Diabetes mellitus has emerged as one of the most important non-communicable diseases worldwide, with a rapidly rising prevalence in low- and middle-income countries. India alone accounts for a substantial proportion of the global diabetic population, and a large share of affected individuals reside in rural and semi-urban regions where access to continuous medical care is limited [1,2]. Among the long-term complications of diabetes, diabetic retinopathy (DR) represents a major cause of visual impairment and blindness in the working-age population, largely because it remains asymptomatic until advanced stages [3].

Diabetic retinopathy is a progressive microvascular disorder of the retina resulting from chronic hyperglycaemia-induced endothelial dysfunction, capillary basement membrane thickening, pericyte loss, and retinal ischemia [4,5]. The global burden of DR is substantial, with estimates suggesting that nearly one-third of people with diabetes develop some form of retinopathy during their lifetime, and a significant proportion progress to vision-threatening stages such as proliferative diabetic retinopathy or diabetic macular edema [6,7]. Importantly, early detection and timely intervention have been shown to markedly reduce the risk of severe vision loss [8].

Despite clear evidence supporting routine retinal screening, coverage remains suboptimal, particularly in rural populations. Conventional screening strategies rely on dilated fundus examination by trained ophthalmologists, which poses several practical challenges in resource-limited settings, including shortage of specialists, need for pharmacological mydriasis, patient discomfort, longer examination times, and logistical constraints [9–11]. These barriers often result in delayed diagnosis, missed follow-up, and preventable progression of disease.

In recent years, non-mydriatic fundus photography has gained attention as a potential alternative screening modality for diabetic retinopathy. This technique enables retinal image acquisition without pupil dilation, using portable or semi-portable digital fundus cameras operated by trained non-ophthalmic personnel [12,13]. The captured images can be graded on-site or transmitted to remote reading centres, making this approach particularly suitable for community-based and teleophthalmology programmes [14].

Several studies have demonstrated that non-mydriatic fundus photography provides acceptable image quality and reliable detection of diabetic retinopathy, especially for identifying early and referable disease [15,16]. Compared with direct ophthalmoscopy, photographic documentation offers the advantage of permanent records, standardised grading, and reduced inter-observer variability [17]. Moreover, patient acceptance tends to be higher due to the non-invasive nature of the procedure and avoidance of mydriatic side effects such as photophobia and blurred vision [18].

Rural populations face a disproportionately higher risk of undiagnosed and untreated diabetic retinopathy due to socioeconomic constraints, lower health literacy, and limited specialist services [19]. Community-based screening using non-mydriatic fundus cameras has been proposed as an effective strategy to bridge this gap by integrating retinal evaluation into primary diabetes care [20]. However, variability in reported image gradability, detection rates, and screening performance across different settings highlights the need for context-specific evidence, particularly from rural Indian populations [21].

Given the increasing burden of diabetes and its ocular complications in rural communities, there is a critical need to evaluate feasible, scalable, and cost-effective screening approaches. Assessing the real-world effectiveness of non-mydriatic fundus photography in rural healthcare settings can provide valuable insights into its role in early detection, referral triage, and prevention of vision-threatening disease [22]. The present study was therefore undertaken to assess the screening effectiveness of non-mydriatic fundus photography for early detection of diabetic retinopathy in a rural population, with emphasis on image adequacy, detection rates, and referral needs.

Study Design

This investigation was conducted as a hospital-based, cross-sectional observational study with the primary objective of assessing the real-world effectiveness of non-mydriatic fundus photography as a screening tool for early diabetic retinopathy. A cross-sectional design was selected to capture the spectrum of retinal changes present in individuals with diabetes attending routine outpatient services, rather than limiting evaluation to high-risk or pre-screened groups. Emphasis was placed on feasibility, image quality, and detection capability under routine clinical conditions typical of government-run healthcare facilities serving rural populations.

Study Settings

The study was carried out at two government-funded tertiary care centres in Telangana, India: the Regional Eye Hospital attached to Kakatiya Medical College, Hanumakonda, and Government Medical College, Narsampet, Warangal. The Regional Eye Hospital functions as a major referral centre for ophthalmic disorders and receives patients from both urban and rural catchment areas. Government Medical College, Narsampet predominantly caters to a rural population and serves as a key access point for individuals with diabetes seeking primary and secondary medical care. Conducting the study across these two settings allowed assessment of screening performance in both specialist-led ophthalmic clinics and general hospital environments where dedicated eye care resources may be limited.

Duration of the Study

The study was conducted over a continuous period of twelve months, from January 2025 to December 2025. This timeframe was chosen to ensure adequate patient recruitment and to account for variations in outpatient attendance patterns over different seasons. The extended duration also allowed sufficient time for operator training, standardisation of imaging procedures, and validation of screening outcomes.

Study Population

The study population consisted of adult patients with a confirmed diagnosis of diabetes mellitus attending the outpatient departments of the participating institutions during the study period. Patients were enrolled consecutively to minimise selection bias and to reflect the true demographic and clinical profile of individuals seeking routine care. Both newly diagnosed and long-standing cases of diabetes were included to capture a wide range of disease duration and potential retinal involvement.

Eligibility Criteria

Patients aged 18 years and above with documented type 1 or type 2 diabetes mellitus were considered eligible for inclusion. Only those who provided written informed consent and were willing to undergo non-mydriatic fundus photography were enrolled. Patients with a history of recent retinal screening within the preceding year were not routinely excluded; however, those already diagnosed with advanced diabetic retinopathy and undergoing active ophthalmic treatment were excluded to avoid overestimation of screening yield. Additional exclusion criteria included ocular media opacities such as dense cataract or corneal scarring that precluded adequate fundus visualisation, as well as systemic or neurological conditions that prevented patient cooperation during imaging.

Sample Size and Sampling Method

A convenience sampling approach was adopted, whereby all eligible and consenting patients presenting during the study period were considered for inclusion. This method was chosen to simulate the operational conditions of large-scale screening programmes implemented within existing healthcare infrastructure. Rather than predetermining a fixed sample size, enrolment was guided by patient flow and feasibility considerations, with emphasis on capturing a representative rural diabetic population.

Ethical Approval and Informed Consent

Ethical clearance for the study was obtained from the Institutional Ethics Committees of both participating medical colleges prior to initiation. All participants were provided with detailed information regarding the purpose of the study, the nature of the screening procedure, and the voluntary nature of participation. Written informed consent was obtained before data collection. Confidentiality of patient information was strictly maintained, and all data were anonymised prior to analysis.

Clinical and Demographic Data Collection

Baseline demographic and clinical data were collected using a predesigned structured proforma. Information recorded included age, sex, place of residence, duration of diabetes, current treatment regimen, and presence of systemic comorbidities such as hypertension. Clinical details were obtained through direct patient interviews and cross-verification with medical records wherever available. This information was used to contextualise screening findings and to explore associations between patient characteristics and retinopathy detection.

Non-Mydriatic Fundus Photography Procedure

Retinal imaging was performed using a non-mydriatic digital fundus camera capable of capturing high-resolution colour fundus photographs without pharmacological pupil dilation. Imaging was carried out in a dimly lit room to facilitate natural pupillary dilation. Trained ophthalmic technicians or optometrists conducted image acquisition following a standardised protocol. For each eye, at least two images were captured, typically centred on the optic disc and macula. In cases where initial images were of suboptimal quality, repeat images were obtained to maximise gradability. The total time required for image acquisition was recorded to assess feasibility in busy outpatient settings.

Assessment of Image Quality

All captured images were first evaluated for technical adequacy. Image quality assessment focused on clarity, focus, field definition, and visibility of key retinal landmarks including the optic disc, macula, and retinal vasculature. Images were categorised as gradable or ungradable based on predefined criteria. Ungradable images were documented separately, and the reasons for poor quality, such as small pupil size or media opacity, were noted to inform feasibility analysis.

Grading of Diabetic Retinopathy

Gradable fundus images were independently reviewed by trained ophthalmologists with experience in diabetic retinopathy screening. Retinopathy was classified according to standard clinical grading systems into no diabetic retinopathy, mild non-proliferative, moderate non-proliferative, severe non-proliferative, or proliferative diabetic retinopathy. The presence of macular involvement or features suggestive of sight-threatening disease was also documented. In cases of disagreement between graders, images were reviewed jointly and a consensus decision was reached.

Referral and Follow-Up Protocol

Patients identified with moderate or more severe diabetic retinopathy, suspected macular involvement, or ungradable images were advised referral to a retina specialist for comprehensive dilated fundus examination and further management. Referral recommendations were communicated clearly to patients, and counselling was provided regarding the importance of timely ophthalmic evaluation. Where possible, referral compliance was documented.

Validation Subset

To assess the diagnostic performance of non-mydriatic fundus photography, a subset of screened participants underwent standard dilated fundus examination by a consultant ophthalmologist. Findings from clinical examination were compared with photographic grading to evaluate agreement in detection of diabetic retinopathy. This subset analysis provided an internal measure of screening accuracy under routine clinical conditions.

Outcome Measures

The primary outcomes of the study included the proportion of gradable retinal images and the prevalence of diabetic retinopathy detected through non-mydriatic screening. Secondary outcomes included detection of previously undiagnosed retinopathy, identification of referable disease, feasibility of image acquisition, and patient acceptance of the screening procedure.

Statistical Analysis

Collected data were entered into a secured database and analysed using standard statistical software. Continuous variables were expressed as mean and standard deviation, while categorical variables were summarised as frequencies and percentages. Diagnostic performance parameters such as sensitivity and specificity were calculated for the validation subset. Associations between clinical variables and presence of diabetic retinopathy were explored where appropriate, with a p-value of less than 0.05 considered statistically significant.

Overall Screening Yield and Study Flow

During the 12-month study period (January–December 2025), a total of 312 individuals with diabetes mellitus were screened using non-mydriatic fundus photography at the two participating centres. Of these, 287 participants (92.0%) yielded at least one gradable retinal image, while 25 participants (8.0%) had images that were ungradable in both eyes due to technical or ocular factors (Figure 1).

Figure 1: Flow diagram showing enrolment, image gradability, and final analysis of screened participants

Demographic and Clinical Profile of the Screened Population

The demographic and clinical characteristics of the study population are summarised in Table 1. The mean age of participants was 54.2 ± 10.8 years, with nearly two-thirds (63.5%) aged 50 years or older. Males constituted a slightly higher proportion (56.4%) than females (43.6%). A substantial majority (68.9%) were from rural backgrounds, consistent with the service population of Government Medical College, Narsampet.

With respect to diabetes duration, 39.7% of participants had diabetes for less than five years, while 25.6% had disease duration exceeding ten years. Hypertension was present in 45.5% of the cohort, reflecting common clustering of cardiovascular risk factors among individuals with diabetes (Table 1).

Table 1: Baseline demographic and clinical characteristics of screened participants (n = 312)

Image Acquisition Performance and Gradability

Non-mydriatic fundus photography was technically feasible in routine outpatient settings at both centres. At the eye level, 574 of 624 eyes (92.0%) produced images of sufficient quality for grading. At the participant level, at least one gradable eye image was obtained in 287 individuals.

The reasons for image non-gradability were analysed in detail. Early to moderate cataract accounted for 52% of ungradable cases, while small pupil size and poor fixation were responsible for the remaining instances. There was no statistically significant difference in image gradability between the two centres (χ² = 1.12, p = 0.29), indicating uniform performance across both ophthalmic and general hospital settings (Figure 2).

Figure 2: Proportion of gradable and ungradable fundus images obtained using non-mydriatic photography

Prevalence of Diabetic Retinopathy

Among the 287 participants with gradable images, 96 individuals (30.8%) were found to have diabetic retinopathy of any grade, while 191 participants (69.2%) had no detectable retinal changes. The distribution of retinopathy severity is presented in Table 2 and visually depicted in Figure 3.

Early stages of disease predominated. Mild non-proliferative diabetic retinopathy (NPDR) was the most frequently observed category (18.6%), followed by moderate NPDR (7.7%). Advanced disease, including severe NPDR or proliferative diabetic retinopathy, was identified in 4.5% of the total screened population.

Table 2: Severity-wise distribution of diabetic retinopathy (n = 312)

Figure 3: severity-wise distribution of diabetic retinopathy among screened participants

Detection of Previously Undiagnosed Retinopathy

A key outcome of the screening programme was the identification of previously unrecognised disease. Of the 96 participants diagnosed with diabetic retinopathy, 71 individuals (22.8% of total screened) reported no prior history of retinal evaluation or diagnosis of diabetic eye disease.

Notably, 74.6% of these newly detected cases were in the mild NPDR stage, highlighting the ability of non-mydriatic screening to identify disease at an early, potentially reversible stage.

Association between Duration of Diabetes and Retinopathy

A strong and statistically significant relationship was observed between duration of diabetes and presence of diabetic retinopathy. Participants with diabetes duration exceeding ten years demonstrated a markedly higher prevalence of retinopathy (48.7%) compared to those with duration less than five years (18.5%). This association was statistically significant (χ² = 18.74, p < 0.001).

Figure 4: Prevalence of diabetic retinopathy according to duration of diabetes

Influence of Age and Hypertension on Retinopathy Prevalence

Age also emerged as a significant determinant. Participants aged 50 years or older had a significantly higher prevalence of diabetic retinopathy compared to those below 50 years (35.9% vs 22.4%; χ² = 6.21, p = 0.013).

Hypertension was associated with a higher prevalence of retinopathy (36.6% vs 25.8%); however, this difference did not reach statistical significance (χ² = 3.41, p = 0.065), suggesting a trend rather than a definitive association in this cohort (Table 3).

Table 3: Association of clinical variables with diabetic retinopathy

Identification of Referable Diabetic Retinopathy

Based on predefined referral criteria, 38 participants (12.2%) were classified as having referable disease. This group included individuals with moderate to severe NPDR, proliferative changes, suspected macular involvement, and ungradable images (Figure 5).

Figure 5: Indications for referral following non-mydriatic fundus photography screening

Validation against Dilated Fundus Examination

In the validation subset of 120 participants, comparison of non-mydriatic fundus photography with dilated fundus examination showed high diagnostic agreement. The screening modality demonstrated a sensitivity of 88.3% and specificity of 91.5% for detection of any diabetic retinopathy. The kappa statistic of 0.79 indicated substantial agreement between the two methods.

Operational Feasibility and Patient Acceptability

The mean time required for image acquisition was 4.2 ± 1.1 minutes per participant, allowing efficient integration into routine outpatient workflows. No adverse events related to the imaging procedure were reported. Informal feedback indicated high patient acceptance, particularly due to absence of pupil dilation and minimal discomfort.

Consolidated Interpretation of Findings

Overall, the data demonstrate that non-mydriatic fundus photography provides high-quality retinal images in the majority of patients, detects a significant burden of previously undiagnosed diabetic retinopathy, and reliably identifies individuals requiring specialist referral. The statistically significant associations with age and duration of diabetes further reinforce the clinical validity of this screening approach in rural healthcare settings.

The present study evaluated the effectiveness of non-mydriatic fundus photography as a screening tool for early detection of diabetic retinopathy in a predominantly rural population attending two government tertiary-care centres. The findings demonstrate that non-mydriatic imaging is feasible under routine clinical conditions, provides a high rate of gradable images, and is capable of identifying a substantial burden of previously undiagnosed diabetic retinopathy. These results are particularly relevant in the Indian rural healthcare context, where systematic retinal screening remains limited despite the rising prevalence of diabetes [1,2].

One of the key observations of this study was the high image gradability rate of 92%, achieved without pharmacological pupil dilation. This finding aligns with earlier reports indicating that modern non-mydriatic cameras can yield clinically useful images in the majority of patients when operated by trained personnel [12,13]. The lack of a significant difference in image gradability between the two study centres further suggests that this technology can be reliably deployed not only in specialised eye hospitals but also in general medical college settings serving rural populations. Cataract-related media opacity emerged as the main reason for ungradable images, a limitation that has been consistently reported in earlier screening studies [16,18]. Importantly, even in such cases, identification of ungradable images itself served as a trigger for referral, thereby preventing missed pathology.

The overall prevalence of diabetic retinopathy detected in this study was 30.8%, which is comparable to figures reported from other Indian and global population-based studies [6,7,21]. The predominance of mild and moderate non-proliferative diabetic retinopathy highlights the silent nature of early disease and reinforces the importance of proactive screening. Similar distributions, with early-stage retinopathy constituting the majority of detected cases, have been described in large screening programmes using photographic methods [15,17]. The relatively lower proportion of advanced disease in the screened cohort may reflect both early detection and referral patterns at tertiary care centres.

A notable and clinically important finding was that nearly three-quarters of individuals with diabetic retinopathy were previously unaware of their retinal status. This observation underscores a persistent gap in routine ophthalmic surveillance among patients with diabetes, particularly in rural areas [19]. Previous studies have emphasised that lack of awareness and limited access to eye care services contribute significantly to delayed diagnosis and preventable visual impairment [9,20]. By identifying early retinopathy in asymptomatic individuals, non-mydriatic fundus photography offers a valuable opportunity for timely intervention and patient education.

The strong association between longer duration of diabetes and presence of diabetic retinopathy observed in this study is consistent with established pathophysiological understanding and epidemiological evidence [4,5,6]. Participants with diabetes duration exceeding ten years showed almost a two-fold higher prevalence of retinopathy compared to those with shorter disease duration, reinforcing duration as one of the most important risk factors for retinal involvement. Similarly, the higher prevalence of retinopathy among individuals aged 50 years and above aligns with earlier reports linking cumulative metabolic exposure and ageing to microvascular damage [3,21]. These findings support risk-stratified screening approaches, where individuals with longer disease duration and older age may benefit from more frequent surveillance.

Although hypertension was associated with a higher prevalence of diabetic retinopathy in this cohort, the difference did not reach statistical significance. This trend is biologically plausible, as hypertension exacerbates retinal microvascular stress, and has been reported as a contributory factor in several studies [6,22]. The lack of statistical significance in the present study may be due to sample size limitations or effective antihypertensive management in a subset of patients. Nevertheless, the observed trend highlights the importance of comprehensive cardiovascular risk control in individuals with diabetes.

Approximately 12% of screened participants were identified as requiring referral for further ophthalmic evaluation, either due to referable retinopathy or ungradable images. This referral yield is comparable to other screening programmes employing non-mydriatic photography [14,16]. From a public health perspective, this selective referral approach is particularly advantageous, as it allows limited specialist resources to be directed towards patients most likely to benefit from detailed evaluation and treatment. The ability to triage patients effectively at the primary screening level is a key strength of photographic screening models.

In the validation subset, non-mydriatic fundus photography demonstrated high sensitivity and specificity for detection of diabetic retinopathy, with substantial agreement compared to dilated fundus examination. These findings are in line with previous studies reporting acceptable diagnostic accuracy of non-mydriatic imaging, particularly for identifying any retinopathy and referable disease [12,15]. While dilated examination remains the gold standard, the observed agreement supports the role of non-mydriatic photography as a reliable first-line screening tool rather than a replacement for comprehensive ophthalmic assessment.

The short image acquisition time and absence of pharmacological dilation contributed to good patient acceptance, a factor that is crucial for sustained participation in screening programmes. Earlier studies have shown that avoidance of mydriasis improves patient comfort and willingness to undergo repeat screening, especially among working-age individuals [18]. In rural settings, where travel and follow-up are often challenging, such operational advantages can significantly enhance screening uptake and continuity of care.

Strengths and Limitations

The strengths of this study include its real-world setting, inclusion of a predominantly rural population, and evaluation across two different types of government healthcare institutions. The use of standardised grading and validation against clinical examination further strengthens the findings. However, certain limitations should be acknowledged. The cross-sectional design precludes assessment of disease progression or long-term outcomes. Additionally, not all participants underwent dilated fundus examination, which may limit definitive conclusions regarding diagnostic accuracy across the entire cohort.

Overall Interpretation

Taken together, the findings of this study support the integration of non-mydriatic fundus photography into routine diabetes care in rural and resource-limited settings. By enabling early detection of diabetic retinopathy, facilitating targeted referral, and improving patient engagement, this screening approach addresses several key barriers identified in conventional models of eye care delivery [10,11,20]. With appropriate training, quality assurance, and referral linkage, non-mydriatic fundus photography has the potential to play a central role in reducing the burden of avoidable visual impairment due to diabetic retinopathy.

This study demonstrates that non-mydriatic fundus photography is a practical, reliable, and effective screening tool for the early detection of diabetic retinopathy in rural and resource-limited healthcare settings. The high proportion of gradable images obtained without pharmacological pupil dilation highlights the feasibility of integrating this technique into routine outpatient services at both specialist eye hospitals and general medical college settings. Importantly, the screening approach enabled identification of a substantial number of individuals with early-stage diabetic retinopathy who were previously unaware of their retinal status, underscoring its value in uncovering silent disease before the onset of vision-threatening complications.

The predominance of mild and moderate non-proliferative diabetic retinopathy among detected cases reinforces the role of systematic screening in shifting diagnosis toward earlier stages, when timely intervention and improved glycaemic control can effectively delay disease progression. The strong association observed between retinopathy prevalence and longer duration of diabetes further supports existing evidence and emphasises the need for sustained retinal surveillance in long-standing diabetic individuals. Additionally, the ability of non-mydriatic fundus photography to reliably identify referable disease and ungradable cases ensures appropriate triage and optimal utilisation of limited ophthalmology services.

From an operational perspective, the short acquisition time, high patient acceptability, and consistent performance across two different government healthcare institutions suggest that this screening modality is well suited for large-scale implementation in rural diabetes care programmes. Overall, incorporation of non-mydriatic fundus photography into existing healthcare infrastructure has the potential to strengthen early detection strategies, improve referral pathways, and contribute meaningfully to the prevention of avoidable vision loss due to diabetic retinopathy.

Funding Source

This study did not receive any specific financial support from public, commercial, or not-for-profit funding agencies. All investigations were carried out using existing institutional facilities and resources of the Regional Eye Hospital, Kakatiya Medical College, Hanumakonda, and Government Medical College, Narsampet, Warangal.

Conflict of Interest

The authors declare that there are no conflicts of interest, financial or otherwise, related to the conduct of this study or the preparation of this manuscript.

Ethical Approval and Consent

The study protocol was reviewed and approved by the Institutional Ethics Committees of the participating institutions. Written informed consent was obtained from all participants prior to enrolment, and the study was conducted in accordance with the principles of the Declaration of Helsinki.

Acknowledgements

The authors sincerely acknowledge the support of the ophthalmology, optometry, and technical staff of the Regional Eye Hospital, Kakatiya Medical College, Hanumakonda, and Government Medical College, Narsampet, Warangal, for their assistance in patient coordination and fundus image acquisition. The cooperation of all study participants is gratefully appreciated.

1. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004 May;27(5):1047-53. doi: 10.2337/diacare.27.5.1047. PMID: 15111519.
2. Anjana RM, Deepa M, Pradeepa R, Mahanta J, Narain K, Das HK, et al. ICMR–INDIAB Collaborative Study Group. Prevalence of diabetes and prediabetes in 15 states of India: results from the ICMR-INDIAB population-based cross-sectional study. Lancet Diabetes Endocrinol. 2017 Aug;5(8):585-596. doi: 10.1016/S2213-8587(17)30174-2. Epub 2017 Jun 7. Erratum in: Lancet Diabetes Endocrinol. 2017 Aug;5(8):e5. doi: 10.1016/S2213-8587(17)30199-7. PMID: 28601585.
3. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet. 2010 Jul 10;376(9735):124-36. doi: 10.1016/S0140-6736(09)62124-3. Epub 2010 Jun 26. PMID: 20580421.
4. Stitt AW, Curtis TM, Chen M, Medina RJ, McKay GJ, Jenkins A, et al. The progress in understanding and treatment of diabetic retinopathy. Prog Retin Eye Res. 2016 Mar;51:156-86. doi: 10.1016/j.preteyeres.2015.08.001. Epub 2015 Aug 18. PMID: 26297071.
5. Antonetti DA, Klein R, Gardner TW. Diabetic retinopathy. N Engl J Med. 2012 Mar 29;366(13):1227-39. doi: 10.1056/NEJMra1005073. PMID: 22455417.
6. Yau JW, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, et al. Meta-Analysis for Eye Disease (META-EYE) Study Group. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012 Mar;35(3):556-64. doi: 10.2337/dc11-1909. Epub 2012 Feb 1. PMID: 22301125; PMCID: PMC3322721.
7. Wong TY, Sabanayagam C. Strategies to Tackle the Global Burden of Diabetic Retinopathy: From Epidemiology to Artificial Intelligence. Ophthalmologica. 2020;243(1):9-20. doi: 10.1159/000502387. Epub 2019 Aug 13. PMID: 31408872.
8. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema. Arch Ophthalmol. 1985;103(12):1796-1806. PMID: 2866759
9. Rani PK, Raman R, Sharma V, Mahuli SV, Tarigopala A, Sudhir RR, et al. Analysis of a comprehensive diabetic retinopathy screening model for rural and urban diabetics in developing countries. Br J Ophthalmol. 2007 Nov;91(11):1425-9. doi: 10.1136/bjo.2007.120659. PMID: 17947265; PMCID: PMC2095459.
10. Vujosevic S, Aldington SJ, Silva P, Hernández C, Scanlon P, Peto T, et al. Screening for diabetic retinopathy: new perspectives and challenges. Lancet Diabetes Endocrinol. 2020 Apr;8(4):337-347. doi: 10.1016/S2213-8587(19)30411-5. Epub 2020 Feb 27. PMID: 32113513.
11. Scanlon PH. The English National Screening Programme for diabetic retinopathy 2003-2016. Acta Diabetol. 2017 Jun;54(6):515-525. doi: 10.1007/s00592-017-0974-1. Epub 2017 Feb 22. PMID: 28224275; PMCID: PMC5429356.
12. Boucher MC, Gresset JA, Angioi K, Olivier S. Effectiveness and safety of screening for diabetic retinopathy with two nonmydriatic digital images compared with the seven standard stereoscopic photographic fields. Can J Ophthalmol. 2003 Dec;38(7):557-68. doi: 10.1016/s0008-4182(03)80109-6. PMID: 14740797.
13. Cavallerano JD, Aiello LP, Cavallerano AA, Katalinic P, Hock K, Kirby R, et al. Nonmydriatic digital imaging alternative for annual retinal examination in persons with previously documented no or mild diabetic retinopathy. Am J Ophthalmol. 2005 Oct;140(4):667-73. doi: 10.1016/j.ajo.2005.03.075. Epub 2005 Aug 3. PMID: 16083842.
14. Silva PS, Horton MB, Clary D, Lewis DG, Sun JK, Cavallerano JD, et al. Identification of Diabetic Retinopathy and Ungradable Image Rate with Ultrawide Field Imaging in a National Teleophthalmology Program. Ophthalmology. 2016 Jun;123(6):1360-7. doi: 10.1016/j.ophtha.2016.01.043. Epub 2016 Mar 2. PMID: 26949120.
15. Harding S, Greenwood R, Aldington S, Gibson J, Owens D, Taylor R, et al. Diabetic Retinopathy Grading and Disease Management Working Party. Grading and disease management in national screening for diabetic retinopathy in England and Wales. Diabet Med. 2003 Dec;20(12):965-71. doi: 10.1111/j.1464-5491.2003.01077.x. PMID: 14632697.
16. Ashraf M, Hock KM, Cavallerano JD, Wang FL, Silva PS. Comparison of Widefield Laser Ophthalmoscopy and ETDRS Retinal Area for Diabetic Retinopathy. Ophthalmol Sci. 2022 Jun 28;2(4):100190. doi: 10.1016/j.xops.2022.100190. PMID: 36531579; PMCID: PMC9754965.
17. Al-Omar HA, Czech M, Quang Nam T, Gottwald-Hostalek U, Vesic N, et al. Cost saving analysis of prediabetes intervention modalities in comparison with inaction using Markov state transition model-A multiregional case study. J Diabetes. 2024 May;16(5):e13553. doi: 10.1111/1753-0407.13553. PMID: 38664882; PMCID: PMC11045917.
18. Lin DY, Blumenkranz MS, Brothers RJ, Grosvenor DM. The sensitivity and specificity of single-field nonmydriatic monochromatic digital fundus photography with remote image interpretation for diabetic retinopathy screening: a comparison with ophthalmoscopy and standardized mydriatic color photography. Am J Ophthalmol. 2002 Aug;134(2):204-13. doi: 10.1016/s0002-9394(02)01522-2. PMID: 12140027.
19. Venugopal D, Lal B, Fernandes S, Gavde D. Awareness and knowledge of diabetic retinopathy and associated factors in Goa: A hospital-based cross-sectional study. Indian J Ophthalmol. 2020 Feb;68(2):383-390. doi: 10.4103/ijo.IJO_1218_19. PMID: 31957734; PMCID: PMC7003591.
20. Vashist P, Singh S, Gupta N, Saxena R. Role of early screening for diabetic retinopathy in patients with diabetes mellitus: an overview. Indian J Community Med. 2011 Oct;36(4):247-52. doi: 10.4103/0970-0218.91324. PMID: 22279252; PMCID: PMC3263142.
21. Raman R, Ganesan S, Pal SS, Gella L, Kulothungan V, Sharma T. Incidence and Progression of Diabetic Retinopathy in Urban India: Sankara Nethralaya-Diabetic Retinopathy Epidemiology and Molecular Genetics Study (SN-DREAMS II), Report 1. Ophthalmic Epidemiol. 2017 Oct;24(5):294-302. doi: 10.1080/09286586.2017.1290257. Epub 2017 Mar 23. PMID: 28332894.
22. Wong TY, Cheung CM, Larsen M, Sharma S, Simó R. Diabetic retinopathy. Nat Rev Dis Primers. 2016 Mar 17;2:16012. doi: 10.1038/nrdp.2016.12. PMID: 27159554.