Eligibility rates and representativeness of the General Medical Services scheme population in Ireland 2016-2021: A methodological report

Background In Ireland, the means tested General Medical Services (GMS) scheme provides access to a range of healthcare services at no or low cost to approximately one third of the population. Individuals eligible for the GMS scheme are often a focus of research, as a population that account for a large proportion of healthcare services use. The aim of this study is to describe the eligibility rates and representativeness of the GMS scheme population over time, with respect to age group, sex, and geographical area in Ireland. Methods Population data was obtained from the Central Statistics Office (CSO), using 2016 Census figures and projected population figures for 2017-2021. GMS eligibility figures for 2016-2021 were obtained from the HSE Primary Care Reimbursement Service (PCRS). GMS eligibility rates and relative rates of eligibility were calculated for 2016-2021 by age group and sex. Additionally, 2016 eligibility rates were calculated by geographical area. Results The crude eligibility rate decreased from 36.4% in 2016 to 31.2% in 2020, with a slight increase to 31.6% in 2021. In the 75+ years age group, 78.2% of the total population were eligible for the GMS scheme in 2021. The age group with the lowest rate of eligible individuals was the 25–34 age group, with 19.5% eligible in 2021. The eligibility rate was higher among females compared to males throughout the study period. The highest eligibility rate was seen in Donegal, with a crude rate of 52.8%. Dublin had the lowest rate, with a crude rate of 29.3%. Conclusions GMS eligibility varies greatly depending on age, sex, and geographical area, and decreased between 2016 and 2021. This study uses the most up-to-date data available to provide age group, sex and area-based figures for GMS eligibility which may inform planning and conduct of research focusing on GMS-eligible individuals.


Amendments from Version 1
Based on reviewer comments, this version has been updated with further detail on medications entitlement of private patients (Introduction), the CSO population forecasting approach, our rationale for geographic area groupings, and direct standardisation approaches (Methods), information on changing income thresholds for the GMS scheme, and how this analysis may inform future research (Discussion), updating figure legends and titles, and minor typographical errors.

Introduction
Ireland currently has a mixed public-private system of healthcare.Entitlement to publicly-funded healthcare services is largely via the General Medical Services (GMS or "medical card") scheme 1 .This scheme provides access to a range of healthcare services at no or low cost to approximately one third of the population 2 .This includes unlimited visits to general practitioners (GP) and other community services, emergency department visits, and hospital inpatient care, free to patients at the point of access.It also includes dispensing of a wide range of prescribed medications which are reimbursed to pharmacies by the Irish health service, where a small prescription charge applies to patients per item dispensed (currently €1.50 or €1 for those aged 70 years and over), with a monthly household cap (€15 or €10 for over 70s).
Entitlement to services under the GMS scheme is mainly based on income, with lower income thresholds for older adults resulting in higher eligibility in this group.The income thresholds have varied over time, based on government budgetary decisions.In a small proportion of cases, eligibility is granted on a discretionary basis if there is evidence of high medical expenses.In 2021, approximately 11% of all medical cards were discretionary 3 .Individuals eligible for the GMS scheme are often a focus of research, as a population that account for a large proportion of healthcare services use 4 .In addition, the management of the GMS scheme results in the generation of administrative data, which represent an important resource for health research 5 , and has been linked with other research data to facilitate pharmacoepidemiology studies [6][7][8] .Those without GMS eligibility can avail of the Drugs Payment Scheme, which caps household prescribed medication spend per month, however this yields inconsistent claims data only covering individuals in months where their household prescribed medication spend exceeds this cap.Therefore, for individuals in the privately funded cohort, there is limited information available on healthcare and medication use.An important consideration in any research focussing on GMS scheme eligible individuals as the study population is their representativeness and the potential for selection bias.Therefore, the aim of this study is to describe the eligibility rates and representativeness of the GMS scheme population over time, with respect to age group, sex, and geographical area in Ireland.

Methods
This is a repeated cross-sectional study.

Data sources
Population data for 2016-2021 were obtained from the Central Statistics Office (CSO) via Ireland's Open Data Portal 9 .The census is a legally mandatory count of the population and is completed by each household every five years.For 2016, census data from April of the same year was used, as this was when the census was last recorded.For 2017-2021, estimated population figures for April each year based on Census 2016 projections were obtained.The CSO produce these by trending forwards Census data, based on the number of males and females in each region by single year of age, each person was aged by one year, births and number of immigrants were added, and deaths and number of emigrants were subtracted).Data included single year of age and sex for the estimated years, while Census 2016 data also included information at county level.The CSO does not provide population estimates by county, age, and sex for non-Census years.Age groups were created to match those reported for the GMS eligibility figures (0-4, 5- For April 2016, information at HSE Local Health Office (LHO) level (i.e.geographical areas for the administration of healthcare entitlements) was also obtained.As LHOs do not correspond directly to counties (i.e.areas for reporting GMS eligibility may incorporate (parts of) multiple counties which is the basis for population numbers), broader areas combining counties and/or LHOs were created to provide equivalence for analysis purposes (see Table 1).

Analysis
The characteristics of the GMS population per year were summarised.Crude eligibility rates were calculated for 2016-2021 by age group and sex, and demographic group (children, 0-15 years; adults, 16-64 years; and older adults, 65+ years), along with the relative rate of eligibility (how much more or less likely a member of a particular age or sex group was to be eligible compared to the general population).The crude eligibility rate per year was calculated, and the yearly rates were also directly standardised to the 2016 population based on age group and sex (i.e.eligibility rates in each age-sex strata were weighted according to the age-sex distribution in 2016 to estimate the eligibility rate had the age-sex structure of population remained the same as 2016).
For the area-level analysis, crude eligibility rates were calculated using Census 2016 data and April 2016 eligibility data and directly standardised to national population based on age group and sex.Standardised eligibility rates were plotted on    4 shows a map of the adjusted eligibility rates across areas.

Discussion
The decreasing trend in eligibility rate from 2017 is consistent with previous analysis, which identified a rise from approximately 30% to over 40% from 2008 to 2013 (during the economic crisis affecting Ireland), followed by a decrease in 2014 and 2015 13 .The income threshold for GMS eligibility remained the same from 2013 to November 2020, suggesting the declining eligibility rate may have been due to improved financial circumstances of individuals.The maximum income threshold was increased in November 2020, which may explain the slight increase in eligibility rate from 2020 to 2021, and this could also be partially attributable to the adverse    financial impact of the coronavirus disease 2019 (COVID-19) pandemic 14 .
As expected given that eligibility is often based on income, the areas we identified as having the highest rates are somewhat similar to those with higher levels of deprivation 15 .However, because of the grouping of areas in our analysis to allow for comparable population and eligibility figures, many of the highest deprivation areas were grouped with others, precluding direct comparison.
Considering representativeness of health data drawn from routine sources (as non-random samples) is important for assessing the external validity of research.A recently published  analysis of the OpenSAFELY data, derived from GP records in England and used extensively since the onset of the COVID-19 pandemic to generate evidence, suggests some geographic variation but otherwise good population representativeness 16 .Even for prospectively collected research data, such as the UK BioBank, consideration of representativeness and the potential for healthy user bias is important 17 .As well as undermining generalisability, selection bias may also impact internal validity, where sampling into the study is affected by both the exposure and outcome of interest, and thus the exposure-outcome association may be biased 18 .As the GMS scheme is means tested, older people and individuals from lower socioeconomic backgrounds are overrepresented, which may have implications for research using this data and the conclusions drawn from it.The analysis presented may inform future research, for example for sample size estimates if including only GMS-eligible patients or extrapolating findings from analysis of GMS-eligible patients to the wider population.The overrepresentation of older adults has important policy and fiscal implications, particularly as Ireland is projected to see significant growth in the 65 years and over age group in the coming years.Financing to support medicines expenditure in this growing older cohort is an important consideration for health care planning.

Limitations
For the year-by-year analysis, the population figures for 2017-2021 are estimated based on projections from the census, which may not fully capture true population changes.There were several areas where I would have liked more description and/or discussion.I would have appreciated more information on the basis of (or assumptions applied for) the projections for the "out years."(I was not sure what the authors meant by "rated were directly standardized" to the 2016 population).I also could have used some more information on the possible economic or other social forces that could have been driving the projected changes in eligibility and geographic differences.Also, I am not sure that the readers were given a reason why they were required to group several counties.I also could have used a brief discussion of how specifically these data could be used in the "planning and conduct of research focusing on GMS-eligible individuals." In general, I found the tables followable and useful.I struggled some with the figures.It took me a while see and understand the white/black bar distinction.
I agree with the feedback provided by the first reviewer.I appreciated the opportunity to review the manuscript.I had not seen previously a system whereby a review is sought after publication.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Yes

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate?I cannot comment.A qualified statistician is required.
Are all the source data underlying the results available to ensure full reproducibility?Yes

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Health services research, implementation science I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Introduction:
Final paragraph; include sentence to explain that there's very limited information on healthcare use, dispensed claims, etc. in the privately funded cohort.

1.
The authors could include a brief description of the Drugs Payment Scheme (DPS) as all persons not eligible for GMS are eligible for DPS.This could sit in the discussion, either.

2.
Results: 'The GMS population decreased from For the population pyramids, the grey legend for cohort (GMS or Population) isn't intuitive.I think this could be removed/changed, as the Figure description is more informative.

4.
Discussion: Authors could include a brief discussion on the practical implications of the difference in eligibility rates across age groups -the aging population and fiscal implications going forward, etc.However, this is just a suggestion -not a requirement. 1.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Yes

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate?Yes Are all the source data underlying the results available to ensure full reproducibility?Yes

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Drug utilisation studies using PCRS pharmacy claims data.Health technology management.
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Figure 1 .
Figure 1.Percentage of males and females in each age group within the General Medical Services (GMS) scheme (light-coloured, black-bordered bars) overlaid with percentage in the full population (bold-coloured, white-bordered bars) in 2016.

Figure 2 .
Figure 2. Percentage of males and females in each age group within the General Medical Services (GMS) scheme (light-coloured, black-bordered bars) overlaid with percentage in the full population (bold-coloured, white-bordered bars) for 2016-2021.

Figure 3 .
Figure 3. Percentage of males and females in each age group within the GMS scheme (light-coloured, black-bordered bars) overlaid with percentage in the full population (bold-coloured, white-bordered bars) for 2016 across geographic areas.

Figure 4 .
Figure 4. Eligibility rate for the General Medical Services (GMS) scheme by geographical area for 2016, standardised to the national population by age group and sex.

Table 1 . Local health offices (LHOs), Central Statistics Office (CSO) areas, and areas used for analysis. Local Health Office (PCRS) County and city (CSO) New area
11, by area.Analyses were conducted using Stata version 1710, and figures were generated in RStudio using the tmap and ggplot2 packages11,12 aResultsCharacteristics of the GMS population over time are included inTable 2. The GMS population decreased from 1,735,524 in 2016 to 1,554,759 in 2020, and increased to 1,581,294 in 2021.The 75+ years age group was the largest group throughout the study period, ranging from 13.0% in 2016 to 15.8% in 2021.Conversely, 0-4 years age group was consistently the smallest, ranging from 5.8% in 2016 to 4.4% in 2021.Females made up the majority of the GMS population throughout the study period, increasing from 53.1% in 2016 to 54.1% in 2021.

Table 3
shows the eligibility rates and relative rates by age group and sex.In the 75+ years age group, 85.2% of the total population were eligible for the GMS scheme in 2016, decreasing to 78.2% in 2021.The age group with the lowest rate of eligible individuals was the 25-34 age group, with 27.0% eligible in 2016 and 19.5% in 2021.The eligibility rate was higher among females compared to males throughout the study period, with 38.2% eligible in 2016 and 33.Crude and adjusted eligibility rates over time are included in Table 4.The crude rate decreased from 36.4% in 2016 to 31.2% in 2020, with a slight increase to 31.6% in 2021.After directly standardising the rate to the 2016 population based on age group and sex, a similar pattern was observed

Table 5
shows the crude eligibility rate by area for 2016, and rates directly standardised to the national population based on age group and sex.Population pyramids by area are shown in Figure3.The highest rate was seen in Donegal, with a crude rate of 52.8% and an adjusted rate of 52.0% (95% CI 51.7% to 52.2%).Dublin had the lowest rate, with a crude rate of 29.3% and an adjusted rate of 30.1% (95% CI 30.0% to 30.2%). Figure

Table 4 . General Medical Services (GMS) scheme crude eligibility rate over time, and rate directly standardised to 2017 population based on age group and sex. Total estimated population Total eligible Crude rate Standardised rate (95% CI)
We also provide the statistical code to import open data (where available) and conduct analysis, along with extracted data from the PCRS portal.Provision of open, interoperable PCRS eligibility data per month via Ireland's Open Data Portal would enhance the usability of this data for research and wider purposes.
1,735,534 in 2016 to 1,554,759 in 2020, and increased to 1,581,294 in 2021.' Double check figures in Table 2 -states the 2016 as 1,735,524.