About NCBI

Screening the older population

Summary

A hypothetical intervention of on an educational program and eye screening for people aged over 70 years would result in:

costs of €31,844 for the public health educational campaign and €165,092 for the screening tests;
482 cases of mild VI being detected through resulting screening tests;
discounted five-year health care costs of €812,952 to treat detected cases;
deadweight efficiency losses of €486,204 from government funding of the campaign, eye tests and health care costs; and
84 DALYs averted as a result of the intervention.

Under a societal perspective, this intervention would be associated with a cost of €17,738 per DALY averted under a societal perspective and €11,974 under a health care perspective. Therefore, the intervention is highly cost effective using WHO thresholds.

7.1.1 Description

Eye tests are important for detecting eye disease and refractive error, particularly for the elderly population who are at higher risk of developing partial sight and blindness. According to the Royal National Institute of Blind People (RNIB, 2007), a large proportion of elderly people neglect to have regular eye tests due various reasons, including:

absence of symptoms;
worries about the cost of glasses and/or lens prescription;
accessibility restrictions, e.g. lack of coverage of opticians in certain geographic areas and associated travel issues for elderly people;
lack of awareness of the impact of eye tests and treatments on eye disease; and
lack of awareness of entitlement to free eye tests and free glasses/lens prescriptions.

The hypothetical intervention assessed in this section is an educational program similar to Vision Initiative, a public health program aimed at preventing avoidable partial sight and blindness for those living in Victoria, Australia (Müller et al, 2007). The cost effectiveness of screening older people in the UK was assessed in Access Economics (2009) and a similar methodology has been applied here.

The hypothetical education program for the ROI is assumed to be similar, would target the population aged over 70 years, and consist of messages and advertisements through national and regional television and radio stations, national and regional newspapers, and alternative publications such as magazines and online media outlets. The aim of the program would be to generate greater awareness and access to services in order to discover mild VI in those who would not have otherwise presented. The education program is expected to increase the numbers of eye tests in the elderly population. It is assumed that people with moderate VI would already be aware of their condition and seek treatment. Detected mild VI is assumed to be treated immediately.
The HSE provides optical services free of charge to Medical card holders, including everyone aged 70 years and over who is normally resident in the ROI (NCBI, 2010). Since this intervention is assumed to target those aged 70 years and over, the increase in resulting eye tests are all assumed to be funded completely by the government.

7.1.2 Target population and reach

There were an estimated 353,298 people aged 70 years and over in the ROI in 2010. The 1998 Survey of Lifestyles, Attitudes and Nutrition in Ireland reported 47.9% of older respondents taking an eye test over the previous five years (National Council on Ageing and Older People, 2004). This translates to an annual non-tested rate of 10.4% ([100%-47.9%]/5). When the number of people aged 70 years and over in 2010 (353,298) is multiplied by this non-tested rate (10.4%) it is estimated that 36,814 older people did not have an eye test in 2010.

Müller et al (2007) reported that the Vision Initiative campaign reached 64% of the target audience, and of the people reached, 27% noted that campaign messages changed their eye health behaviours. In the context of this intervention, this would translate to getting eyes tested. These data were applied to the target population of 36,814 people in 2010 to estimate that 6,361 people aged 70 years and over would get their eyes tested specifically as a result of the public campaign.

7.1.3 Cases of undetected mild VI diagnosed

It is assumed that the VI detected by screening tests would only be in its mild stages (6/18≤VA<6/12). As explained above, it is assumed that a person with moderate VI or blindness would visit an optometrist or ophthalmologist since their vision loss would affect their daily life. Some people with comorbidities such as dementia or limited mobility may not fall into this category. Therefore, the cost effectiveness of the screening program reported here should therefore be considered an under-estimate.

Prevalence rates of mild VI by condition and age (Section 2.4) were applied to the additional number of eye tests generated by the campaign to estimate the additional number of mild VI cases (diagnosed and undiagnosed) within this population subset. However, since it is assumed that screening would detect undiagnosed cases of mild VI, an adjustment was needed to estimate the proportion of cases that would be undiagnosed prior to screening.

A UK study was used to estimate the number of undiagnosed mild VI cases detected through screening (Reidy et al, 1998) was used. This study surveyed people aged 65 and over in the North London area and set out the proportion of people with vision loss not in touch with eye care services. These proportions were also reported by Access Economics (2009).

The proportions of people with partial sight and blindness not in touch with eye care services (based on UK data) are presented in Table 7.1.

Table 7.1: People with partial sight and blindness not in touch with eye care services, UK

Eye disease	Proportion not in touch with eye care services (%)
AMD	86.0%
Cataract	88.0%
Diabetic Retinopathy	88.0%
Glaucoma	74.0%
Other	88.0%

Source: Access Economics (2009) from analysis of Reidy et al (1998).
These proportions were applied to the estimated number of mild VI cases to estimate the portion of cases previously undetected that were picked up by screening.

Estimates of eye tests and previously undetected mild VI diagnosed through screening are presented by condition and age in Table 7.2. Overall, it is estimated that 482 cases of mild VI would be diagnosed through additional eye screening tests resulting from the public campaign based on 2010 population estimates.

Table 7.2: Number of eye tests and undetected mild VI cases diagnosed by screening

Age	Tests	AMD	Cataract	DR	Glaucoma	Other	Total
Male
70-74	1,107	8	1	6	6	29	50
75-79	810	6		2	4	16	28
80-84	492	13	1	1	5	17	37
85-89	248	14	1	1	5	13	34
90+	90	16			7	12	36
Total	2,746	57	4	10	27	87	185
Female
70-74	1,195	8	1	5	3	32	50
75-79	976	12	1	2	3	18	35
80-84	744	26	1	1	4	24	56
85-89	467	37	1	1	5	28	72
90+	232	40	2	1	6	35	85
Total	3,615	124	6	10	22	136	298
TOTAL	6,361	181	10	20	48	223	482

Source: Deloitte Access Economics calculations using CSO (2008), Müller et al (2007), National Council on Ageing and Older People (2004), Reidy et al (1998) and prevalence rates from Section 2.4.

7.1.4 Cost of campaign and eye tests

Following a similar analysis for a UK educational campaign (Access Economics, 2009), the cost of the educational campaign in the ROI was derived from costs of the Vision Initiative in Australia (Müller et al, 2007). Based on the data reported by Muller, Access Economics (2009) estimated the program to have cost $A0.116 per person targeted in 2005. This cost would convert to $A0.130 in 2010 prices (RBA, 2011) and to €0.090 per person targeted at the average exchange rate in 2010, being €1 = $A1.44 (ECB, 2011). This per person cost was multiplied by the number of people aged 70 and over in the ROI (353,298) to estimate a total campaign cost of €31,844.

It should be considered that there may be economies of scale for education campaigns. That is, the cost of a campaign per person may be lower when the fixed costs of a campaign are attributed to a greater population. The relative population sizes (70+ years) in the ROI (353,298 people in 2010 from CSO, 2008) and Australia (2.1 million people from Deloitte Access Economics’ Demographic Model) differ, therefore suggesting that per person costs may be higher in the ROI than estimated using Australian campaign data, and hence cost effectiveness results should be interpreted with caution.

The cost of a screening test was estimated as the average of the fees for an eye examination by an ophthalmologist/ophthalmic medical practitioner (€26.50) and ophthalmic optician (€23.35) in 2008 (from Table 3.6, Section 3.3). These were assumed to be the most representative fees for a standard eye test for the elderly amongst all eye examinations listed in Table 3.6 (Section 3.3).

This cost (€24.93 in 2008 prices) was inflated to 2010 prices using the annual change in the health consumer price index in the ROI, which was 3.5% between 2008 and 2009, and 0.6% between 2009 to 2010 (CSO, 2011). The resulting fee was €25.95 per examination. This fee was multiplied by the 6,361 additional screening tests in the elderly population to estimate a total screening cost of €165,092 generated by the campaign.

7.1.5 Treatment effectiveness and compliance

Effectiveness is defined here as the percentage of vision loss that can be avoided through timely clinical intervention, given the current prevalence of VI. Compliance is the percentage of people who are likely to adhere to recommended timely clinical intervention.

Estimates of compliance were needed to calculate the health care treatment costs of mild VI resulting from screening. Estimates of treatment effectiveness were also needed to determine the number of DALYs averted by successful treatment.

In the absence of ROI specific data, treatment effectiveness and compliance were estimated from Access Economics’ study on eye care interventions in Australia (Access Economics, 2005). As reported previously by Access Economics (2005), compliance and treatment parameters were derived from applicable international and Australian studies detailed below.

The effectiveness of treatments for cataract, refractive error and other causes, were derived from Australian MVIP data on presenting VA at follow-up and expert opinion from Eye Research Australia ophthalmologists on conservative estimates.
The effectiveness of DR treatment was derived from a UK study of the proportions of severe, moderate and mild DR prevented by control of blood pressure to avoid diabetic complications (Turner et al, 1998).
The effectiveness of glaucoma treatment was derived from a US study (Kass et al, 2002) and a Swedish study (Heijl et al, 2002), which both assessed the effectiveness of medication interventions for primary open angle glaucoma.

Similar effectiveness was reported for surgery in later stage glaucoma patients in the US (Van Veldhuisen et al, 2000).

Standard treatment for AMD is currently ranibizumab. The effectiveness of AMD treatment was measured by the proportion of MARINA trial patients who were no longer vision impaired after starting ranibizumab treatment, using the definition of VI in this study (VA > 20/40 as a proxy for VA > 6/12). Rosenfeld et al (2006) reported these proportions to be 40% and 42.1% at 12 months and 24 months of treatment, respectively. Ranibizumab’s effectiveness at two years was assumed to remain constant to the end of year five.
Compliance parameters for each condition were derived from expert opinion from Eye Research Australia ophthalmologists.

The five-year effectiveness parameter was applied to estimate DALYs, since this report assumes that each elderly person with undetected mild VI would seek an eye test and subsequent treatment of their own accord after five years (Taylor et al, 2004). Table 7.3 presents the treatment effectiveness and compliance parameters by eye condition.

Table 7.3: Treatment effectiveness and compliance parameters

	AMD	Cataract	DR	Glaucoma	Other cause
Effectiveness - 1 year	40.0%	99.0%	95.9%	99.7%	50.0%
Effectiveness - 2 year	42.1%	98.0%	91.8%	98.5%	50.0%
Effectiveness - 3 year	42.1%	97.0%	87.8%	97.6%	50.0%
Effectiveness - 5 year	42.1%	95.0%	79.6%	95.2%	50.0%
Effectiveness - lifetime	42.1%	95.0%	53.0%	82.9%	50.0%
Compliance	95.0%	90.0%	80.0%	66.7%	80.0%

Source: Access Economics (2005).

7.1.6 Treatment costs for diagnosed mild VI

The health care cost per person diagnosed with mild VI was calculated from total health care costs in 2010 (Section 6.1). Health care costs of €116.8 million were divided by the total prevalence of VI and blindness in 2010 (224,832 from Section 2.5) to estimate a per person cost of €519.29. This would be an overestimate for the cost of mild VI only and hence is likely to overestimate the cost per DALY averted through screening.

Similar to Access Economics’ UK study (2009), each elderly person with undetected mild VI is assumed to seek an eye test and subsequent treatment within the next five years of their own accord (Taylor et al, 2004). This is because the condition will worsen to a point where vision loss starts affecting daily life.

Therefore, the assumed impact of the intervention is to bring an eye test forward by five years. Treatment costs resulting from the intervention were thus calculated over a five year period, by applying the per person annual health care cost (€519.29) and 4% discount rate commonly used in CEAs of public sector projects in the ROI (Department of Finance, 2010).

Discounted five-year treatment costs incorporated life expectancy (CSO, 2009) using the assumed age of each diagnosed case (the midpoint of each age group) and compliance to treatment by condition. Compliance was incorporated by multiplying the compliance rate by condition (from Table 7.3) by the discounted treatment costs.

Discounted five-year treatment costs by condition are presented in Table 7.4. Overall, the total five-year health care cost of treating mild VI diagnosed through the intervention is estimated to be €812,952 based on the population size in 2010.

Table 7.4: Discounted five-year treatment costs for all diagnosed cases

Age Group	AMD	Cataract	DR	Glaucoma	Other	Total
Male
70-74	€16,859	€2,790	€10,984	€8,939	€53,857	€93,429
75-79	€13,207	€960	€3,556	€6,180	€29,446	€53,349
80-84	€28,391	€1,750	€2,121	€7,828	€30,545	€70,635
85-89	€25,351	€1,176	€1,394	€6,107	€19,753	€53,781
90+	€21,997	€621	€368	€6,641	€14,262	€43,890
Total	€105,804	€7,298	€18,424	€35,696	€147,862	€315,084
Female
70-74	€18,615	€2,776	€9,871	€5,188	€58,300	€94,751
75-79	27,194	1,412	€3,069	€4,594	€32,639	€68,908
80-84	€57,006	€1,588	€1,694	€5,935	€45,028	€111,251
85-89	€65,830	€1,950	€14,617	€6,477	€41,822	€117,696
90+	€55,252	€2,439	€1,265	€6,014	€40,293	€105,264
Total	€329,702	€17,463	€35,939	€63,903	€365,944	€812,952

Source: Deloitte Access Economics calculations using Access Economics (2005), Department of Finance (2010), Taylor et al (2004) and health care cost estimate from Section 3.5.

7.1.7 DWL

As additional tax revenue would need to be raised to fund additional sight tests, the campaign and treatment, there will be an associated DWL to the economy. Using a MCPF of 57 cents for every euro raised from taxation (Kleven and Kreiner, 2003, see Section 4.3), the DWL was estimated to be €486,204. This assumes that 80.7% of treatment costs are government funded (OECD, 2009), and that all eye tests are government funded as the targeted group solely consists of people aged over 70 years (NCBI, 2010).

7.1.8 DALYs averted from successfully treated cases

There is a personal health cost attached to those who fail to take a regular eye test as their health will deteriorate without detection and treatment. This reduction in health is measured using DALYs. The 0.02 disability weight for mild sight loss (Stouthard et al, 1997) was multiplied by the number of undetected mild VI cases diagnosed through screening tests in 2010 to estimate the gain in DALYs specifically due to the screening program.

This annual number of DALYs was then applied over a five-year period and multiplied by five-year effectiveness parameters (from Table 7.3) to estimate total DALYs averted as a result of the intervention each year. A discount rate of 4% was applied to DALYs (Department of Finance, 2010).

It is assumed that after five years, a person with undetected mild VI would seek any eye test and treatment on their own accord (Taylor et al, 2004). Thus DALYs avoided through screening only apply for the first five years after detection by screening.

Overall, it is estimated that 84 DALYs would be averted as a result of the intervention.

7.1.9 Cost effectiveness results

Societal costs of the intervention including DWL are estimated to total nearly €1.5 million. Excluding DWL, total costs of the intervention would be just over €1.0 million (health care perspective).

The total costs associated with the older population screening intervention described above are summarised in Table 7.5.

Table 7.5: Total costs associated with intervention

Eye screening tests: €165,092
Public health education campaign: €31,844
Discounted treatment costs: €812,952
DWL: €486,204
Total costs (healthcare perspective): €1,009,888
Total costs (societal perspective): €1,496,092

Source: Deloitte Access Economics estimates.

Under a societal perspective, the hypothetical screening program for the older population is associated with a cost of €17,738 per DALY averted. Under a health care perspective, cost effectiveness is estimated to be €11,974 per DALY averted. The intervention is therefore highly cost effective under both perspectives, using WHO thresholds.