Recent advances in neonatal care have led to an increase in the survival rates of preterm infants or those with low birth weight in Korea [1]. The average birth weight is declining, and the incidence of preterm births is on the increase [1,2]. Because preterm or low birth weight infants are at high risk of developmental delays or disorders, early and regular assessments of neurodevelopmental outcomes of graduates of neonatal intensive care units (NICUs) should continue immediately after discharge [3]. However, there is significant heterogeneity in the neurodevelopmental assessment follow-up policies in different countries [4]. A variety of development assessment tools are available for each domain and age range; however, a consensus gold standard is still lacking in terms of defining the best neurodevelopmental assessment and follow-up program for the early diagnosis of developmental delay [5,6].

Early identification of infants at high risk of developmental delays or disorders is critical for timely referral for appropriate intervention and family counseling. Proper surveillance of neurodevelopmental outcomes of infants is necessary due to the following: (1) early detection or diagnosis of developmental delay or neurodevelopmental disorders; (2) timely intervention and provision of individualized care within critical periods for better outcomes; (3) to educate family/caregivers regarding the developmental status, prognosis, and any possible problems or dangers of infants to prevent further deterioration; and (4) to improve functional outcomes of these infants and the well-being and quality of life of the entire family. Earlier involvement of parents in the care of babies with neurodevelopmental impairments before hospital discharge from the NICU is known to be effective in improving the parent-infant relationship, providing a nurturing environment, and targeting the intervention for individualized infant and family needs [7]. There should be sufficient parental education on useful and safe home exercises or play, information on proper feeding, positioning, sleep, and any available social services. In the future, these developmental interventions beyond the NICU should be updated with evidence-based intervention techniques for individual diagnoses.

Regular hospital visits for neurodevelopmental assessments after NICU discharge is widely acknowledged; however, there should be a systematic follow-up program for both appropriate diagnosis of neurodevelopmental delay and assessment of the efficacy of developmental interventions [5]. Developmental surveillance programs for NICU graduates should include all domains of neurological, motor, language, cognition, perception, and social skills. Moreover, this program should be encouraged to consider each child’s developmental status, caregiver’s socioeconomic status, individualized therapeutic program, healthcare resources, and social services or welfare [8,9]. This review summarizes the current evidence of available neurodevelopmental assessment tools for each domain and suggests appropriate Korean surveillance guidelines for NICU graduates.

Developmental surveillance should consider appropriate timing and intervals in terms of cost-effectiveness and availability of healthcare resources. If NICU graduates have more severe risk factors, they would be at an even higher risk of various developmental problems [10-12]. Several risk factors that must be evaluated during NICU stay are summarized in Table 1 for clinicians not to delay neurodevelopmental evaluation for referral to the Department of Pediatric Rehabilitation Medicine. According to the numbers and grades of risk factors (Table 1), a corrected age (CA) to visit for neurodevelopmental surveillance and follow-up periods are suggested in Fig. 1 [3,5,10,13,14].

Early developmental screening is recommended within less than 1 month after discharge if there is at least one high risk factor; any organic brain lesion, such as grade 3 or 4 intraventricular hemorrhage, cystic periventricular leukomalacia, infarction, hypoxic ischemic encephalopathy, neonatal meningitis or encephalitis, and congenital brain malformation, ventriculomegaly, etc.; any feeding disorders associated with malnutrition; neonatal sepsis; bronchopulmonary dysplasia with mechanical ventilation until gestational age of 36 weeks; hyperbilirubinemia; any congenital or neuromuscular disorder, confirmed with gene study; extremely preterm (<28 weeks); extremely low birth weight (<1,000 g); high social risk such as any domestic violence or child abuse, severe poverty or homelessness, no antenatal care provided, caregivers’ intellectual disability or psychological problems; any tone abnormality of hyper/hypotonia or fluctuating tones are observed; or a history of infantile spasm or status epilepticus (Table 1). If any neurodevelopmental delay is suspected at the initial immediate follow-up, next follow-up visits or further evaluations should be determined at the physician’s discretion according to the individual infant’s medical and neurological conditions. Otherwise, later visits can be scheduled as routine follow-ups for low-risk NICU graduates. Furthermore, an immediate intervention plan and/or education can be suggested for those with high-risk factors while still in hospital, rather than waiting for a confirmative diagnosis of developmental impairments [3].

Afterwards, moderate risk factors should be screened: very preterm (28–32 weeks) or very low birth weight (1,000–1,500 g) neonates; multiple pregnancies more than twins or discordant twins who show significantly different birth weight between twins; diagnosis of sensory abnormality, such as hearing or visual impairment, and severe retinopathy of prematurity; small for gestational age, that is, birth weight less than 10th percentile for gestational age; major perinatal surgery in the brain, heart, thorax, or abdomen including necrotizing enterocolitis operation; moderate to late preterm (32–37 weeks) or low birth weight (1,500–2,500 g) with any clinical perinatal event like epilepsy or feeding problems (Table 1). If two or more moderate risk factors are present, NICU graduates are required to be followed-up for neurodevelopmental screening within less than 1 month after discharge, similar to the existence of one high-risk factor. On the other hand, if there is only 0–1 moderate risk factor, it is recommended that NICU graduates should have regular checkups at a CA of 3–4 months for the first visit. Thereafter, further follow-up visits for neurodevelopmental assessment are recommended at CA of 8–9 months, 12–18 months, 24 months, and 36 months (Fig. 1).

However, the follow-up schedule should be refined by clinicians based on the functional and/or medical status of each infant. For example, if any special diagnosis is made, such as genetic or neurodegenerative diseases, the follow-up schedule should be individualized through experts’ and multidisciplinary care plans. Although the Bayley Scales of Infant Development (BSID) is an extensive formal developmental assessment tool for diagnosing developmental delays in early childhood for 1 to 42 months old babies (Table 2), it cannot predict long-term outcomes of development, especially when assessed at a young age such as before CA of 24 months old [13,15]. For those still undiagnosed with extremely preterm birth (<28 weeks) or extremely low birth weight (<1,000 g), BSID is strongly recommended at a CA of 36 months. Likewise, individual decision-making regarding which and when each neurodevelopmental assessment tool to choose would enrich better clinical practice and more accurate assessments.

Currently available neurodevelopmental assessment tools are extremely varied at each age band. A regular neurodevelopmental follow-up program should include all developmental domains for more accurate surveillance and diagnosis, including motor, sensory perception, cognition, and language. Irrespective of how comprehensive neurodevelopmental assessment tools are employed, they are often insufficient, and clinicians should decide on additional specialized diagnostic tools for specific domains regarding individual functional status. Based on a comprehensive history taking and physical/neurological examination, including growth, primitive reflexes, postural reactions, developmental history, social/family history, and musculoskeletal evaluation, experienced clinicians should be able to decide any necessary further evaluations, including blood tests, genetic studies, or imaging modalities [5].

Categorical neurodevelopment assessment tools are summarized and compared in detail in Table 2. Each assessment tool is characterized by its target age range, test type characteristic about whether it is norm-referenced based on standard score or criterion-referenced, suggesting a clear-cut cut-off score for diagnosis, evaluation of components within domains, diagnostic criteria, average time to administer, and immediate availability of the Korean-translated version and/or education for evaluators. Among them, the Denver Development Screening Tool (DDST), BSID, Korean-Developmental Screening Test (K-DST), Peabody Developmental Motor Scales (PDMS), Korean-Wechsler Preschool and Primary Scale of Intelligence-IV, and Developmental Test of Visual-Motor Integration-6 (VMI-6) are currently covered by National Insurance in Korea.

Furthermore, currently accumulated evidence on predictive accuracy regarding the reliability, internal consistency, and validity of each development assessment tool is searched and gathered in the Supplementary Tables S1-S4 to help healthcare professionals make a more convenient decision. Reliability is the extent to which patients can be distinguished from normal despite measurement errors and is evaluated through inter-/intra-rater intraclass correlation coefficient (ICC) or Cohen’s weighted kappa values. It is “+” if ICC or kappa ≥0.70, “–” for <0.70, and “0” if no available information is found for reliability. Internal consistency is the extent to which items within a domain are inter-correlated to measure the same construct; it is “+” if factor analysis was provided with adequate sample size and Cronbach’s alpha values are ≥0.70, “–” if Cronbach’s alpha <0.70, and “0” if no available information was found for consistency. Content validity is the comprehensiveness of items in the assessment tools for the domain of interest: “+” if a clear description about the measurement aim, target population, and item selection while target population and evaluators or experts were involved in this item selection, “–” if target population or experts did not involve item selections, “0” if no information was found, and “?” if the description of these aspects is lacking. Criterion validity is the extent to which the test scores are related to a gold standard, and is demonstrated through the correlation coefficient: “+” if coefficient ≥0.70, “–” if <0.70, and “0” if no information is suggested. Construct validity is the extent to which scores on a specific domain measure the intended theoretical construct or concept. It is assessed as “+” if specific hypotheses were formulated and ≥75% of the results are in accordance, “–” if <75% of hypotheses were confirmed, “0” if no information was available, and “?” only if doubtful hypotheses or method exists [16].

This guideline summarizes neurodevelopmental surveillance methods for patients who have been cared for and discharged from the NICU until the age of approximately 3 years. This is based on risk factor stratification and currently available assessment tools for each development domain. This surveillance program aims to enable early diagnosis and timely intervention for people with developmental disorders to support their functions and quality of life. Although there is still a lack of evidence-based early treatment guidelines for NICU graduates, this standardized post-discharge neuromotor development surveillance program would lead to a more concrete database for identifying those who need early rehabilitation interventions in the future.