Ophthalmology. Retina · 2026
Global Validation of the Postnatal Growth and Retinopathy of Prematurity Screening Model: A Systematic Review and Meta-Analysis.
СтатьяСистематический обзорМетаанализНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Chi Sheng-Chu, Chang Hsin-Ho, Lin Tai-Chi, Chen Shih-Jen, Chen Jin-Hua
Аннотация
TOPIC: The Postnatal Growth and Retinopathy of Prematurity (G-ROP) criteria, incorporating postnatal weight gain, are widely validated for retinopathy of prematurity (ROP) screening. We assessed the G-ROP algorithm's diagnostic accuracy across various outcomes and explored heterogeneity sources.
CLINICAL RELEVANCE: If G-ROP maintains sensitivity with fewer examinations, it would optimize screening for this blinding disease.
METHODS: We searched PubMed, Embase, and Cochrane databases (through September 2024) for studies using G-ROP to predict type 1, type 2, treated, or any-stage ROP. We assessed risk of bias using Quality Assessment of Diagnostic Accuracy Studies-2 and certainty of evidence using the Grading of Recommendations Assessment, Development and Evaluation approach. The protocol was prospectively registered with International Prospective Register of Systematic Reviews (CRD42024571794).
RESULTS: Twenty-three studies were included. For type 1 ROP (17 cohorts; 1406 infants), sensitivity was 0.99 (95% confidence interval [CI]: 0.99-1.00; I2 = 44.5%) and specificity (9 cohorts; 2522 infants) was 0.34 (95% CI: 0.32-0.36; I2 = 96.3%). The diagnostic odds ratio (DOR) was 6.22 (95% CI: 2.34-16.53; I2 = 40.7%), and the area under the curve (AUC) was 0.87 (95% CI: 0.83-0.90). Certainty of evidence was high for sensitivity and low for specificity and DOR. For type 2 ROP, pooled sensitivity from 9 cohorts (871 infants) was 0.98 (95% CI: 0.97-0.99; I2 = 41.4%), whereas specificity from 3 cohorts (621 infants) was 0.25 (95% CI: 0.22-0.29; I2 = 92.2%). The DOR was 3.21 (95% CI: 1.24-8.28; I2 = 0%), and the AUC was 0.76 (95% CI: 0.72-0.80). Certainty was high for sensitivity, very low for specificity, and moderate for DOR. For any-stage ROP, pooled sensitivity from 16 cohorts (3674 infants) was 0.87 (95% CI: 0.86-0.88; I2 = 90.2%), whereas specificity from 13 cohorts (2846 infants) was 0.45 (95% CI: 0.43-0.47; I2 = 95.8%). The DOR was 5.88 (95% CI: 3.89-8.89; I2 = 75.3%), and the AUC was 0.78 (95% CI: 0.74-0.81). Certainty was low for sensitivity, low for specificity, and moderate for DOR.
CONCLUSION: The G-ROP algorithm demonstrates consistently high sensitivity, particularly for type 1 and treated ROP, supporting its use as a screening tool for severe disease. Although the model reduces examinations compared with current standards, its low specificity may limit its usefulness in low-resource settings, where minimizing unnecessary screening is essential.
FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
CLINICAL RELEVANCE: If G-ROP maintains sensitivity with fewer examinations, it would optimize screening for this blinding disease.
METHODS: We searched PubMed, Embase, and Cochrane databases (through September 2024) for studies using G-ROP to predict type 1, type 2, treated, or any-stage ROP. We assessed risk of bias using Quality Assessment of Diagnostic Accuracy Studies-2 and certainty of evidence using the Grading of Recommendations Assessment, Development and Evaluation approach. The protocol was prospectively registered with International Prospective Register of Systematic Reviews (CRD42024571794).
RESULTS: Twenty-three studies were included. For type 1 ROP (17 cohorts; 1406 infants), sensitivity was 0.99 (95% confidence interval [CI]: 0.99-1.00; I2 = 44.5%) and specificity (9 cohorts; 2522 infants) was 0.34 (95% CI: 0.32-0.36; I2 = 96.3%). The diagnostic odds ratio (DOR) was 6.22 (95% CI: 2.34-16.53; I2 = 40.7%), and the area under the curve (AUC) was 0.87 (95% CI: 0.83-0.90). Certainty of evidence was high for sensitivity and low for specificity and DOR. For type 2 ROP, pooled sensitivity from 9 cohorts (871 infants) was 0.98 (95% CI: 0.97-0.99; I2 = 41.4%), whereas specificity from 3 cohorts (621 infants) was 0.25 (95% CI: 0.22-0.29; I2 = 92.2%). The DOR was 3.21 (95% CI: 1.24-8.28; I2 = 0%), and the AUC was 0.76 (95% CI: 0.72-0.80). Certainty was high for sensitivity, very low for specificity, and moderate for DOR. For any-stage ROP, pooled sensitivity from 16 cohorts (3674 infants) was 0.87 (95% CI: 0.86-0.88; I2 = 90.2%), whereas specificity from 13 cohorts (2846 infants) was 0.45 (95% CI: 0.43-0.47; I2 = 95.8%). The DOR was 5.88 (95% CI: 3.89-8.89; I2 = 75.3%), and the AUC was 0.78 (95% CI: 0.74-0.81). Certainty was low for sensitivity, low for specificity, and moderate for DOR.
CONCLUSION: The G-ROP algorithm demonstrates consistently high sensitivity, particularly for type 1 and treated ROP, supporting its use as a screening tool for severe disease. Although the model reduces examinations compared with current standards, its low specificity may limit its usefulness in low-resource settings, where minimizing unnecessary screening is essential.
FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.