Diabetic Retinopathy Screening Programme Research Articles

Transition probabilities of diabetic retinopathy and death in an Asian population with diabetes

PurposeTo evaluate the dynamic transitions in diabetic retinopathy (DR) severity over time and associated risk factors in an Asian population with diabetes. DesignLongitudinal cohort study MethodsWe analyzed data from 9481 adults in the Singapore Integrated Diabetic Retinopathy Screening Program (2010–2015) with linkage to death registry. A multistate Markov model adjusted for age, sex, systolic blood pressure (SBP), diabetes duration, HbA1c, and body mass index (BMI) was applied to estimate annual transition probabilities between four DR states (no, mild, moderate, and severe/proliferative) and death, and the mean sojourn time in each state. ResultsThe median assessment interval was 12 months, with most patients having 3 assessments. Annual probabilities for DR progression (no-to-mild, mild-to-moderate and moderate-to-severe/proliferative) were 6.1 %, 7.0 % and 19.3 %, respectively; and for regression (mild-to-no, moderate-to-mild and severe-to-moderate) were 55.4 %, 17.3 % and 4.4 %, respectively. Annual mortality rates from each DR state were 1.2 %, 2.0 %, 18.7 %, and 30.0 %. The sojourn time in each state were 8.2, 0.8, 0.8 and 2.2 years. Higher HbA1c and SBP levels were associated with progression of no-mild and mild-moderate DR, and diabetes duration with no-to-mild and moderate-to-severe/proliferative DR. Lower HbA1c levels were associated with regression from mild-to-no and moderate-to-mild, and higher BMI with mild-to-no DR. ConclusionsOur results suggest a prolonged duration (∼8 years) in developing mild DR, with faster transitions (within a year) from mild or moderate states. Moderate/above DR greatly increases the probability of progression and death as compared to mild DR/below. HbA1c was associated with both progression as well as regression.

Novel Technologies in Artificial Intelligence and Telemedicine for Glaucoma Screening.

Please note that this content has temporarily been taken down as it was inadvertently posted before it was fully ready for publication.

Attention-based deep learning framework for automatic fundus image processing to aid in diabetic retinopathy grading

Background and objectiveEarly detection and grading of Diabetic Retinopathy (DR) is essential to determine an adequate treatment and prevent severe vision loss. However, the manual analysis of fundus images is time consuming and DR screening programs are challenged by the availability of human graders. Current automatic approaches for DR grading attempt the joint detection of all signs at the same time. However, the classification can be optimized if red lesions and bright lesions are independently processed since the task gets divided and simplified. Furthermore, clinicians would greatly benefit from explainable artificial intelligence (XAI) to support the automatic model predictions, especially when the type of lesion is specified. As a novelty, we propose an end-to-end deep learning framework for automatic DR grading (5 severity degrees) based on separating the attention of the dark structures from the bright structures of the retina. As the main contribution, this approach allowed us to generate independent interpretable attention maps for red lesions, such as microaneurysms and hemorrhages, and bright lesions, such as hard exudates, while using image-level labels only. MethodsOur approach is based on a novel attention mechanism which focuses separately on the dark and the bright structures of the retina by performing a previous image decomposition. This mechanism can be seen as a XAI approach which generates independent attention maps for red lesions and bright lesions. The framework includes an image quality assessment stage and deep learning-related techniques, such as data augmentation, transfer learning and fine-tuning. We used the architecture Xception as a feature extractor and the focal loss function to deal with data imbalance. ResultsThe Kaggle DR detection dataset was used for method development and validation. The proposed approach achieved 83.7 % accuracy and a Quadratic Weighted Kappa of 0.78 to classify DR among 5 severity degrees, which outperforms several state-of-the-art approaches. Nevertheless, the main result of this work is the generated attention maps, which reveal the pathological regions on the image distinguishing the red lesions and the bright lesions. These maps provide explainability to the model predictions. ConclusionsOur results suggest that our framework is effective to automatically grade DR. The separate attention approach has proven useful for optimizing the classification. On top of that, the obtained attention maps facilitate visual interpretation for clinicians. Therefore, the proposed method could be a diagnostic aid for the early detection and grading of DR.

Compound Scaling Encoder-Decoder (CoSED) Network for Diabetic Retinopathy Related Bio-marker Detection.

Biomedical image segmentation plays an important role in Diabetic Retinopathy (DR)-related biomarker detection. DR is an ocular disease that affects the retina in people with diabetes and could lead to visual impairment if management measures are not taken in a timely manner. In DR screening programs, the presence and severity of DR are identified and classified based on various microvascular lesions detected by qualified ophthalmic screeners. Such a detection process is time-consuming and error-prone, given the small size of the microvascular lesions and the volume of images, especially with the increasing prevalence of diabetes. Automated image processing using deep learning methods is recognized as a promising approach to support diabetic retinopathy screening. In this paper, we propose a novel compound scaling encoder-decoder network architecture to improve the accuracy and running efficiency of microvascular lesion segmentation. In the encoder phase, we develop a lightweight encoder to speed up the training process, where the encoder network is scaled up in depth, width, and resolution dimensions. In the decoder phase, an attention mechanism is introduced to yield higher accuracy. Specifically, we employ Concurrent Spatial and Channel Squeeze and Channel Excitation (scSE) blocks to fully utilise both spatial and channel-wise information. Additionally, a compound loss function is incorporated with transfer learning to handle the problem of imbalanced data and further improve performance. To assess performance, our method is evaluated on two large-scale lesion segmentation datasets: DDR and FGADR datasets. Experimental results demonstrate the superiority of our method compared to other competent methods. Our codes are available at https://github.com/DeweiYi/CoSED-Net.

AI APPLICATIONS IN SCREENING AND DIAGNOSIS OF DIABETIC RETINOPATHY IN RURAL SETTINGS

Diabetic retinopathy (DR) remains a significant cause of vision impairment and blindness, particularly in rural settings where access to specialized healthcare services is limited. The integration of artificial intelligence (AI) holds promise in revolutionizing the screening and diagnosis of DR, offering a scalable solution to bridge the gap in healthcare disparities. This systematic review synthesizes existing literature on AI applications tailored for screening and diagnosing diabetic retinopathy in rural areas. Through a comprehensive search across various databases, including PubMed, IEEE Xplore, and Google Scholar, a total of 88 studies meeting the inclusion criteria were identified. These studies encompassed a range of AI techniques, including deep learning algorithms, machine learning models, and image processing methods, deployed in diverse rural healthcare settings globally. The findings reveal that AI-based systems demonstrate high accuracy, sensitivity, and specificity in detecting diabetic retinopathy from fundus images, thereby enabling early identification and timely intervention. Moreover, the scalability and cost-effectiveness of these AI solutions make them particularly suitable for resource-constrained rural environments. However, several challenges persist, including the need for robust validation studies, integration with existing healthcare infrastructure, and addressing ethical and regulatory concerns. Additionally, considerations regarding data privacy, patient acceptance, and healthcare provider training are crucial for the successful implementation of AI-driven DR screening programs in rural settings. This systematic review underscores the transformative potential of AI technologies in improving access to diabetic retinopathy screening and diagnosis in rural areas. Future research should focus on addressing the identified challenges and optimizing AI systems to enhance their efficacy and accessibility in underserved communities.
 Keywords: AI, Rural, Diagnosis, Diabetic, Retinopathy, Rural, Review.

Addressing Technical Failures in a Diabetic Retinopathy Screening Program [Letter

Addressing Technical Failures in a Diabetic Retinopathy Screening Program [Letter

Addressing Technical Failures in a Diabetic Retinopathy Screening Program [Response to Letter

Addressing Technical Failures in a Diabetic Retinopathy Screening Program [Response to Letter

The Economic Burden of Diabetic Retinopathy in Jordan: Cost Analysis and Associated Factors.

Diabetic retinopathy (DR) is the leading cause of visual loss worldwide in patients with diabetes mellitus (DM). The aims of our study are to describe the costs associated with (DR) and to evaluate its economic impact in Jordan. Retrospectively, we included all patients with DM and classified them according to the severity of DR. Data regarding medical history, ophthalmic history, stage of DR, presence of DME, and the ophthalmic procedures and operations were collected. The total DR-related cost was measured as a direct medical cost for the outpatient and inpatient services. Two hundred and twenty-nine patients were included in the study. Only 49.7% of the patients presented without DR, and 21% presented with diabetic macular edema (DME) unilaterally or bilaterally. The DR-related cost was significantly associated with insulin-based regimens, longer duration of DM, higher HbA1c levels, worse stage of DR at presentation, the presence of DME at presentation, the presence of glaucoma, and increased mean number of intravitreal injections, laser sessions, and surgical operations. Multivariate analysis should the presenting stage of DR, presence of DME, and the presence of DME be the independent factors affecting the DR-related cost. This study is the first study to be conducted in Jordan and encourages us to establish a screening program for DR for earlier detection and treatment. DM control and treatment compliance will reduce the heavy costs of the already exhausted healthcare and financial system.

The use of social simulation modelling to understand adherence to diabetic retinopathy screening programs

The success of screening programs depends to a large extent on the adherence of the target population, so it is therefore of fundamental importance to develop computer simulation models that make it possible to understand the factors that correlate with this adherence, as well as to identify population groups with low adherence to define public health strategies that promote behavioral change. Our aim is to demonstrate that it is possible to simulate screening adherence behavior using computer simulations. Three versions of an agent-based model are presented using different methods to determine the agent’s individual decision to adhere to screening: (a) logistic regression; (b) fuzzy logic components and (c) a combination of the previous. All versions were based on real data from 271,867 calls for diabetic retinopathy screening. The results obtained are statistically very close to the real ones, which allows us to conclude that despite having a high degree of abstraction from the real data, the simulations are very valid and useful as a tool to support decisions in health planning, while evaluating multiple scenarios and accounting for emergent behavior.

Economic evaluation for medical artificial intelligence: accuracy vs. cost-effectiveness in a diabetic retinopathy screening case

Artificial intelligence (AI) models have shown great accuracy in health screening. However, for real-world implementation, high accuracy may not guarantee cost-effectiveness. Improving AI’s sensitivity finds more high-risk patients but may raise medical costs while increasing specificity reduces unnecessary referrals but may weaken detection capability. To evaluate the trade-off between AI model performance and the long-running cost-effectiveness, we conducted a cost-effectiveness analysis in a nationwide diabetic retinopathy (DR) screening program in China, comprising 251,535 participants with diabetes over 30 years. We tested a validated AI model in 1100 different diagnostic performances (presented as sensitivity/specificity pairs) and modeled annual screening scenarios. The status quo was defined as the scenario with the most accurate AI performance. The incremental cost-effectiveness ratio (ICER) was calculated for other scenarios against the status quo as cost-effectiveness metrics. Compared to the status quo (sensitivity/specificity: 93.3%/87.7%), six scenarios were cost-saving and seven were cost-effective. To achieve cost-saving or cost-effective, the AI model should reach a minimum sensitivity of 88.2% and specificity of 80.4%. The most cost-effective AI model exhibited higher sensitivity (96.3%) and lower specificity (80.4%) than the status quo. In settings with higher DR prevalence and willingness-to-pay levels, the AI needed higher sensitivity for optimal cost-effectiveness. Urban regions and younger patient groups also required higher sensitivity in AI-based screening. In real-world DR screening, the most accurate AI model may not be the most cost-effective. Cost-effectiveness should be independently evaluated, which is most likely to be affected by the AI’s sensitivity.

Present and future screening programs for diabetic retinopathy: a narrative review.

Diabetes is a prevalent global concern, with an estimated 12% of the global adult population affected by 2045. Diabetic retinopathy (DR), a sight-threatening complication, has spurred diverse screening approaches worldwide due to advances in DR knowledge, rapid technological developments in retinal imaging and variations in healthcare resources.Many high income countries have fully implemented or are on the verge of completing a national Diabetic Eye Screening Programme (DESP). Although there have been some improvements in DR screening in Africa, Asia, and American countries further progress is needed. In low-income countries, only one out of 29, partially implemented a DESP, while 21 out of 50 lower-middle-income countries have started the DR policy cycle. Among upper-middle-income countries, a third of 59 nations have advanced in DR agenda-setting, with five having a comprehensive national DESP and 11 in the early stages of implementation.Many nations use 2-4 fields fundus images, proven effective with 80-98% sensitivity and 86-100% specificity compared to the traditional seven-field evaluation for DR. A cell phone based screening with a hand held retinal camera presents a potential low-cost alternative as imaging device. While this method in low-resource settings may not entirely match the sensitivity and specificity of seven-field stereoscopic photography, positive outcomes are observed.Individualized DR screening intervals are the standard in many high-resource nations. In countries that lacks a national DESP and resources, screening are more sporadic, i.e. screening intervals are not evidence-based and often less frequently, which can lead to late recognition of treatment required DR.The rising global prevalence of DR poses an economic challenge to nationwide screening programs AI-algorithms have showed high sensitivity and specificity for detection of DR and could provide a promising solution for the future screening burden.In summary, this narrative review enlightens on the epidemiology of DR and the necessity for effective DR screening programs. Worldwide evolution in existing approaches for DR screening has showed promising results but has also revealed limitations. Technological advancements, such as handheld imaging devices, tele ophthalmology and artificial intelligence enhance cost-effectiveness, but also the accessibility of DR screening in countries with low resources or where distance to or a shortage of ophthalmologists exists.

Addressing Technical Failures in a Diabetic Retinopathy Screening Program.

Diabetic retinopathy (DR) is a preventable cause of blindness detectable through screening using retinal digital photography. The Irish National Diabetic Retina Screening (DRS) programme, Diabetic RetinaScreen, provides free screening services to patients with diabetes from aged 12 years and older. A technical failure (TF) occurs when digital retinal imaging is ungradable, resulting in delays in the diagnosis and treatment of sight-threatening disease. Despite their impact, the causes of TFs, and indeed the utility of interventions to prevent them, have not been extensively examined. Primary analysis aimed to identify factors associated with TF. Secondary analysis examined a subset of cases, assessing patient data from five time points between 2019 and 2021 to identify photographer/patient factors associated with TF. Patient data from the DRS database for one provider were extracted for analysis between 2018 and 2022. Information on patient demographics, screening results, and other factors previously associated with TF were analyzed. Primary analysis involved using mixed-effects logistic regression models with nested patient-eye random effects. Secondary analysis reviewed a subset of cases in detail, checking for causes of TF. The primary analysis included a total of 366,528 appointments from 104,407 patients over 5 years. Most patients had Type 2 diabetes (89.2%), and the overall TF rate was 4.9%. Diabetes type and duration, dilate pupil status, and the presence of lens artefacts on the camera were significantly associated with TF. The Secondary analysis identified the primary cause of TF was found to be optically dense cataracts, accounting for over half of the TFs. This study provides insight into the causes of TF within the Irish DRS program, highlighting cataracts as the primary contributing factor. The identification of patient-level factors associated with TF facilitates appropriate interventions that can be put in place to improve patient outcomes and minimize delays in treatment and diagnosis.

Implementing and evaluating a fully functional AI-enabled model for chronic eye disease screening in a real clinical environment.

Artificial intelligence (AI) has the potential to increase the affordability and accessibility of eye disease screening, especially with the recent approval of AI-based diabetic retinopathy (DR) screening programs in several countries. This study investigated the performance, feasibility, and user experience of a seamless hardware and software solution for screening chronic eye diseases in a real-world clinical environment in Germany. The solution integrated AI grading for DR, age-related macular degeneration (AMD), and glaucoma, along with specialist auditing and patient referral decision. The study comprised several components: (1) evaluating the entire system solution from recruitment to eye image capture and AI grading for DR, AMD, and glaucoma; (2) comparing specialist's grading results with AI grading results; (3) gathering user feedback on the solution. A total of 231 patients were recruited, and their consent forms were obtained. The sensitivity, specificity, and area under the curve for DR grading were 100.00%, 80.10%, and 90.00%, respectively. For AMD grading, the values were 90.91%, 78.79%, and 85.00%, and for glaucoma grading, the values were 93.26%, 76.76%, and 85.00%. The analysis of all false positive cases across the three diseases and their comparison with the final referral decisions revealed that only 17 patients were falsely referred among the 231 patients. The efficacy analysis of the system demonstrated the effectiveness of the AI grading process in the study's testing environment. Clinical staff involved in using the system provided positive feedback on the disease screening process, particularly praising the seamless workflow from patient registration to image transmission and obtaining the final result. Results from a questionnaire completed by 12 participants indicated that most found the system easy, quick, and highly satisfactory. The study also revealed room for improvement in the AMD model, suggesting the need to enhance its training data. Furthermore, the performance of the glaucoma model grading could be improved by incorporating additional measures such as intraocular pressure. The implementation of the AI-based approach for screening three chronic eye diseases proved effective in real-world settings, earning positive feedback on the usability of the integrated platform from both the screening staff and auditors. The auditing function has proven valuable for obtaining efficient second opinions from experts, pointing to its potential for enhancing remote screening capabilities. Institutional Review Board of the Hamburg Medical Chamber (Ethik-Kommission der Ärztekammer Hamburg): 2021-10574-BO-ff.

Analysis of the photographers' quality control in the diabetic retinopathy screening at the IOBA's Reading Center

Aims/Purpose: Diabetic retinopathy (DR) is the most frequent ocular complication of diabetes and one of the leading causes of avoidable blindness. Teleophthalmology screening by taken retinal images is an effective tool in its early detection and delivering prompt treatment. IOBA's Reading Center (RC) of the University of Valladolid is in charge for the Regional Health System (SACYL) DR screening program. Retinographies are taken at Primary Care Centers, under mydriasis, following the three‐field Joslin Vision Network protocol by certified photographers. Quality control (QC) is an indispensable process in a RC to ensure an accurate retinography reading. An upper limit of 5% of photographs taken with low quality has been set as acceptable. This study aimed to analyse the quality of the retinographies taken in the DR screening at the IOBA's RC during 2021 and 2022.Methods: A retrospective study was designed analysing the quality outcomes of DR screening. Certified in image reading Optometrists determined if images received at the RC had enough quality according to seven factors (halo >1/3, inadequate vessels/parenchyma sharpness, inadequate central, nasal, temporal or fundus reflex image and inadequate disk sharpness). Statistical analysis was performed using SPSS 26 (Chicago, USA). Chi‐square test was used to compare frequencies between 2021 and 2022 outcomes.Results: 4474 retinographies were taken during 2021 and 6345 during 2022. 14 photographers took part in the program (8 in both years, 2 left in 2021, and 4 joined in 2022). All analysed factors showed less than 3% of incorrect images. 6.1% of images showed at least one error in 2021 and 4.4% in 2022 (p < 0.01). Every analysed factor showed a significant improvement (p < 0.02) between 2021 and 2022 except for disk sharpness in both eyes (p > 0.05).Conclusions: The quality of the images of the IOBA RC is adequate for the SACYL DR screening program showing values below the fixed upper limit in all the analysed factors with a significant decrease in almost all values that can be attributed to increased experience of capture personnel.

Certification process in capturing retinographies in Castilla y Leon diabetic retinopathy screening program

Aims/Purpose: Certification processes are critical for tele‐ophthalmology and Reading Centers. In the diabetic retinopathy Castilla y Leon screening program capture is performed by nurses accredited in the three‐field protocol of the Joslin Vision Network (JVN). The purpose of this work was to evaluate the certification process including the times, difficulties and satisfaction degree.Methods: Type II diabetics attend their primary care centre. Retinographies of both eyes under pharmacological mydriasis are acquired by nurses after inform consent. Images are sent, interpreted and categorized to the IOBA's Reading Center (University of Valladolid). Training consisted of an on‐site lecture, on‐line theory course (University patented) and on‐site practices. Trainees uploaded retinographies from 10 real patients to an online platform for evaluation. Certification required at least 5 correctly performed. The time and intents required to obtain the certification were analysed.Results: 23 trainees from to 5 health areas were included. None of them had ever used a retinograph before. All of them successfully completed the program. The 21.74% of the trainees obtained their certification with the first 5 practices and completed the evaluation process in 1 day. 13.04% repeated 1 practice, 34.78% repeated 2, another 13.04% repeated 3 and 7.39% repeated 5. The certification process was completed after 17, 11, 20 and 8 days, respectively. Satisfaction of trainees with the theoretical lessons with a score of 8.67/10 and the practical training with a score of 9/10 was evaluated through an ad‐hoc survey.Conclusions: Retinography capture certification was achieved between 1 and 20 days and the majority (69.52%) by repeating none to two practices. The trainees were highly satisfied with the theoretical and practical content of the course.

Transition from an in-person to a telemedicine diabetic retinopathy screening program.

Timely screening and treatment are essential for preventing diabetic retinopathy blindness. Improving screening workflows can reduce waiting times for specialist evaluation and thus enhance patient outcomes. This study assessed different screening approaches in a Brazilian public healthcare setting. This retrospective study evaluated a telemedicine-based diabetic retinopathy screening implemented during the COVID-19 pandemic and compared it with in-person strategies. The evaluation was conducted from the perspective of a specialized referral center in an urban area of Central-West Brazil. In the telemedicine approach, a trained technician would capture retinal images by using a handheld camera. These images were sent to specialists for remote evaluation. Patient variables, including age, gender, duration of diabetes diagnosis, diabetes treatment, comorbidities, and waiting time, were analyzed and compared. In total, 437 patients with diabetes mellitus were included in the study (mean age: 62.5 ± 11.0 years, female: 61.7%, mean diabetes duration: 15.3 ± 9.7 years, insulin users: 67.8%). In the in-person assessment group, the average waiting time between primary care referral and specialist evaluation was 292.3 ± 213.9 days, and the referral rate was 73.29%. In the telemedicine group, the average waiting time was 158.8 ± 192.4 days, and the referral rate was 29.38%. The telemedicine approach significantly reduced the waiting time (p<0.001) and significantly lowered the referral rate (p<0.001). The telemedicine approach significantly reduced the waiting time for specialist evaluation in a real-world setting. Employing portable retinal cameras may address the burden of diabetic retinopathy, especially in resource-limited settings.

Implementation of a Teleophthalmology Screening Program for Diabetic Retinopathy in New York City.

Purpose: To examine the implementation of a teleophthalmology program for diabetic retinopathy (DR) screening at a metropolitan hospital system and identify the challenges that the clinical teams encountered using the program. Methods: The study was conducted in 2 parts. The first was a pilot retrospective chart review of 300 consecutive patients screened for DR by the teleophthalmology screening program. The baseline variables, DR capture rate and staging, and continuity of care for those diagnosed with DR were analyzed. The second was a web-based survey identifying the barriers encountered by 36 physicians and clinical staff as they participated in the teleophthalmology screening program. Results: Part 1: Of the patients evaluated, 57 (19.0%) were diagnosed with DR; 42 (73.7%) had mild nonproliferative DR (NPDR), 7 (12.3%) had moderate NPDR, none had severe NPDR, and 8 (14.0%) had PDR. Thirty-one patients (54.4%) with retinopathy diagnoses were referred for an in-person follow-up at the clinic while the rest continued monitoring via the program. Of this subset, 22 (71.0%) completed the follow-up visit. Part 2: The survey respondents comprised 28 physicians (77.8%), 6 licensed nurse practitioners (16.7%), and 2 medical assistants (5.6%). Twenty-two providers (71.0%) preferred initiating referrals for in-person annual examinations over teleophthalmology screening referrals. The most common barriers described were related to workflow interruption, time constraints, and staff shortages. Conclusions: The teleophthalmology DR screening program allowed identification of early or absent DR at clinics in an urban setting (New York City). The findings suggest areas for targeted improvement in the screening program to better complement internal referral practices' workflows.

A comprehensive strategy of diabetic retinopathy screening in a public health system: Identifying and overcoming obstacles for implementation

AimsThis study reports the implementation of a diabetic retinopathy (DR) screening program in Aracaju, Brazil, emphasizing the challenges and premature termination. MethodsThe program, a collaboration between local health authorities, national DR screening experts, private clinics, trained retinal imaging technicians, a portable retinal camera with artificial intelligence (AI), telemedicine, and AI-assisted image analysis, was established. Screening occurred at primary care centers, with free specialized treatment for high-risk DR cases. ResultsAfter a public tender, two clinics were selected for screening, and four for secondary management. Initially, diabetic patient attendance at primary care clinics varied considerably, posing a challenge. Nonetheless, the screening continued, with 3,561 patients screened within six months. However, due to complaints and disputes between providers and authorities, the program was prematurely terminated, achieving less than a quarter of the initial screening target. ConclusionsThe DR screening program faced significant challenges, including low adherence and organizational issues at primary care centers. This experience highlights the need to standardize clinical workflows, enhance organization and integration among primary care units, and address cultural factors to improve adherence rates. These insights are valuable for implementing screening programs, particularly in low- and middle-income countries.

Real-world evaluation of smartphone-based artificial intelligence to screen for diabetic retinopathy in Dominica: a clinical validation study

ObjectiveSeveral artificial intelligence (AI) systems for diabetic retinopathy screening have been validated but there is limited evidence on their performance in real-world settings. This study aimed to assess the performance...

Cost-effectiveness analysis of artificial intelligence-based diabetic retinopathy screening in rural China based on the Markov model.

This study assessed the cost-effectiveness of different diabetic retinopathy (DR) screening strategies in rural regions in China by using a Markov model to make health economic evaluations. In this study, we determined the structure of a Markov model according to the research objectives, which required parameters collected through field investigation and literature retrieval. After perfecting the model with parameters and assumptions, we developed a Markov decision analytic model according to the natural history of DR in TreeAge Pro 2011. For this model, we performed Markov cohort and cost-effectiveness analyses to simulate the probabilistic distributions of different developments in DR and the cumulative cost-effectiveness of artificial intelligence (AI)-based screening and ophthalmologist screening for DR in the rural population with diabetes mellitus (DM) in China. Additionally, a model-based health economic evaluation was performed by using quality-adjusted life years (QALYs) and incremental cost-effectiveness ratios. Last, one-way and probabilistic sensitivity analyses were performed to assess the stability of the results. From the perspective of the health system, compared with no screening, AI-based screening cost more (the incremental cost was 37,257.76 RMB (approximately 5,211.31 US dollars)), but the effect was better (the incremental utility was 0.33). Compared with AI-based screening, the cost of ophthalmologist screening was higher (the incremental cost was 14,886.76 RMB (approximately 2,070.19 US dollars)), and the effect was worse (the incremental utility was -0.31). Compared with no screening, the incremental cost-effectiveness ratio (ICER) of AI-based DR screening was 112,146.99 RMB (15,595.47 US dollars)/QALY, which was less than the threshold for the ICER (< 3 times the per capita gross domestic product (GDP), 217,341.00 RMB (30,224.03 US dollars)). Therefore, AI-based screening was cost-effective, which meant that the increased cost for each additional quality-adjusted life year was merited. Compared with no screening and ophthalmologist screening for DR, AI-based screening was the most cost-effective, which not only saved costs but also improved the quality of life of diabetes patients. Popularizing AI-based DR screening strategies in rural areas would be economically effective and feasible and can provide a scientific basis for the further formulation of early screening programs for diabetic retinopathy.