The mathematical algorithm, which is at the core of the Retina Risk App, is based on extensive research on risk factors known to affect the progression of diabetic retinopathy, including the UK Prospective Diabetes Study (UKPDS) and Wisconsin data. The risk factors incorporated in the model are the duration of diabetes, gender, blood glucose levels, HB1Ac and the type of diabetes. With the aid of these clinical outcomes, the algorithm can estimate the risk of developing sight-threatening retinopathy over a set period of time.
Validations studies on the algorithm have been performed in Danish, Dutch, Spanish and British cohorts of over 20.000 diabetics.
Collaborating institute: Department of Ophthalmology, Aarhus University Hospital.
Main collaborators: Toke Bek and Jesper Mehlsen,
Aims/hypothesis: The aim of this study was to reduce the frequency of diabetic eye-screening visits, while maintaining safety, by using information technology and individualised risk assessment to determine screening intervals.
Methods: A mathematical algorithm was created based on epidemiological data on risk factors for diabetic retinopathy. Through a website, www.risk.is, the algorithm receives clinical data, including type and duration of diabetes, HbA1c or mean blood glucose, blood pressure and the presence and grade of retinopathy. These data are used to calculate risk for sight-threatening retinopathy for each individual’s worse eye over time. A risk margin is defined and the algorithm recommends the screening interval for each patient with standardised risk of developing sight- threatening retinopathy (STR) within the screening interval. We set the risk margin so that the same number of patients develop STR within the screening interval with either fixed annual screening or our individualised screening system. The database for diabetic retinopathy at the Department of Ophthalmology, Aarhus University Hospital, Denmark, was used to empirically test the efficacy of the algorithm. Clinical data exist for 5,199 patients for 20 years and this allows testing of the algorithm in a prospective manner.
Results: In the Danish diabetes database, the algorithm recommends screening intervals ranging from 6 to 60 months with a mean of 29 months. This is 59% fewer visits than with fixed annual screening. This amounts to 41 annual visits per 100 patients.
Conclusion: Information technology based on epidemiological data may facilitate individualised determination of screening intervals for diabetic eye disease. Empirical testing suggests that this approach may be less expensive than conventional annual screening, while not compromising safety. The algorithm determines individual risk and the screening interval is individually determined based on each person’s risk profile. The algorithm has potential to save on healthcare resources and patients’ working hours by reducing the number of screening visits for an ever increasing number of diabetic patients in the world.
Access article here
Collaborating institute: EMGO Institute for Health and Care research, VU University Medical Center, Amsterdam.
Authors: A.A.W.A. van der Heijden, I. Walraven and G. Nijpels.
Objective: Validation of a model for determination of a personalized screening frequency for diabetic retinopathy.
Research design and methods: A model calculating a personalized screening interval to monitor diabetic retinopathy based on patients risk profile was validated using clinical data of 3,347 type 2 diabetes patients of the Diabetes Care System West-Friesland in the Netherlands. Two-field fundus photographs were graded according to the EURODIAB coding system. Sight threatening retinopathy (STR) was considered grade 3 to 5. Validity of the model was assessed using calibration and discrimination measures. We investigated whether model based time of screening was before or after diagnosis of STR and calculated differences in number of fundus photographs using the model compared to annual or biennial screening.
Results: During 60 months of follow-up, 89 (3.8%) patients developed STR. Using the model, mean recommended screening interval was 31 months, leading to a reduced screening frequency of 61% compared to annual screening and 22% compared to bi-annual screening. In the total population, STR incidence occurred after a mean of 23 months after the recommended time of screening in 78 patients (87.6%). In 11 patients (12.4%) STR had developed before the model-based recommended time of screening and was therefore detected later (mean: 22 months) compared to current care. Discriminatory ability of the model was good (area under the curve: 0.77, 95% CI 0.71 to 0.82). Calibration showed that the model slightly overestimated STR risk.
Conclusions: Using the model, screening frequency to monitor retinopathy can be reduced while efficacy and safety were hardly compromised which may help to reduce retinopathy screening costs.
Access article here
Collaborating institute: Health Intelligence, Cheshire.
Main collaborators: Paul Dodson, Dominic Clarke and Phil Kirby.
Objective: To validate an algorithm for diabetic retinopathy risk assessment in a UK diabetic population with UK staging (R0-4;M1,2) of diabetic retinopathy. Also, to estimate the possible reduction in screening frequency in this cohort, while maintaining safety standards unchanged.
Research design and methods: Calibrated risk estimates of R2, R3 and M1 were calculated and the risk distribution reported. The cohort is 9,690 individuals with diabetes in England, followed up for two years. Screening intervals were calculated such that the average risk of developing each stage of retinopathy within the screening period was kept the same as the observed frequency of incidences of the corresponding retinopathy grading. ROC curves were drawn and AUC calculated to estimate the prediction capability. Risk profiles for progression of retinopathy were plotted for the entire cohort.
Results: The algorithm predicts the occurrence of any retinopathy grading with 80% accuracy or type II diabetic patients (CI 0.78-0.81). The mean recommended screening interval was 22.5 months, which maintains average risk levels for the group unchanged at next visit. A great majority of the cohort is at less than 1% risk of progression within two years.
Conclusions: The algorithm predicts the occurrence of any retinopathy grading with 78-91% accuracy. The frequency of screening can be reduced by up to 47% without compromising safety. The algorithm also identifies patients in high risk of developing advanced stages of diabetic retinopathy, including R2 and R3. Relatively few patients have more than 1% risk of progression between stages in two years.
Access article here
Objective: To validate a sight-threatening diabetic retinopathy (STDR) risk assessment model to adjust the frequency of eye-screening visits in patients with diabetes mellitus.
Methods: Retrospective follow-up study of patients with diabetes mellitus attending a diabetes center. Anonymized data on gender, type and duration of diabetes, HbA1c, blood pressure and the presence and grade of diabetic retinopathy were gathered to estimate risk for STDR for each individual’s worse eye over time by means of a prediction model. Receiver operating characteristics (ROC) analysis was performed to determine the diagnostic ability of the model and a calibration graph was done to see the model fit.
Results: 508 screening intervals were analyzed, median diabetes duration was 10 years, 87% were type 2 diabetes mellitus, and 3.1% developed STDR before the next screening visit. The area under the ROC curve was 0.74, and the calibration graph showed that model had a good fit. The reduction in screening frequency was 40% compared with fixed annual screening.
Conclusion: Current prediction model used to estimate the risk of developing STDR in patients with diabetes performed well. A personalized screening frequency for diabetic retinopathy could be implemented in practice.
Keywords: diabetes mellitus, diabetic retinopathy, diabetes management algorithm,
External link here