Beta	164.36	4265.84
MI 0-1 year	4.69%	3.99%	5.39%	±15%	Beta	162.68	3305.92
MI 1-2 year	7.57%	6.43%	8.71%	±15%	Beta	157.73	1925.91
MI stable	3.71%	3.15%	4.27%	±15%	Beta	164.36	4265.84
Stroke 0-1 year	1.68%	1.43%	1.93%	±15%	Beta	167.85	9823.01
Stroke 1-2 year	1.68%	1.43%	1.93%	±15%	Beta	167.85	9823.01
Stroke stable	1.68%	1.43%	1.93%	±15%	Beta	167.85	9823.01
To UA
FH primary (No event)	0.24%	0.20%	0.28%	±15%	Beta	170.32	70796.08	DeSC analysis
Post revasc	2.71%	2.30%	3.12%	±15%	Beta	166.08	5962.25
UA 0-1 year	13.48%	11.46%	15.50%	±15%	Beta	147.58	947.24
UA 1-2 year	7.56%	6.43%	8.69%	±15%	Beta	157.75	1928.87
UA stable	3.68%	3.13%	4.23%	±15%	Beta	164.41	4303.30
MI 0-1 year	13.48%	11.46%	15.50%	±15%	Beta	147.58	947.24
MI 1-2 year	7.56%	6.43%	8.69%	±15%	Beta	157.75	1928.87
MI stable	3.68%	3.13%	4.23%	±15%	Beta	164.41	4303.30
Stroke 0-1 year	0.85%	0.72%	0.98%	±15%	Beta	169.27	19745.06
Stroke 1-2 year	0.85%	0.72%	0.98%	±15%	Beta	169.27	19745.06
Stroke stable	0.85%	0.72%	0.98%	±15%	Beta	169.27	19745.06
To MI
FH primary (No event)	0%	-	-	-	-	-	-	DeSC analysis
Post revasc	1.01%	0.86%	1.16%	±15%	Beta	169.00	16563.38
UA 0-1 year	8.54%	7.26%	9.82%	±15%	Beta	156.07	1671.40
UA 1-2 year	0%	-	-	-	-	-	-
UA stable	2.42%	0.86%	1.16%	±15%	Beta	169.00	16563.38
MI 0-1 year	8.54%	7.26%	9.82%	±15%	Beta	156.07	1671.40
MI 1-2 year	0%	-	-	-	-	-	-
MI stable	2.42%	2.06%	2.78%	±15%	Beta	166.58	6716.71
Stroke 0-1 year	0.88%	0.75%	1.01%	±15%	Beta	169.22	19060.36
Stroke 1-2 year	0.88%	0.75%	1.01%	±15%	Beta	169.22	19060.36
Stroke stable	0.88%	0.75%	1.01%	±15%	Beta	169.22	19060.36
To Stroke
FH primary (No event)	3.31%	2.81%	3.81%	±15%	Beta	165.05	4821.27	DeSC analysis
Post revasc	4.12%	3.50%	4.74%	±15%	Beta	163.66	3808.58
UA 0-1 year	5.69%	4.84%	6.54%	±15%	Beta	160.96	2667.86
UA 1-2 year	2.49%	2.12%	2.86%	±15%	Beta	166.46	6518.50
UA stable	1.20%	1.02%	1.38%	±15%	Beta	168.67	13887.22
MI 0-1 year	5.69%	4.84%	6.54%	±15%	Beta	160.96	2667.86
MI 1-2 year	2.49%	2.12%	2.86%	±15%	Beta	166.46	6518.50
MI stable	1.20%	1.02%	1.38%	±15%	Beta	168.67	13887.22
Stroke 0-1 year	24.54%	20.86%	28.22%	±15%	Beta	128.59	395.41
Stroke 1-2 year	24.54%	20.86%	28.22%	±15%	Beta	128.59	395.41
Stroke stable	24.54%	20.86%	28.22%	±15%	Beta	128.59	395.41
To CV death
FH primary (No event)	1.64%	1.39%	1.89%	±15%	Beta	167.92	10070.81	DeSC analysis
Post revasc	4.13%	3.51%	4.75%	±15%	Beta	163.64	3798.56
UA 0-1 year	11.92%	10.13%	13.71%	±15%	Beta	150.26	1110.32
UA 1-2 year	6.02%	5.12%	6.92%	±15%	Beta	160.39	2503.95
UA stable	2.41%	2.05%	2.77%	±15%	Beta	166.59	6745.97
MI 0-1 year	11.92%	10.13%	13.71%	±15%	Beta	150.26	1110.32
MI 1-2 year	6.02%	5.12%	6.92%	±15%	Beta	160.39	2503.95
MI stable	2.41%	2.05%	2.77%	±15%	Beta	166.59	6745.97
Stroke 0-1 year	10.97%	9.32%	12.62%	±15%	Beta	151.89	1232.73
Stroke 1-2 year	10.97%	9.32%	12.62%	±15%	Beta	151.89	1232.73
Stroke stable	10.97%	9.32%	12.62%	±15%	Beta	151.89	1232.73
Without DM
To Revasc
FH primary (No event)	0.36%	0.31%	0.41%	±15%	Beta	170.11	47083.57	DeSC analysis
Post revasc	12.70%	10.80%	14.61%	±15%	Beta	148.92	1023.69
UA 0-1 year	5.23%	4.45%	6.01%	±15%	Beta	161.75	2930.98
UA 1-2 year	1.40%	1.19%	1.61%	±15%	Beta	168.33	11855.05
UA stable	0.78%	0.66%	0.90%	±15%	Beta	169.39	21547.53
MI 0-1 year	5.23%	4.45%	6.01%	±15%	Beta	161.75	2930.98
MI 1-2 year	1.40%	1.19%	1.61%	±15%	Beta	168.33	11855.05
MI stable	0.78%	0.66%	0.90%	±15%	Beta	169.39	21547.53
Stroke 0-1 year	0.62%	0.53%	0.71%	±15%	Beta	169.67	27195.94
Stroke 1-2 year	0.62%	0.53%	0.71%	±15%	Beta	169.67	27195.94
Stroke stable	0.62%	0.53%	0.71%	±15%	Beta	169.67	27195.94
To UA
FH primary (No event)	0.12%	0.10%	0.14%	±15%	Beta	170.53	141934.00	DeSC analysis
Post revasc	1.87%	1.59%	2.15%	±15%	Beta	167.52	8790.78
UA 0-1 year	11.52%	9.79%	13.25%	±15%	Beta	150.95	1159.36
UA 1-2 year	2.78%	2.36%	3.20%	±15%	Beta	165.96	5803.73
UA stable	1.57%	1.33%	1.81%	±15%	Beta	168.04	10534.85
MI 0-1 year	11.52%	9.79%	13.25%	±15%	Beta	150.95	1159.36
MI 1-2 year	2.78%	2.36%	3.20%	±15%	Beta	165.96	5803.73
MI stable	1.57%	1.33%	1.81%	±15%	Beta	168.04	10534.85
Stroke 0-1 year	0.44%	0.37%	0.51%	±15%	Beta	169.98	38460.91
Stroke 1-2 year	0.44%	0.37%	0.51%	±15%	Beta	169.98	38460.91
Stroke stable	0.44%	0.37%	0.51%	±15%	Beta	169.98	38460.91
To MI
FH primary (No event)	0.12%	0.10%	0.14%	±15%	Beta	170.53	141934.00	DeSC analysis
Post revasc	1.15%	0.98%	1.32%	±15%	Beta	168.76	14505.73
UA 0-1 year	7.85%	6.67%	9.03%	±15%	Beta	157.25	1845.94
UA 1-2 year	1.37%	1.16%	1.58%	±15%	Beta	168.38	12122.04
UA stable	1.51%	1.28%	1.74%	±15%	Beta	168.14	10966.85
MI 0-1 year	7.85%	6.67%	9.03%	±15%	Beta	157.25	1845.94
MI 1-2 year	1.37%	1.16%	1.58%	±15%	Beta	168.38	12122.04
MI stable	1.51%	1.28%	1.74%	±15%	Beta	168.14	10966.85
Stroke 0-1 year	0.56%	0.48%	0.64%	±15%	Beta	169.77	30146.27
Stroke 1-2 year	0.56%	0.48%	0.64%	±15%	Beta	169.77	30146.27
Stroke stable	0.56%	0.48%	0.64%	±15%	Beta	169.77	30146.27
To Stroke
FH primary (No event)	1.43%	1.22%	1.64%	±15%	Beta	168.28	11599.26	DeSC analysis
Post revasc	3.51%	2.98%	4.04%	±15%	Beta	164.70	4527.71
UA 0-1 year	4.23%	3.60%	4.86%	±15%	Beta	163.47	3701.01
UA 1-2 year	1.34%	1.14%	1.54%	±15%	Beta	168.43	12401.00
UA stable	3.01%	2.56%	3.46%	±15%	Beta	165.56	5334.85
MI 0-1 year	4.23%	3.60%	4.86%	±15%	Beta	163.47	3701.01
MI 1-2 year	1.34%	1.14%	1.54%	±15%	Beta	168.43	12401.00
MI stable	3.01%	2.56%	3.46%	±15%	Beta	165.56	5334.85
Stroke 0-1 year	22.31%	18.96%	25.66%	±15%	Beta	132.42	461.12
Stroke 1-2 year	22.31%	18.96%	25.66%	±15%	Beta	132.42	461.12
Stroke stable	22.31%	18.96%	25.66%	±15%	Beta	132.42	461.12
To CV death
FH primary (No event)	1.35%	1.15%	1.55%	±15%	Beta	168.41	12306.63	DeSC analysis
Post revasc	4.05%	3.44%	4.66%	±15%	Beta	163.78	3880.08
UA 0-1 year	10.68%	9.08%	12.28%	±15%	Beta	152.39	1274.49
UA 1-2 year	5.37%	4.56%	6.18%	±15%	Beta	161.51	2846.12
UA stable	2.88%	2.45%	3.31%	±15%	Beta	165.79	5590.66
MI 0-1 year	10.68%	9.08%	12.28%	±15%	Beta	152.39	1274.49
MI 1-2 year	5.37%	4.56%	6.18%	±15%	Beta	161.51	2846.12
MI stable	2.88%	2.45%	3.31%	±15%	Beta	165.79	5590.66
Stroke 0-1 year	9.82%	8.35%	11.29%	±15%	Beta	153.87	1413.01
Stroke 1-2 year	9.82%	8.35%	11.29%	±15%	Beta	153.87	1413.01
Stroke stable	9.82%	8.35%	11.29%	±15%	Beta	153.87	1413.01

^a Beta (α, β)

Abbreviations: CV: cardiovascular, DM: diabetes mellitus, DSA: deterministic sensitivity analysis, FH: familial hypercholesterolaemia, MI: myocardial infarction, PSA: probabilistic sensitivity analysis, Revasc: revascularization, UA: unstable angina.

The rate ratio per unit change in LDL-C (1.0 mmol/L) was derived from values reported by the Cholesterol Treatment Trialists’ Collaboration¹⁴⁾. The event risk adjustment coefficient according to age was set to 1.05/year for cardiovascular (CV) death and 1.03/year for other events^{9, 15)}. The minimum age applied to event risk adjustment was 20 years [assumption]. The initiation rates of SOC treatment in the FH(+) and no-SOC treatment groups were set to 1.04%/year [DeSC analysis].

Screening-related costs were calculated based on the medical treatment fee points. FH treatment costs were set based on drug prescriptions collected by experts’ opinions and the frequency of prescription [DeSC analysis]. The SOC for patients ≤ 14 years old was set to 1 mg/day of pitavastatin, and that for patients ≥ 15 years old was estimated based on the prescription ratio per standard and daily cost of drugs (including both original and generic drugs) of 6 statins (atorvastatin, simvastatin, rosuvastatin, pravastatin, pitavastatin, and fluvastatin). The state costs (costs for each CVD event) were derived from Kodera 2019 ¹⁶⁾.

The utility score for each state was derived from NICE TA^{9, 15)}, and that of the general population was derived from Shiroiwa et al.¹⁷⁾.

Analyses

• Base-Case Analysis

Case yields were estimated for FH screening (US+RCS) performed for 100,000 children 9–10 years old. Cost-effectiveness was evaluated for FH screening compared to no screening. An incremental cost-effectiveness ratio (ICER) of JPY 5,000,000 (USD 34,722)/quality-adjusted life year (QALY) was used as a threshold to determine whether the screening strategy was cost-effective¹⁸⁾ (USD 1 = JPY 144, as of October 1, 2024). The QALY is one of the measures of effect in a cost-effectiveness analysis, calculated by multiplying the life years by the utility score. A utility score of 1 indicates full health, whereas a score of 0 indicates death as measured by the standard gamble method, time-trade-off method, or EuroQol 5 dimension. ICER is the incremental cost divided by the incremental effectiveness, using the following equation¹⁹⁾:

ICER = IC/IE = C_A−C_B/E_A−E_B

IC: incremental cost

IE: incremental effect (QALY etc.)

C_A: expected cost of treatment A

C_B: expected cost of treatment B

E_A: expected effectiveness of treatment A

E_B: expected effectiveness of treatment B

• Scenario Analyses

Eight scenario analyses were conducted in this study. The conditions for each scenario are listed in Supplementary Table 4 .

Supplementary Table 4.Results of scenario analyses

Strategy	Effectiveness (QALYs)	Incremental effectiveness (QALYs)	Cost (JPY)	Incremental cost (JPY)	ICER (JPY/QALY)
Scenario 1: Consultation rate of a doctor at a local medical facility (Base-case: 60%)
i. Consultation rate of a doctor at a local medical facility: 80%
Target strategy (US+RCS)	33.148	0.005	2,307,254 (USD 16,023)	-53 (USD -0.4)	Dominant
Comparator (No screening)	33.142	-	2,307,307 (USD 16,023)	-	-
ii. Consultation rate of a doctor at a local medical facility: 90%
Target strategy (US+RCS)	33.145	0.006	2,307,764 (USD 16,026)	-390 (USD -3)	Dominant
Comparator (No screening)	33.139	-	2,308,154 (USD 16,029)	-	-
Scenario 2: Genetic testing costs: 38,800 JPY/test ^a (Base-case: 50,000 JPY/test)
Target strategy (US+RCS)	33.154	0.004	2,306,137 (USD 16,015)	525 (USD 4)	129,657 (USD 900)
Comparator (No screening)	33.150	-	2,305,612 (USD 16,011)	-	-
Scenario 3: Child health examination costs: 1,430 JPY/ examination ^b (Base-case: 2,870 JPY/examination)
Target strategy (US+RCS)	33.154	0.004	2,304,907 (USD 16,006)	-705 (USD -5)	Dominant
Comparator (No screening)	33.150	-	2,305,612 (USD 16,011)	-	-
Scenario 4: RCS is not performed for adult relatives of child with “suspected FH(+) & no genetic testing (Base-case: RCS performed)
Target strategy (US+RCS)	33.170	0.001	2,303,133 (USD 15,994)	1,981 (USD 14)	1,641,843 (USD 11,402)
Comparator (No screening)	33.169	-	2,301,152 (USD 15,980)	-	-
Scenario 5: The number of patients with PV(+) detected by RCS (Base-case: 2)
i. The number of patients with PV(+) detected by RCS: 0.5
Target strategy (US+RCS)	33.168	0.002	2,303,587 (USD 15,997)	1,794 (USD 12)	1,110,973 (USD 7,715)
Comparator (No screening)	33.166	-	2,301,793 (USD 15,985)	-	-
ii. The number of patients with PV(+) detected by RCS: 6.1
Target strategy (US+RCS)	33.115	0.011	2,313,434 (USD 16,066)	-2,574 (USD -18)	Dominant
Comparator (No screening)	33.105	-	2,316,008 (USD 16,083)	-	-
Scenario 6: Initiation rates of SOC treatment in FH(+) & no SOC treatment (per year) (Base-case: 1.04%)
i. Initiation rates of SOC treatment (per year): 2%
Target strategy (US+RCS)	33.155	0.003	2,306,017 (USD 16,014)	860 (USD 6)	254,515 (USD 1,767)
Comparator (No screening)	33.151	-	2,305,157 (USD 16,008)	-	-
ii. Initiation rates of SOC treatment (per year): 4%
Target strategy (US+RCS)	33.155	0.002	2,305,775 (USD 16,012)	1,197 (USD 8)	494,664 (USD 3,435)
Comparator (No screening)	33.153	-	2,304,578 (USD 16,004)	-	-
Scenario 7: The percentage of cases that do not immediately start SOC treatment among FH children diagnosed with FH(+) who were with “suspected FH(+) & genetic testing”: 30% ^c (Base-case: 0%)
Target strategy (US+RCS)	33.155	0.004	2,306,119 (USD 16,015)	697 (USD 5)	177,790 (USD 1,235)
Comparator (No screening)	33.151	-	2,305,423 (USD 16,010)	-	-
Scenario 8: The loss of productivity after the occurrence of CVD event was analyzed in patients aged 20–64 years as a limited social position ^d
Target strategy (US+RCS)	33.154	0.004	7,318,786 (USD 50,825)	-5,540 (USD -38)	Dominant
Comparator (No screening)	33.150	-	7,324,325 (USD 50,863)	-	-

^a 3,880 points of genetic testing D006-4 (1) Easy process covered by the National Health Insurance were calculated.

^b Breakdown of base-case analysis values: blood collection (37 points) + biochemical test I fee (106 points) + assessment fee (144 points). The assessment fee was excluded from scenario analyses.

^c All subjects start SOC treatment at 20 years, and the initiation rate of SOC treatment due to aging (1.04%) was considered from the first year.

^d The mean annual salary (307,400 JPY (USD 2,135)/month ´ 12 = 3,688,800 JPY (USD 25,617)) was used18).

USD 1 = JPY 144, as of October 1, 2024.

Abbreviations: CVD: cardiovascular disease, FH: familial hypercholesterolaemia, ICER: incremental cost-effectiveness ratio, PV: pathogenic variant, QALY: quality-adjusted life year, RCS: reverse cascade screening, SOC: standard of care, US: universal screening.

• Sensitivity Analysis

○A deterministic sensitivity analysis (DSA)

For parameters for which confidence interval (CI) information was available, the 95% CI was used for the lower and upper limits of the sensitivity analysis. Otherwise, a set value distribution of ±15% was used. The top 15 variables are summarized in a tornado diagram.

○A probabilistic sensitivity analysis (PSA)

The shape of the distribution suitable for each parameter was considered. The distribution was set using the standard error (SE) of the estimated value for each parameter, but the distribution was set to ±15% if the SE was not reported or could not be estimated. Random numbers were generated per probability distribution for each parameter to conduct the simulation by extracting the data 10,000 times.

Results

Base-Case Analyses

The results of the base case analysis are presented in Tables 1 and 2. A series of FH screenings (US + RCS) based on the Kagawa model estimated the case yields. FH(+) children detected as FH(+) (with genetic testing, considering SOC treatment) or suspected FH(+) (without genetic testing, diet, and exercise therapies) were 110 per 100,000 children, and FH(+) adult relatives detected by RCS were 154 per 100,000 children (Table 1).

Table 1.Case yields and cost-effectiveness of FH screening for all (US in children + RCS in adults)

Strategy	FH cases identified per 100,000 screened			Effectiveness (QALYs)	Incremental effectiveness (QALYs)	Cost (JPY)	Incremental cost (JPY)	ICER (JPY/ QALY)
Strategy	Children diagnosed with FH(+)	Children diagnosed with suspected FH	Adult relative FH patient detected by RCSa	Effectiveness (QALYs)	Incremental effectiveness (QALYs)	Cost (JPY)	Incremental cost (JPY)	ICER (JPY/ QALY)
Target strategy (US+RCS)	14	96	154	33.154	0.004	2,306,233 (USD 16,016)	621 (USD 4)	153,254 (USD 1,064)
Comparator (No screening)	0	0	0	33.150	-	2,305,612 (USD 16,011)	-	-
-Note: a Aggregation was conducted based on a rule that RCS was performed for adult relatives of a child diagnosed with suspected FH(+) even if FH patients were not found by RCS performed for adult relatives of a FH(-) and PV(+) child.

Breakdown of costs

Item

Breakdown of costs (JPY)

US+RCS

No screening

Incrementa

Screening-related

costs

3,083

(USD 21)

3,083

(USD 21)

FH treatment costs

(costs for SOC,

diet and exercise therapies)

1,362

(USD 9)

648

(USD 5)

714

(USD 5)

CV event costs

2,301,788

(USD 15,985)

2,304,965

(USD 16,007)

-3,176

(USD -22)

USD 1 = JPY 144, as of October 1, 2024.

Abbreviations: CV: cardiovascular, FH: familial hypercholesterolaemia, ICER: incremental cost-effectiveness ratio, PV: pathogenic variant, RCS: reverse cascade screening, QALY: quality-adjusted life year, SOC: standard of care, US: universal screening.

Table 2.Cost-effectiveness of FH screening for US in children and RCS in adults

Strategy	Effectiveness (QALYs)	Incremental effectiveness (QALYs)	Cost (JPY)	Incremental cost (JPY)	ICER (JPY/QALY)
US in children
Target strategy (US)	33.173	0.001	2,302,704 (USD 15,991)	2,186 (USD 15)	2,724,837 (USD 18,922)
Comparator (No screening)	33.172	-	2,300,517 (USD 15,976)	-	-
RCS in adults
Target strategy (RCS)	20.864	2.111	4,596,478 (USD 31,920)	-1,015,228 (USD -7,050)	Dominant
Comparator (No screening)	18.753	-	5,611,705 (USD 38,970)	-	-

USD 1 = JPY 144, as of October 1, 2024.

Abbreviations: FH: familial hypercholesterolaemia, ICER: incremental cost-effectiveness ratio, RCS: reverse cascade screening, QALY: quality- adjusted life year, US: universal screening.

The total cost of the screening strategy and no screening were estimated to be JPY 2,306,233 and 2,305,612 (USD 16,016 and 16,011)/person, respectively. The obtained QALYs were estimated to be 33.154 and 33.150 QALYs/person, respectively. The screening strategy had a higher cost and obtained a QALY of JPY 621 (USD 4)/person and 0.004 QALY/person. Based on this, the ICER was estimated as JPY 153,254 (USD 1,064)/QALY (Table 1). This was considered acceptable, as it was below the ICER threshold of JPY 5,000,000 (USD 34,722)/QALY ¹⁸⁾ . The ICER of US in children without RCS was estimated to be JPY 2,724,837 (USD 18,922)/QALY, which was also considered acceptable (Table 2).

Scenario Analyses

The results of the scenario analysis are presented in Supplementary Table 4. In Scenario 1, the consultation rate of a doctor at a local medical facility was ≥ 79%. In Scenario 3, it was dominant if the child health examination cost was JPY 1,430 (USD 10).

Sensitivity Analyses

The DSA results are shown in Supplementary Fig.1. A one-way sensitivity analysis revealed that the largest effect was observed in the “event risk adjustment coefficient according to age (per year).” The results were below JPY 5,000,000 (USD 34,722)/QALY for all analyses. The PSA results are shown in Supplementary Figs.2. For the ICER threshold of JPY 5,000,000 (USD 34,722)/QALY, the probability of the target strategy becoming cost-effective was 95.8%.

Supplementary Fig.1. Result of deterministic sensitivity analysis

Abbreviations: CV: cardiovascular, FH: familial hypercholesterolaemia, ICER: incremental cost-effectiveness ratio, LDL-C: low-density lipoprotein-cholesterol, MI: myocardial infarction, PV: pathogenic variant, QALY: quality-adjusted life year, RCS: reverse cascade screening, Revasc: revascularization, RR: rate ratio, UA: unstable angina.

Supplementary Fig.2 Results of probabilistic sensitivity analysis

Abbreviations: ICER: incremental cost-effectiveness ratio, QALY: quality-adjusted life year.

Discussion

In this study, cost-effectiveness was evaluated for a series of FH screenings (US+RCS) based on the Kagawa model in Japan compared with no screening. The estimated ICER was JPY 153,254 (USD 1,064)/QALY. In addition, the ICER of the US in children without RCS was estimated to be JPY 2,724,837 (USD 18,922)/QALY. Under both conditions, the estimated ICER is considered acceptable. While heterozygous FH is asymptomatic in youth, a long-term simulation showed that FH screening is cost-effective, as quality of life (QOL) can be improved by reducing the occurrence of an event by providing early intervention as a result of FH screening.

Due to differences in various conditions affecting the results of the analysis, such as healthcare systems and social environments among countries, screening methods, and age range for US, the results should be compared with caution, but the ICER compared with no screening for “cholesterol screening followed by diagnostic genetic-testing and reverse cascade testing” in the analysis by McKay et al., which is the closest to the flow of the Kagawa model in their analysis, was estimated to be £12,480/QALY and found to be cost-effective⁷⁾.

In a systematic review of the cost-effectiveness of screening strategies for FH by Marquina et al., 21 studies were identified, and most concluded that screening for FH compared with no screening was cost-effective, regardless of the screening strategy²⁰⁾.

The health economic benefit of the combination with RCS was indicated because the ICER of US in children without RCS was approximately 18 times higher than that of US+RCS. A total of 110 children (27.8%) were detected among 396 potential FH(+) children per 100,000 children 9–10 years old (0.396%). Of the 286 undetected children, 129 (45.1%) were not detected because of the failure to consult (consultation failure rate: 40%), indicating the importance of public education regarding consultation for the effective screening of potential FH groups. In this analysis, the target population for the US was assumed to be children 9-10 years old, which is considered reasonable because the guideline recommends this age for initiation of treatment¹⁾.

Scenario analyses revealed that it was expected to be dominant (cost-saving) depending on the improvement in consultation rate and the settings of child health examination costs. The present study provides information for necessary interventions through policies and their priorities. The study revealed that if 79% of the children followed the health examination recommendations and visited a doctor at a local medical facility, a reduction in medical costs could theoretically be achieved. It was shown that improvement in consultation rates and reduction of health examination costs as possible policy interventions are the most effective means to activate this system from the perspective of health economics.

Several limitations associated with the present study warrant mention. First, the estimated baseline risk of the DeSC database analysis set adjusted for age and LDL-C levels was used as the event risk. The uncertainty of the data cannot be denied, as the DeSC database analysis set was a group defined by public health insurance claims data in which the diagnosis is not very precise. In addition, event risk was estimated based on the assumption that the variation factors were limited to age and plasma LDL-C levels. The bias caused by this limitation was examined using DSA, and no marked influence was observed. Second, the target relatives of the RCS performed were assumed to be the primary prevention group, consisting of people who were 42 years old. Although siblings of a child diagnosed with FH could be eligible for RCS, only adult relatives were considered because of the lack of sufficient epidemiological information for these siblings, such as the age at RCS, LDL-C level, and their percentage in the total target population. Adult relatives for secondary prevention (those with a history of CVD events) were not included. This was because from a clinical practice perspective, the majority of adult relatives with FH(+) were considered to be in the primary prevention group. Because the secondary prevention group was considered to have an increased risk of CVD, a cost-effectiveness evaluation excluding them was considered conservative. Third, only statins were assumed to be used as therapeutic drugs without considering switching to other drugs (e.g. ezetimibe and evolocumab). The present analysis aimed to evaluate the cost-effectiveness of early detection of FH in patients through FH screening; the avoidance of the risk of CVD by switching drugs was therefore not assumed. As these settings were applied to both FH screening and the comparator, the comparability of incremental cost and effectiveness between the two groups was fairly well maintained. Finally, some of the parameter sets were based on expert opinion. However, these values were discussed and set based on evidence from the Kagawa model⁶⁾ and a previous study⁷⁾, and their impact on the results of the analysis was addressed as much as possible by confirmation through a sensitivity analysis. However, values involving social interventions, such as the consultation rate of a doctor at a local medical facility and screening-related costs, cannot be set on the basis of evidence. For these values, scenario analyses (Supplementary Table 4) were conducted to verify their impact on the analysis results. Thus, a cost-effectiveness analysis is a scientific tool that can be used for policy recommendations, and this analysis demonstrates its practical potential.

In this study, the model was built based on the Japanese Atherosclerosis Society Guideline 2017 (JAS2017), but recently, the JAS2022 was created and reported to increase sensitivity while maintaining specificity compared with JAS2017 ²¹⁾. This means that the widespread use of this guideline is likely to further improve the health economic benefits of screening compared to this analysis and should be re-evaluated in the future. It may be important to continue and replicate such studies, as examining and quantifying the health economic benefits in real-world clinical practice may facilitate the implementation of public US for FH.

Conclusion

A cost-effectiveness analysis revealed that a series of FH screenings (US in children + RCS in adult relatives) based on the Kagawa model was cost-effective.

Acknowledgements

None.

Financial Support

This study was funded by Novartis Pharma K.K., Tokyo, Japan.

Conflict of Interest

AU and KM are employees of Novartis Pharma K.K. (Tokyo, Japan). The authors have received lecture and/or consultation fees from the following sources: MS: Amgen, Medpace Japan, Liid Pharmaceuticals, Recordati Rare Diseases Japan, SY: Otsuka Pharmaceutical, Skylight Biotech, Hayashihara, Kowa, Novartis, IK: Crecon Medical Assessment, CMIC, Takeda Pharmaceutical, Becton Dickinson Japan, TM: Astellas Pharma, Otsuka Pharmaceutical, Omron, A&D, Kyowa Kirin, Sanofi, Matsutani Chemical, Melody International. MY and SI are employees of CRECON Medical Assessment, Inc.

Author Contributions

All authors were involved in the conception and design of this study. MY and SI were involved in the data analysis. All authors were involved in the interpretation of the data. MY and SI drafted the manuscript, and all authors revised it critically for intellectual content. All authors were involved in the final approval of the version to be published and agreed to be accountable for all aspects of this work.

Supplementary References

1)Matsunaga K, Mizobuchi A, Ying Fu H, Ishikawa S, Tada H, Kawashiri MA, Yokota I, Sasaki T, Ito S, Kunikata J, Iwase T, Hirao T, Yokoyama K, Hoshikawa Y, Fujisawa T, Dobashi K, Kusaka T, Minamino T: Universal Screening for Familial Hypercholesterolemia in Children in Kagawa, Japan. J Atheroscler Thromb, 2022; 29: 839-849

2)Toritani S, Matsunaga K, Tani R, Inoue T, Fu HY, Minamino T: Consideration of Achilles tendon thickness in adult patients diagnosed with familial hypercholesterolaemia in reverse cascade screening. 120th Annual Meeting of the Japanese Society of Internal Medicine Abstract, 2023;

3)Ministry of Health, Labour and Welfare Research Project of Overcoming Intractable Diseases: primary dyslipidaemia research: Appendix 1. Transition stage medicine in primary dyslipidaemia [Internet]. [accessed 11 October 2023]. Available from: https: //nanbyo-lipid.com/wp/wp-content/themes/nanbyo_lipid/pdf/disease10_02_2.pdf

4)Information Center for Specific Pediatric Chronic Diseases, Japan: Diabetes mellitus type 1 [Internet]. [accessed 8 May 2023]. Available from: https: //www.shouman.jp/disease/details/07_01_001/

5)Harada-Shiba M, Arai H, Ohmura H, Okazaki H, Sugiyama D, Tada H, Dobashi K, Matsuki K, Minamino T, Yamashita S, Yokote K: Guidelines for the Diagnosis and Treatment of Adult Familial Hypercholesterolemia 2022. J Atheroscler Thromb, 2023; 30: 558-586

6)Bujo H, Takahashi K, Saito Y, Maruyama T, Yamashita S, Matsuzawa Y, Ishibashi S, Shionoiri F, Yamada N, Kita T: Clinical features of familial hypercholesterolemia in Japan in a database from 1996-1998 by the research committee of the ministry of health, labour and welfare of Japan. J Atheroscler Thromb, 2004; 11: 146-151

7)Cholesterol Treatment Trialists' Collaboration: Efficacy and safety of statin therapy in older people: a meta-analysis of individual participant data from 28 randomised controlled trials. Lancet, 2019; 393: 407-415

8)Japan Atherosclerosis Society (JAS): Guidelines for Prevention of Atherosclerotic Disease 2022 [Internet]. 2022 [accessed 8 May 2023]. Available from: https: //www.j-athero.org/jp/jas_gl2022/

9)Kelley GA, Kelley KS, Roberts S, Haskell W: Comparison of aerobic exercise, diet or both on lipids and lipoproteins in adults: a meta-analysis of randomized controlled trials. Clin Nutr, 2012; 31: 156-167

10)McKay AJ, Hogan H, Humphries SE, Marks D, Ray KK, Miners A: Universal screening at age 1-2 years as an adjunct to cascade testing for familial hypercholesterolaemia in the UK: A cost-utility analysis. Atherosclerosis, 2018; 275: 434-443

11)Nordestgaard BG, Chapman MJ, Humphries SE, Ginsberg HN, Masana L, Descamps OS, Wiklund O, Hegele RA, Raal FJ, Defesche JC, Wiegman A, Santos RD, Watts GF, Parhofer KG, Hovingh GK, Kovanen PT, Boileau C, Averna M, Borén J, Bruckert E, Catapano AL, Kuivenhoven JA, Pajukanta P, Ray K, Stalenhoef AF, Stroes E, Taskinen MR, Tybjærg-Hansen A: Familial hypercholesterolaemia is underdiagnosed and undertreated in the general population: guidance for clinicians to prevent coronary heart disease: consensus statement of the European Atherosclerosis Society. Eur Heart J, 2013; 34: 3478-3490a

12)Baigent C, Blackwell L, Emberson J, Holland LE, Reith C, Bhala N, Peto R, Barnes EH, Keech A, Simes J, Collins R: Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet, 2010; 376: 1670-1681

13)Wilson PW, D'Agostino R, Sr., Bhatt DL, Eagle K, Pencina MJ, Smith SC, Alberts MJ, Dallongeville J, Goto S, Hirsch AT, Liau CS, Ohman EM, Röther J, Reid C, Mas JL, Steg PG: An international model to predict recurrent cardiovascular disease. Am J Med, 2012; 125: 695-703.e691

14)Medical treatment point table, April 2020/ April 2021 Enlarged Edition (Medical), Igakutsushinsha Co., Ltd., Tokyo, 2021

15)Ministry of Health, Labour and Welfare: Information about items listed in National Health Insurance Drug Price Standard and generic drugs (until March 31, 2023) [Internet]. [accessed 8 May 2023]. Available from: https: //www.mhlw.go.jp/topics/2022/04/tp20220401-01.html

16)Kodera S, Morita H, Kiyosue A, Ando J, Komuro I: Cost-Effectiveness of Percutaneous Coronary Intervention Compared With Medical Therapy for Ischemic Heart Disease in Japan. Circ J, 2019; 83: 1498-1505

17)National Institute for Health and Care Excellence: Alirocumab for treating primary hypercholesterolaemia and mixed dyslipidaemia (TA393). [Internet]. 2016 [accessed 8 May 2023]. Available from: https: //www.nice.org.uk/guidance/ta393

18)Ministry of Health, Labour and Welfare: Basic statistical survey of wage structure 2021: Overview of results [Internet]. [accessed 11 October 2023]. Available from: https: //www.mhlw.go.jp/toukei/itiran/roudou/chingin/kouzou/z2021/dl/01.pdf

References

Corresponding author

Register with J-STAGE for free!