Retinopathy of prematurity (ROP) is a proliferative retinal vascular disease that affects premature infants.1 Infants born at low gestational age (GA) and/or low birth weight (BW) have a risk of ROP.2 Without timely intervention, severe ROP can progress to retinal detachment and blindness. Currently, ROP is one of the leading preventable causes of childhood blindness worldwide.3

Successful management of ROP relies on appropriate screening for early detection of high-risk disease, along with prompt treatment to prevent disease progression and visual loss. The United Kingdom (UK) Guidelines (published in 2008 by the Royal College of Paediatrics and Child Health, the Royal College of Ophthalmologists, and the British Association of Perinatal Medicine) recommend that all infants born at GA ≤31 weeks and 6 days or BW <1501 g undergo ROP screening.4 On the other hand, the United States (US) Guidelines (published in 2013 and 2018 by the American Academy of Pediatrics, American Academy of Ophthalmology, and American Association for Pediatric Ophthalmology and Strabismus) use narrower criteria; they recommend that all infants born at GA ≤30 weeks and 0 days or BW ≤1500 g undergo ROP screening.5 6

In Hong Kong, many hospitals use the UK screening criteria to guide ROP detection.7 8 9 Although the UK screening criteria are appropriate for ROP detection in many countries,10 11 12 they are not universally appropriate.13 14 15 16 17 18 In India14 15 19 and China,17 18 some infants with GA and BW above the UK screening thresholds also developed severe ROP requiring treatment. Thus, there is a need to understand the epidemiology of ROP in Hong Kong and evaluate the utility of current international guidelines for ROP detection in Hong Kong infants.

In the Early Treatment for Retinopathy of Prematurity study,20 type 1 ROP was defined as: (1) zone I, any stage of ROP, with plus disease; (2) zone I, stage 3 ROP, without plus disease; or (3) zone II, stage 2 or 3 ROP, with plus disease. Type 1 ROP requires treatment.4 6 Although it is important not to miss any infants who develop type 1 ROP requiring treatment, it is also important to avoid unnecessarily screening a large number of infants because the ROP screening procedure is painful and distressful for premature infants; it can lead to oxygen desaturation, tachycardia, and apnea.2 21 22 There is also a need to limit the systemic absorption of dilating eye drops that may cause adverse events.23 24 An effective strategy would reduce the number of infants unnecessarily screened without missing any cases of severe ROP requiring treatment. This study was conducted to determine whether the UK or the US screening criteria are more appropriate for Hong Kong, in terms of sensitivity for detecting type 1 ROP and the number of infants requiring screening.

In this retrospective cohort study, we reviewed the medical records of all premature infants who underwent ROP screening between 1 January 2009 and 31 December 2018 in Prince of Wales Hospital, Hong Kong. During the study period, all infants born at GA ≤31 weeks and 6 days or BW <1501 g (ie, UK screening criteria) underwent ROP screening. Infants with GA and BW above the UK screening threshold who had a high risk of ROP because of an unstable clinical course also underwent ROP screening at the request of the attending neonatologist. Analyses were performed to determine the numbers of ROP and type 1 ROP cases that would have been detected and missed if the US screening criteria (GA ≤30 weeks & 0 days or BW ≤1500 g) had been used.

All infants who underwent ROP screening in Prince of Wales Hospital were included. Infants were excluded if they died or were transferred to other institutions before completion of ROP screening without a known ROP outcome. Data were recorded concerning GA, BW, most severe ROP stage, any treatment, and treatment outcome. ROP findings were classified in accordance with the International Classification of ROP25 (Table 1). Treatment was indicated for infants with type 1 ROP. If the ROP stage differed between eyes in an individual infant, the more severe ROP stage was used for analysis.

The primary outcome measure was the sensitivity of the US screening criteria, compared with the UK screening criteria, for detection of type 1 ROP. The secondary outcome measure was the number of infants requiring screening.

R software (R version 3.6.1) was used for statistical analysis. All demographic data were expressed as medians and interquartile ranges (IQRs).

Of the 857 infants who underwent ROP screening in the study period, 61 were excluded because they died or were transferred to other hospitals before the completion of ROP screening. Thus, the remaining 796 infants (404 boys [50.8%] and 392 girls [49.2%]) were included in the study. The median GA was 30 weeks and 2 days (IQR=7 weeks & 3 days; range, 23 weeks & 4 days to 37 weeks & 4 days), and the median BW was 1320 g (IQR=471; range, 470-2550).

In total, 238 infants (29.9%) developed ROP, including 38 infants (4.8%) who developed type 1 ROP requiring treatment. The median GA and BW of infants who developed ROP were 27 weeks and 4 days (IQR=3 weeks & 0 days; range, 23 weeks & 4 days to 35 weeks & 5 days) and 943 g (IQR=366; range, 470-2550), respectively. The median GA and BW of infants who developed type 1 ROP were 26 weeks and 0.5 days (IQR=2 weeks & 2.5 days; range, 23 weeks & 4 days to 32 weeks & 0 days) and 781 g (IQR=315; range, 510-1240), respectively. Among the infants who developed type 1 ROP requiring treatment, 81.6% were extremely preterm (GA <28 weeks) infants and 100% were extremely low BW (<1000 g) infants. Of the treated infants, 13 had stage 2 ROP and 25 had stage 3 ROP. No infants had stage 4 or 5 ROP.

In total, 795 infants underwent ROP screening in accordance with the UK screening criteria. One infant had a GA above the UK screening threshold; however, the infant continued to undergo screening because he was only 1 day older than the screening threshold, and the attending neonatologist concluded that he had a risk of ROP. The UK screening criteria detected all cases of ROP (n=238) and type 1 ROP requiring treatment (n=38) [Table 2].

If the US screening criteria had been used, the number of infants receiving ROP screening would have decreased to 627 (21.1% reduction compared with the UK screening criteria) [Table 2]. The use of the US screening criteria would have detected 234 cases of ROP (98.3% of cases detected using the UK criteria, 234/238) and 38 cases of type 1 ROP (100% of cases detected using the UK criteria, 38/38) [Table 2]. Of the 168 infants who would not have been screened using the US screening criteria, only 4 of them (2.4%) had developed ROP (Table 3) and all cases were mild (maximum stage 1 only); all affected infants displayed spontaneous resolution of ROP without the need for treatment. No cases of type 1 ROP were missed by the US screening criteria (ie, 100% sensitivity) [Table 4].

This study showed that if the US screening criteria had been used, instead of the UK screening criteria, the number of infants screened in our population would have decreased by 21.1% without missing any case of type 1 ROP requiring treatment. The number of ROP cases that would have been missed was very small (n=4), and all cases were mild (maximum stage 1).

Previous studies showed that many hospitals in Hong Kong follow the UK screening criteria for ROP screening7 8 9,26,27; consistent with our findings, the reported incidences of ROP and type 1 ROP in Hong Kong were 16% to 28%7 8 9 and 3.4% to 3.8%,7 8 9 respectively. In the present study, type 1 ROP mainly developed in extremely preterm infants with a median GA of 26 weeks and 0.5 days (IQR=2 weeks & 2.5 days), suggesting that low GA was an important predictor of type 1 ROP in our population. Because the GA criterion is lower in the US screening criteria (≤30 weeks & 0 days) than in the UK screening criteria (≤31 weeks & 6 days), the US screening criteria may be more appropriate for Hong Kong.

Our findings were also consistent with the results of a study conducted in another hospital in Hong Kong7; in that study, 12.4% of infants would not have required ROP screening if the US screening criteria had been used, rather than the UK criteria, none of those infants would have developed ROP. Our results suggest similar outcomes in different hospitals across Hong Kong.

In a study conducted in Shanghai in mainland China, the screening thresholds were GA of 34 weeks and BW of 2000 g. The mean GA and BW of infants requiring ROP treatment were 29.3 weeks (range, 24-35) and 1331 g (range, 750-2550), respectively17; these infants were more mature and heavier than the infants in our study. The Shanghai study showed that 9% of severe ROP cases requiring treatment would have been missed if the UK screening criteria were used; 26% would have been missed if the US screening criteria were used.17 Another study conducted in Beijing in mainland China showed that 17% of treatment-requiring ROP cases would have been missed if the UK screening criteria were used; 21% would have been missed if the US screening criteria were used.18 Therefore, despite sharing the same Chinese ethnicity, infants with severe ROP differed in maturity between Hong Kong and mainland China. This discrepancy could be the result of variations in comorbidities, perinatal risk factors, standard of neonatal healthcare, and level of supplemental oxygen therapy used. Long oxygen duration, mechanical ventilation, and high level of supplemental oxygen are known risk factors for ROP.2 Therefore, the results of our study are not generalisable to regions outside of Hong Kong.

There is evidence that the UK and the US screening criteria are not appropriate for many low- and middle-income countries.15 19 28 29 In North India, 17% of severe ROP cases would have been missed if the US screening criteria were used; 22% would have been missed if the UK screening criteria were used.15 In South India, 8% of treatment-requiring ROP cases would have been missed if the US screening criteria were used; all of these cases were aggressive posterior ROP.19 In Saudi Arabia, 35% of infants older than the UK screening threshold developed ROP; one infant developed severe ROP (stage 3).28 In Turkey, severe ROP developed in 3.8% of infants born at ≥32 weeks and 6.5% of infants born at ≥1500 g.29

Although it is important not to miss any severe ROP cases, it is also preferable to avoid missing mild ROP cases because the detection of early ROP (even mild cases) can influence decisions regarding systemic management (eg, level of supplemental oxygen), thereby reducing the rate of ROP progression. In the present study, only four cases of mild ROP would have been missed by the US screening criteria; this number was very small, compared with the 168 infants (21.1%) who could have been excluded from screening. The number of screened infants required to detect one additional case of ROP was 42 (ie, 168/4). Considering that few mild ROP cases were missed in exchange for the exclusion of a large number of infants from screening, we conclude that it is acceptable and appropriate to use the US screening criteria for ROP screening in Hong Kong.

There are several benefits to reducing the number of infants screened without compromising the detection of severe ROP. First, this modified approach minimises unnecessary stress and the potential for ROP screening-related adverse events among infants. Previous studies revealed significant elevation of blood pressure, increase in pulse rate, and decrease in oxygen saturation, which persisted after ROP screening.30 A significant increase in the number of apnoea events was also observed after screening.31 Approximately half of infants develop bradycardia from the oculocardiac reflex caused by scleral depression during screening.32 Second, this modified approach can reduce hospital expenses. The estimated cost of ROP screening is approximately US$230 per infant in the US33 and US$198.9 per infant in India.34 Third, the approach can reduce the length of hospitalisation related to delays in the completion of ROP screening.35 Finally, it may minimise unnecessary parental stress and anxiety. For example, one study showed that parents of infants undergoing ROP screening had significantly higher anxiety and depression scores compared with the general population.36

In recent decades, several ROP prediction models have been developed to improve screening sensitivity and specificity, including WINROP,37,38 ROPScore,39 CHOP ROP,40 41 CO-ROP,42 STEP-ROP, 43 and G-ROP.44 45 However, these prediction models have many limitations. First, they require the collection of postnatal data such as postnatal weight gain and insulin-like growth factor 1 level, which may not be available to ophthalmologists. Second, the mechanisms by which these predictive factors would interact to affect ROP outcome are not fully understood. Third, these models were all derived from Western countries and may not be appropriate for Asian populations.46 Finally, none of these models have been validated in Hong Kong. Considering our findings in the present study, we suggest narrowing the GA screening criterion to ≤30 weeks and 0 days, consistent with the US screening criteria; this simple and straightforward approach avoids the need for calculations required by prediction models.

This study had several limitations. First, its retrospective design hindered the assessment of other risk factors (eg, supplemental oxygen level and comorbidities) that may affect ROP outcomes. Second, because of the retrospective design, we could not determine whether the use of a narrower GA screening criterion would reduce the number of screenings in real-world clinical practice. A prospective cohort study is needed to confirm our findings. Third, although the G-ROP screening criteria are more sensitive and specific than the current US screening criteria for populations in the US,44 45 we could not evaluate the suitability of G-ROP criteria in our population because we lacked data concerning postnatal weight gain. Finally, data were missing regarding infants who died or were transferred to other hospitals without a known ROP outcome. Despite these limitations, our findings are robust because the present study revealed consistent results when the same screening practices were applied to a large number of infants over a study period of 10 years.

Compared with the UK screening criteria, the US screening criteria appeared to be more appropriate for our population because they could greatly reduce the number of infants screened without compromising sensitivity for the detection of type 1 ROP. Thus, we suggest narrowing the GA criterion for consistency with the US screening criteria during ROP screening in Hong Kong. A prospective cohort study is needed to further explore the impact of changes to the screening criteria.

Concept or design: LPL Iu, WWK Yip.
Acquisition of data: LPL Iu, LTY Cheung, THM Wu.
Analysis or interpretation of data: LPL Iu, WWK Yip, JYC Lok.
Drafting of the manuscript: LPL Iu.
Critical revision of the manuscript for important intellectual content: WWK Yip, JYC Lok, M Ho, AL Young.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This research was approved by the Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee (Ref No.: 2020.176) and was performed in accordance with the tenets of the Declaration of Helsinki. A waiver of obtaining patient consent has been approved by the Research Ethics Committee for this retrospective study.

1. Hartnett ME, Penn JS. Mechanisms and management of retinopathy of prematurity. N Engl J Med 2012;367:2515-26. Crossref

2. Kim SJ, Port AD, Swan R, Campbell JP, Chan RV, Chiang MF. Retinopathy of prematurity: a review of risk factors and their clinical significance. Surv Ophthalmol 2018;63:618-37. Crossref

3. Solebo AL, Teoh L, Rahi J. Epidemiology of blindness in children. Arch Dis Child 2017;102:853-7. Crossref

4. Wilkinson AR, Haines L, Head K, Fielder AR. UK retinopathy of prematurity guideline. Early Hum Dev 2008;84:71-4. Crossref

5. Fierson WM; American Academy of Pediatrics Section on Ophthalmology; American Academy of Ophthalmology; American Association for Pediatric Ophthalmology and Strabismus; American Association of Certified Orthoptists. Screening examination of premature infants for retinopathy of prematurity. Pediatrics 2013;131:189-95. Crossref

6. Fierson WM; American Academy of Pediatrics Section on Ophthalmology; American Academy of Ophthalmology; American Association for Pediatric Ophthalmology and Strabismus; American Association of Certified Orthoptists. Screening examination of premature infants for retinopathy of prematurity. Pediatrics 2018;142:e20183061. Crossref

7. Iu LP, Lai CH, Fan MC, Wong IY, Lai JS. Screening for retinopathy of prematurity and treatment outcome in a tertiary hospital in Hong Kong. Hong Kong Med J 2017;23:41-7. Crossref

8. Yau GS, Lee JW, Tam VT, et al. Incidence and risk factors of retinopathy of prematurity from 2 neonatal intensive care units in a Hong Kong Chinese population. Asia Pac J Ophthalmol (Phila) 2016;5:185-91. Crossref

9. Luk AS, Yip WW, Lok JY, Lau HH, Young AL. Retinopathy of prematurity: applicability and compliance of guidelines in Hong Kong. Br J Ophthalmol 2017;101:453-6. Crossref

10. Amer M, Jafri WH, Nizami AM, Shomrani AI, Al-Dabaan AA, Rashid K. Retinopathy of prematurity: are we missing any infant with retinopathy of prematurity? Br J Ophthalmol 2012;96:1052-5. Crossref

11. Chaudhry TA, Hashmi FK, Salat MS, et al. Retinopathy of prematurity: an evaluation of existing screening criteria in Pakistan. Br J Ophthalmol 2014;98:298-301. Crossref

12. Ugurbas SC, Gulcan H, Canan H, Ankarali H, Torer B, Akova YA. Comparison of UK and US screening criteria for detection of retinopathy of prematurity in a developing nation. J AAPOS 2010;14:506-10. Crossref

13. Akman I, Demirel U, Yenice O, Ilerisoy H, Kazokoğlu H, Ozek E. Screening criteria for retinopathy of prematurity in developing countries. Eur J Ophthalmol 2010;20:931-7. Crossref

14. Jalali S, Matalia J, Hussain A, Anand R. Modification of screening criteria for retinopathy of prematurity in India and other middle-income countries. Am J Ophthalmol 2006;141:966-8. Crossref

15. Vinekar A, Dogra MR, Sangtam T, Narang A, Gupta A. Retinopathy of prematurity in Asian Indian babies weighing greater than 1250 grams at birth: ten year data from a tertiary care center in a developing country. Indian J Ophthalmol 2007;55:331-6. Crossref

16. Dogra MR, Katoch D, Dogra M. An update on retinopathy of prematurity (ROP). Indian J Pediatr 2017;84:930-6. Crossref

17. Xu Y, Zhou X, Zhang Q, et al. Screening for retinopathy of prematurity in China: a neonatal units-based prospective study. Invest Ophthalmol Vis Sci 2013;54:8229-36. Crossref

18. Chen Y, Li XX, Yin H, et al. Risk factors for retinopathy of prematurity in six neonatal intensive care units in Beijing, China. Br J Ophthalmol 2008;92:326-30. Crossref

19. Hungi B, Vinekar A, Datti N, et al. Retinopathy of prematurity in a rural neonatal intensive care unit in South India—a prospective study. Indian J Pediatr 2012;79:911-5. Crossref

20. Early Treatment for Retinopathy of Prematurity Cooperative Group. Revised indications for the treatment of retinopathy of prematurity: results of the early treatment for retinopathy of prematurity randomized trial. Arch Ophthalmol 2003;121:1684-94. Crossref

21. Kandasamy Y, Smith R, Wright IM, Hartley L. Pain relief for premature infants during ophthalmology assessment. J AAPOS 2011;15:276-80. Crossref

22. Cohen AM, Cook N, Harris MC, Ying GS, Binenbaum G. The pain response to mydriatic eyedrops in preterm infants. J Perinatol 2013;33:462-5. Crossref

23. Mitchell A, Hall RW, Erickson SW, Yates C, Lowery S, Hendrickson H. Systemic absorption of cyclopentolate and adverse events after retinopathy of prematurity exams. Curr Eye Res 2016;41:1601-7. Crossref

24. Alpay A, Canturk Ugurbas S, Aydemir C. Efficiency and safety of phenylephrine and tropicamide used in premature retinopathy: a prospective observational study. BMC Pediatr 2019;19:415. Crossref

25. International Committee for the Classification of Retinopathy of Prematurity. The International Classification of Retinopathy of Prematurity revisited. Arch Ophthalmol 2005;123:991-9. Crossref

26. Chow PP, Yip WW, Ho M, Lok JY, Lau HH, Young AL. Trends in the incidence of retinopathy of prematurity over a 10-year period. Int Ophthalmol 2019;39:903-9. Crossref

27. Yau GS, Lee JW, Tam VT, Liu CC, Chu BC, Yuen CY. Incidence and risk factors for retinopathy of prematurity in extreme low birth weight Chinese infants. Int Ophthalmol 2015;35:365-73. Crossref

28. Binkhathlan AA, Almahmoud LA, Saleh MJ, Srungeri S. Retinopathy of prematurity in Saudi Arabia: incidence, risk factors, and the applicability of current screening criteria. Br J Ophthalmol 2008;92:167-9. Crossref

29. Araz-Ersan B, Kir N, Akarcay K, et al. Epidemiological analysis of retinopathy of prematurity in a referral centre in Turkey. Br J Ophthalmol 2013;97:15-7. Crossref

30. Jiang JB, Zhang ZW, Zhang JW, Wang YL, Nie C, Luo XQ. Systemic changes and adverse effects induced by retinopathy of prematurity screening. Int J Ophthalmol 2016;9:1148-55.

31. Mitchell AJ, Green A, Jeffs DA, Roberson PK. Physiologic effects of retinopathy of prematurity screening examinations. Adv Neonatal Care 2011;11:291-7. Crossref

32. Schumacher AC, Ball ML, Arnold AW, Grendahl RL, Winkle RK, Arnold RW. Oculocardiac reflex during ROP exams. Clin Ophthalmol 2020;14:4263-9. Crossref

33. Yanovitch TL, Siatkowski RM, McCaffree M, Corff KE. Retinopathy of prematurity in infants with birth weight>or=1250 grams-incidence, severity, and screening guideline cost-analysis. J AAPOS 2006;10:128-34. Crossref

34. Kelkar J, Kelkar A, Sharma S, Dewani J. A mobile team for screening of retinopathy of prematurity in India: cost-effectiveness, outcomes, and impact assessment. Taiwan J Ophthalmol 2017;7:155-9. Crossref

35. Zupancic JA, Ying GS, de Alba Campomanes A, Tomlinson LA, Binenbaum G; G-ROP Study Group. Evaluation of the economic impact of modified screening criteria for retinopathy of prematurity from the Postnatal Growth and ROP (G-ROP) study. J Perinatol 2020;40:1100-8. Crossref

36. Xie W, Liang C, Xiang D, Chen F, Wang J. Resilience, anxiety and depression, coping style, social support and their correlation in parents of premature infants undergoing outpatient fundus examination for retinopathy of prematurity. Psychol Health Med 2021;26:1091-9. Crossref

37. Löfqvist C, Andersson E, Sigurdsson J, et al. Longitudinal postnatal weight and insulin-like growth factor I measurements in the prediction of retinopathy of prematurity. Arch Ophthalmol 2006;124:1711-8. Crossref

38. Löfqvist C, Hansen-Pupp I, Andersson E, et al. Validation of a new retinopathy of prematurity screening method monitoring longitudinal postnatal weight and insulinlike growth factor I. Arch Ophthalmol 2009;127:622-7. Crossref

39. Eckert GU, Fortes Filho JB, Maia M, Procianoy RS. A predictive score for retinopathy of prematurity in very low birth weight preterm infants. Eye (Lond) 2012;26:400-6. Crossref

40. Binenbaum G, Ying GS, Quinn GE, et al. A clinical prediction model to stratify retinopathy of prematurity risk using postnatal weight gain. Pediatrics 2011;127:e607-14. Crossref

41. Binenbaum G, Ying GS, Quinn GE, et al. The CHOP postnatal weight gain, birth weight, and gestational age retinopathy of prematurity risk model. Arch Ophthalmol 2012;130:1560-5. Crossref

42. Cao JH, Wagner BD, McCourt EA, et al. The Colorado-retinopathy of prematurity model (CO-ROP): postnatal weight gain screening algorithm. J AAPOS 2016;20:19-24. Crossref

43. Ricard CA, Dammann CE, Dammann O. Screening tool for early postnatal prediction of retinopathy of prematurity in preterm newborns (STEP-ROP). Neonatology 2017;112:130-6. Crossref

44. Binenbaum G, Bell EF, Donohue P, et al. Development of modified screening criteria for retinopathy of prematurity: primary results from the postnatal growth and retinopathy of prematurity study. JAMA Ophthalmol 2018;136:1034-40. Crossref

45. Binenbaum G, Tomlinson LA, de Alba Campomanes AG, et al. Validation of the postnatal growth and retinopathy of prematurity screening criteria. JAMA Ophthalmol 2019;138:31-7. Crossref

46. Iu LP, Yip WW, Lok JY, Fan MC, Lai CH, Ho M, Young AL. Prediction model to predict type 1 retinopathy of prematurity using gestational age and birth weight (PW-ROP). Br J Ophthalmol 2023;107:1007-11. Crossref