# Population‐Based Cohort Study for Development of National Retinopathy of Prematurity Screening Criteria

## TL;DR

This study proposes new ROP screening criteria in Switzerland that could reduce screening burden while ensuring early detection.

## Contribution

A population-based cohort study identifies optimal ROP screening criteria to reduce screening burden while maintaining safety.

## Key findings

- 168 out of 11,354 very preterm infants required ROP treatment.
- Applying new criteria could reduce screening burden by 56% while detecting all ROP cases.
- A predictive model with an AUC of 0.963 accurately forecasts ROP treatment need.

## Abstract

Screening criteria for retinopathy of prematurity (ROP) vary among countries. Early detection of ROP and minimising the burden of screening are important.

We analysed data from very preterm infants born in Switzerland between 2006 and 2022. Logistic regression models were fitted to evaluate 17 potential risk factors for ROP treatment.

168/11354 patients (median (range) gestational age (GA) 29.6 (23.0–31.9) weeks) required ROP treatment. All would have been detected and screening burden would have been reduced by 56% if screening had required meeting ≥ 1 of the following criteria: GA < 27 weeks (89.3%), birth weight < 1000 g (97.0%), intraventricular haemorrhage≥II° (24.0%), congenital tumour (1.2%). We identified six statistically significant risk factors for ROP: GA (adjusted odds ratio (aOR) 0.46, 95% CI 0.40–0.52, p < 0.001), birth weight z‐score (aOR 0.58, 95% CI 0.46–0.73, p < 0.001), duration of supplemental oxygen (aOR 1.01 95% CI 1.01–1.02, p < 0.001), duration of mechanical ventilation (OR 1.01, 95% CI 1.00–1.02, p = 0.018), caesarean section (OR 1.84, 95% CI 1.06–3.36, p = 0.038), and congenital tumour (OR 26.3, 95% CI 2.71–189, p = 0.002). The model allowed for excellent prediction of ROP treatment (AUC 0.963, 95% CI 0.944–0.981).

Safely reducing the burden of ROP screening appears achievable in Switzerland.