Due to the recognised risk associated with preterm birth, several strategies are in place to evaluate outcomes in very preterm children (Kallioinen et al., 2017; NICE,
35
2017). Before the release of the updated NICE guidelines in August 2017, standard practice in the UK following preterm birth of before 32 weeks gestation required infants to be followed up at 4 time points after discharge from hospital; at 3, 6, 12 and 24 months of age. The most recent guidelines include an additional follow-up within the fourth year (NICE, 2017). The most frequently used standardised assessment in preterm follow-up clinics in the first two years after discharge is the Bayley-scales of Infant and Toddler Development (current version, edition III) (Bayley, 2006). These follow-ups are recommended to all clinics, however, variations in practical and economic allowances across difference hospital trusts may mean assessments such as these are not always performed. Tracking the development of these children in this crucial and the nature of these assessments should work in theory, but this does not always translate into clinical practice. Multiple hour-long assessments are expensive and require well-structured comprehensive follow-up services. The difficulties in implementing these practices suggest a need to address the ease of follow-up in this population, potentially calling for faster and more efficient methods of assessment.
In addition to these clinical implications, many studies have reviewed the current edition of the Bayley following concerns regarding the sensitivity and predictive validity of the tool (Milne, McDonald and Comino, 2012; Aylward, 2013; Spittle et
al., 2013; Johnson, Moore and Marlow, 2014; Mansson and Stjernqvist, 2014;
Spencer-Smith et al., 2015; Anderson and Burnett, 2017). The overwhelming conclusion of the majority of these studies finds the current edition not sensitive enough to detect children with mild cognitive impairments. Although not originally designed to predict IQ, the Bayley-III is commonly reported to under-identify those likely to later present with cognitive impairments, and exhibits particular difficulty in recognising those that fall within the mild impairment range that are likely to require additional assistance before starting school (Anderson and Burnett, 2017). Prior to the Bayley-III, the Bayley-II was the most frequently used infant developmental assessment and yielded much professional confidence (Johnson and Marlow, 2006). Upon an update to the test, the re-standardisation procedure used different reference population strategies, including seeding the population with
36
poorly performing groups. Final scores were found to be 7 points higher than Bayley-II according to the publisher (Bayley, 2006; Moore, Johnson, et al., 2012). This lower sensitivity at lower scores and a non-linear relationship with the Bayley- III brings challenges when targeting the detection of mild impairments as in preterm cohorts (Moore, Johnson, et al., 2012). The proposition from a number of studies to overcome this insensitivity in the first instances is to use 1 standard deviation below the norm, or to formulate a local control reference for the purposes of research (Marlow, 2013; Spencer-Smith et al., 2015).
Assessments at 2 years of age also show poor predictive validity of cognitive performance at 5 (Potharst et al., 2012). Development of better measures is therefore essential before successful interventions can be put in place that work to improve later outcomes. This emphasizes the importance of obtaining a greater understanding of the domains that underpin the cognitive deficits later observed, and the subsequent need to identify the development profile and trajectory of specific domains. This will allow the necessary predictions to be made regarding which infants may benefit from intervention services. More broadly, early interventions have shown to have advantageous effects on the developmental profile within infancy and leading on to improved cognitive performance at preschool age, including in preterm populations (Spittle et al., 2007; Hadders-Algra, 2011). However, evidence is limited that interventions significantly improve longer term cognitive performances within ex-preterm children, supporting the requirement of more extensive research (Spittle et al., 2007).
A key factor to consider in the assessment of premature cohorts is the adjustment for age at birth. Accounting for prematurity in developmental assessments has been a long running discussion within the literature. The disagreements stem from two viewpoints: the biological opinion and the environmental opinion. The biological perspective states development takes a set time from conception. Originally proposed by Gesell and Amatruda (1947) (for review see Wilson and Cradock, 2004), they stated development was dictated by time itself and was not influenced by external factors. As such, a preterm child would lag behind a term born child on a
37
developmental basis, at least initially, due to the underdeveloped central nervous system. Once the nervous system was fully developed the opinion within the literature employed the concept that ‘catch up’ growth would occur and the preterm child would meet the developmental stage of a term born within the first few years after birth. On the contrary, the environmental perspective suggests that development is primarily driven by external influences and the exposure to the outside world, with factors such as medical care and parental stimulation, advances and aids development (Wilson and Cradock, 2004). In support of this is the research into the benefits of higher social economic status effects on developmental achievement (Tideman, 2000; Hack, 2006). In any regard, the set clinical practice following preterm birth is to utilise correction for gestational age as proportionately using a biological basis, the assessment at chronological age puts preterm children at a disadvantage, and this disadvantage increases with lower gestational age (Wilson-Ching et al., 2014). Therefore in the current investigation adjusted ages will be utilised.