Cerebral palsy and newborn care: I, II, and III (1981).

Full text

Page 1

Cerebral palsy and
newborn care: I, II,
and III (1981)
Another in our series of commentaries on notable papers
from the DMCN archives. The full papers are available at
http://www.mackeith.co.uk
Kiely JL, Paneth N, Stein Z, Susser M. Cerebral palsy and newborn
care. I: Secular trends in cerebral palsy. Dev Med Child Neurol
1981; 23: 533–38.
Kiely JL, Paneth N, Stein Z, Susser M. Cerebral palsy and newborn
care. II: Mortality and neurological impairment in low-birthweight
infants. Dev Med Child Neurol 1981; 23: 650–59.
Kiely JL, Paneth N, Stein Z, Susser M. Cerebral palsy and newborn
care. III: Estimated prevalence rates of cerebral palsy under
differing rates of mortality and impairment of low-birthweight
infants. Dev Med Child Neurol 1981; 23: 801–07.
I chose these three historical papers by Kiely et al. for several
reasons. I wanted to acknowledge the New York City Colum-
bia University Gertrude Sergievsky group under Mervyn Sus-
ser,withhiswifeZenaStein,andcolleagues,NigelPanethand
JohnKielybecausetheymadesuchoutstandingcontributions
to neuroepidemiology (they still do!). They were particularly
interestedincausalpathwaysandinobtainingthebestdatato
answer important questions in neurodevelopment. I remem-
ber that my first ever seminar in the US in 1981 was to their
groupand theywereveryencouraging of this young epidemi-
ologist setting up the first population-based registers of the
cerebralpalsiesandofbirthdefectsinAustralia.
Intheearly1980stherewasconsiderableconcernaboutthe
impact of neonatal intensive care on the survival of low birth-
weight (LBW) brain-damaged newborns. In fact it was one of
themajorreasonsweestablishedourWestAustraliancerebral
palsy (CP) registry. The major question was whether more
aggressiveinterventionswhichseemedsoeffectiveinreducing
mortalitywouldresultinmoredamagedinfantssurviving.
These three papers were rigorous in summarizing the best
epidemiological studies at that time, relying heavily on the
wonderful Swedish studies of Bengt and Gudrun Hagberg
et al.1as well as other European studies. They included fol-
low-up studies from Neonatal Intensive Care Unit (NICU)
populations, the best being from Ann Stewart et al.2in Lon-
don and Pam Fitzhardinge et al.3in Toronto, and the work of
Karin Nelson and Jonas Ellenberg4from the National Insti-
tutes of Health and the famous US perinatal project.
Several of the methodological issues raised in these papers
have influenced research around the world since then includ-
ing: (1) the importance of obtaining total population data on
deaths and impairments by birthweight and linkable back to
NICU experience – they note in these papers how challeng-
ing it is to compare different follow-up data from newborn
studies which have such different antecedents (e.g. propor-
tions in- or out-born); (2) the consideration of ‘prevalence’
or ‘incidence’ rates for CP rates with the realization that true
incidence is unattainable due to deaths of those before CP
diagnosis; this led on to (3) the understanding that pathways
to preterm cerebral damage can cause death and CP and that
CP can only be ascertained if the child survives – and hence
death and CP are competing outcomes; and (4) suggesting
that various possible scenarios to explain how these path-
ways may be influenced by NICU included: (i) severely dam-
aged infants who would previously have died were surviving
to increase the rates of CP, (ii) good care may also shift
infants from impairment to non-impairment, and (iii)
increases in impairment could only be avoided if the rate of
morbidity declined more rapidly than the rate of mortality.
As deaths are more common an outcome than CP, large data
sets are needed to monitor trends precisely. And of course
population data on the patterns and trends of preterm and
LBW infants was important to ascertain the extent to which
any changes in mortality or CP might be due to falling rates
of those births at high risk.
They correctly predicted from the limited epidemiological
data theymanagedto obtain at that time, that CPratesin LBW
infants would rise as survival increased. However, they also
commented that the overall increase in survival of non-
impaired children was also considerable. Since these careful
and thoughtful reviews were published, we have better and
more data on both deaths and impairments in larger popula-
tions ofchildren with standardmethodsofascertainmentand
diagnosis which these authors also suggested was important.
We have shown that CP has increased along with increased
survival and particularly so in those under 30 weeks’ gesta-
tional age. These papers left a legacy of better data and more
rigorous analysis to enable us to continue perinatal neuro-
epidemiologicalanalysestoachievethebestoutcomes.
Fiona Stanley MD
Telethon Institute for Child Health Research, Centre for
Child Health Research, The University of Western Australia
References
1. Hagberg B. The epidemiology panorama of major neuropediatric
handicaps in Sweden. In: Apley J (Ed) Care of the handicapped
child. Clinics in Developmental medicine, No. 67. London.
S.I.M.P. with Heinemann Medical: Philadelphia: Lippincott.
1978: pp 111–124.
2. Stewart AL, Turcan DM, Rawlings G, Reynolds EOR. Prognosis
for infants weighing 1000g or less at birth. Arch Dis Child
1977; 52: 97–104.
3. Fitzhardinge PM, Kalman E, Ashby S, Pope KE. Present status of
the infant of very low birthweight treated in a referral neonatal
intensive care unit in 1974. In: Major mental handicap:
methods and costs of prevention. Ciba Foundation Symposium
59. Amsterdam: Elsevier, 1978: 139–44.
4. Nelson KB, Ellenberg JH. Epidemiology of cerebral palsy. Adv
in Neurol 1978; 19: 419–32.
886
Developmental Medicine & Child Neurology 2008, 50: 886