Do NICU developmental care improve cognitive and motor outcomes for preterm infants? A systematic review and meta-analysis

Every year, an estimated 15 million infants are born preterm [1]. The highest rate of preterm births is in sub-Saharan Africa and South Asia (over 60%) [1]. The preterm birth rate in developing countries varies widely and has a different pattern than developed countries [2].

In recent years, the mortality rate in preterm infants were reduced by advanced perinatal care, but the developmental morbidity is remarkably high [3, 4]. In preterm infants, in addition to cerebral palsy, hearing loss, visual impairment, and growth retardation, long-term follow-up studies have identified important developmental disorders [5]. A recent study showed that more than 25% of neonates born between 28 and 32 weeks of gestation have developmental disorders at the age of 2 years old, and this ratio reaches 40% at the age of 10 [6].

Each sensory experience, proportionate or disproportionate, in the infant’s brain causes a behavioral response, which itself results in another sensory experience. When the preterm infant has sensory experiences that are disproportionate to its developmental stage, its neurodevelopment will be different from when it is in the protective environment of the uterus. Therefore, it is not surprising to see different neurodevelopmental outcomes in preterm infants compared with term ones [7]. In addition, fetal nervous system is in a very active stage of development during the third trimester. Therefore, the nervous system is vulnerable due to the immature and rapid growth in preterm infants [8].

Neurological care includes strategies that can prevent the neuronal death [9]. These strategies are interventions that protect the evolving brain or help the brain to reduce the death of neurons after trauma and improve their performance by creating new communication pathways. The more immature is the infant, the more vulnerable is its brain, and the more necessary is neurological care for it [10]. In order to prevent these complications, various methods have been proposed and implemented by researchers over the past few decades. Most of these methods are developmental interventions or cares in infants admitted to the NICU (Neonatal Intensive Care Unit). Developmental cares are methods that are intended to adjust the NICU environment to diminish the stress, support the behavioral organization, improve physiological stability, keep sleep rhythms, and promote neural growth and maturation of infant [7, 11]. In this type of care, the training and participation of parents or caregivers are critical for the social, emotional and physical health of the infant, and are important factors in the family-based care process [12]. The goals of developmental care for the family are to encourage and support parents in the primary caregiver role, and enhance family emotional and societal well-being [11, 13].

It appears that intense sensory impacts of NICU have devastating effects on neurodevelopmental outcomes in preterm infants. However, according to research, it is unclear which of the severe [8] or mild stimuli or inappropriate stimuli [14] can be most harmful to neonatal development.

The effectiveness of developmental care in preterm infants has been investigated in a number of previous systematic reviews [13, 15‐21]. Symington et al. investigated the effects of five core measures on preterm infants during NICU admission, emphasizing the short-term medical and neurobehavioral development outcomes. Symington et al. concluded that there is limited evidence to support the benefits of developmental care in improving the cognitive, motor and behavioral development in infants at the time of discharge, but negative effects have not been reported in this regard and improved neurodevelopmental outcomes to 24 months corrected age [13, 21].

Lavallée et al. examined the Developmental Care Interventions (DCI) outcomes by presenting a comprehensive narrative review of recent findings on the effectiveness of DCI on stress relief, sleep promotion, and neurodevelopmental outcomes. The main limitation of this study was the lack of a systematic method of review. Therefore, the quality of the papers examined has not been evaluated with clear criteria or standard tools and there might be bias in their report [17].

Orton, Spittle, and their colleagues investigated the effects of interventions after discharge on the cognitive and motor development of premature infants in three age groups; infancy (0 to < 3 years old), pre-school (3–5 years old), and school (5–13 and 13–18 years old) and concluded that early developmental interventions improved cognitive outcomes at infant age, and at pre-school age. However, the benefit was not sustained at school age. In addition, developmental interventions had little effect on motor outcome at infancy or school age [15, 16].

Jacobs et al. studied the long-term developmental effects during school and short-term medical and developmental effects of NIDCAP in comparison with routine care in preterm and low-birth-weight infants. The study concluded that there is insufficient evidence to support NIDCAP in improving short-term medical results and neurodevelopmental enhancement of preterm infants at school [18].

In the case of family-centered interventions, Vanderveen et al. have recently conducted a systematic review of early intervention programs in preterm infants focusing on training the parents. The review identified studies that used various interventions, including training the parents, infant’s stimulation by parents, home visits, and individual developmental care. The meta-analysis of the studies showed that early interventions had improved cognitive and motor performance of preterm infants at 12 and 24 months of age, but their effects are not sustainable until the school age [19].

Among these reviews, only Symington et al. examined effects of developmental interventions during NICU setting; however, they only searched for interventions such as control of external stimuli (vestibular, auditory, visual, tactile), clustering of nursery care activities, positioning or swaddling of the preterm infant and individual strategies such as the ‘Newborn Individualized Developmental Care and Assessment Program’ (NIDCAP), and did not use an extensive search method [13, 21].

The results of the two reviews by Symington et al. indicate that there is very limited evidence that NIDCAP have positive long-term effects on the behavior and movement of preterm children at 5 years corrected age, but there is no effect on their cognition at 5 years corrected age. Also, other individualized developmental care interventions have demonstrated some effects in enhancing neuro-developmental outcome [13].

In addition, in other systematic reviews, the developmental outcomes of post-NICU discharge interventions have been addressed [15‐20].

Various assessment tools are used to assess the effectiveness of developmental outcomes in infants and children. These tools have different accuracy and validity, and there is concept that these tools may not have enough sensitivity to identify or monitor the improvement of minor problems as long-term neuro-developmental outcomes [15, 22].

The Bayley Scales of Infant Development (BSID) [23‐25] is the best measure for the assessment of infants and the most widely used measure to assess developmental improvement. BSID is frequently viewed as the end point of follow-up in high-risk infants [26]. The BSID-I/II includes two scales, including the Mental Developmental Index (MDI) and Psychomotor Developmental Index (PDI), The Bayley-III comprises three scales, including a cognitive, language, and motor Composite.

When an intervention begins for infants at risk of developmental disorders, the intervention has a preventive focus and has strategies to minimize developmental complications. Thus, it is important for the care provider to assess the effectiveness of these programs in at risk infants. Before being able to support this developmental care as one of the care goals in patients we need evidence to show that the developmental care has proper effects on short- and long-term developmental outcomes. However, the effectiveness of these developmental care programs in preterm infants in the NICU has not been fully approved.

Considering the importance of developmental care on reduction of neuro-developmental disorders in preterm infants, and various and more precise clinical trials conducted in this field in recent years, we decided to investigate the effect of developmental care in NICU setting on first 2 years of mental and motor development of preterm children by a systematic review, with an analysis of the risk of bias and an extensive search method.

In this review, developmental care was considered as a program that begins in the NICU with the aim of support and promoting development, and we analyzed papers that tracked the intervention effects of various types of developmental care on the neonatal development, examining them with different editions of the BSID.

The aim of this systematic review was to assess the effects of developmental care in NICU setting on mental and motor development of preterm neonates at 6–12 and 13–24 months of age. The interventions in this review included a wide range of interventions and were also very different from one another, so the results could not be combined for an estimate of overall effectiveness.

To our knowledge, this is the first meta-analysis of NICU developmental care interventions for premature infants with a rigorous search strategy, review methodology, inclusion of Randomized controlled trials and a broad range of developmental care and the unique methods of measurement outcomes. Using GRADE with low risk of bias studies, we evaluated the certainty of the evidence to be high-to-moderate for described outcomes in meta-analysis.

MDI and PDI of preterm infants at 12 months of age were significantly better in developmental care group with moderate effect size. Furthermore, MDI at 24 months of age were better with minimal effect size and no statistically significant result. PDI at 24 months of age were also significantly better with minimal effect size.

The span of this review is 46 years, with tremendous changes in the management and care of preterm neonates in the NICU and changes in survival rate of them. Therefore, we used subgroup analysis by intervention date prior and after 2000 year to decrease the clinical heterogeneity.

Furthermore, developmental interventions were highly variable. Therefore, we performed a subgroup analysis by intervention type to minimize this effect. However because of the small number of studies, we could not have divided them in more similar interventions. In addition, we chose only studies that did the developmental care in NICU setting and excluded studies that implemented the developmental care in NICU and Post NICU setting to compare more similar interventions with regard to duration of the interventions.

In subgroup analysis of MDI at 12 months of age, heterogeneity decreased with intervention date, environmental intervention type, and delivery and assessment time. Furthermore, MDI at 12 months of age were better in subgroup analyses with regard to the difference in effect size, intervention dates prior to 2000 year, NIDCAP intervention type, 9 month assessment time with large effect size, and MDI at 12 months of age were better in subgroup analyses by nurse/mother delivery method, birth weight ≤ 1250 g, gestational age ≤ 28 w, with medium effect size in intervention groups.

In subgroup analyses of PDI at 12 months of age, heterogeneity decreased with environmental and other intervention type, and environmental intervention delivery.

In subgroup analyses of PDI at 12 months of age, with regard to the difference in effect size, PDI of studies with low risk of bias, intervention dates prior to 2000 year, NIDCAP, nurse and nurse/mother delivery method and 9 months assessment time were better with medium effect size in intervention groups. In addition, PDI of studies with gestational age > 28 weeks and other intervention types were better with small effect size in intervention groups.

In intervention type, the NIDCAP at 12 months of age, MDI (SMD 0.83, [CI] 0.39–1.32) and PDI (SMD 0.52, [CI] 0.14–0.90) have large and medium effect size respectively on developmental outcomes, which was different with Symington and Pinelli, who concluded that the neuro-developmental outcome results (up to 12 months corrected age) of the NIDCAP trials were conflicting with respect to benefit [13].

In subgroup analyses of MDI at 24 months of age, with regard to the difference in effect size, intervention dates prior to 2000 year, with medium effect size have difference in intervention group. MDI at 24 months of age were also better in subgroup analyses by Birth weight ≤ 1250 g, gestational age ≤ 28 w with small effect size in intervention groups.

In subgroup analyses of PDI at 24 months of age, PDI were better in low risk and other intervention type’s method with small effect size.

With regard to the magnitude and precision of relation, the intervention date and assessment time may be an important factor in assessing intervention effectiveness. The results of studies that occurred before 2000, at 12 months of age, [MDI (SMD 0.91, [CI] 0.63–1.19) and PDI (SMD 0.47, [CI] 0.07–0.87)]; and at 24 months of age, [MDI (SMD 0.53, [CI] 0.12 to 0.93)] were significantly better. The better outcome may be due to the lower survival rate of very premature and VLBWs, and healthier preterm neonates who survived before 2000; therefore, the developmental care in NICU could have more effect on those healthier preterm neonates compared to more fragile preterm neonates after 2000 [15]. Furthermore, routine cares in NICUs are better in recent years and they obviously differ across studies; therefore, pooling results across periods may result in comparing different groups of infants with respect to outcomes. As more recent studies are published it may be possible to group studies according to their time of perinatal care. Also MDI (SMD 0.85, [CI] 0.45–1.25) and PDI (SMD 0.53, [CI] 0.15–0.90) decreased after 9 months of age. This may be due to the increased effects of social and environmental variables with increasing age on development, or perhaps the effect of intervention is more apparent at 9 months of age. Subgroup analyses at 18 and 24 months of age on MDI and PDI were not different according to the effect size.

Previous systematic reviews that investigated the effects of developmental care during NICU admission [13, 24] concluded that there is limited evidence of the long-term positive effect of NIDCAP on behavior and movement of preterm children at 5 years corrected age, but there is no effect on their cognition. Other individualized developmental care interventions have also demonstrated some effect in enhancing neurodevelopmental outcome [13].

For publication bias assessment, Egger’s test showed that there was publication bias in MDI at 12 months of age. The publication bias may be due to small number of studies.

Meta-analysis was limited in this review due to the large variation in interventions and limited number of randomized trials that were included in each intervention category. A limitation with developmental care trials was that it is not possible to mask the recipient of the intervention (in this case the mother and infant) or the person applying the intervention, unless the study includes a comparison group getting another intervention rather than no treatment or by cluster RCTs, that we have no cluster RCTs in this study .

Also, there are limitations to meta-analysis, particularly if there is evidence of publication bias, if the outcomes are too dissimilar, or if the studies are at risk of bias. We have tried to minimize these difficulties by representing that there was little evidence of publication bias, by matching comparable outcomes within narrow age ranges and unique assessment tool, and by evaluating the quality of the studies and highlighting the higher-quality studies.

In this review, BSID were considered for developmental assessment. BSID is a comprehensive developmental measure with high sensitivity to identify or monitor the major improvement of long-term neurodevelopmental outcomes and do not exactly assess minor problems [16]. Use BSID could decrease the heterogeneity of studies, but it may limit the number of studies reviewed.

Our primary outcome was the long-term outcome of developmental care in NICU setting. However, the deficiency of available data above 24 months old limits the capacity to compare results between the studies. Diversity and difficulty in identifying the most effective interventions in this review have been emphasized. It is important to aspects of the programs that are mostly useful, and neonates and families who benefit from such programs, more effectively targeting interventions. Long-term benefits for the child and family should also be considered during the intervention. Future studies may include assessments that evaluate functional outcomes, such as educational, behavioral and social problems.

In this review, we highlighted that NICU developmental care could have a significant effect on mental and motor development of preterm neonates, especially at 9–12 months of age.

The strengths of this systematic review were a rigorous search strategy, review methodology, inclusion of a broad range of developmental care interventions, and the unique methods of measurement outcomes.

Because of high cost, many low and middle income countries (LMIC) cannot implement post NICU interventions; thus, more research is needed for understanding the actual benefit of the administration of developmental care in NICU setting in LMIC for prevention of developmental morbidity in preterm newborns.