Optimizing point-of-care testing strategies for diagnosis and treatment of hepatitis C virus infection in Australia: a model-based cost-effectiveness analysis

Summary

Background

Timely diagnosis and treatment of hepatitis C virus (HCV) is critical to achieve elimination goals. This study evaluated the cost-effectiveness of point-of-care testing strategies for HCV compared to laboratory-based testing in standard-of-care.

Methods

Cost-effectiveness analyses were undertaken from the perspective of Australian Governments as funders by modelling point-of-care testing strategies compared to standard-of-care in needle and syringe programs, drug treatment clinics, and prisons. Point-of-care testing strategies included immediate point-of-care HCV RNA testing and combined point-of-care HCV antibody and reflex RNA testing for HCV antibody positive people (with and without consideration of previous treatment). Sensitivity analyses were performed to investigate the cost per treatment initiation with different testing strategies at different HCV antibody prevalence levels.

Findings

The average costs per HCV treatment initiation by point-of-care testing, from A$890 to A$1406, were up to 35% lower compared to standard-of-care ranging from A$1248 to A$1632 depending on settings. The average costs per treatment initiation by point-of-care testing for three settings ranged from A$1080 to A$1406 for RNA, A$960–A$1310 for combined antibody/RNA without treatment history consideration, and A$890–A$1189 for combined antibody/RNA with treatment history consideration. When HCV antibody prevalence was <74%, combined point-of-care HCV antibody and point-of-care RNA testing were the most cost-effective strategies. Modest increases in treatment uptake by 8%–31% were required for immediate point-of-care HCV RNA testing to achieve equivalent cost per treatment initiation compared to standard-of-care.

Interpretation

Point-of-care testing is more cost-effective than standard of care for populations at risk of HCV. Testing strategies combining point-of-care HCV antibody and RNA testing are likely to be cost-effective in most settings.

Funding

National Health and Medical Research Council.

Research in context.

Evidence before this study

We searched several databases, including MEDLINE, Scopus, Web of Science, Cochrane Central Register of Controlled Trials, Embase, Cost-Effectiveness Analysis Registry, International HTA database, Econlit, and PsycINFO, up to 7 April 2022 by the search terms “Hepatitis C” OR “HCV” OR “hepacivirus”, and “screen∗ OR “test” OR “tested” OR “testing” OR “treat∗” OR “diagnos∗”, and “costs and cost analysis” OR “cost∗” OR “healthcare resource∗” OR “economic∗” OR “QALY∗ OR “ICER∗ OR “life year∗” OR “LY” OR “DALY”. Collectively, four protocols of randomized controlled trials, feasibility and cohort studies involving point-of-care HCV testing were identified. Before the empirical evidence and trial-based cost-effectiveness were available, many modelling cost-effectiveness analyses suggested that HCV point-of-care testing could identify more cases, achieve more cures, lead to greater life expectancy with QALY gains. Most modelling studies indicated point-of-care testing was more expensive in the short-term but has the potential to lower the long-term lifetime costs by reducing HCV incidence and prevalence. Two papers reported results of HCV point-of-care testing combined with HIV testing or offered concomitantly to COVID-19 vaccination showing the prospect of point-of-care testing to identify undiagnosed infections at a reasonable cost and a feasible way to implement in practice.

Added value of this study

This study evaluated the cost-effectiveness of point-of-care testing strategies compared to standard of care in key Australian service settings for HCV elimination, including prisons, needle and syringe programs, and drug treatment clinics. Our analyses determined the cost-effective point-of-care testing strategies for different HCV antibody prevalence levels and the treatment uptake required for point-of-care testing to be more cost-effective than standard of care. This study identifies which point-of-care testing strategies are more cost-effective to implement across service settings and HCV prevalence and provides economic evidence to guide point-of-care testing to optimize treatment outcomes. This information is critical for the implementation and scale-up of point-of-care testing for HCV infection globally and will therefore inform clinical guidelines, clinical practice, and health policy, including national and international guidelines.

Implications of all the available evidence

Our findings suggested that all point-of-care HCV testing strategies had a lower average cost per treatment initiation compared to standard of care across all settings evaluated and HCV antibody prevalence. Based on HCV prevalence, different testing strategies should be used to optimize the testing and treatment outcomes. When HCV antibody prevalence was <74%, combined point-of-care HCV antibody and point-of-care RNA testing (with and without consideration of treatment history) is shown to be the most cost-effective strategies. Modest increases in treatment uptake by 8%–31% were required for immediate point-of-care HCV RNA testing to achieve equivalent cost per treatment initiation compared to standard of care.

Combining point-of-care HCV antibody and RNA testing could optimise testing outcomes, enhance cost-effectiveness by linking more people with current HCV to treatment, and improve patient acceptability by reducing time to receipt of results. This study is important for informing the future implementation and evaluation of point-of-care testing strategies to achieve HCV elimination targets.

Introduction

The ability to meet the World Health Organization's hepatitis C virus (HCV) elimination targets is threatened by low diagnosis (23%)¹ and treatment (5% of people diagnosed) globally.² Increasing HCV testing uptake is hampered by current diagnostic pathways requiring multiple visits, contributing to loss to follow-up.³ In Australia, modelling suggests that HCV RNA testing must increase by 50% annually to achieve elimination by 2030.⁴

Traditional testing pathways involve an HCV antibody test to confirm exposure and an HCV RNA test to detect active infection. Point-of-care HCV RNA testing for detection of current HCV infection in 1 h has changed HCV clinical management,^5,6 enabling diagnosis and treatment in a single-visit,7, 8, 9, 10, 11 increasing testing acceptability,^12,13 reducing loss to follow-up7, 8, 9, 10, 11 and increasing treatment uptake.7, 8, 9, 10, 11 Point-of-care HCV RNA tests have good technical accuracy^5,6,14 and clinical utility, with high treatment uptake following testing in needle and syringe programs (NSPs) (81%),⁷ supervised consumption sites (89%),⁹ mobile outreach models (74%),⁸ and prisons (control: 22%; intervention 93%).¹¹ However, point-of-care HCV RNA testing is more expensive than point-of-care antibody testing and has a longer time to result (60 min for an RNA result vs. 1–20 min for a negative antibody result and 1–15 min for a positive antibody result). Combining point-of-care HCV antibody and RNA testing could optimise testing outcomes, enhance cost-effectiveness and acceptability, without impact on effectiveness.

Many modelling cost-effectiveness studies suggested that HCV point-of-care testing could identify more cases, achieve more cures, lead to greater life expectancy with QALY gains.15, 16, 17, 18, 19, 20, 21, 22, 23, 24 However, studies exploring the cost-effectiveness of HCV point-of-care testing strategies are limited to particular populations (mostly low- and middle-income countries),^18,19,24 have lacked detailed costing information,17, 18, 19, 20 and have not included real-world epidemiological and clinical data to inform analyses.17, 18, 19, 20 Understanding the HCV prevalence and settings where different point-of-care testing strategies should be used is critical to inform implementation, funding, and integration into policy/practice. There is a lack of economic evidence to guide implementation of different testing strategies to optimize the outcomes.

This study evaluated the cost-effectiveness of point-of-care testing strategies including immediate point-of-care HCV RNA testing and combined point-of-care HCV antibody and RNA testing for HCV antibody positive people (with and without consideration of previous treatment) compared to standard of care in key service settings for HCV elimination. This study determined the most cost-effective point-of-care testing strategies for different HCV antibody prevalence levels and the treatment uptake required for point-of-care testing to be more cost-effective than standard of care.

Methods

Economic evaluation

This economic evaluation followed the Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) Statement.²⁵ The PICO (Population, Intervention, Comparator, Outcomes) statement and evaluation parameters are presented in Supplementary Table S1.

Study perspective

Cost-effectiveness analysis was undertaken from the perspective of the Australian Governments as the funders responsible for the delivery and funding of HCV testing and treatment, recognising settings where healthcare provision is segregated (e.g., testing in prison is paid by the State government only and testing in the community is paid by the State and Federal governments).

Study population and setting

The target population was people at risk of HCV infection. Study settings included prisons, NSPs, and drug treatment clinics.

Diagnostic strategies

• A.Point-of-care HCV RNA testing.
• B.Combined point-of-care HCV antibody and reflex point-of-care RNA testing (for antibody positive people).
• C.Combined point-of-care HCV antibody and reflex point-of-care RNA testing (for antibody positive people) in those self-reporting no previous HCV treatment and immediate point-of-care RNA testing in people self-reporting previous HCV treatment.
• D.Standard of care (on-site sample collection by venepuncture and laboratory-based HCV antibody followed by RNA testing if antibody positive).

We assumed that people with detectable HCV RNA following point-of-care testing were offered same-visit treatment and people receiving standard of care required 2–3 visits for testing and treatment.²⁶

Outcome measured

Outcomes included: 1) current HCV infection (detectable RNA); and 2) direct-acting antiviral (DAA) treatment initiation in people with current HCV infection. The time horizon of analysis was one year, thus discounting was not applied.

Analytic methods

Decision analytic models were constructed to assess the four HCV diagnostic strategies using the TreeAge software (Fig. 1).²⁷ Further modelling details are provided in the Supplementary Materials. Costs and outcomes for point-of-care HCV testing strategies were compared to standard of care, considering HCV prevalence, treatment history, and test accuracy. Costs and outcomes among the three point-of-care testing strategies were also compared to investigate the most efficient way to conduct point-of-care HCV testing in each setting. Cost-effectiveness analyses estimated the average costs of diagnosis and treatment initiation by the four testing strategies in prisons, NSPs, and drug treatment clinics.

Fig. 1.

Decision analysis structure of four testing strategies. (a) POC HCV RNA for all. (b) POC HCV-Antibody for all with Reflex HCV-RNA. (c) POC HCV-Antibody for Tx-naive with Reflex HCV-RNA. (d) Standard of Care.

Model parameters for outcomes

Model parameters and estimate sources are presented in Table 1. HCV antibody and RNA prevalence, and self-reported treatment history were estimated based on Australian studies evaluating HCV testing, diagnosis and treatment in prisons, NSPs, and drug treatment clinics.^28,29 Health outcomes were assessed under each testing strategy, including true positives identified and people with current infection linked to treatment. Assay diagnostic performance was sourced from peer-reviewed literature and manufacturer specifications.

Table 1.

Decision analytical model epidemiological parameters.

Parameter	Setting			Range for sensitivity analyses	Distribution in probabilistic sensitivity analyses	Source
	Prison (95% CI)	Needle Syringe Program (95% CI)	Drug treatment clinic (95% CI)
Proportion of people tested	1	1	1	0.50–1.00	Triangular	Assumptiona
HCV Antibody prevalence	0.37 (0.33, 0.40)	0.61 (0.54, 0.68)	0.69 (0.66, 0.72)	0.10–0.90	Beta	28,29
HCV RNA (+) in HCV Antibody (+)	0.32 (0.26, 0.38)	0.30 (0.22, 0.38)	0.20 (0.17, 0.23)	0.20–0.40	Beta	28,29
HCV RNA prevalence	0.12 (0.09, 0.14)	0.18 (0.13, 0.24)	0.14 (0.12, 0.16)	0.10–0.20	Beta	28,29
HCV Antibody (+) in HCV RNA (−)	0.28 (0.25, 0.32)	0.53 (0.45, 0.60)	0.64 (0.61, 0.68)	0.20–0.70	Beta	28,29
Point-of-care RNA invalid test due to operation error	0.03	0.03	0.03	0.01–0.1	Triangular	Assumption
Sensitivity of test
- Point-of-care HCV antibody	0.993	0.993	0.993	0.961–0.999	Beta	Bioline HCV, Abbottb
- Point-of-care HCV RNA	1.00	1.00	1.00	0.84–1.00	Beta	5,6
- Laboratory HCV antibody	1.00	1.00	1.00
- Laboratory HCV RNA	1.00	1.00	1.00
Specificity of test
- Point-of-care HCV antibody	0.981	0.981	0.981	0.945–0.994	Beta	Bioline HCV, Abbottb
- Point-of-care HCV RNA	1.00	1.00	1.00	0.81–1.00	Beta	5,6
- Laboratory HCV antibody	1.00	1.00	1.00			Assumptionc
- Laboratory HCV RNA	1.00	1.00	1.00			Assumptionc
Proportion lost to follow-up
- Point-of-care RNA re-test	0.25	0.25	0.25	0.20–0.30	Triangular	Assumption
- Standard of care RNA test	0.31	0.31	0.31	0.30–0.32	Beta	3,11
Treatment uptake
- Point-of-care	0.93 (0.78, 0.99)	0.81 (0.62, 0.94)	0.81 (0.62, 0.94)	0.70–0.99	Beta	7,11
- Standard of care	0.60 (0.57, 0.63)	0.59 (0.56, 0.61)	0.59 (0.56, 0.61)	0.20–0.65	Beta	3,11
Self-report history of HCV treatment history in RNA (+)	0.14 (0.11, 0.17)	0.18 (0.07, 0.33)	0.26 (0.19, 0.34)	0.10–0.30	Beta	28,29
Self-report history of HCV treatment history in RNA (−)	0.09 (0.08, 0.10)	0.39 (0.31, 0.46)	0.43 (0.39, 0.46)	0.05–0.50	Beta	28,29
Treatment-naïve in HCV Antibody (+)	0.74 (0.72, 0.77)	0.43 (0.35, 0.52)	0.42 (0.38, 0.46)	0.30–0.80	Beta	28,29
Treatment-naïve in HCV RNA (+)	0.86 (0.83, 0.88)	0.83 (0.67, 0.93)	0.74 (0.66, 0.81)	0.60–0.90	Beta	28,29
Self-report history of HCV treatment	0.10 (0.09, 0.11)	0.35 (0.28, 0.41)	0.40 (0.37, 0.43)	0.10–0.40	Beta	28,29
HCV Antibody (+) and RNA (−) in Tx-Naïve	0.17 (0.15, 0.18)	0.23 (0.15, 0.32)	0.38 (0.33, 0.42)	0.10–0.40	Beta	28,29

^aProportion of people tested: Assume to be 100% in the base case and tested in the ranges 0.5–1.0 in sensitivity analysis.

^bSensitivity and specificity of POC HCV-Antibody test: by Bioline HCV, Abbott, product specification insert.

^cSensitivity and specificity of lab HCV Antibody and RNA tests: assume 100% as gold standard tests.

Costs

An ingredient approach of costing was used to identify all resources required to perform testing and clinical assessment prior to treatment initiation. Costs of point-of-care RNA testing were estimated at three levels: 1) variable costs only (device, testing cartridge, consumables, and staff time); 2) direct costs including variable costs and fixed costs of IT/connectivity, quality assurance, and training; 3) total program costs including direct variable and fixed costs, and indirect program coordination/communication costs (Table 2 and Supplementary Table S2). Direct costs (variable and fixed costs for the operation of HCV detection, diagnosis, assessment and reporting as per clinical and public health requirements) were used in the base case analysis and sensitivity analysis were performed considering costs ranging from variable costs only (testing) to total program costs.

Table 2.

Cost estimates for point-of-care HCV RNA, point-of-care anti-HCV and standard of care laboratory HCV tests. (Costs in 2021 A$).

Test	Costs			Source
	Base case estimate (low, high)
Point-of-care RNA test	Testing variable cost	Direct cost (including variable and fixed costs)	Total cost
	$96 ($77, $165)	$129 ($109, $199)	$153 ($133, $222)	Cepheid 2019, 2020a Cepheid 2019b National Programc TEMPOd
Point-of-care combined HCV Ab and RNA tests	Anti-HCV only	Anti-HCV Reflex RNA	RNA onlye
	$34 ($30, $86)	$163 ($139, $285)	$129 ($109, $199)	7
Standard of care laboratory HCV tests	Anti-HCV only	Anti-HCV & RNA	RNA onlye
	$38 ($34, $90)	$153 ($117, $191)	$115 ($103, $174)	31
Treatment initiation assessment	$180 ($162, $198)			28,31

^aGeneXpert IV 4sites Laptop model with lifespan 10 years for 375 screened annually per site, including annual service and maintenance fees. Discount 20%∼50% applied for sensitivity analysis.

^bCosts include cartridge, reagent, delivery fee, and consumables.

^cNational Point-of-care Testing Program budget 2021.

^dRanges are based on TEMPO pilot data - 2 patients per hour by CNS, additional 15 min for reflex RNA after positive HCV antibody.

^eImmediate RNA tests with treatment history.

Variable costs (tests and labour costs involved in testing/clinical assessment) were measured from clinical studies.^7,11,28 Labour costs included nursing staff time (at different seniority levels according to tasks performed),³⁰ infectious disease specialists, and correctional officers (in prisons). Capital cost for point-of-care testing device was proportionated to unit cost per test (annuitization discounted 3% per annum with 10 years effective life). Testing kit, testing cartridge, and consumables for point-of-care testing and IT/connectivity, quality assurance, training, and program coordination were estimated based on the financial information from the National Australian Hepatitis C Point-of-Care Testing Program (Supplementary Table S2).

Costs associated with tests and clinical assessments prior to treatment initiation were considered, thus DAA pharmaceutical costs were excluded (Table 2). Clinical assessments costs included labour costs and costs of serology tests for hepatitis B virus and human immunodeficiency virus, liver function tests, full blood count, urea, electrolytes and creatine, pregnancy test (women only) and FibroScan. The Australian Medical Benefit Schedule was used for testing costs and services rendered by private healthcare providers, and government employment pay rates were used for public healthcare providers.^31,32 Costs were valued in 2021 Australian Dollar (A$).

Cost-effectiveness analysis

Incremental analysis

Costs and outcomes were compared to standard of care laboratory-based HCV diagnosis for each point-of-care testing strategy. Pairwise comparisons (testing strategy A vs. B, A vs. C, B vs. C) were carried out for the three point-of-care testing strategies in three settings. Incremental costs were compared to incremental outcomes to generate incremental cost-effectiveness ratios (ICERs).

Sensitivity analysis

Deterministic univariate sensitivity analysis was performed to investigate the cost per treatment initiation for each testing strategy by HCV antibody prevalence and treatment uptake. The most efficient and cost-effective testing strategies by HCV antibody prevalence were identified for each setting. Threshold analyses were undertaken to identify the minimum treatment uptake for point-of-care testing strategies to achieve equivalent average cost per treatment initiation compared to standard of care. One-way sensitivity analysis was performed with all model parameters with value ranges presented in Table 1 (effectiveness variables) and Supplementary Table S2 (cost variables). Probabilistic sensitivity analysis with 1000 iterations was conducted to illustrate the impact of parameter variability on costs and outcomes.

Role of the funding source

The study was supported by the Point of Care Research Consortium for Infectious Disease in the Asia Pacific (RAPID) which is funded through a National Health and Medical Research Council Centre for Research Excellence Grant in Australia. The funders did not have any role in study design, data collection, data analysis, interpretation, writing of the report.

Results

Average cost-effectiveness for diagnosis of HCV infection

With standard of care HCV testing, the average costs per HCV-infected person detected were $748 in prisons, $630 in NSPs, and $857 in drug treatment clinics (base case) (Table 3). The expected average cost of HCV diagnosis by point-of-care testing varied by setting and testing strategy, A$689–A$1140 per person detected in prisons, A$575–A$729 in NSPs, and A$817–A$957 in drug treatment clinics, depending on testing strategies (Table 3). Across all settings, diagnosis by immediate point-of-care HCV RNA testing was the most expensive testing strategy (A$729–A$1140), while combining point-of-care HCV antibody and reflex point-of-care RNA testing with consideration of treatment history achieved the lowest average cost per HCV-infected person detected (A$575–A$817).

Table 3.

Average cost per person tested, HCV detected, and people treated by three point-of-care testing strategies and incremental cost-effectiveness ratio compared to standard of care by laboratory testing.

Strategy	Average cost per person tested	Incremental cost	Effectiveness (HCV diagnosed or treated)	Incremental effectiveness	ICER	Average cost per person diagnosed or treated	Number to test for one person diagnosed or treated
Detection of active HCV infection
Prison setting
POC HCV RNA for all	$132	$66	0.1159	0.0274	$2409	$1140	8.6
POC HCV-Antibody for all with reflex HCV-RNA	$83	$17	0.1151	0.0266	$650	$725	8.7
POC HCV Antibody for treatment-naive with Reflex HCV-RNA	$79	$13	0.1153	0.0268	$493	$689	8.7
Standard of Care	$66	–	0.0885	–	–	$748	11.3
Needle Syringe Program Setting
POC HCV RNA for all	$132	$47	0.1813	0.0453	$1027	$729	5.5
POC HCV-Antibody for all with reflex HCV-RNA	$115	$29	0.1800	0.0441	$669	$640	5.6
POC HCV Antibody for treatment-naive with reflex HCV-RNA	$104	$18	0.1802	0.0443	$408	$575	5.5
Standard of Care	$86	–	0.1359	–	–	$630	7.4
Drug Treatment Setting
POC HCV RNA for all	$132	$40	0.1381	0.0308	$1308	$957	7.2
POC HCV-Antibody for all with Reflex HCV-RNA	$126	$34	0.1371	0.0299	$1127	$916	7.3
POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$112	$20	0.1740	0.0301	$676	$817	7.3
Standard of Care	$92	–	0.1072	–	–	$857	9.3
HCV detected and linked to treatment
Prison setting
POC HCV RNA for all	$152	$76	0.1078	0.0547	$1386	$1406	9.3
POC HCV-Antibody for all with Reflex HCV-RNA	$103	$27	0.1071	0.0539	$500	$960	9.3
POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$99	$23	0.1072	0.0541	$424	$921	9.3
Standard of Care	$76	–	0.0531	–	–	$1426	18.8
Needle Syringe Program Setting
POC HCV RNA for all	$159	$59	0.1468	0.0666	$879	$1080	6.8
POC HCV-Antibody for all with Reflex HCV-RNA	$141	$41	0.1458	0.0656	$630	$970	6.9
POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$130	$30	0.1460	0.0658	$455	$890	6.9
Standard of Care	$100	–	0.0802	–	–	$1248	12.5
Drug Treatment Setting
POC HCV RNA for all	$152	$49	0.1118	0.0486	$1010	$1362	8.9
POC HCV-Antibody for all with Reflex HCV-RNA	$146	$42	0.1111	0.0478	$884	$1310	9.0
POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$132	$29	0.1113	0.0480	$605	$1189	9.0
Standard of Care	$103	–	0.0633	–		$1632	15.8

HCV: hepatitis C, ICER: incremental cost-effective ratio, POC: point-of-care, Tx: treatment.

– The comparator to generate incremental cost/effectiveness.

Average cost-effectiveness for HCV treatment initiation

With standard of care HCV testing, the average cost per treatment initiation was $1426 in prisons, $1248 in NSPs, and $1632 in drug treatment clinics (base case) (Table 3). Across all settings, the average cost per treatment initiation with point-of-care testing was up to 35% less than the average cost in standard of care. The expected average cost per treatment initiation by point-of-care testing ranged from A$921–A$1406 in prisons, A$890–A$1080 in NSPs, to A$1189–A$1362 in drug treatment clinics, depending on testing strategies (Table 3). The average cost of immediate point-of-care HCV RNA testing per treatment initiation in the base case analysis was A$1406 in prison, A$1080 in NSP, and A$1362 in drug treatment clinics. The average cost per treatment initiation with combined point-of-care HCV antibody and reflex RNA testing was lower than immediate point-of-care RNA testing, with the lowest average cost per treatment initiation when considering previous treatment history (prisons A$921 vs. A$960, NSP A$890 vs. A$970, and drug treatment clinics A$1189 vs. A$1310).

Incremental cost-effectiveness ratios

Compared to standard of care, point-of-care HCV testing strategies detected more infections and increased treatment uptake with higher cost per person tested, resulting in ICERs from A$408 to A$2409 per detectable RNA and A$424–A$1386 per treatment initiation. Among the three point-of-care testing strategies, combined antibody and reflex RNA testing with consideration of treatment history was always the most efficient (lowest average cost per treatment initiation). Immediate RNA testing was more expensive but detected more people with HCV and associated with higher treatment uptake, resulting in ICERs from A$33,764 to A$89,757 for each additional person initiating HCV treatment, compared to the most efficient combined point-of-care testing strategy considering treatment history (strategy A vs. C in Table 4). When compared to the combined point-of-care antibody and reflex RNA tests irrespective of treatment history, immediate RNA testing produced ICERs from A$8698 in drug treatment clinics to A$64,706 in prisons (strategy A vs. B in Table 4). Pairwise comparisons of the two combined point-of-care antibody and reflex RNA testing showed that, in all settings, the testing strategy without consideration of treatment history was dominated (more expensive and less people treated) by the combined tests with consideration of treatment history. Cost-effectiveness frontier presented in a cost-effectiveness plane in Fig. 2 illustrates the points of base case corresponding to four testing strategies with the gradient of the line representing the ICER of the testing strategy comparison between two alternatives. The testing strategy of point-of-care antibody testing for everyone and reflex RNA test for antibody positives sit on the left side of frontier lines, meaning this strategy was dominated by the expansion of intervention undominated lines and would not be a choice for consideration from cost-effectiveness point of view.

Table 4.

Pairwise comparison between point-of-care testing strategies for HCV detection and treatment initiation for the base cases in three settings.

Pairwise comparison	Strategy	Average cost per person tested	Incremental cost	Effectiveness (HCV detected or treated)	Incremental effectiveness	ICER
	Prison setting
	Detection of active HCV infection
A	POC HCV RNA for all	$132	$49	0.1159	0.0008	$60,000
B	POC HCV-Antibody for all with Reflex HCV-RNA	$83	–	0.1151	–	–
B	POC HCV-Antibody for all with Reflex HCV-RNA	$83	4	0.1151	−0.0002	dominateda
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$79	–	0.1153	–	–
A	POC HCV RNA for all	$132	$53	0.1159	0.0006	$83,310
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$79	–	0.1153	–	–
	HCV detected & linked to treatment
A	POC HCV RNA for all	$152	$49	0.1078	0.0008	$64,700
B	POC HCV-Antibody for all with Reflex HCV-RNA	$103	–	0.1071	–	–
B	POC HCV-Antibody for all with Reflex HCV-RNA	$103	$4	0.1071	−0.0002	dominated
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$99	–	0.1072	–	–
A	POC HCV RNA for all	$152	$53	0.1078	0.0006	$89,800
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$99	–	0.1072	–	–
	Needle Syringe Program Setting
	Detection of active HCV infection
A	POC HCV RNA for all	$132	$17	0.1813	0.0013	$13,400
B	POC HCV-Antibody for all with Reflex HCV-RNA	$115	–	0.1800	–	–
B	POC HCV-Antibody for all with Reflex HCV-RNA	$115	$11	0.1800	−0.0002	dominated
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$104	–	0.1802	–	–
A	POC HCV RNA for all	$132	$28	0.1813	0.0010	$27,200
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$104	–	0.1802	–	–
	HCV detected & linked to treatment
A	POC HCV RNA for all	$159	$17	0.1468	0.0010	$16,800
B	POC HCV-Antibody for all with Reflex HCV-RNA	$141	–	0.1458	–	–
B	POC HCV-Antibody for all with Reflex HCV-RNA	$141	$11	0.1458	−0.0002	dominated
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$130	–	0.1460	–	–
A	POC HCV RNA for all	$159	$29	0.1468	0.0008	$33,800
C	POC HCV-Antibody for Tx-naive with Reflex HCV-RNA	$130	–	0.1460	–	–
	Drug Treatment Setting
	Detection of active HCV infection
A	POC HCV RNA for all	$132	$7	0.1381	0.0010	$6900
B	POC HCV-Antibody for all with Reflex HCV-RNA	$126	–	0.1371	–	–
B	POC HCV-Antibody for all with Reflex HCV-RNA	$126	$13	0.1371	−0.0003	dominated
C	POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$112	–	0.1374	–	–
A	POC HCV RNA for all	$132	$20	0.1381	0.0007	$27,900
C	POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$112	–	0.1374	–	–
	HCV detected & linked to treatment
A	POC HCV RNA for all	$152	$7	0.1118	0.0008	$8700
B	POC HCV-Antibody for all with Reflex HCV-RNA	$146	–	0.1111	–	–
B	POC HCV-Antibody for all with Reflex HCV-RNA	$146	$13	0.1111	−0.0002	dominated
C	POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$132	–	0.1113	–	–
A	POC HCV RNA for all	$152	$20	0.1118	0.0006	$34,600
C	POC HCV Antibody for Tx-naive with Reflex HCV-RNA	$132	–	0.1113	–	–

ICER: incremental cost-effective ratio.

– The comparator to generate incremental cost/effectiveness.

^aDominated: meaning more costly but less effective.

Fig. 2.

Cost-effectiveness planes for the base case costs and effectiveness by treatment initiation of four testing strategies in three settings. (a) Prison. (b) Needle Syringe Program. (c) Drug treatment clinics.

Sensitivity analysis

To assess the impact of HCV antibody prevalence on cost-effectiveness of different testing strategies, one-way sensitivity analysis was performed (Fig. 3). Combined point-of-care HCV antibody testing with reflex point-of-care RNA testing with consideration of previous treatment was the most cost-effective testing strategy across all settings up to an HCV antibody prevalence of 79% in prisons and 86% in NSPs and drug treatment clinics. When HCV antibody prevalence was <74%, combined point-of-care HCV antibody and point-of-care RNA testing, both with and without consideration of treatment history, were more cost-effective than immediate point-of-care RNA testing and lab-based standard of care testing. At an HCV antibody prevalence <35% in prison, <45% in NSP, and <49% in drug treatment clinics, immediate point-of-care RNA testing was the least cost-effective strategy; while at an antibody prevalence >79% in prisons and >86% in NSPs and drug treatment clinics, immediate point-of-care RNA testing became the most cost-effective testing strategy because the cost of the HCV antibody test is not wasted on people who are highly likely to be HCV antibody positive.

Fig. 3.

Average cost per HCV treatment initiation in relation to HCV antibody prevalence in three settings (a) Prison (b) Needle Syringe Program and (c) Drug treatment clinics.

The key factor determining the cost-effectiveness per person treated was the difference in treatment uptake between point-of-care and standard of care testing. Threshold analyses demonstrated that for immediate point-of-care RNA testing, treatment uptake levels at 67%–91% would achieve an equivalent average cost per treatment initiation as those in standard of care (59%–60%, depending on setting) (Table 5 and Fig. 4). For combined point-of-care HCV antibody and reflex RNA testing, treatment uptake of 54%–57% with consideration of prior treatment and 58%–64% without consideration of prior treatment, were required to achieve an equivalent average cost per treatment initiation as in standard of care.

Table 5.

Threshold of point-of-care treatment uptake for equal average cost per person treated compared to standard of care in three settings.

Setting	Treatment uptake in the base case		Threshold of point-of-care treatment uptake for equal average cost per person treated compared to standard of care
	SOC (95% CI)	POC (95% CI)	Combined POC HCV antibody and reflex RNA with treatment history consideration, base case (range)	Combined POC HCV antibody and reflex RNA without treatment history consideration, base case (range)	Immediate POC HCV RNA for all, base case (range)
Prison	0.60 (0.57, 0.63)	0.93 (0.78, 0.99)	0.55 (0.52, 0.59)	0.58 (0.55, 0.59)	0.91 (0.88, 0.95)
Needle Syringe Program	0.59 (0.56, 0.61)	0.81 (0.62, 0.94)	0.54 (0.52, 0.56)	0.60 (0.57, 0.62)	0.68 (0.66, 0.70)
Drug treatment clinic	0.59 (0.56, 0.61)	0.81 (0.62, 0.94)	0.57 (0.53, 0.59)	0.64 (0.60, 0.66)	0.67 (0.62, 0.68)

POC: Point-of-Care, SOC: Standard of Care.

Difference in treatment linkage between POC and SOC: base case 0.93–0.60, worst case 0.78–0.63, best case 0.99–0.57 in prison; base case 0.81–0.59, worst case 0.62–0.61, best case 0.94–0.56 in NSP and drug treatment clinics.

Fig. 4.

Threshold analysis of equivalent cost per treatment initiation for point-of-care treatment uptake in three settings (a) Prison, (b) Needle Syringe Program, and (c) Drug treatment clinics.

One-way sensitivity analyses suggested that point-of-care RNA test costs, number of annual tests per site, and number of samples collected in 1 h played a role in determining the ICERs of point-of-care testing compared to standard of care (Supplementary Fig. S1). Standard of care treatment uptake also impacted ICERs by determining the magnitude of improvements in treatment uptake with point-of-care testing.

Accounting for model input variability by simultaneously changing all parameters with plausible ranges (including effectiveness and costs variables) in the probabilistic sensitivity analysis, the three point-of-care testing strategies showed an increased treatment initiation compared to standard of care at a higher cost, albeit with great uncertainly (scatterplots of cost and effectiveness in Supplementary Fig. S2). Additionally, cost-effectiveness planes of incremental cost and incremental effectiveness of pairwise comparisons between the three point-of-care testing strategies are presented in Supplementary Fig. S3, which further validates the robustness of our analyses and strengthen the base case results presented in Table 4 and Fig. 2. That is, strategy B was dominated by strategy C, i.e., POC antibody testing for everyone without treatment history consideration and reflex with RNA testing (strategy B) less effective and more costly than the same testing with treatment history consideration (strategy C).

Discussion

This study demonstrated that all point-of-care HCV testing strategies had a lower average cost per treatment initiation compared to standard of care across settings and HCV antibody prevalence. When HCV antibody prevalence was below 74%, combining point-of-care HCV antibody and reflex point-of-care RNA testing (with or without consideration of previous treatment experience) is most cost-effective. Only modest improvements in treatment uptake are required for point-of-care RNA testing to be cost-effective compared to standard of care. Our study findings are important to inform optimal strategies for the scale-up of point-of-care testing in settings critical to support HCV elimination.

Point-of-care RNA testing was more cost-effective than standard of care, consistent with studies in the US, Canada, and Africa.15, 16, 17, 18, 19, 20, 21 Limitations of previous studies included limited cost inclusions,^15,17 underestimated point-of-care RNA testing costs,²⁰ and simplified testing pathways assuming no loss to follow-up.¹⁹ The strengths of this study are the availability of detailed costing data, the inclusion of costs for critical aspects of point-of-care testing (e.g., operator training, quality assurance, and IT/connectivity), and real-world data for key parameters related to HCV prevalence and treatment. The superior cost-effectiveness of point-of-care testing was primarily driven by a higher treatment uptake compared to standard of care (despite the higher costs related to point-of-care testing). Most costs related to standard of care were attributed to wasted testing costs due to loss of follow-up in the care cascade compared to point-of-care testing.

When the HCV antibody prevalence was below 74%, combined point-of-care HCV antibody with reflex point-of-care HCV RNA testing was most cost-effective, with consideration of previous treatment history further improving cost-effectiveness. This testing strategy was more cost-effective because people who were HCV antibody negative received a less expensive antibody test rather than an RNA test. The improved cost-effectiveness of this strategy is primarily driven by the decreased costs of point-of-care antibody testing (A$34 including test kit, consumables, labour) compared to immediate point-of-care RNA testing (A$96 including GeneXpert machine, test cartridge, consumables, labour). Point-of-care HCV antibody testing also has a shorter time to result compared to point-of-care HCV RNA testing (60 min). Combined point-of-care HCV antibody and RNA testing strategies are likely to improve patient and staff acceptability by reducing time to result for those who are HCV antibody negative (1–20 min vs. 60 min), while not considerably increasing the total time to result for those who are positive (61–65 min in total). It is unclear whether a combined point-of-care and RNA testing strategy has any impact on treatment uptake or response to therapy. Further research is needed to understand the effectiveness, acceptability, implementation challenges, and long-term cost-effectiveness to inform implementation, funding, and integration into policy and practice.

This study suggested that only modest increases in treatment uptake from 59% to 60% in standard of care to 67%–91% with immediate point-of-care HCV RNA testing is needed for the higher-cost point-of-care HCV RNA testing to become a preferred strategy compared to standard of care. Clinical studies have demonstrated that high treatment uptake (81%–93%) following point-of-care HCV RNA testing can be achieved in various settings.7, 8, 9, 10, 11 In our analyses, treatment uptake in the standard of care was based on the proportion of diagnosed people in 2016–2017 in New South Wales, providing an accurate estimate of HCV treatment uptake.³ The benefit of point-of-care testing with same-visit treatment is the ability to detect current infection and initiate treatment in a timely manner.^7,11 This study demonstrates that if real-world treatment uptake following point-of-care RNA testing is lower than pilot studies to date, point-of-care testing is still likely to be more cost-effective than standard of care. Lower treatment uptake is required by the point-of-care combined testing with treatment history consideration to achieve equivalent average costs per treatment initiation in standard of care, suggesting the relative advantage of point-of-care HCV antibody testing in cost terms.

This study has several limitations. Our models are based on empirical data from clinical trials and studies in Australian service settings. Thus, diagnosis outcomes for the base cases in each setting are mainly built on the epidemiological profiles of the population in these Australian settings.^28,29 The results are driven by the attributes of these studies, particularly the HCV antibody and RNA prevalence and treatment history among the study population. The findings may not be comparable to other populations with different epidemiological characteristics from these Australian studies. Interpretation and application of our study results for other populations and settings should be undertaken cautiously as HCV prevalence and testing costs would vary widely (e.g., homeless people compared to patients of mental health services or attendees of general practice).

Treatment costs for people diagnosed with HCV were excluded as the end of testing pathway for this cost-effectiveness analysis was initiation, rather than completion, of DAA treatment. The DAA reimbursement agreement between the Australian government and the pharmaceutical manufacturers stipulates the prices by a fixed budget regardless of number of patients treated.^33,34 Thus, the additional benefits of increasing treatment did not incur additional costs; therefore, the DAA pharmaceutical cost was regarded as a sunk cost and was excluded from this analysis. Lastly, lack of modelling the long-term outcomes for people living with HCV is a key limitation of the current analysis. The one-year time horizon means that potential long-term costs and benefits of treatment with respect to prevention of HCV transmission and reductions in morbidity/mortality are not captured. Our work in the near future will include the measurement of quality-adjusted life-years (QALYs) and mathematical modelling to estimate expected transmission and long-term disease burden benefits to guide implementation and enable comparison with other public health interventions and programs. In the interim, this study lays the foundations for future modelling by providing useful information on costs and benefits that can be expected in the short term and by identifying uncertain estimates requiring more detailed investigation for future extrapolation beyond the trial follow-up period. This study also provides a model that could be used to assess scale-up of point-of-care testing strategies in other countries that could be parametrised to the local epidemiology, including the Western Pacific. This is important given the recent inclusion of point-of-care testing in the new recommendations for simplified service delivery and diagnostic testing by the World Health Organization.³⁵

Conclusion

Point-of-care HCV testing enables diagnosis and treatment in a single-visit, increases testing acceptability, and reduces loss to follow-up, addressing the drop-off in the HCV care cascade. Combining point-of-care HCV antibody and RNA testing could optimise testing outcomes, enhancing cost-effectiveness and patient acceptability. This study is important for informing the future implementation of point-of-care testing strategies to achieve HCV elimination targets.

Contributors

S.T.F.S., V.W., and J.G. conceived the study. S.T.F.S., R.T.G., A.R.L., V.W., and J.G. obtained funding. S.T.F.S., Q.C., and R.T.G. contributed to model development. S.T.F.S. and J.G. designed the analyses, interpreted the results, and drafted the manuscript. J.C., H.V., Y.S., E.B.C., J.A.K., A.R.L., G.J.D., and J.G. provided data and validated model inputs. V.W. and J.G. supervised the project. All authors reviewed, edited, and approved the manuscript.

Data sharing statement

Models parameters are presented in Table 1 and Supplementary Table S2. Data sources for model parameters will be made available on request.

Declaration of interests

R.T.G. has received funding for his research from WHO and has provided non-funded project advice to Gilead and ViiV. Y.S. is a co-investigator on investigator-initiated research grants from AbbVie and Gilead Sciences. A.R.L. has received investigator-initiated research grants from AbbVie, Gilead Sciences, and Sequiris. G.J.D. is a consultant/advisor and has received research grants from Abbvie, Abbot Diagnostics, Gilead Sciences, Bristol Myers Squibb, Cepheid, GlaxoSmithKline, Merck, Janssen and Roche. J.G. is a consultant/advisor and has received research grants from AbbVie, Biolytical, Camurus, Cepheid, Gilead Sciences, Hologic, Indivor, and Merck/MSD and has received honoraria from AbbVie, Cepheid, Gilead Sciences, and Merck. No input into this work was provided by any of the above listed organisations or institutions.

All other authors have no conflicts of interest to declare.