The health care and societal costs of inherited retinal diseases in Australia: a microsimulation modelling study

Abstract Objectives To estimate the health care and societal costs of inherited retinal diseases (IRDs) in Australia. Design, setting, participants Microsimulation modelling study based on primary data — collected in interviews of people with IRDs who had ophthalmic or genetic consultations at the Children's Hospital at Westmead or the Save Sight Institute (both Sydney) during 1 January 2019 – 31 December 2020, and of their carers and spouses — and linked Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Schedule (PBS) data. Main outcome measures Annual and lifetime costs for people with IRDs and for their carers and spouses, grouped by payer (Australian government, state governments, individuals, private health insurance) and type (health care costs; societal costs: social support, National Disability Insurance Scheme (NDIS), income and taxation, costs associated with caring for family members with IRDs); estimated annual national cost of IRDs. Results Ninety‐four people (74 adults, 20 people under 18 years; 55 girls and women [59%]) and 30 carers completed study surveys (participation rate: adults, 66%; children, 66%; carers, 63%). Total estimated lifetime cost was $5.2 million per person with an IRD, of which 87% were societal and 13% health care costs. The three highest cost items were lost income for people with IRDs ($1.4 million), lost income for their carers and spouses ($1.1 million), and social spending by the Australian government (excluding NDIS expenses: $1.0 million). Annual costs were twice as high for people who were legally blind as for those with less impaired vision ($83 910 v $41 357 per person). The estimated total annual cost of IRDs in Australia was $781 million to $1.56 billion. Conclusion As the societal costs associated with IRDs are much larger than the health care costs, both contributors should be considered when assessing the cost‐effectiveness of interventions for people with IRDs. The increasing loss of income across life reflects the impact of IRDs on employment and career opportunities.

I nherited retinal diseases (IRDs) comprise a group of heterogeneous monogenic disorders characterised by retinal dysfunction and degeneration, often leading to progressive vision loss and blindness. 1 Genomic variant prevalence data suggest that IRDs affect about one in 1000 people, 2,3 and IRDs are now the leading cause of blindness in working age adults. 4 As early onset of disease often causes significant vision deficits during childhood and throughout life, their impact, both on those directly affected and on their families, is considerable. 5 The limited treatment options in Australia have gradually improved since voretigene neparvovec (Luxturna, Novartis) was approved for people with RPE65 retinopathy in 2020, the first publicly subsidised in vivo gene therapy in Australia; 6 the first patient was treated in the same year. 7 The Medical Services Advisory Committee (MSAC) acknowledged the clinical need for an effective treatment, but also noted some limitations, including in the cost-effectiveness analysis for the application (which did not take benefits for carers into account), and required a full review within three years of its approval. 8 The impact of vision loss includes increased risk of poverty, reduced quality of life and employment opportunities, and greater expenses for informal care. 9-11 Information on the total cost of IRDs, including use of health and social services, education, employment and income, is limited. 10, 12 In the United States, mean annual health care costs for people with IRDs were US$7317 higher than for the general population, 13 but this estimate, like those of other overseas studies, 14,15 did not take all health care (eg, allied health, vision aids) and other costs into account.
Two recent overseas investigations found that the annual national costs of IRDs are considerable: US$13.4-31.8 billion in the United States and CAN$1.6-6.7 billion in Canada, 16 and £42.6 million in Ireland and £523.3 million in the United Kingdom. 17 The costs associated with reduced workforce participation and quality of life were substantial, health care expenses comprising only 2-7% of total costs. However, these studies included people with non-IRD vision impairment, included large loss of quality of life components in addition to disease costs, did not compare income losses for people with IRDs with mean population incomes, did not control for factors such as age and education, and relied on mean welfare payments and tax rates rather than collecting primary data on these items. Further, their online surveys of people with vision impairment were subject to selection biases (including vision capacity). A health care system and societal perspective is needed when assessing the costs and benefits of new clinical interventions, including genomic testing and targeted therapies. However, no complete overview of the costs associated with IRDs in Australia has been reported. We therefore estimated the Total estimated lifetime cost was $5.2 million per person with an IRD, of which 87% were societal and 13% health care costs. The three highest cost items were lost income for people with IRDs ($1.4 million), lost income for their carers and spouses ($1.1 million), and social spending by the Australian government (excluding NDIS expenses: $1.0 million). Annual costs were twice as high for people who were legally blind as for those with less impaired vision ($83 910 v $41 357 per person). The estimated total annual cost of IRDs in Australia was $781 million to $1.56 billion.

Conclusion:
As the societal costs associated with IRDs are much larger than the health care costs, both contributors should be considered when assessing the cost-effectiveness of interventions for people with IRDs. The increasing loss of income across life reflects the impact of IRDs on employment and career opportunities.
The known: Inherited retinal diseases (IRDs) have significant cost implications, both for people with these disorders and for their families. Comprehensive information about the costs of IRDs in Australia, however, have not been published.
The new: Using a microsimulation model, we estimated an overall lifetime cost of $5.2 million per person with an IRD. Societal costs, including government support and lost income for people with impaired sight and their families, accounted for 87% of all costs, health care costs for only 13%.
The implications: To correctly assess the cost-effectiveness of IRD treatments, including genomic testing and gene therapies, the substantial societal costs of IRDs must be taken into account.
health and societal costs of IRDs, analysing both primary and administrative health care data for a group of people with IRDs and their carers. We used microsimulation modelling to estimate income losses, important flow-on costs (eg, subsidies for housing), and National Disability Insurance Scheme (NDIS) costs.

Methods
Our microsimulation modelling study is one component of the Economic and Psychosocial Impact of Caring for Families Affected by Visual Impairment (EPIC-Vision) study. All people with clinical IRD diagnoses who had ophthalmic or genetic consultations at the Children's Hospital at Westmead or the Save Sight Institute (Sydney Eye Hospital campus, University of Sydney) during 1 January 2019 -31 December 2020 and their carers or partners were invited to participate. Specialist clinicians interviewed participants using the EPIC-Vision questionnaire, a tailored IRD questionnaire designed in consultation with ophthalmologists, clinical geneticists, genetic counsellors, and health economists; it includes questions on social services use (government support, NDIS) and other costs (education, aids, and modifications). When appropriate, carers were also interviewed on behalf of people with IRDs, including all those under 18 years of age.

Costs analysis
Costs data for our simulation model were derived from the EPIC-Vision questionnaire responses and from linked Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Schedule (PBS) data. This information was provided to Services Australia, which performed the data linkage and supplied de-identified data to the investigators. The health care-related, income and employment, and societal parameters included and their data sources, together with the indices used for adjustment to December 2022 prices, are summarised in the Supporting Information. All analyses were undertaken in SAS 9.4.

Income and tax effects
We estimated the income and tax effects of IRDs in a static income model (STINMOD) based on a representative sample of people and households in Australia (Box 1). People with IRDs and their carers and partners were matched in STINMOD with respect to age, gender, and highest level of education, and 1000 simulations run for each person. The mean outcomes were then used as the counterfactuals for each participant (ie, the estimated value if they did not have an IRD), and the effect of the IRD defined as the difference between the simulation outcomes and EPIC-Vision data for the participant. STINMOD data reflect income and tax levels at the end of 2019. 18

Household and lifetime cost
Participants were grouped by age (0-5, 6-18, 19-29, 30-39, 40-49, 50-59, 60-69, 70-80 years). Total lifetime costs were calculated from the mean costs for each age group. As none of the participants were over 80 years of age, we used mean annual costs for the 70-80 year age group to project costs to the mean life expectancy age in Australia (84 years). 19 The estimated total costs for each person diverge from the sum of the component costs because of an adjustment that reflects the fact that as income declines for an individual, they receive increasing welfare support.

Visual acuity and inherited retinal disease diagnosis
To assess whether costs differed by visual acuity, we estimated a model in which people were grouped as to whether they were eligible to be registered as being legally blind in Australia (Snellen visual acuity poorer than 6/60) or had measured visual acuity of better than normal to moderate vision impairment (6/60 or better).

Health and societal costs
Health costs are costs related directly to the health system and use of health products such as pharmaceuticals. Societal costs refer to all other costs, including income losses, as well as government service and NDIS costs. We report the proportions of all costs attributed to health and societal costs.

Total national costs
Because of uncertainty regarding the prevalence of IRDs, we used two estimated values for estimating total national costs: 0.1% 1 and 0.05%. 20

Ethics approval
The EPIC-Vision study was approved by the Sydney Children's Hospitals Network Human Research Ethics Committee (HREC/18/SCHN/292). Carers completed surveys on behalf of ten adults with IRDs; the mean age of these adults was 31.7 years (SD, 17.6 years), and visual acuity less than 6/60 vision for nine. Seven of these adults also completed surveys themselves; their responses, not the proxy responses of their carers, were included in our analysis.
Of thirty people under 18 years invited to participate, the carers of twenty agreed on their behalf and completed the surveys (66%). Their mean age was 11.0 years (SD, 4.6 years) and eight were girls (40%). The visual acuity of ten children (50%) was less than 6/60 (legally blind); twelve had NDIS plans (60%) (Box 2).
Of 47 carers invited to participate in the study, thirty agreed to do so and completed the surveys (63%). Their mean age was 47.0 Research years (SD, 11.0 years) and 25 were women (83%). Nineteen carers had university degrees (63%); ten had full-time employment (33%), nine had part-time work (30%), and eleven (30%) were not employed (Box 2).

Overall costs
The total estimated lifetime cost was $5.2 million per person with an IRD and $5.6 million per household. The mean annual cost was highest for the 50-59-year age group ($96 167), attributable to income loss for both people with IRDs and for their spouses and carers. The mean annual cost was lowest for the 18-29 (smaller income losses) and 70-80-year age groups (not eligible for NDIS plans) (Box 3). Of total costs, 87% were societal costs and 13% health care costs. The three highest cost items were $1.4 million in lost income for people with IRDs, $1.1 million in lost income for their carers and spouses, and $1.0 million in social spending by the Australian government (excluding NDIS expenses) (Box 4). The mean annual costs were about 40% higher for male than female participants (Box 5), including greater loss of income and NDIS expenses (data not shown).

Health care costs
Total lifetime health care costs were estimated to be $690 725 per person with an IRD: $324 429 borne by the Australian government (47%), $206 614 by state governments (30%), $120 789 by people with IRDs (17%), and $38 893 by private insurance (6%) (Box 4). Lifetime health care costs were highest for the 0-5 year age group (higher health service use) and the 30-39 year age group (explained by extended hospital stays for one person with an IRD) (Box 3).

Income and taxes
Total estimated lifetime loss of income was $2.6 million per person with an IRD (people with

Federal government social and welfare support
Total estimated lifetime Australian government social support costs (excluding the NDIS) was $1.0 million per person with an IRD; total lifetime NDIS costs for a person with an IRD were $749 462 (Box 4). Social support costs were generally higher for older age groups (Box 3). The estimated lifetime cost to the federal government for social support for carers and spouses was $264 113 per person with an IRD (Box 4), primarily associated with caring for a person under 18 years of age with an IRD (Box 3).

Visual acuity
The total estimated annual cost for people with visual acuity poorer than 6/60 (ie, legally blind) were $83 910 per person, more than twice those for people with IRDs but whose visual acuity was at least 6/60 (Box 5); the difference was attributable to greater lost income and higher social support costs (including NDIS costs). However, total health care costs were slightly lower for the people with visual acuity poorer than 6/60 (data not shown).

Total national annual cost
The estimated total annual cost attributable to IRDs in Australia was $781 million to $1.56 billion. After adjusting for the larger proportion of female participants with IRDs in our study, the estimated annual cost was $830 million to $1.66 billion (data not shown).

Discussion
On the basis of primary and linked administrative data, we estimate that the lifetime cost associated with an IRD is $5.2 million per person, or a total of $781 million to $1.56 billion per year. Of these costs, 87% are societal costs and 13% health care costs.
Previous studies have found that vision loss in general is associated with substantial costs. 11, 12,23,24 Interpreting the findings of two recent studies specifically concerned with IRDs 16,17 is made difficult by their data limitations and imprecise estimates of income losses. In our study, micro simulation modelling overcame these limitations by matching a group of people with diagnosed IRDs and their carers with people in the general population with similar characteristics rather than relying on comparisons with mean population income. This approach provides more precise estimates of the income and tax values we would expect for our participants if they did not have IRDs. Further, we found that estimated costs were highest for the 50-59-year age group, primarily because of lost income.
People with IRDs receive a mixture of NDIS services, and their lifetime NDIS costs are greater than their health care costs. As people with NDIS plans before the age of 65 years can choose to remain in the NDIS beyond this age, our modelling would underestimate costs for people over 65 were the NDIS to support more services in the future than are currently available to people over 65 years of age.
We found that IRDs are associated with substantial income losses, exceeding $44 000 per year for people aged 50-59 years. This finding was consistent with comments by survey participants that they worked less because of their IRD or had left the workforce entirely. This suggests that helping people with IRDs with employment could improve societal outcomes. The impacts on both carers and spouses were also notable. For carers, lost income was largely linked with caring for children with IRDs. Although spouses might not be full-time carers, the burden could nevertheless affect their ability to undertake paid work.
Health care costs were slightly higher for people with better vision, but overall costs were much higher for those who were legally blind. Societal costs must be considered when assessing the costs and benefits of IRDs, as health care cost estimates may not capture all costs.
Out-of-pocket health expenses comprised a small portion of total health costs, consistent with the lack of treatment options Research available for people with IRDs. They also spend substantial amounts on housing and transport modifications, and they may not be aware of all support provided by health care, the NDIS, and other agencies and consequently not receive appropriate support. Further, as some data were collected during the early phases of the coronavirus disease 2019 pandemic, our estimates of service use and related costs may be conservative.

Limitations
The data for our study were collected from a large group of people with IRDs, allowing a comprehensive assessment of costs. Further, our simulation model was designed to be adaptable and applicable to other medical conditions. However, as the proportion of women in our study was larger than in the Australian population, income loss may have been underestimated, given the higher employment and income rates of men and the greater use of health services by women. Finally, we did not include any people near the end of life living in institutions such as nursing homes. The health care and other costs for people with IRDs near the end of life require further investigation.

Conclusion
In our simulation study, we identified that IRDs are associated with substantial health care and societal costs. Given that 87% of the overall costs were societal, largely related to lower rates of employment for both patients and carers and their greater need for social support, it is crucial that societal costs are considered by cost-effectiveness evaluations of future IRD interventions, including genomic testing and targeted therapies.