Diabetic retinopathy screening in Australian general practice: persistently low uptake despite national guidelines and equity initiatives

Abstract

Purpose

To examine the geographic and demographic variability in diabetic retinopathy (DR) screening uptake in Australian general practice (GP) and assess factors influencing screening disparities.

Methods

A retrospective, population-based observational study was conducted using Medical Benefits Schedule (MBS) claims data for DR screening using non-mydriatic fundus cameras, under items 12,325 and 12,326 from 2016 to 2024 across Australia. MBS data was obtained from Services Australia and the demographic and socioeconomic indicators from the Australian Bureau of Statistics (ABS). Screening uptake was stratified by state, age, gender, and socioeconomic indicators. The association between screening uptake and sociodemographic factors was assessed.

Results

A total of 15,817 DR screenings were conducted under the two MBS items, representing only 1.2% of Australia’s estimated diabetic population, highlighting a substantial unmet need. New South Wales and Western Australia had the highest total screening, while Tasmania recorded the lowest uptake. Screening uptake was highest in the < 25 age group (due to statistical inflation from low diabetes prevalence) and lowest in those aged ≥ 75 years. he COVID-19 pandemic contributed to a significant decline in screening uptake during 2020–2022, with partial recovery in 2023–2024. DR screening uptake showed no significant association with socio-demographic factors.

Conclusion

Despite national guidelines, GP-led DR screening uptake remains low, with substantial geographic, demographic and temporal disparities in Australian general practice settings. Addressing systemic barriers, strengthening GP-led screening, and implementing targeted interventions are crucial for improving DR screening uptake in Australia. The inability to calculate per-capita screening among Aboriginal and Torres Strait Islander populations due to data limitations restricts equity analysis; future studies should address this gap.

Introduction

Diabetic retinopathy (DR) is a leading cause of vision impairment and blindness among working-age adults worldwide [1, 2]. With diabetes prevalence projected to increase substantially in Australia due to an ageing population [3], the associated health and economic burden is also expected to rise [4]. According to Australia’s Eye Health Survey (2016), the prevalence of any DR among adults with self-reported diabetes, aged ≥ 40 years, was 39.4%, with vision-threatening DR (VTDR) at 9.5% [5]. Without timely intervention, over half of individuals with proliferative DR may develop blindness within five years, underscoring the critical importance of early detection [6].

Beyond direct healthcare costs, DR imposes significant indirect economic burdens, including lost productivity, reduced quality of life, and increased dependence on social support systems [7, 8]. DR-related vision impairment disproportionately affects individuals from lower socio-economic backgrounds and rural or remote areas, where access to specialized eye care services is often limited [9]. Given its public health impact, DR requires effective, accessible and sustainable screening strategies.

Despite clinical guidelines recommending routine DR screening for individuals with diabetes, screening uptake in Australia remains suboptimal [5]. Barriers such as limited access to eye care professionals, inadequate integration of screening into primary care, and low patient awareness contribute to low screening rates, particularly in rural and underserved communities [10–12]. Currently, DR is primarily detected through opportunistic screening by optometrists, rather than through systematic programs within primary care. This over-reliance on optometrists may delay diagnosis and reduce opportunities for early intervention. In Australia, most DR screening is currently conducted by optometrists, often outside the Medicare Benefits Schedule (MBS) framework.

Primary care plays a central role in managing chronic diseases, including diabetes, in Australia, with general practitioners (GPs) serving as the first point of contact for most patients [13]. According to the Australian Institute of Health and Welfare (AIHW), in 2022–23, approximately 86% of Australians received at least one Medicare-subsidized GP attendance [14]. The Australian primary care system is primarily funded through Medicare, which reimburses general medical services, including diabetes-related care [15]. To enhance DR screening accessibility, the Australian Government introduced MBS items 12,325 and 12,326 in 2016, allowing GPs to perform DR screening using retinal photography [16]. MBS item 12,325 is designated for Aboriginal and Torres Strait Islander patients, while item 12,326 applies to the general population (Box 1). Additionally, the nationwide “Provision of Eye Health Equipment and Training” program allocated approximately $33 million over five years to equip non-eye care professionals with the necessary tools and training needed for retinal photography. Despite these efforts, the uptake of DR screening in primary care remains low [5].

Description of MBS items for diabetic retinopathy screening

MBS Item	Description
MBS 12,325	GP-led DR screening for Aboriginal and Torres Strait Islander patients with medically diagnosed diabetes, involving visual acuity assessment and non-mydriatic retinal photography
MBS 12,326	GP-led DR screening for patients with medically diagnosed diabetes, involving visual acuity assessment and non-mydriatic retinal photography

MBS Medicare Benefit Schedule

Although retinal photography is a proven and effective screening tool, its implementation in primary care is hindered by multiple challenges, such as the need for trained personnel for high-quality image acquisition and reliance on ophthalmologists or optometrists for interpretation. These factors limit the scalability and sustainability of point-of-care DR screening in general practice. This issue is further underscored by the low number of claims submitted under the designated MBS items, with only 1,236 claims recorded between November 2016 and March 2018, despite the widespread availability of retinal cameras in primary care settings [16].

The persistently low uptake of GP-led DR screening suggests the presence of underlying systemic and operational barriers. However, no recent systematic study has comprehensively evaluated the geographic and demographic distribution of DR screening uptake across Australia. Understanding these screening patterns is essential for informing targeted policy interventions and optimizing resource allocation. In this context, this study aimed to analyze the geographic and demographic distribution of DR screening coverage in Australian primary care using data from the Australian Bureau of Statistics (https://www.abs.gov.au/) and Services Australia (https://www.servicesaustralia.gov.au/).

Methods

We conducted a retrospective, population-based observational study to analyze the uptake of diabetic retinopathy screening services in Australia over an eight-year period (July 2016 to June 2024) using MBS claims data. Data on service utilization for two MBS items related to GP-led DR screening were obtained from Services Australia [16]. These data reflect service claims rather than confirmed clinical outcomes; thus, the results should be interpreted as an assessment of healthcare utilization, not diagnostic effectiveness. According to the Medicare Benefits Schedule guidelines for items 12,325 and 12,326, DR screening should include visual acuity assessment and retinal photography using a non-mydriatic fundus camera, by suitably trained general practitioners. When diabetic retinopathy is detected, the patient should be referred to an optometrist or ophthalmologist for further evaluation and management, in line with the National Health and Medical Research Council (NHMRC) guidelines for diabetic retinopathy management. Similarly, if the images are of insufficient quality for interpretation, referral is also recommended. These clinical pathways ensure that MBS-claimed DR screenings align with nationally endorsed best-practice protocols [17]. Demographic and socioeconomic indicators were sourced from the Australian Bureau of Statistics (ABS), including the estimated diabetes prevalence, socioeconomic index and education and occupation index [18]. This study used publicly available, de-identified data from Services Australia and ABS and hence received ethical approval through a non-Human Research Ethics Committee (non-HREC) pathway from the Centre for Eye Research Australia, University of Melbourne.

Definition of DR screening uptake

The DR screening uptake was defined as the number of claims of MBS Items 12,325 and 12,326 submitted by general practitioners for DR screening, divided by the estimated diabetic population in each state, derived from the 2021 Australian Census. Given the relatively low uptake of DR screening services, in a few instances, the uptake has been scaled per 100,000 individuals with diabetes to facilitate meaningful comparisons across different population subgroups and to improve interpretability. The denominator was derived from state-wise diabetes prevalence data based on the 2021 Australian Census. The state-wise population and the corresponding diabetes prevalence are presented in Table 1.

Table 1.

Number of people with diabetes in Australia by state

	Total population	Prevalence of diabetes (%)	Number of people with diabetes
New South Wales	8,072,163	5.1	411,680
Victoria	6,503,491	5.0	325,175
Queensland	5,156,138	5.1	262,963
South Australia	1,781,516	7.2	128,269
Western Australia	2,660,026	5.2	138,321
Tasmania	557,571	6.1	34,012
Northern Territory	232,605	5.7	13,258
Australian Capital Territory	454,499	3.4	15,453

Socioeconomic indicators

The Australian Bureau of Statistics provides a set of indicators known as Socioeconomic Indexes for areas (SEIFA), which classify relative levels of socioeconomic advantage and disadvantage across different geographic regions, including states and territories [18]. This study used two SEIFA indicators: ‘Index of Relative Socioeconomic Disadvantage’ and ‘Index of Education and Occupation’. A lower score indicates that an area is relatively disadvantaged compared to an area with a higher score.

Statistical analysis

This study analyzed service utilization based on claims data submitted under MBS items 12,325 and 12,326, and did not include individual clinical outcomes or diagnostic confirmation of diabetic retinopathy. Descriptive statistics were used to analyze service uptake trends across states, gender and age groups from 2016 to 17 to 2023-24. State-level uptake was analyzed and mapped using geo-spatial visualization to illustrate disparities in DR screening coverage across Australia. To assess whether gender differences in DR screening uptake varied over time, a Poisson regression model was used to analyze the interaction between year and gender. A multivariable regression model was used to assess the correlation between DR screening uptake (dependent variable) among the diabetic population and three independent variables: socioeconomic index, education occupation index and diabetes prevalence. Median scores for each state were derived using suburb- and locality-level data, allowing for geographic comparisons. A log transformation was applied to the dependent variable to address the skewness in the data and to improve the normality of the distribution. Results are reported as β coefficients with corresponding standard errors, and statistical significance was set at p < 0.05. All statistical analyses were conducted using STATA version 17.0 (StataCorp, College Station, TX, USA) and the spatial map was generated using GeoDa 1.20 (Chicago, IL, USA).

Results

A total of 15,817 DR screenings were conducted under MBS items 12,325 and 12,326 across Australia during the study period, representing only 1.2% of the country’s estimated diabetic population. Of these, 50.8% (n = 8,040) were billed under MBS 12,325 (for Aboriginal and Torres Strait Islander patients) and 49.2% (n = 7,777) were billed under MBS 12,326 (for the general population) (Table 2). Screening was nearly evenly distributed by gender, with 50.9% (n = 8,057) in males and 49.1% (n = 7,760) in females. Uptake was highest among individuals aged 45–64 years (47.2%; n = 7,465) and lowest in those aged < 25 years (2.4%; n = 374), followed by ≥ 75 years (11.3%; n = 1,794).

Table 2.

Age and gender distribution of people who underwent DR screening

Characteristics	People with diabetes screened n (%)
Total screenings = 15,817
MBS number
12,326 12,325	7777 (49.2) 8040 (50.8)
Gender
Male Female	8057 (50.9) 7760 (49.1)
Age group
< 25 years 25–44 years 45–64 years 65–74 years ≥75 years	374 (2.4) 2776 (17.6) 7465 (47.2) 3408 (21.5) 1794 (11.3)

Geographic distribution of DR screening

The state-wise distribution of DR screening is shown in Fig. 1. New South Wales (NSW) and Western Australia (WA) recorded the highest total screenings, followed by Victoria (VIC) and Queensland (QLD). In contrast, screening uptake was lowest in Tasmania (TAS), the Northern Territory (NT) and the Australian Capital Territory (ACT). WA had a greater proportion of 12,325 claims, consistent with its higher indigenous population, while Victoria and New South Wales had more MBS 12,326 claims.

Fig. 1.

Geographic distribution of DR screening across Australian states, showing total number of screenings and proportion by MBS item (12325 vs. 12326)

Figure 1 also presents a line graph on the secondary axis, depicting the diabetic population estimates. While NSW and VIC had high absolute screening numbers, their screening uptake relative to the diabetic population was not the highest. Notably, NT exhibited a higher relative screening uptake, while SA and TAS recorded lower rates, suggesting gaps in screening access and engagement.

Temporal trends in DR screening

Screening increased steadily from 2016 to 2019, peaking in 2019–2020, followed by a marked decline in 2020–2022, and partial recovery in 2023–2024 (Fig. 2). NSW, VIC and WA consistently recorded the highest total screenings. Trends in NT showed distinct fluctuations, possibly reflecting variations in screening program implementation over time. The post-2019 decline was more pronounced in certain states, likely due to Covid-19-related disruptions.

Fig. 2.

Temporal trends in DR screening uptake (2016–2025) by state, showing variation in uptake patterns across years and regions

Gender disparity in DR screening uptake

Annual screening uptake by gender shows that females consistently had higher uptake under item 12325, while males had relatively higher screening uptake under item 12326 (Table 3). The highest overall uptake was observed in 2019-20 (238 per 100,000 females vs. 210 per 100,000 males). Poisson regression revealed that females had a higher screening uptake compared to males (IRR = 1.24 (95% CI: 1.06–1.45, p = 0.007), but this varied by year. In 2021–2022 and 2022–2023, female uptake was 19% (IRR = 0.81, 95% CI: 0.67–0.97, p = 0.025) and 29% (IRR = 0.71, 95% CI: 0.59–0.85, p < 0.001 lower than males, respectively.

Table 3.

Overall Temporal trend in DR screening by gender and MBS item per 100,000 people with diabetes

	MBS no. 12,326		MBS no. 12,325		Total		IRR (95% CI)	p-value
	Male	Female	Male	Female	Male	Female
2016 to 2017	20	20	20	31	40	51	Ref*
2017 to 2018	57	49	67	99	124	148	0.95 (0.80–1.15)	0.620
2018 to 2019	76	66	93	143	169	209	0.99 (0.83–1.18)	0.910
2019 to 2020	132	134	78	104	210	238	0.91 (0.77–1.09)	0.310
2020 to 2021	93	88	86	113	179	201	0.91 (0.76–1.08)	0.267
2021 to 2022	75	68	52	60	127	128	0.81 (0.67–0.97)	0.025
2022 to 2023	105	73	47	62	152	135	0.71 (0.59–0.85)	< 0.001
2023 to 2024	44	41	62	81	106	122	0.93 (0.77–1.12)	0.442

*For interaction term, uptake of males in the year 2016–2017 was taken as reference

Age-stratified DR screening uptake

Age-specific screening uptake per 100,000 individuals with diabetes is presented in Table 4. Due to the lack of age-specific diabetes prevalence data, TAS, NT and ACT were excluded from this analysis. The < 25 age group had the highest relative screening uptake, increasing from 160 per 100,000 in 2016-17 to a peak of 1,118 in 2018–2019, before declining to 718 in 2023–2024. This exceptionally high screening uptake is largely a statistical artifact resulting from the very low diabetes prevalence in this age group, which inflates screening uptake when expressed per 100,000 individuals. Uptake was consistently lowest in the ≥ 75 years age group. Most age groups showed a peak around 2019-20, with a subsequent decline and partial recovery by 2023-24.

Table 4.

Temporal trend for DR screening by MBS item across age groups per 100,000 people with diabetes*

	2016 to 2017	2017 to 2018	2018 to 2019	2019 to 2020	2020 to 2021	2021 to 2022	2022 to 2023	2023 to 2024
< 25 years
MBS 12,326	40	220	459	499	180	279	439	339
MBS 12,325	120	459	659	579	758	279	299	379
Total	160	679	1118	1078	938	558	738	718
25–44 years
MBS 12,326	16	46	84	182	137	119	116	52
MBS 12,325	42	140	183	133	149	82	90	99
Total	58	186	267	315	286	201	206	151
45–64 years
MBS 12,326	23	56	85	151	102	81	97	44
MBS 12,325	33	93	158	122	117	69	69	89
Total	56	149	243	273	219	150	166	133
65–74 years
MBS 12,326	21	67	67	139	96	74	99	50
MBS 12,325	12	53	64	51	60	40	37	40
Total	33	120	131	190	156	114	136	90
≥ 75 years
MBS 12,326	18	39	40	73	41	33	54	26
MBS 12,325	3	21	21	19	17	14	10	11
Total	21	60	61	92	58	47	64	37

*Three states (TAS, NT & ACT) were excluded as age-group-wise diabetes prevalence data were not available for those states

Socio-demographic factors and screening uptake

Multivariable regression (Table 5) showed no significant association between DR screening uptake and the independent factors: socioeconomic index, education/occupation index, and diabetes prevalence. All variables had minimal effect sizes and non-significant p-values, further underscores the negligible influence of socio-demographic factors. These findings suggest that structural or operational systems-level barriers, rather than individual-level socio-demographic predictors, likely account for the low screening uptake.

Table 5.

Association of sociodemographic factors with DR screening uptake

Predictors	β coefficient	Standard error	p-value
Socio-economic index	0.004	0.03	0.906
Education and occupation	−0.012	0.03	0.630
DM prevalence	−0.288	1.43	0.851

Discussion

This study highlights significant geographic, demographic, and temporal disparities in DR screening uptake across Australia, despite national guidelines recommending routine screening of individuals with diabetes [5, 6]. The National Health and Medical Research Council (NHMRC) guidelines (2008) recommend annual eye examinations for First Nations people with diabetes, and at least every 2 years for non-Indigenous Australians with diabetes [17]. International bodies such as the International Council of Ophthalmology (ICO; https://icoph.org/eye-care-delivery/diabetic-eye-care/) and American Diabetes Association (ADA; https://diabetes.org/health-wellness/eye-health) also recommend annual or biennial screening based on risk. Screening peaked in 2019–2020, followed by a decline in 2020–2022, coinciding with the COVID-19 pandemic, and a partial recovery in 2023–2024. However, the extremely low overall DR screening uptake in GP clinics (only 1.2% of the estimated diabetic population) is concerning. DR affects an estimated 28% to 32% of people with diabetes in Australia, and vision-threatening DR (VTDR) affects approximately 4.5% [3, 5]. This disparity suggests a large proportion of individuals with diabetes remain undiagnosed and at risk of progressive vision loss due to delayed detection and intervention. Importantly, as the study was based on claims data, it measures service use rather than confirmed diagnosis or outcomes. This distinction is vital for interpreting screening coverage vs. clinical effectiveness.

A key finding is the difference in screening uptake between MBS items 12,325 and 12,326. Although Aboriginal and Torres Strait Islander people constitute a small proportion of the diabetic population [19], MBS item 12,325 accounted for 50.8% of total DR screening, likely due to targeted outreach programs. This aligns with the AIHW’s annual report on eye health measures for Aboriginal and Torres Strait Islander people, which noted that approximately 47% of First Nations individuals who had a diabetes test had also undergone an eye examination in 2021-22 [20]. In contrast, MBS item 12,326, which applies to the general Australian population, showed lower uptake despite serving a larger diabetic population. This underscores the need for similar targeted interventions in non-indigenous populations to address barriers to screening in primary care settings.

The low uptake of DR screening in primary care indicates underutilization of MBS-funded GP-led DR screening items introduced in 2016 [16]. Barriers such as low GP referral rates, a shortage of trained personnel and insufficient screening incentives likely contribute to this underuse. Even after training, some primary care clinicians did not refer patients requiring specialist care, likely due to image quality or interpretation [21, 22]. Additionally, time constraints within general practice might hinder DR screening [11]. Given that most individuals with diabetes visit a GP at least once a year [14], integrating systematic DR screening into routine consultations could significantly improve uptake. Automated reminders, incentivized reimbursement models and task-sharing with trained allied health professionals could enhance screening uptake. Emerging artificial intelligence-assisted screening also offers a promising approach to improving accessibility [21].

Geographic variability in DR screening is consistent with disparities in healthcare access, infrastructure, and policy implementation across Australian states [23]. NSW and VIC recorded the highest total screening, though their relative screening uptakes were not the highest. Conversely, NT exhibited a higher relative screening uptake despite its smaller diabetic population, possibly due to targeted screening initiatives and community-based outreach programs. In contrast, TAS and SA recorded lower screening uptake, suggesting potential gaps in screening availability and patient engagement. These findings emphasize the need for state-specific interventions to address regional barriers to DR screening.

Gender-based differences in screening uptake were also observed. Females had consistently higher screening uptake under MBS item 12,325, possibly due to greater engagement in preventive healthcare among Aboriginal and Torres Strait Islander women [24]. Males had higher screening uptake under MBS item 12,326, potentially due to differences in primary care utilization and employment-related health checkups [25]. Further research is needed to identify gender-specific barriers and facilitators to ensure equitable access to DR screening.

The decline in DR screening during 2020–2022 aligns with global reports of COVID-19-related health care disruptions, which significantly affected preventive screening services [26, 27]. Lockdowns, travel restrictions, and the diversion of healthcare resources likely contributed to reduced access to DR screening during this period [28]. Similar declines were reported in other preventive services, such as breast and cervical cancer screenings, due to reduced GP visits, telehealth use, and public health restrictions. These disruptions likely affected DR screening access and prioritization [29, 30]. The uneven post-pandemic recovery across states suggests structural challenges in restoring routine screening services to pre-pandemic levels. Outreach programs, expanded DR screening in primary care, and use of teleophthalmology, may help mitigate these effects [23, 31, 32].

The exceptionally high screening uptake in the < 25 years age group is primarily due to low diabetes prevalence (< 6%) in this age cohort [18], leading to inflated uptake when expressed per 100,000 individuals with diabetes. This does not necessarily indicate higher engagement in DR screening among younger individuals. On the contrary, older adults (≥ 75 years) consistently had the lowest screening uptake, despite being at higher risk for DR progression and vision-threatening complications. This may be attributed to limited mobility, reduced healthcare-seeking behavior, or lack of awareness about DR screening recommendations [10, 33]. Targeted interventions, such as reminder-based systems, mobile screening programs, and integrating DR screening with routine diabetes care, could help improve uptake in this high-risk group.

Contrary to expectations, none of the sociodemographic factors were significantly associated with DR screening uptake. While previous studies suggest that low socioeconomic status and limited education may be barriers to DR screening [10, 12], our findings suggest that other systemic and operational factors, such as the availability of trained healthcare personnel, the integration of screening services in routine primary care and policy-driven initiatives, may have a stronger role in determining screening uptake [34, 35]. These findings highlight the need for system-level improvements, including expanding workforce training, embedding DR screening within routine diabetes care, and improving referral pathways. While postcode-level Socio-Economic Indexes for Areas (SEIFA) data could provide greater granularity, DR screening data from Services Australia were only available at the state level. Our approach is aligned with previous national studies using aggregated data to assess health service disparities across Australian jurisdictions [36–38].

To address the persistent underutilization of GP-led screening, future policy efforts could include reimbursement incentives and integration of AI-based decision-support tools into general practices workflows. Autonomous AI-based DR screening tools could be embedded within GP clinics, supported by clear referral protocols [39]. This would allow opportunistic screening without overburdening clinic staff. Financial incentives, embedded within chronic disease care plans may motivate general practices to prioritize DR screening. In addition, team-based models involving nurses or practice assistants in the screening process could enhance capacity and support the scale-up of GP-led screening initiatives. These strategies may help reduce workload barriers and enable timely identification and management of at-risk individuals.

Strengths and limitations

This study provides comprehensive national-level insights into the patterns of GP-led DR screening uptake across Australia. By utilizing longitudinal MBS data over an 8-year period, the study captures temporal trends and programmatic changes since the introduction of GP-led screening items in 2016. The integration of state-wise comparisons and age- and gender-stratified analyses enables a detailed understanding of geographic and demographic disparities. In addition, the use of publicly available national datasets ensures transparency and replicability. The application of both Poisson and multivariable regression models strengthens the analytical rigour by exploring associations across multiple dimensions, including socio-economic factors.

However, certain limitations must be acknowledged. The reliance on MBS claims data may underestimate true screening uptake, as the dataset excludes privately funded screenings or screenings conducted outside the MBS system. DR screening by optometrists - who conduct the majority of primary-level screenings - was not included, as they are typically billed under general optometric services or private health cover. This might underrepresent the true extent of national DR screening coverage and hence, should be considered when interpreting the reported 1.2% uptake rate. The exclusion of TAS, NT and ACT from age-specific analysis due to the unavailability of age-specific diabetes prevalence data from the ABS limits comprehensive state-level comparisons. This may introduce bias by underrepresenting trends in regions with smaller populations, distinct healthcare access models, or higher proportions of indigenous communities.

Moreover, the lack of association between screening uptake and sociodemographic indicators must be interpreted in light of our use of state-level aggregated SEIFA scores. Although this approach was necessitated by data availability, it may mask important intra-state variations and dilute localized effects. Additionally, while MBS item 12,325 (specific to Aboriginal and Torres Strait Islander people) accounted for 50.8% of the total DR screenings, this apparent high uptake may be misleading. The absence of indigenous-specific diabetes prevalence data by state limited our ability to calculate per capita screening rates for this group. Thus, this high proportion does not necessarily reflect better program coverage or equity, and should be interpreted with caution. This limitation highlights a critical gap in national health data infrastructure. Urgent investment in Indigenous-specific diabetes prevalence data is needed to enable equity focused health service planning and performance monitoring. Future research should also incorporate postcode- or individual-level data and draw on multiple data sources to provide a more complete and equity-informed understanding of DR screening uptake and its determinants.

Conclusion

This study highlights significant geographic, demographic and temporal disparities in GP-led DR screening uptake across Australia. Despite national screening guidelines, overall participation remains suboptimal, with notable variations across states, age groups, and gender. that the lack of association with socioeconomic and educational indicators suggest that systemic barriers, such as limited workforce capacity, fragmented service delivery, and underutilized funding mechanisms, play a greater role. Without comprehensive policy reform, GP-led DR screening is unlikely to significantly reduce diabetes-related vision loss. Urgent national investment in workforce support, reimbursement reform, and integration of AI-enabled DR screening into routine chronic disease care is needed. Tailored state-specific strategies will be critical to closing the screening gap and achieving equitable and sustainable improvements in DR screening coverage.

Author contributions

SJ obtained the data, conducted data analysis, wrote the first draft of the manuscript and incorporated feedback from co-authors and reviewers to finalize the manuscript; WH and CJ provided additional inputs for data analysis and reviewed the manuscript draft; ZZ, XS, HT, TR and MH provided supervision and guidance, reviewed the manuscript and provided intellectual inputs. All authors reviewed and approved the final draft of the manuscript.

Funding

This project received grant funding from the Australian Government: the National Critical Research Infrastructure Initiative, Medical Research Future Fund (MRFAI00035) and the NHMRC Investigator Grant (APP1175405). The contents of the published material are solely the responsibility of the Administering Institution, a participating institution or individual authors and do not reflect the views of the NHMRC. This work was supported by the Global STEM Professorship Scheme (P0046113), the Fundamental Research Funds of the State Key Laboratory of Ophthalmology, and Project of Investigation on Health Status of Employees in Financial Industry in Guangzhou, China (Z012014075). The Centre for Eye Research Australia receives Operational Infrastructure Support from the Victorian State Government. SJ is supported by the Melbourne Research Scholarship established by the University of Melbourne. The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data availability

The data that supports the findings of this study were obtained from two publicly accessible websites: Services Australia (https://www.servicesaustralia.gov.au/) and Australia Bureau of Statistics (https://www.abs.gov.au/).

Declarations

Ethical approval

Ethical approval or informed consent from participants was not required as the study was completely based on analysis of publicly available, de-identified data.

Human ethics and consent to participate declarations

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.