Intended for healthcare professionals


The outcomes of neonatal intensive care

BMJ 1995; 310 doi: (Published 18 March 1995) Cite this as: BMJ 1995;310:681
  1. Christine Sanderson,
  2. D M B Hall
  1. Lecturer in community child health Professor of community paediatrics Children's Hospital, Sheffield S10 2TH

    Let's use the results of research into early intervention

    Deform'd, unfinish'd, sent before my time Into this breathing world, scarce half made up, And that so lamely and unfashionable That dogs bark at me, as I halt by them;1

    Shakespeare identified the association between prematurity and disability 300 years before Little's classic description of spastic diplegia in 1862. Disability in surviving premature infants is still an important problem, particularly in those of extremely low birth weight.2 Most of the data on the incidence of disability and the costs and benefits of neonatal intensive care units come from a few research centres. A broader picture could be obtained by a standardised nationwide approach to the collection of data. There are two separate issues. Firstly, what are the outcomes of interest and how should they be monitored? Secondly, what proportion of childhood disability is attributable to the increased survival of babies nursed in neonatal intensive care units and how is the epidemiology of disability changing as a result?

    A recent report highlighted the inadequacies of the collection of data currently and proposed that both issues could be addressed by collecting and linking two minimum datasets.3 The first would be of neonatal data, including demographic, obstetric, and perinatal data for each neonate and additional items for infants who had been admitted to neonatal intensive care units. The second would entail evaluating every 2 year old for disability.

    Could a minimum dataset on all 2 year olds be collected during routine child health surveillance and be stored on a community child health computer? Unfortunately, obstacles exist to this apparently simple task. Firstly, child health surveillance in the 2 year age group has never been complete,4 and considerable extra investment would be needed to improve it. Secondly, in epidemiological surveillance the condition, event, or procedure in question must be defined precisely. The rigorous definition of cognitive and language impairment is notoriously difficult, particularly in 2 year olds, and assessment needs considerable skill. Reasonable reliability between observers is hard to achieve but would be essential to ensure comparable results between districts. Furthermore, although computer systems manage immunisation and biochemical screening programmes successfully, measures of child development and disability are harder to adapt to the constraints of existing computer software. Thirdly, in child health surveillance there has been a move away from routine standardised checking for “defects” in favour of a more flexible health promoting approach.5 Purchasers and providers would view reversal of this trend as regressive.

    Most of these difficulties could be overcome, but how much would it cost to do the job properly? Collecting information on every child is expensive. Two or three extra professional and clerical staff would probably be needed in each district. The cost in England and Wales might be around £5-10 million a year. This money could be spent in ways that would benefit children and their parents more directly.

    Is the second birthday the optimal age for a formal reexamination of every graduate of neonatal intensive care? In well run neonatal intensive care services with adequate follow up6 retinopathy of prematurity (the main visual deficit) and sensorineural hearing loss should be detected within the first three months, and most of the remaining severe impairments, such as cerebral palsy and severe intellectual deficits, can be found before the first birthday. With increasing knowledge about the predictive importance of neonatal encephalopathy and of imaging7 before discharge, the age at which follow up is performed can be targeted at the babies at high risk more precisely than was possible 20 years ago.

    The aim of neonatal intensive care is no longer merely survival or avoidance of severe disability but rather the preservation of normal brain function. Other indicators of good care have become more important, such as the incidence of learning deficits,8 minor movement disorders,9 behaviour problems,10 and child abuse. These problems often need additional educational support11 and probably have an adverse long term prognosis for employment—but they cannot be identified reliably at age 2 even by experts. The true outcomes and economic costs of intensive care, and the impact of disability, cannot be assessed unless progress is monitored at least until the early school years.

    How else could better monitoring be achieved? Firstly, a consensus among neonatologists on the role and function of follow up clinics for those at high risk is needed. Secondly, each neonatal intensive care unit could employ someone trained in counselling, child development, and psychological assessment to follow up its graduates. They would provide anticipatory guidance and support to the parents12 and collaborate with education departments in developing early intervention services for specific groups of infants.13 Lastly, research could be commissioned at national level, involving health, education, and social services, to improve computer systems, develop better registers of children with special needs, and simplify the sharing of information among these three key agencies. These registers could then be used more readily than at present to monitor the contribution of graduates from neonatal intensive care to the total burden of disability.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.