Validation of DIGIROP models and decision support tool for prediction of treatment for retinopathy of prematurity on a contemporary Swedish cohort

Predicting the risk of treatment-requiring retinopathy of prematurity in preterm infants in Greece. External validation of DIGIROP prognostic models

PURPOSE

To assess the predictive performance of DIGIROP-v1.0 models in identifying treatment-requiring ROP among infants undergoing ROP screening at a tertiary neonatal intensive care unit in Greece.

METHODS

Retrospective cohort analysis of 640 consecutive screened preterm infants with gestational age (GA) 240/7 to 306/7 weeks and known ROP outcome in the 2nd Neonatology Department of Aristotle University of Thessaloniki (2009-2021). The primary outcome was the development of type 1 ROP according to the Early Treatment of ROP criteria or treatment based on the ophthalmologist's judgement. Sensitivity, specificity, area under the curve (AUC) with corresponding 95% confidence intervals (CI) and calibration plots for the DIGIROP-v1.0 models were displayed.

RESULTS

The DIGIROP-Birth-v1.0 model correctly identified 35/43 treated infants (sensitivity 81.4% [95% CI, 66.6%-91.6%], specificity 61.5% [95% CI, 57.4%-65.4%], AUC 0.82 [95% CI, 0.75-0.90]). During the postnatal weeks 6-14 the sensitivity of the DIGIROP-Screen-v1.0 model ranged from 82.6% to 100%. Eleven infants, all with severe comorbidities, that is, congenital malformation(s), syndrome(s), hydrocephalus or history of intestinal surgery, that were treated, were missed by the model, but met criteria for screening according to DIGIROP-v1.0 models' recommendations, and to our unit's routine standards.

CONCLUSION

The DIGIROP-v1.0 models resulted in lower sensitivity and higher specificity in this Greek cohort compared with the Swedish development group. Despite higher GA and BW, infants in our cohort had higher incidence of treated ROP than in Sweden, resulting in an under-estimation of their risk for treatment-requiring ROP. Further validation of the DIGIROP-v2.0 models and potential adjusting are recommended to maximize generalizability in populations with different characteristics.

Nutritional interventions to prevent retinopathy of prematurity

Very preterm infants are at high risk of growth failure. Poor weight gain is a prominent risk factor for retinopathy of prematurity (ROP) and optimizing nutrition could potentially promote growth and reduce ROP. Most infants at risk of ROP need parenteral nutrition initially and studies of enhanced parenteral provision of lipids and amino acids have suggested a beneficial effect on ROP. Higher amino acid intake was associated with lower incidence of hyperglycemia, a risk factor for ROP. For very preterm infants, providing unpasteurized fortified raw maternal breast milk appears to have a dose-dependent preventive effect on ROP. These infants become deficient in arachidonic acid (ArA) and docosahexaenoic acid (DHA) after birth when the maternal supply is lost. Earlier studies have investigated the impact of omega-3 fatty acids on ROP with mixed results. In a recent study, early enteral supplementation of ArA 100 mg/kg/d and DHA 50 mg/kg/d until term equivalent age reduced the incidence of severe ROP by 50%. IMPACT: Previous reviews of nutritional interventions to prevent morbidities in preterm infants have mainly addressed bronchopulmonary dysplasia, brain lesions and neurodevelopmental outcome. This review focusses on ROP. Neonatal enteral supplementation with arachidonic acid and docosahexaenoic acid, at levels similar to the fetal accretion rate, has been found to reduce severe ROP by 50% in randomized controlled trials.

Artificial Intelligence and Ophthalmic Clinical Registries

PURPOSE

The recent advances in artificial intelligence (AI) represent a promising solution to increasing clinical demand and ever limited health resources. Whilst powerful, AI models require vast amounts of representative training data to output meaningful predictions in the clinical environment. Clinical registries represent a promising source of large volume real-world data which could be used to train more accurate and widely applicable AI models. This review aims to provide an overview of the current applications of AI to ophthalmic clinical registry data.

DESIGN AND METHODS

A systematic search of EMBASE, Medline, PubMed, Scopus and Web of Science for primary research articles that applied AI to ophthalmic clinical registry data was conducted in July 2024.

RESULTS

Twenty-three primary research articles applying AI to ophthalmic clinic registries (n = 14) were found. Registries were primarily defined by the condition captured and the most common conditions where AI was applied were glaucoma (n = 3) and neovascular age-related macular degeneration (n = 3). Tabular clinical data was the most common form of input into AI algorithms and outputs were primarily classifiers (n = 8, 40%) and risk quantifier models (n = 7, 35%). The AI algorithms applied were almost exclusively supervised conventional machine learning models (n = 39, 85%) such as decision tree classifiers and logistic regression, with only 7 applications of deep learning or natural language processing algorithms. Significant heterogeneity was found with regards to model validation methodology and measures of performance.

CONCLUSIONS

Limited applications of deep learning algorithms to clinical registry data have been reported. The lack of standardized validation methodology and heterogeneity of performance outcome reporting suggests that the application of AI to clinical registries is still in its infancy constrained by the poor accessibility of registry data and reflecting the need for a standardization of methodology and greater involvement of domain experts in the future development of clinically deployable AI.

The Effects of Early Enteral and Parental Nutrition on Retinopathy of Prematurity: A Systematic Review

The management of preterm newborns must consider the severe problem of retinopathy of prematurity (ROP). A systematic review has been conducted to effectively acknowledge how enteral and parenteral early nutrition affect the growth and progression of ROP. The study summarizes recent findings from various sources to give insight into the relationship between dietary practices and ROP risks. When untreated, retinopathy of prematurity (ROP) may cause severe vision loss or blindness in premature newborns. The latter two phases of ROP progression are the most serious. A child's early nutrition, both orally and intravenously, significantly impacts the severity and progression of ROP. This systematic review aims to examine the evidence linking early nutrition to ROP in premature infants. The study used Embase, Scopus, and PubMed to conduct our search. ROP, premature newborns, and nutrition were keywords used to find relevant papers. Nine research studies made it through the screening process and offered important information on the impact of diet on ROP. These studies support the idea that poor nutrition is a driving force behind the onset of ROP. The risk of ROP has been associated with postnatal development, hyperglycemia, polyunsaturated fatty acid levels, and the presence of breast milk. The outlook for ROP has also been discovered to be affected by the length of time the patient has received parenteral feeding. The incidence and severity of ROP may be mitigated by providing better nutrition to premature newborns. This comprehensive study concludes that early nutrition, both enteral and parenteral, substantially influences the development and progression of ROP in premature newborns. The significance of nutrition in newborn care is highlighted by the possibility that improved dietary methods might aid in preventing and treating this vision-threatening illness.

Prognostic Value of Parenteral Nutrition Duration on Risk of Retinopathy of Prematurity: Development and Validation of the Revised DIGIROP Clinical Decision Support Tool

Importance

The prognostic impact of parenteral nutrition duration (PND) on retinopathy of prematurity (ROP) is not well studied. Safe prediction models can help optimize ROP screening by effectively discriminating high-risk from low-risk infants.

Objective

To evaluate the prognostic value of PND on ROP; to update and validate the Digital ROP (DIGIROP) 2.0 birth into prescreen and screen prediction models to include all ROP-screened infants regardless of gestational age (GA) and incorporate PND; and to compare the DIGIROP model with the Weight, IGF-1, Neonatal, and ROP (WINROP) and Postnatal Growth and ROP (G-ROP) models.

Design, Setting, and Participants

This retrospective study included 11 139 prematurely born infants from 2007 to 2020 from the Swedish National Registry for ROP. Extended Poisson and logistic models were applied. Data were analyzed from August 2022 to February 2023.

Main Outcomes and Measures

Any ROP and ROP requiring treatment were studied in relation to PND. ROP treatment was the outcome in DIGIROP models. Sensitivity, specificity, area under the receiver operating characteristic curve, and adjusted OR (aOR) with 95% CI were the main measures. Internal and external validations were performed.

Results

Of 11 139 screened infants, 5071 (45.5%) were girls, and the mean (SD) gestational age was 28.5 (2.4) weeks. ROP developed in 3179 infants (29%), treatment was given in 599 (5%), 7228 (65%) had PND less than 14 days, 2308 (21%) had PND for 14 days or more, and 1603 (14%) had unknown PND. PND was significantly correlated with ROP severity (Spearman r = 0.45; P < .001). Infants with 14 days or more of PND vs less than 14 days had faster progression from any ROP to ROP treatment (adjusted mean difference, -0.9 weeks; 95% CI, -1.5 to -0.3; P = .004). Infants with PND for 14 days or more vs less than 14 days had higher odds of any ROP (aOR, 1.84; 95% CI, 1.62-2.10; P < .001) and of severe ROP requiring treatment (aOR, 2.20; 95% CI, 1.73-2.80; P < .001). Among all 11 139 infants, the DIGIROP 2.0 models had 100% sensitivity (95% CI, 99.4-100). The specificity was 46.6% (95% CI, 45.6-47.5) for the prescreen model and 76.9% (95% CI, 76.1-77.7) for the screen model. G-ROP as well as the DIGIROP 2.0 prescreen and screen models showed 100% sensitivity on a validation subset (G-ROP: sensitivity, 100%; 95% CI, 93-100; DIGIROP prescreen: sensitivity, 100%; 95% CI, 93-100; DIGIROP screen: sensitivity, 100%; 95% CI, 93-100), whereas WINROP showed 89% sensitivity (95% CI, 77-96). Specificity for each prediction model was 29% (95% CI, 22-36) for G-ROP, 38% (95% CI, 32-46) for DIGIROP prescreen, 53% (95% CI, 46-60) for DIGIROP screen at 10 weeks, and 46% (95% CI, 39-53) for WINROP.

Conclusion and Relevance

Based on more than 11 000 ROP-screened infants born in Sweden, PND of 14 days or more corresponded to a significantly higher risk of having any ROP and receiving ROP treatment. These findings provide evidence to support consideration of using the updated DIGIROP 2.0 models instead of the WINROP or G-ROP models in the management of ROP.

Latest Trends in Retinopathy of Prematurity: Research on Risk Factors, Diagnostic Methods and Therapies

Retinopathy of prematurity (ROP) is a vasoproliferative disorder with an imminent risk of blindness, in cases where early diagnosis and treatment are not performed. The doctors' constant motivation to give these fragile beings a chance at life with optimal visual acuity has never stopped, since Terry first described this condition. Thus, throughout time, several specific advancements have been made in the management of ROP. Apart from the most known risk factors, this narrative review brings to light the latest research about new potential risk factors, such as: proteinuria, insulin-like growth factor 1 (IGF-1) and blood transfusions. Digital imaging has revolutionized the management of retinal pathologies, and it is more and more used in identifying and staging ROP, particularly in the disadvantaged regions by the means of telescreening. Moreover, optical coherence tomography (OCT) and automated diagnostic tools based on deep learning offer new perspectives on the ROP diagnosis. The new therapeutical trend based on the use of anti-VEGF agents is increasingly used in the treatment of ROP patients, and recent research sustains the theory according to which these agents do not interfere with the neurodevelopment of premature babies.

Retinopathy of prematurity: Metabolic risk factors

At preterm birth, the retina is incompletely vascularized. Retinopathy of prematurity (ROP) is initiated by the postnatal suppression of physiological retinal vascular development that would normally occur in utero. As the neural retina slowly matures, increasing metabolic demand including in the peripheral avascular retina, leads to signals for compensatory but pathological neovascularization. Currently, only late neovascular ROP is treated. ROP could be prevented by promoting normal vascular growth. Early perinatal metabolic dysregulation is a strong but understudied risk factor for ROP and other long-term sequelae of preterm birth. We will discuss the metabolic and oxygen needs of retina, current treatments, and potential interventions to promote normal vessel growth including control of postnatal hyperglycemia, dyslipidemia and hyperoxia-induced retinal metabolic alterations. Early supplementation of missing nutrients and growth factors and control of supplemental oxygen promotes physiological retinal development. We will discuss the current knowledge gap in retinal metabolism after preterm birth.