The fiscal consequences of public health investments in disease-modifying therapies for the treatment of multiple sclerosis in Sweden

Abstract

Background and aims: The economic consequences of multiple sclerosis (MS) are broader than those observed within the health system. The progressive nature suggests that people will not be able to live a normal productive life and will gradually require public benefits to maintain living standards. This study investigates the public economic impact of MS and how investments in disease-modifying therapies (DMTs) influence the lifetime costs to government attributed to changes in lifetime tax revenue and disability benefits based on improved health status linked to delayed disease progression.

Methods: Disease progression rates from previous MS Markov cohort models were applied to interferon beta-1a, peginterferon beta-1a, dimethyl fumarate, and natalizumab using a public economic framework. The established relationship between expanded disability status scale and work-force participation, annual earnings, and disability rates for each DMT were applied. Subsequently, we assessed the effect of DMTs on discounted governmental costs consisting of health service costs, social insurance and disability costs, and changes in lifetime tax revenues.

Results: Fiscal benefits attributed to informal care and community services savings for interferon beta-1a, peginterferon beta-1a, dimethyl fumarate, and natalizumab were SEK340,387, SEK486,837, SEK257,330, and SEK958,852 compared to placebo, respectively. Tax revenue gains linked to changes in lifetime productivity for interferon beta-1a, peginterferon beta-1a, dimethyl fumarate, and natalizumab were estimated to be SEK27,474, SEK39,659, SEK21,661, and SEK75,809, with combined fiscal benefits of cost savings and tax revenue increases of SEK410,039, SEK596,592, SEK326,939, and SEK1,208,023, respectively.

Conclusion: The analysis described here illustrates the broader public economic benefits for government attributed to changes in disease status. The lifetime social insurance transfer costs were highest in non-treated patients, and lower social insurance costs were demonstrated with DMTs. These findings suggest that focusing cost-effectiveness analysis only on health costs will likely underestimate the value of DMTs.

Keywords:

• Public economics
• fiscal analysis
• multiple sclerosis
• disease-modifying therapies
• macroeconomic
• economics

JEL CLASSIFICATION CODES:

• E62
• C54

Introduction

Multiple sclerosis (MS) is an autoimmune-mediated disorder that affects the central nervous system and is the most prevalent, demyelinating neurological disease worldwide with a prevalence in North America and Europe of greater than 100 per 100,000 inhabitants¹. MS is mostly detected between the ages of 20 and 40 years, with less than 1% occurring in childhood and ∼ 2–10% after 50 years of age².

MS mostly afflicts individuals of working-age and often results in severe physical and cognitive disability with broad economic consequences that extend beyond the health service. Many studies recognize that foregone productivity is predominantly due to absenteeism, reductions in workforce participation and productive life years loss due to premature retirement linked to health status^3–8. Increasing disability amongst MS patients therefore raises the need for social insurance support from governments^8,9. In the United States, MS-attributed absenteeism and disability claims were reported to be 26% of the total MS attributable costs¹⁰. The effect of MS-attributable health shocks on workforce participation and early retirement can be translated into tax revenue loss and additional governmental costs¹¹.

Although there is no cure for MS, the introduction of disease-modifying therapies (DMTs) has improved the management of the disease, and offered profound benefits to patients that may result in substantial reductions of both the clinical and the economic burdens of MS, thus partially or totally offsetting their acquisition cost^12,13. The aim of this study was to estimate the effect of DMTs on government public economics by quantifying lost tax revenue and additional spending on social benefit transfer programs, i.e. transfers attributed to disability progression and preventable by DMTs. The model captures a broader set of cost parameters that are not included in traditional cost-effectiveness by applying a “government perspective” to understand the broader public economic consequences for investing in DMTs. This entails exploring the individual transactions between people with MS and the government and how DMTs influence these transactions in terms of taxes paid and transfer benefits received. In our analysis we compare each product with placebo, which is not a realistic treatment option considering the range of treatments available, to illustrate how public economic parameters have influenced changes in government transfer payments and taxes since the introduction of DMTs. This approach is not used to guide technical efficiency questions, rather it makes the investment case for public health investment costs in MS.

Methods

Disease simulation model

To simulate the natural history of disease by disease states, a previously described cohort model, developed to quantify the cost-effectiveness of MS DMTs, was used^12,13. The disease simulation model is a cohort-based Markov chain, tracking patients annually through EDSS levels whereby every year patients can remain at the same EDSS state, transition to a better or worse EDSS health state, have a relapse, transition from RRMS to SPMS and progress within SPMS, and/or die. If a patient transitions from RRMS to SPMS, it is assumed that the patient cannot transition back to RRMS. DMTs were assumed to reduce the rate of EDSS progression within RRMS and the annual relapse rate, but not the rate of transition from RRMS to SPMS. The model simulates the natural history of cohorts receiving no treatment (placebo) or one of the following DMTs:

1. Interferon beta-1a 30 µg (IFN ß-1a 30 µg);

2. Pegylated interferon beta-1a 125 µg (Peg IFN ß-1a 125 µg);

3. Dimethyl fumarate 240 mg (DMF 240 mg); or

4. Natalizumab 300 mg. (NTZ 300 mg).

A stopping rule was applied in the model whereby patients discontinued treatment at EDSS 7 for all treatments evaluated.

Estimation of public economic consequences

In this study we apply a modified generational accounting framework often used by central governments for policy planning to evaluate the broader public economic costs of MS that includes non-healthcare related costs and to understand how DMTs influence public spending by government^14,15.

The model compared fiscal outcomes of a MS cohort receiving one of the four DMTs vs a cohort receiving placebo. The comparative analysis simulated individuals of different ages diagnosed with RRMS at early EDSS levels and the resulting clinical and survival outcomes attributed to each DMT. The public economic consequences of DMTs and placebo were estimated alongside the outcomes of the MS cost-effectiveness model including lifetime direct and indirect medical costs. In addition, the model quantified the pharmaceutical, monitoring, and adverse events (AE) management costs of each DMT.

For each EDSS level a specific earnings profile and hence tax paying potential was assigned to individuals in each level based on the reported workforce participation by EDSS⁸. Furthermore, for each EDSS health state we applied the previously reported likelihood of receiving social insurance and disability benefits¹⁶. The analytic method employed assumes that lost productivity is attributed both to morbidity in more severe EDSS levels that impairs work activity and decreased life-expectancy based on severity¹⁷. The methods used for this analysis originated from the traditional human capital theory and the theory of generational accounting¹⁴, originally developed to explore the cross-sectorial and intertemporal impact of government policy on the value of current and future taxes a person is likely to pay over their lifetime. The methods for quantifying fiscal benefits are similar to the methods used for conventional cost-effectiveness analyses and are based on cumulative, discounted flows.

Equations (1)(1) $${\mathit{Gross}\mathrm{ }\mathit{Tax}}_{{\mathit{dmt}}_z}=M\times {\sum }_{t=0}^T{N}_j\times (Y_{ij}^{\mathit{Tax}}\times {\mathit{Tax}}^D+{{\mathrm{ }Y}_{ij}^{\mathit{Disp}}\times {\mathit{Tax}}^{\mathit{InD}})\times (1+d)}^{-t}$$(1) and (2)(2) $${\mathit{Transfers}}_{{\mathit{dmt}}_z}=M\times {\sum }_{t=0}^T{N}_j\times \left({\mathit{Dis}}_{ij}\right.+\left.{\mathit{Sic}}_{ij}+{HC}_{ij}\right)\mathrm{ }{\times \left(1+d\right)}^{-t}$$(2) illustrate the calculations of public economic outcomes. Equation (1)(1) $${\mathit{Gross}\mathrm{ }\mathit{Tax}}_{{\mathit{dmt}}_z}=M\times {\sum }_{t=0}^T{N}_j\times (Y_{ij}^{\mathit{Tax}}\times {\mathit{Tax}}^D+{{\mathrm{ }Y}_{ij}^{\mathit{Disp}}\times {\mathit{Tax}}^{\mathit{InD}})\times (1+d)}^{-t}$$(1) describes the calculation of gross taxes paid by MS patients for the selected time horizon. Gross tax is a function of age-specific and EDSS-specific taxable and disposable income. Taxable earnings refer to earnings from wages and investments, whereas disposable income includes transfers, activity, and non-activity income and represents the proportion of income which is allocated to consumption and hence indirectly taxed. Equation (2)(2) $${\mathit{Transfers}}_{{\mathit{dmt}}_z}=M\times {\sum }_{t=0}^T{N}_j\times \left({\mathit{Dis}}_{ij}\right.+\left.{\mathit{Sic}}_{ij}+{HC}_{ij}\right)\mathrm{ }{\times \left(1+d\right)}^{-t}$$(2) illustrates the calculation for the aggregation of age- and EDSS-specific social insurance allowances and disability pensions paid by government. (1) $${\mathit{Gross}\mathrm{ }\mathit{Tax}}_{{\mathit{dmt}}_z}=M\times {\sum }_{t=0}^T{N}_j\times (Y_{ij}^{\mathit{Tax}}\times {\mathit{Tax}}^D+{{\mathrm{ }Y}_{ij}^{\mathit{Disp}}\times {\mathit{Tax}}^{\mathit{InD}})\times (1+d)}^{-t}$$(1) (2) $${\mathit{Transfers}}_{{\mathit{dmt}}_z}=M\times {\sum }_{t=0}^T{N}_j\times \left({\mathit{Dis}}_{ij}\right.+\left.{\mathit{Sic}}_{ij}+{HC}_{ij}\right)\mathrm{ }{\times \left(1+d\right)}^{-t}$$(2) where ${\mathit{dmt}}_z$ = DMT $z$ ($z=1,\dots ,4$) or placebo; $M$ = number of Markov cycles; $t$ = year of analysis; $i$ = age; $j$ = EDSS level; $T$ = time horizon; $N_j\mathrm{ }$= number of patients in EDSS level $j;$ $d$ = discount rate; $Y^{\mathit{Tax}}\mathrm{ }$= annual age-specific taxable earnings (average for age $i,$ in EDSS $j$); $Y^{\mathit{Disp}}\mathrm{ }$= annual age-specific disposable income (average for age $i,$ in EDSS $j$); ${\mathit{Tax}}^D\mathrm{ }$= average direct tax rate; ${\mathit{Tax}}^{\mathit{InD}}\mathrm{ }$= average indirect tax rate, i.e. VAT; ${\mathit{Dis}}_{ij}\mathrm{ }$= annual disability costs (allowances/benefits, average for age $i,$ in EDSS $j$); ${\mathit{Sic}}_{ij}\mathrm{ }$= annual social insurance costs (allowances/benefits, average for age $i,$ in EDSS $j$); and ${HC}_{ij}\mathrm{ }$= annual healthcare costs in Markov state EDSS $j.$

The public economic framework affords the opportunity to measure: (1) cumulative discounted gross tax for each DMT and incremental difference with placebo; (2) cumulative discounted public transfers by DMT and incrementally compared to placebo; and (3) cumulative discounted health costs by DMT and incrementally compared to placebo. The costs per DMT investment are evaluated using fiscal benefit-cost ratios (fBCR) which consider all fiscal benefits (i.e. the costs savings and additional tax resulting from DMTs compared to placebo) in relation to DMT costs (i.e. drug costs including AE management costs). The fBCR provides a measure of fiscal benefit in relation to DMT costs with fBCR > 1.0, indicating fiscal surplus in relation to DMT costs. When the fBCR < 1.0 this indicates the amount of cost off-sets of the DMT investment from fiscal savings. For example, a fBCR of 0.59 indicates that 59% of the DMT treatment costs are being off-set by other government cost-savings or gains in tax revenue from improved productivity. In the presentation of results, all taxes and transfer values are reported, for simplicity, as positive. Hence, positive incremental (vs placebo) gross tax and negative incremental transfers signify a fiscal benefit for the government. Positive incremental healthcare costs represent a fiscal benefit (cost-savings), whereas negative increments represent fiscal costs. Fiscal outcomes for the four DMTs were evaluated separately in comparison with placebo. Fiscal outcomes are estimated for a time horizon of 50 years and all costs are discounted at 3%. The fiscal outcomes are continuously adjusted for survival in the model, therefore the timeframe has very limited impact on the results as few people are alive over the 50 year time horizon. Deterministic one-way sensitivity analysis was conducted for parameters which may have an impact on the fBCR applied to excess risk of morality for EDSS, changes in future earnings, transfers received and time horizon. Conducting sensitivity analysis on the fBCR ratio was considered most appropriate as this reflects how a single parameter will influence future outcomes, earnings, and transfers received from government.

Treatment outcomes and health service costs

The analysis simulated individuals (mean age 36 years old, 70% female) diagnosed with RRMS at early EDSS levels, consistent with the AFFIRM clinical trial^18,19. The same clinical study was used to obtain transition probabilities between EDSS levels within RRMS. Relapse rates by EDSS level and transitions from RRMS to SPMS were also obtained from the literature^20–22. Annualized relapse rate ratios and 6-month disability progression hazard ratios relative to placebo were obtained from a network meta-analysis and mixed-treatment comparison, discontinuation rates and AEs obtained from the clinical trials of each treatment^18,23–26. The full range of medical costs used in the model for the DMTs under study and the direct and informal care costs are provided in the Appendix.

Earnings, transfers, and tax

Direct income and indirect consumption (i.e. VAT) taxes paid by individuals were based on earnings and consumable spending, respectively, therefore the starting point for estimating tax revenue was derived from age-specific earnings profiles. Individuals with MS earn progressively lower wages than non-MS individuals, have deteriorating workforce participation rates, and face higher unemployment rates as their disability increases. To estimate the direct and indirect taxes paid by individuals with MS in Sweden, direct taxes were linked to age- and EDSS-specific earnings and indirect taxes were linked to consumption attributed to disposable income which represents 75% of all income²⁷.

To estimate earnings and transfer data of MS patients by age, data were obtained from a study in Sweden comparing the sources and levels of income of MS patients with the general population²⁸. To transform the annual earnings and benefits into EDSS-specific the estimators from Kavaliunas et al.¹⁶ were applied to the reported average age-specific earning and benefits²⁸. Specifically, the study by Kavaliunas et al.¹⁶, using cross-sectional Swedish data, reported (1) earnings, (2) disability pension, (3) sickness absence, (4) disability allowance, (5) unemployment compensation, and (6) assistance. The study stratified earnings and benefits into four disability groups: those with mild (EDSS 0–3.5), moderate mild (EDSS 4–5.5), moderate severe (EDSS 6–6.5), and severe MS (EDSS 7–9.5). The findings showed that MS patients with EDSS 4–5.5 have reduced earnings by −21.10%, patients with EDSS 6–6.5 receive earnings reduced by −31.30%, and finally that EDSS 7–9.5 results in a reduction of earnings of −58.80% compared to those with mild disability. Benefits increased by 49.3%, 73%, and 92% for patients with EDSS 4–5.5, EDSS 6–6.5, and EDSS 7–9.5, respectively. A linear interpolation of the resulting combined data was used to smooth and reflect earnings and benefits by discrete EDSS levels rather than strata.

To convert data to 2017 prices an inflation rate of 0.90% was applied²⁵. To estimate direct and indirect tax, earnings were linked to tax revenues. Specifically, the lowest average tax burden of 41% reported for Sweden by OECD was applied to earnings in order to quantify direct tax²⁹. The effective VAT rate of 25% was applied to the disposable income, as a proxy of the indirect tax paid through consumption²⁷.

Results

The per person disaggregated public economic costs for the four active DMTs and placebo are provided in Table 1. In all cases the estimated average per patient expenditure for informal care and community services were the predominant public costs at more than SEK5.9 million per person, followed by disease management costs that ranged from SEK2,159,483 (NTZ 300 mg) to SEK2,238,190 (placebo). For active treatment, DMT costs were approximately the second highest expenditure category. The lifetime social insurance transfers costs were highest in placebo treated subjects (SEK1,671,765), and lower social insurance costs were demonstrated with DMTs IFN ß-1a 30 µg (SEK1,660,576), Peg IFN ß-1a 125 µg (SEK1,655,245), DMF 240 mg (SEK1,663,974), and NTZ 300 mg (SEK1,633,829). Projected lifetime tax revenue per person was highest with NTZ 300 mg (SEK893,052) (Table 2).

Table 1. Per person incremental fiscal revenue/savings and treatment costs per DMT intervention (50-year horizon, discounted, SEK 2017).

	IFN ß-1a 30 µg	Peg IFN ß-1a 125 µg	DMF 240 mg	NTZ 300 mg
Fiscal benefits vs placebo	410,039	596,592	326,939	1,208,023
Savings for the health service (direct disease management costs (non-drug-related), informal care & community services costs)	371,377	540,413	297,488	1,094,278
Savings for social insurance (disability allowances, pensions and all other social benefits/transfers)	11,188	16,519	7,790	37,936
Fiscal benefits from tax revenue gains	27,474	39,659	21,661	75,809
Treatment costs (drug related costs) of DMTs vs placebo	392,360	354,952	448,277	1,735,970
Fiscal Benefit-Cost Ratio (fBCR)	1.05	1.68	0.73	0.69

Table 2. Public economic outcomes disaggregated (tax revenues and transfers) per person by treatment (50-year horizon, discounted, SEK 2017).

Outcome	IFN ß-1a 30 µg	Peg IFN ß-1a 125 µg	DMF 240 mg	NTZ 300 mg	Placebo
Total costs of health service including medical costs, informal care & community services	8,810,896	8,604,451	8,940,703	9,431,606	8,789,913
^†Disease Management	2,215,361	2,203,031	2,220,240	2,159,483	2,238,190
^†Relapses	261,858	251,601	247,811	213,300	270,018
^†Informal care & community services	5,932,657	5,786,207	6,015,714	5,314,192	6,273,044
^‡DMT therapy	350,148	353,716	447,302	1,263,817	–
^‡Administration and monitoring	44,682	7,215	7,867	311,561	–
^‡AE management	6,191	2,682	1,768	169,254	8,661
Social insurance costs (disability allowances, pensions and all other social benefits/transfers)	1,660,576	1,655,245	1,663,974	1,633,829	1,671,765
Tax revenue (fiscal revenues from direct & indirect tax)	844,717	856,902	838,905	893,052	817,243

^†Disease management costs; ^‡Drug related costs.

To estimate the fiscal return to government attributed to DMTs, we compare incremental fiscal benefits attributed to healthcare cost savings, social insurance savings, and increased tax revenue in relation to DMT treatment costs compared with placebo (Table 1). The sum of all incremental fiscal benefits per person for the four DMTs ranged from SEK326,939 for DMF 240 mg to SEK1,208,023 for NTZ 300 mg. The majority of cost savings were attributed to direct health costs, which varied from SEK297,488 for DMF 240 mg to SEK1,208,023 with NTZ 300 mg. The minimum lifetime increase in tax revenue was SEK21,661 with DMF 240 mg, but was as high as SEK75,809 for NTZ 300 mg. The measure of fiscal benefit in relation to costs varied from 0.69 for NTZ 300 mg to 1.68 for Peg IFN ß-1a 125 µg.

Sensitivity analysis

A sensitivity analysis was performed on the fBCR as this captures the impact of changing a single parameter on both future tax revenue and government transfer payments and reflected in a single ratio. Changing excess mortality by EDSS, future earnings, and benefits by ±20% had limited impact on the fBCR for all four products. Changing the time horizon to shorter duration generally improved the fBCR (Table 3).

Table 3. Sensitivity analysis of fiscal benefit cost ratio (fBCR) for changes to key parameters.

Parameter variation	−20%	20%
Sensitivity analysis for the EDSS RR
IFN ß-1a 30 µg	1.07	1.05
Peg IFN ß-1a 125 µg	1.73	1.68
DMF 240 mg	0.75	0.73
NTZ 300 mg	0.72	0.70
Sensitivity analysis for earnings
IFN ß-1a 30 µg	1.03	1.06
Peg IFN ß-1a 125 µg	1.66	1.70
DMF 240 mg	0.72	0.74
NTZ 300 mg	0.69	0.70
Sensitivity analysis for benefits
IFN ß-1a 30 µg	1.04	1.05
Peg IFN ß-1a 125 µg	1.67	1.69
DMF 240 mg	0.73	0.73
NTZ 300 mg	0.69	0.70
Sensitivity analysis for the time horizon
Time horizon in years	20 years	30 years	40 years
IFN ß-1a 30 µg	0.98	1.11	1.07
Peg IFN ß-1a 125 µg	1.57	1.78	1.73
DMF 240 mg	0.70	0.77	0.75
NTZ 300 mg	0.62	0.73	0.72

Discussion

The results described here illustrate several findings that can inform treatment access, funding, and policy in MS. Firstly, we demonstrate that the major cost driver in MS is attributed to informal care and community services, supporting previous findings in Sweden³⁰. These costs were the largest cost driver for all active DMTs and placebo (i.e. no treatment), and costs associated with disease management were the second highest cost category. Second, of all public costs, those attributed to social insurance payments were the third highest cost category. It is important to recognize that the different cost categories described here occur at different stages over the duration of the model and that discounting can distort the overall costs. For example, DMTs are mostly given in the early stages of the disease, therefore less discounting has occurred over shorter time periods in the model. In contrast, social insurance payments occur at later stages of life and would be further discounted in today’s value. The third observation is that DMTs can provide cost savings for government beyond the health service. For example, we demonstrate a reduction in social insurance costs associated with all four DMTs. The magnitude of the savings is associated with the disease modifying efficacy profile. DMT’s ability to slow progression will positively influence tax revenue for government as people can remain active and employed for longer and reduce dependence on social insurances.

As anticipated, all DMTs result in higher drug and drug-related costs. However, to put costs of DMT treatment into perspective we present the fiscal benefit-cost ratio (fBCR) for each intervention. These costs represent the investment of governments in DMTs of the reimbursed products enabling one to understand costs including AE costs, administration, and monitoring costs for each therapy in relation to all public benefits. Although DMTs are not curative, in return for the initial investment, the government receives benefits arising from the efficacy of DMTs in delaying disease progression and morbidity as well as the occurrence of relapses. These include: health service cost-savings, social insurance, and disability (transfers) cost-savings. With IFN ß-1a 30 µg and Peg IFN ß-1a 125 µg we demonstrate that more public economic benefit is gained in relation to the overall treatment costs with fBCRs of 1.05 and 1.68, respectively. The fBCRs for DMF 240 mg and NTZ 300 mg were 0.73 and 0.69, which indicates that 73% and 69% of treatment costs are offset in fiscal savings, respectively. These savings represent the fiscal implication of delaying MS progression and corresponding disability. Furthermore, all DMTs result in higher tax revenues compared to placebo due to the higher productivity and earnings.

The fBCR for each product is dependent on a range of factors including the population treated, clinical efficacy, and how improvements in efficacy translate into future productivity gains and fiscal impact in relation to the costs of therapy. To gauge the size of return we compared each DMT with placebo, which is not realistic in clinical practice, however it does illustrate the incremental fiscal gain for government from its initial investment for each therapy. We don’t encourage comparing across the different products which is more appropriately addressed using conventional cost-effectiveness analysis. Although the example described here is applied to MS, we believe that fiscal analysis is an appropriate approach to compare any healthcare investment¹⁴.

Governments project future tax revenues based on population averages for wages, labor force participation, tax rates, and growth adjusted for population level survival rates¹⁵. Consequently, health conditions that influence one’s ability to live a normal, active, and productive life will have economic consequences for government attributed to spending/transfers providing income support and reduced lifetime tax revenues that can be much greater than those experienced in the health service¹⁴. Previous government assessment of the broader impact of ill-health in working aged populations have illustrated that lost tax revenue and government disability payments account for more than 80% of costs of ill-health, and health costs represent ∼ 15% of overall costs to government³¹. While the main aim of health services is not to generate revenue for government, the analysis described provides some guidance as to the broader public economic consequences and how changing the course of a chronic condition can positively influence future tax revenue and save money across different government sectors. Perhaps more importantly we believe our work highlights the importance of looking beyond only healthcare costs and, therefore, advocates the use of a broader societal or fiscal perspective when evaluating new or existing technologies in MS.

The MS gender imbalance which afflicts higher percentages of women could influence fiscal models due to possible differences in labour force participation and earnings. However, as this model reflects Sweden, which has one of the highest female labour force participation rates globally, the differences are likely to be minimal. Current data published by Statistics Sweden indicate male and female labour force participation rates of 80.8% and 78.1%, respectively, in 2018³². Furthermore, the salary data reported by Wiberg et al.²⁸ used in our model, on which taxes were derived, was obtained in an MS cohort consisting of 71% females, consequently the model already reflects the gender-based earnings reported by women in Sweden. Additionally, whilst it is possible to reflect gender differences in fiscal outcomes, this would require two separate models constructed with gender-specific variables throughout the model, which would be difficult to obtain. Furthermore, as DMT efficacy is observed in both males and females, providing a gender-based analysis is not likely to inform treatment practices and policy-making.

The projected tax revenue described here is likely an underestimation of the overall tax losses attributed to MS. In many situations informal care becomes necessary in more advanced disease stages, which is often performed by family members³³. Family members that forego work to care for family members experience income loss that would result in further reduced income tax revenue. Furthermore, people experiencing wage penalties associated with their condition are likely to have fewer lifetime savings and wealth in retirement, which further exacerbates the fiscal situation. This highlights the mutual dependency of government and citizens and the importance of health, which underpins this relationship. This relationship between health and wealth is well known – although often obscured – and has been the basis for many countries establishing national health services. Furthermore, the focus of this analysis has been on fiscal costs and revenue. It is important to recognize that individuals have a much broader impact on the economy that is not captured here. It is likely that through increased lifetime earnings people generate additional demand for goods and services that can have additive economic benefits not described within the fiscal framework.

The analysis described here highlights the broader public economic benefits for government attributed to changes in disease status. The findings demonstrate that many cross-sectorial cost-savings can be achieved with DMTs in treating MS. In some instance the costs of some DMTs (IFN ß-1a 30 µg and peginterferon beta-1a) are fully paid for from reduced health cost savings, informal care, community services, social insurance benefits, and increases in tax revenue. We demonstrate a considerable rebate for DMF 240 mg and NTZ 300 mg are observed by adopting a broader government fiscal analysis. This suggests that governments will only pay ∼ 20% of the treatment costs and 80% is paid for through savings across other government departments over time. This framework can be used to communicate with relevant government sectors regarding the value of DMT and help secure funding for treatments.

Transparency

Declaration of funding

The work performed by NK and MPC was supported by a grant from Biogen. NK and MPC declare no financial interests or holdings in the sponsoring company. The authors retained full editorial control over published content and final submitted manuscript. TD was previously employed by the sponsoring company and received stock options from the sponsoring company as part of his previous remuneration package. Due to unfortunate circumstances the author TD died in the final stages of manuscript preparation. AK worked as an advisor and contributor to the work described here; specifically, in relation to the data sources, results and manuscript preparation.

A peer reviewer on this manuscript has disclosed that they have previously received reimbursement for speaking and/or research from Biogen, Teva, Sanofi-Genzyme, and Celgene. The peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Declaration of financial/other interests

The study described here is a modeling study of existing data sources and does not use any individual patient data.

Previous presentations

The work described here was previously presented as an abstract at the European Committee for Treatment and Research in Multiple Sclerosis (ECTRIMS) conference in 2019.

Acknowledgement

We would like to thank Mattia Gianinazzi for helping us to prepare the manuscript following the sudden death of our co-author and friend Thibaut Dort.