Which babies should be screened?
‘Which babies should be screened?’ is an important question, particularly as which babies are at risk of sight-threatening retinopathy of prematurity (ROP) varies considerably from unit to unit and from country to country. For example, in units where neonatal care is less than ideal, more mature, bigger babies may develop sight-threatening ROP than would in units providing high quality care.
Another challenge is that in many low- and middle-income countries the baby’s gestational age may not be reliable. The mother may not know the date of her last menstrual period and foetal ultrasound is not routinely provided. In these circumstances, gestational age is assessed using clinical signs such as the Ballard score (Lee at el. 2017) but this can be unreliable as it tends to overestimate. To add further complexity, in many low income countries a high proportion of newborns have foetal growth restriction. This means that birth weight is not a good guide for assessing gestational age.
ROP screening criteria
A number of countries have national guidelines indicating which babies should be screened for ROP. These usually include a combination of birth weight and gestational age. Some countries also include additional criteria, for example the United States of America (USA) includes a ‘sickness criteria’.
In neonatal units providing very high quality care in high-income countries, only the most preterm babies are at risk of ROP and therefore need to be screened, for example:
In the United Kingdom (UK), babies with a gestational age <31 weeks or a birth weight <1251 grams (g) should be screened. Screening should also be undertaken for slightly more mature, heavier infants, that is with a gestational age of 31 weeks to less than 32 weeks or a birth weight of 1251 - 1500g.
In the USA, the screening criteria are a birth weight ≤1500g or a gestational age ≤30 weeks. Infants with a birth weight of 1500 - 2000g should also be screened if they have had an ‘unstable clinical course.’
In Sweden, there is only one criterion, a gestational age of less than 32 weeks.
In low- and middle-income countries the screening criteria include more mature, bigger babies (Gilbert 2016), for example:
In China, babies with a gestational age <34 weeks or a birth weight <2000g should be screened.
In India, babies with a gestational age <34 weeks or a birth weight <1700g and babies with a birth weight <2000g are screened if the gestational age is not known. Sick preterm babies are screened at the neonatalogist’s discretion
In Mexico, babies with a gestational age ≤34 weeks or a birth weight ≤1750g should be screened. Babies who are sick are screened at the neonatalogist’s discretion.
In Russia, babies with a gestational age ≤33 weeks or a birth weight ≤1800g should be screened. In addition, babies who have been exposed to risk factors are screened at the neonatalogist’s discretion.
In Kenya, the screening criteria are a gestational age <32 weeks and a birth weight of <1500g. In addition, sick babies with a gestational age >32 - 35 weeks and a birth weight of 1501 - 2000g are screened at the neonatalogist’s discretion.
In South Africa, the screening criteria are a gestational age <32 weeks or a birth weight of <1500g. Infants who are exposed to risk factors and whose birth weight is between 1500g and 2000g are screened at the neonatalogist’s discretion.
Ideally, studies should be carried out in a range of neonatal units in each country to determine the characteristics of babies developing sight-threatening ROP as this informs which babies should be screened. Misleading data may be obtained if, for example, data from one unit which is providing high quality neonatal care is used to extrapolate to all units. The guidelines must also be reviewed regularly as sometimes they need to be modified to include more mature infants, as has happened in Saudi Arabia, for example.
Variation in the criteria for screening recommended in published guidelines from different countries, by level of economic development (see reference below Gilbert 2016).
Whichever criteria are used, it is the responsibility of the neonatologist to identify which babies should be screened and a neonatal nurse who prepares the babies for screening.
When should screening start?
Preterm babies are not born with ROP as it only develops during the first few weeks after birth. The scheduling of ROP screening examinations should ensure that eyes likely to need treatment are identified in a timely manner and minimise the number of examinations for babies at low risk of ROP.
In high-income settings, where the quality of neonatal care is high, ROP takes longer to develop in very immature infants.
In babies less than 27 weeks gestational age the screening is planned when they are at a postmenstrual age of 30 to 31 weeks.
For babies born between 27 to 32 weeks gestational age the first screening should ideally should take place by four to five weeks after birth.
For babies over 33 weeks gestational age screening is planned for the fourth week after birth.
In low- and middle-income settings, where the quality of care is less good, babies may have been exposed to hyperoxia (excessive oxygen supply) for several days and in situations where the gestational age is often unreliable or unknown, it is useful to have guidelines for the timing of the first screening which are easy to remember and implement. For example, in India the first screening must be by 30 days of life or at four weeks after birth.
If the baby is very premature or has been very sick or received a lot of oxygen, earlier screening should be considered, such as between 21 and 25 days of life, but more research is needed.
After screening: Management decision
At each screening, including the first, one of three management decisions is made:
- Urgent treatment is needed, or
- Follow up screening is needed, or
- It is safe to stop screening.
If treatment is needed, this must be delivered within 48 to 72 hours, or even earlier if aggressive posterior ROP is detected.
Follow-up screening is needed if the retinal blood vessels are not mature, whether ROP is present or not. The timing of follow up screening depending on the findings is shown in the table below. If further screening is needed, the date of the next screening examination must be documented and explained to parents. If the baby is likely to be discharged before the next screening is due, it is important to tell parents where the screening will take place, and the date and time. This should also be given to them in writing so they have a record.
Stopping screening: All eligible babies must be screened before discharge or transfer to another neonatal unit. Making sure this happens improves the parents’ awareness of ROP and, if the retinal blood vessels are mature at the first examination, they can be told that they do not need to bring their baby back for further screening. Examination before discharge has been shown to increase compliance with screening after discharge, if needed.
Management decisions are shown in the practical Job Aid below, which you can download and use in your setting as a guide
Documenting and communicating findings and management decisions
It is very important that accurate records are kept at every screening for all babies who have been screened for ROP. The following should be entered into the infant’s medical records:
- Date of screening
- Findings on both eyes
- Management decision, including when the next screening should take place, and the signature of the screener
- If parents refuse screening this should also be documented
- One of the parents and the screener should both sign the statement.
This helps to ensure that babies are screened at the right time, that follow-up screening is done when needed and for medico-legal reasons.
Additional advances taking place: Does weight gain predict ROP?
At the moment gestational age and birth weight are the most frequently used criteria for screening and some countries add ‘sickness’ or ‘exposure to risk factors’ criteria. However, as we have seen above, there are limitations and challenges with all of these criteria. This has led researchers to explore other factors which may help to predict which babies are at the greatest risk and, conversely, those at low risk.
Several studies have investigated change in insulin-like growth factor 1 during the first few weeks of life, and although the findings were encouraging in predicting babies most at risk, the assays are expensive and need to be repeated. This has led researchers to investigate weight gain during the first few weeks of life as a surrogate for insulin-like growth factor 1, and several different groups have developed statistical models. These studies show that slow, early postnatal weight gain does predict sight-threatening ROP, but the sensitivity can be less than 100% (Lin and Bibenbaum 2019). This means that weight gain alone cannot be used to reliably identify all babies who subsequently developing sight threatening ROP. Another limitation is that all the statistical models have been developed in high income settings using data from very preterm babies who have received excellent neonatal care, and studies show that they do not predict sight threatening ROP in settings where babies are more exposed to risk factors.
In an ideal world, only babies at very high risk of developing sight-threatening ROP would be screened. This would reduce stress to the babies screened unnecessarily, and reduce the cost of screening for service providers and parents. However, we are still a long way from this ideal.
© London School of Hygiene & Tropical Medicine CC BY-NC-SA 4.0