UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
Early neurological delopment, growth and nutrition in very preterm infants
Maas, Y.G.H.
Publication date
1999
Link to publication
Citation for published version (APA):
Maas, Y. G. H. (1999). Early neurological delopment, growth and nutrition in very preterm
infants.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
CHAPTER 7
Predictive value of neonatal neurological tests for 2 years
developmental outcome of very preterm infants
Yolanda G.H. Maas, Majid Mirmiran, Augustinus A.M. Hart, Janna G. Koppe and Henk Spekreijse
7.1 Abstract 7.2 Introduction 7.3 Subjects and methods
7.3.1 Subjects
7.3.2 Cranial ultrasound findings 7.3.3 Neurologic examination at term
7.3.4 Quality of general movements at corrected term age 7.3.5 Quality of behavioural states at corrected term age 7.3.6 Neurologic examination at 2 years corrected age 7.3.7 Statistical analysis
7.4 Results 7.5 Discussion 7.6 References
Revised version under consideration of publication, submitted 10 September 1999, Journal of Pediatrics
127
Predictive value of neonatal neurological tests
for 2 years developmental outcome of very preterm infants
Yolanda G.H. Maas', Majid Mirmiran3, Augustinus A.M. Hart2, Janna G. Koppe' and
Henk Spekreijse4
'Department of Neonatology and
department of Clinical Epidemiology and Biostatistics
Academical Medical Center, University of Amsterdam, Emma Childrens' Hospital Netherlands Institute for Brain Research
4The Netherlands Ophthalmic Research Institute and Laboratory of Medical Physics
7.1 Abstract
Background There is a need in neonatology to identify as early as possible infants who are
at risk of long-term neurological morbidity.
Methods To predict neurodevelopmental outcome of very low birth weight preterm infants
(<30w gestational age) in a population of 100 infants, we have used several neonatal and neurobehavioural tests. These tests include: cranial ultrasound, Prechtl neurological test, quality of spontaneous general movements and quality of sleep-wake organization.
Results The Prechtl neurological test at corrected term age is the best predictor of
neurological and Bayley tests at 2 years corrected age.
Conclusions The Prechtl score at term corrected age and cranial ultrasounds both have high
specificity but the Prechtl's score has better overall predictive power. Neurobehavioural tests combined with new advancements in quantitative magnetic resonance imaging may improve the ability to prognosticate outcome.
7.2 Introduction
Although there has been a significant improvement in the survival of very low birth weight preterm infants less than 30 weeks gestation with neonatal intensive care, there continues to be a higher incidence of long term neurological abnormalities. (1-4). There is a need for dependable diagnostic test(s) to identify infants at high risk for long term neurological morbidity. Early identification would allow earlier intervention and possible improvement
in outcome. This investigation was done at Newborn Intensive Care Unit at the Academic Medical Center Hospital at the University of Amsterdam. One hundred preterm infants
<30 weeks gestation at birth were evaluated by cranial ultrasound, a Prechtl neurologic test at corrected term age (5), quality of general movement (6-10) and quality of behavioural state development (i.e., the degree of sleep wake organization) (11-14). The purpose of the study was to determine which test(s) at corrected term age would predict the neurological and developmental outcome of these preterm infants at 2 years of age.
7.3 Subjects and methods
7.3.1 Subjects
The infants in this study are a subset of a parent study which had 200 infants who were less than 30 weeks gestation at birth . The results of this double blind prospective controlled trial on the effects of thyroxine supplementation on neurologic development have been published (15).
The consent was approved by the Medical Ethical Committee of the Academic Medical Center, Amsterdam. Infants born at a gestational age of less than 30 weeks, who were admitted to the Intensive Care Unit of the Academic Medical Center were enrolled if informed consent from at least one parent was obtained. Gestational age was determined by the maternal menstrual history and was confirmed by an ultrasound examination during early pregnancy and/or by physical examination using the Dubowitz score (16). Infants were excluded if they had a congenital abnormality which would affect growth or for history of illicit maternal drug abuse or endocrinologie disease.
Since the period of this investigation was two years of 2.5 years(1991-1993), one hundred and seventy nine (179) infants were eligible for this portion of the study. Of these, 50 infants died, 47 prior to term age and 3 before 2 years CA. Parents refused consent for 16 infants and consent was not obtained in 13 infants due to language problems. Thus 100 infants were available for study. One infant did not have a cranial ultrasound, 6 did not have a neurological optimality score, all had a Prechtl neurological test at term. Fifty five infants were recorded for behavioural states and 80 for GM quality at corrected term age. Seven infants were not assessed for GM quality and behavioural states, one infant was transferred at 6 days of age, one too critically ill to test in the first postnatal weeks and five
Predictive value of preterm infants ' neurobehavioural tests 129
due to scheduling problems.
Clinical data for these infants until discharge are shown in table 1.
Table 7.1 Clinical data until discharge (n= 100).
gestation (w + d), mean + SD 28 + 9
gestation below 27 weeks n 23 birth weight (g), mean ± SD 1072 ± 227
sex, male/female 45/55 birth weight < 10th centile n 10
antenatal glucocorticoids treatment 71 dexamethasone therapy n (%) 2 APGAR score at 5', mean ± SD 8.4 ± 1.7
intubation at birth n 32 respiratory distress syndrome n 50
surfactant rescue therapy n 31 intrauterine infection n* 7 cranial ultrasound results
Normal 44 Moderately abnormal 43
Severely abnormal 12 * proven by positive bacterial culture
7.3.2 Cranial ultrasound findings
Cranial ultrasounds were done within 24 hours after birth and at 5, 14, 28 and 42 post natal days (more often if clinically indicated). The scoring of the head ultrasounds for
hemmorrhage, ischemia and ventriculomegaly are described in detail in the earlier publication (15). Haemorrhage was scored according to Volpe (17); ischemic lesions were classified based on de Vries et al (18); classification of ventriculomegaly was performed according to Levene (19). Three clinicians (see acknowledgement), blinded to results of the
neonatal tests and outcome results at 2 years, scored the cranial ultrasounds. Normal: no haemorrhage, no ischemia and no ventriculomegaly; moderately abnormal: a grade 1 or 2 hemorrhage and/or grade 1 ischemia and/or grade 1 ventriculomegaly; severly abnormal: grade 3 or 4 hemorrhage and/or grade 2 or 3 ischemia and/or grade 2 ventriculomegaly..
7.3.3 Neurologic examination at corrected term age
Neurodevelopment was assessed using the method described by Prechtl (5) and the Neurological Optimality Score (NOS) as described by Touwen et. al. (20). The Prechtl neurodevelopmental test was scored as normal, suspect or abnormal, using the classification as described by Jurgens-van der Zee et al. (21).
7.3.4 Quality of general movements at corrected term age
At term age a videotape recording of at least 15 minutes was made with infants dressed in a diaper in the supine or semi-lateral position under a radiant warmer. All recordings were done in the hospital between noon and 6 PM after the infant was fed. Video tapes were replayed and three representative general movements were selected from each recording session which excluded periods of crying, fussing and sucking. The quality of the GMs was then assessed from these tapes by two independent trained observers (YGHM and MHA). The movements were scored using the global (normal, suspect and abnormal ) and the Ferrari methods (6,7).
7.3.5 Quality of Behavioural States at corrected term age
For infants who were discharged prior to 38 weeks PMA, the parents were asked to bring their infant back for a 2-hour observation at term age (40 ± 2 weeks PMA). If the parents agreed, the observations, recording procedures and assessment of behavioural states were done as described in the literature (11,12).
Five "behavioural states" were identified and behaviour which did not meet the state criteria became "indeterminate sleep" (IS). Longitudinal studies of behavioural states in preterm infants from our research group as well as others have shown that IS is highly correlated with neurological maturation (13,14). Therefore, we used the amount of time spent in IS at corrected term age as another indicator of neurological maturity in this study.
Predictive value of preterm infants ' neurobehavioural tests 131
7.3.6 Neurologic examination and Bayley at 2 years corrected age
At the corrected age of 24 months neurodevelopment was assessed according to Hempel/-Touwen (22) and scored as normal, suspect or abnormal. Abnormal development was defined as severe abnormality of tone, posture and movement leading to functional impairment and/or a delay in motor development. A suspect neurodevelopmental outcome was defined as a moderate abnormality of tone, posture and movement leading to only minor functional impairment or a minor developmental delay. Developmental assessments were also carried out at the corrected age of 24 months. The Bayley Mental Developmental Index (MDI), the Non-Verbal Developmental Index (NVDI, a mental sub-score excluding the speech and hearing items) and the Psychomotor Developmental Index (PDI) were determined according to the Dutch norms for the Bayley Scales (15,23,24).
Neurodevelopment was assessed by a pediatrician (AGW*) at 24 months CA. Mental and motor development at 24 months were assessed by a developmental psychologist (JMB*) for all infants. Both examiners were unaware of the behavioural states, GM quality and clinical data of the infants. The Prechtl test and the NOS were done by AGW* and HSH* who were also unaware of other test results and clinical data (*see ackowledgements).
7.3.7 Statistical analysis
The predictive value of neonatal cranial ultrasounds, behavioural states, quality of GMs, neurological examination at term according to Prechtl was assessed for agreement with the neurological outcome according to Touwen/Hempel at two years CA. A polychotomous logistic regression with backward elimination was performed using the cumulative proportional odds model (25). The predictive value of these neonatal tests for agreement with the Bayley Mental Developmental Index (MDI), the Non-Verbal Developmental Index (NVDI) and the Psychomotor Developmental Index (PDI) outcomes at two years CA was evaluated using analysis of covariance (25). The exclusion P-value limit was 0.05. We were primarily interested in determining whether the amount of indeterminate state (IS)
(minutes), the GM score (normal, mildly and definitely abnormal), the Prechtl score (normal, suspect and abnormal) and cranial ultrasound findings (normal, moderately and severely abnormal) had predictive power for neurological and developmental outcome at 2
years. The following clinical covariables: gestational age, gender, APGAR score at 5 minutes, surfactant rescue therapy, antenatal glucocorticoids and weight percentiles at birth, were also tested for their effect on predicting 2 year outcome. During each step of the analysis the covariables were used.
7.4 Results
At two years CA the Touwen/Hempel neurological examination classified 75 infants normal, 16 suspect and 9 abnormal. Using the Bayley scales for the MDI 74 infants were classified as normal (J> 84), 14 suspect (83-68), and 9 abnormal (< 68); for the NVDI 80,8 andl2 respectively; for the PDI 55,29 and 14, respectively. The results of the two test have differences which may be due in part to the fact that the Touwen test evaluates neurolgical function whereas the Bayley scales primarily assess mental/psychomotor development. In table 7.2 and 7.3, the results of repeated cranial ultrasounds, Prechtl neurological test, global GM quality score, Neurological Optimality Score, Ferrari optimality score and amount of IS at term compared to the Touwen/Hempel and Bayley outcome at 2 years of age are shown. Considered as single test the Prechtl neurological test and cranial ultrasounds have some predictive value for the 2 years neurological and
mental/psychomotor developmental outcome while almost none was seen for the other tests. The backward elimination polychotomous logistic regression analysis on the neurological outcome according to Touwen/Hempel (table 4) at two years of age showed that the only significant predictive test in the final step was the Prechtl score at term CA (p = 0.002). The covariables did not contribute to this predictive value. In the first step of this analysis all parameters are given equal status, however the test with the highest p value, i.e. least significant, is eliminated at each of the next steps of the analysis. A backward elimination analysis of covariance was done to evaluate the predictive value of neonatal tests and covariables with respect to the Bayley Mental Developmental Index (MDI), the Non-Verbal Developmental Index (NVDI) and the Psychomotor Developmental Index (PDI) outcome (table 7.4) at two years CA. The Prechtl score was significant at the final step analysis for MDI (p= 0.026), NVDI (p = 0.008), and PDI (p=0.007).
00 o
n
o n 3 r, CU X> Z < cl) ~a , 3 r. ca ld n <i> ft , N 3 Q 00 oo . r +1 «a
ca o Z CU e CA Ti u * c 3 CS CA Fi U i CU CU X j fa • a c cu ca O CS E ca 3e
i— er CU o 3 - O S ta < O 00 c CO ca X> 4 - -o O ob e CU C O -C A 3 oo • a c ca C/3 CU 3 O E ca • a ui ca ^^ ca E t-H O ca '5b cu u O Z o O CA O CA CU o U i >> E o o 3 CU CJ
J
"ca "ca 3 O •S O CU H. •S O CU o "3 U i , o cu U i ca ' M U o C/J U I •a CU o a UH M 3 <D 3 C/3 Ui 3 o M 3 <D 3 C/3 Ui CA ca Ó0" O M 3 <D 3 C/3 Ui £Z Z ca Cü D Z >, , , CU ca U i O ' S ca u O ' S ca u CU 00 U >% tN t-^ cu "ca CU ca S 2 ca ft ft H cu uo
CU ft E CU SS o o> O ^ s-~* ^-~* ••^ i o 6? SS SS ^o 8? SS <N SS S? SS SS s§ SS +1 Cn +1 _ l _ ^- w> -ó- V O ^D en O >n +1 Cn ^D o tN C N V-} ^t •—' ^ O (N m i r > od cn 6? t R -vo oo S^ --I —i oo 6? SR SS O CS 00 —' CN cn SR SR SR cn ™ o en cN —i T t - ^^ ("TN en ^-^ * I O T l - ^ H +1 +1+
•* >/-> rt >/-> O C N SS SR 6? O SR SR SR / - N SR SR O r- 6? O ^t- C N SR SR •>*• l~-O Ü t -6? O 0 0 r- t—i r~- r~ r--s—' v—' 6? O v—' •— en \o v-> en oo en *— ^^ 0 0 o C l en V-) ' ' •* <N es en •a e 3 O e ca Ui u <u > O H K a w os a, ca 3 O x> o5
0Û o CU Z o o ft o CU ^ - v >o oo en t-~ \o *•—' , v-i en i - H +1 +1 +1 r» •* <n u-> o C N •* en II o oo II ^^ oo CT 1 II c 3 _c -* •* II \o 1—I II en en CT 1 II ••—' s—' II II e II II ^c II II c Q Q Q e Ç „ ^ c e , ! e c l t/5 on oo ca ca c d "ca og
l a o E "03 os
+1 +1+
E u O e CU ft CA 3 CA o 3 X I ca E u O c CU ft CA 3 CA u O c x> ca E u O a CU ft CA 3 CA u O e X) ca a ca CU e 3 • ca CU S c ca cu E ci, .ts o cfl C ti -oe u i O e -o < U i CO CU -o cz> e C o c/i 3 CO CU ' 5 SP H-H CU O < e U i o c/i 3 CO CJ CU „ a ^ j o CO co Ë ja CO U i O Z CU 1— w o H-H o o T 3 c 3 O ' u o C CO Ë ü C/3 < co T 3 C/3 U C C/3 cU C cw S U i co CU o 3 ^c 4 J O O c/l #g 4 - . O O ta CU u O O c/l > 1 3 o CO O I-H 0 0 O c/l co e 0Q >> U i 1) O la o o ,—N 3 U i < er CO Ui U S U I CU C/5 O IL, J 3 ^^ r. CO X I a X I CO CO o S CO X I O o S 0 0 Z CS e CU U i O cd cd CU U i O o Ç" C/1 CU CO C M •^ >> A l "cö "co lo 'Sb
e
e U i o e U i o o O U i 3 CU O Z U i 3 CU c 'co 6S O •a £ '5b o o ca CU u "3 u - • — V a . 3 Q CJ CU co CJ Z +1 crt CU 4 3 c T 3 w ca C T 3 tu 3 Ce,
O C/3 CO U i CO t / ï O C/3 CO U i CO CU tu CJ tu o ; D T3 •3 ö c "cO M > 1 5 CO U i JU >> >-. u CO CO C Q TD 03 CU CU c o J 3 CO t-~ CU o tg u CU CU u tg3 . o
vo i/-i vo en v o Tj- o oo vo .- £ CS o\ o o Ov ,—1 t N ^ H CS CS CS ~* C S CS ~* CS +1 CS CS a\ +1 +1 +1 +1 +1 +1 +1 +1 +1 en en VO U I CS CS V) O O t - CS O N 0 0 r- O s OO f - OV 0 0 o o >/-> r- r~ vo r-< t -C S C S o o f - V O oo V O V O V O 0 0 r-es r- vo m vo VO en •st en CS c- m CS r- CS O es VO ^ cs es cs C S l/-) es ON +1 +| +| o V O +1 +1 o O +1 CS +1 CS +1 ON O O +1 en •<t CS ON ON cs~ ON 0 0 m O s ON c5T en 0 -* en II ON 0 O II f N 0 0 II •* t i - II VO ' — 1 II en en II ,—^ II ll c II II 13 II II e _G e G ^ H c e , e O, ^^ u •>— CO CO C0 "co H O CU O Hs
Ui O •3e
0 CU 0s
Ui O l a S *—> 0 CU 0 0 3 O C/1 Ui Ui CZ) ä u e« E3 u C/3 c CO O 3 X 5 0 3 - O 0 3 X I CU C c/J C3 C C/l CO C C/l co CU '—| — 1 U e 3 0 co C/2 u CO U i > ri=J O 3 _ l K U w O H ca 3 CTa
X , çj _9 o a b^ 0 0 o •Of o C S t-H es ^t v-i ^^ es cs 0 es O t - ^o- CS 0 0 T t 0 en en es cs ON 0 0 +1 +1 +1 +1 +1 0 >o 0 +1 >n +1 +1 m cs +1 en es « 0 0 0 0 en O r-~ •* 0 t -en es , 0 ^H ^H a\ ^^ OV 0 0 *^ ^ a\ ö 0 0 o 0 0 vo ö e o co Ui cO CJ O H CU U i O 3 CO CJ 0 >> '5b 0 CO O Uie
3 \ 2 CJ 0z
0 CU P H > ï s u3 H ca u fci CU 'n CU D H •0 O '-' VO es en es' en •«cfO l 0 ) I-I O > r i c j e o O •r"! CO ca B ^ Ë o S cf-l ^ •a O i-j ca A £5. fi CL) u , Cu B CU cu O) 33 73 G fi 1 1 !S T ) i-I n U H o o . co H o .2 i--= •« .2 l-H •£> S ? E Si") g o o o «Ö s CO 0 ) 1 ) G 3 Ï H co o > c o a, CO >> < a U CS — B C/l t - l O Kl co 1) CO CU O C i ) 1* (1) 1 -.,_» c o CTJ 3 _^—v - 5
s
Us
O H o ,__ 43 Q >, > o Z o , CUa
s s
o co cu « û i > i -* ca CQ r-' cu i» ,r. X CS L -w os <u ;>> o •o B B U Ë CL, O CU > CU Q CQ , O C3 T f5
^fi Oa.
O ON CL, co1
H D-3 O £ CU t - , ca X >, o 'Eli 3 CU o o >* o 3 co 3 H O , ^ CL, CU ca 3 f » Ö CN i n co CL, C O 4> ^-; CL, ^ 22 Ti-co *~^ —, NO ca \o Cu voa 6
CN r -o o co CO O r-o o LD CO O NO CN O CN CN O O O N o CN ^, T £ CN N O o 0 0 r -O o o T t -C O o o o O N CN O o O NO CN CN O N OO 1/-) IT) Os •* •* -cf CO co 00 NO CN Ö r-o a\ -*• 0 0 r-o oo -*• co o O o o 0 0 NO N O ca E fi >. !_, CU CO ' c o o E - O Ë CL, O CU > CU O ca E L -"ca 3 O * ca B ca ca ~iï Ë CL, O CU > CU CU O G X 3 cao
B _ o ca B co O O fi CU > CU c o 73 O 3 CU c d E o o t/3 CU "cu 'S s o CU | 'S •o t _ ca '•3 B c C G 3 O co ca >, B ca CU 7 3 O B Ë I-, B O CU L -o B ca o CU CL, CO 3 co "ca E S. O 13 t _ o o co >, "ca 3 O * CU T 3 2 o o co >> "caJ
V i o ca X) 3 Ë O O CO o B ca o CU CL, CO 3 co '5b oo
E O - o .53 e "ca "ca l a "C - o .53 .3 Ë - C Ë 2 o "ca E 1—l ^ — V C3 CL, 'S t _ o l-H =3 \ 0 X w- Ui J l C3 cu ca o CU O <U ' o o in ia
0 0 L -o .C L-, Pu fi__z c^
M 3 £ ssa
C N O C N oo co V) 0 0 O ö ö o o o ö oo r -o oo C N o \ oo o oo r -o i n •* C3N ^-' <o o o O o o o oo C N O C O C N O N »^ C N vo C O C N C O O O o O O O C N O >T) r - C N C N C N C N C N ^ H O O C N O O O O o O co •* o co o r-~ C O C N co •* o ^ \ o WO C N O S ö o o o O O oo O o co <r> r~- ON oo O o CN C N C N vq O ö O o ' O o \D >n CN o T j - o C N " O ^J- t ~ O O co O o o o o o •* oo C N •<t o T i -C N u-> •* oo oo C O O o O o co CU1
Ë o o o co CU1
Ë n . ca co "2 'o o co CU1
Ë CU . o 'S >n ä Ui <u 0 0 co CU B CU o CU l-H CU 3 O co O CU o CJ ca co CU B CU o O o CU I-H _3 00 "w 1 - , co _3 00 B o CU o , CH B ca "ca cus
< o "ca B CUH
•o fi CU 0 0 60 O Cl-, "ca B CU co CU OO •o fi CU 0 0 'S C M < ! - i 3 co B caIn a separate analysis on the 50 children who had complete data no association was found between the Touwen/Hempel test (p=0.41) or the Bayley scales (MDI, p= 1.0; NVDI, p = 0.53; PDI, p=0.17) at two years CA and the amount of IS at term age (see table 7.2 and 7.3).
Univariate data on sensitivity (the percentage of infants with abnormal outcome at 2 years CA who had abnormal test results at term CA), specificity (the percentage of infants with normal outcome at 2 years CA who had normal test results at term CA), positive (the percentage of infants with abnormal tests at term CA, correctly predicting abnormal outcome at 2 years CA) and negative (the percentage of infants with normal tests at term CA correctly predicting normal outcome at 2 years) predictive value for cranial
ultrasounds, Prechtl neurological test at term, quality of spontaneous general movements and quality of sleep-wake organization are presented in table 7.5. Although the Prechtl test was the only neonatal test which had a significant predictive value note in table 5 that both cranial ultrasound and Prechtl test had high (>90%) specificity and negative predictive values. This means that normal results in both tests are equally predictive for normal 2 years outcome. Although negative predictive values for the quality of GM and amount of IS at term CA were also high, these neonatal tests had a low specificity and very low positive predictive value (table 7.5).
7.5 Discussion
The main purpose of this study was to investigate the value of cranial ultrasound results from the early neonatal period and several neurobehavioural tests at term CA as predictors of neurological and developmental outcome at 2 years CA. . The Prechtl score at term CA was the only test (independent of the other tests) which could predict neurological and developmental outcome at 2 years CA. Eighty four percent of all infants who scored normal for the Prechtl test at term were also normal for the Touwen/Hempel neurological
examination at 2 years of life (table 2). Only 6% of the infants that scored normal for the Prechtl test were abnormal at 2 years and of the 13 infants that scored abnormal, only 4 were considered abnormal at 2 years CA (table 2). The outcome of the Prechtl test at term CA was significantly correlated with all 3 indices (MDI, NVDI and PDI) of the Bayley test. The Prechtl test had excellent specificity and negative predictive value (table 7.5).
Predictive value of preterm infants ' neurobehavioural tests 137
Table 7.5 Sensitivity, specificity, positive and negative predictive value for the neurological tests performed (with normal and suspect outcomes both scored as normal): repeated Cranial Ultrasounds (CU), neurological score according to Prechtl (Pr), global GM quality score (GMs) and the amount of IS at term (IS) based on the outcome of the
Touwen/Hempel and the Bayley MDI, NVDI and PDI indices (Normal (> 84), Suspect (68<x<84), Abnormal (< 68)) at 24 months (CA).
CU
Touwen/Hempel MDI NVDI PDI
CU 56 18 25 29 SENSITIVITY Pr 44 25 33 29 GMs 71 60 55 55 IS 67 60 60 63 CU 92 89 90 92 SPECIFICITY Pr 90 89 90 90 GMs 53 53 52 52 IS 52 52 52 54 CU 42 17 25 36 POSITIVE Pr 31 23 31 33 PREDICTIVE VALUE GMs 13 15 15 16 IS 7.4 11 11 19 CU 95 90 90 88 NEGATIVE Pr 94 90 91 88 PREDICTIVE VALUE GMs 95 90 88 88 IS 96 93 93 89
Cranial ultrasound results (CU) obtained from examinations between 30 and 40 weeks PMA did not contribute to the predictive power of Prechtl's test at term CA in predicting the neurological and developmental outcomes at 2 years CA. The CU results were comparable to Prechtl's score for identification of infants who later tested abnormal on the Touwen/Hempel and Bayley at 2 years CA (table 7.3).
In our earlier study on the development of behavioural states in very low birth weight infants we have found a significant developmental change in the amount of time spent in IS
between 30-40 w PMA. The amount of IS declines from initial value of about 70% at 30 weeks to less than 15% by corrected term age. However using the amount of IS as a measure of developmental maturity we could not find any significant differences between normal and abnormal infants at 2 years CA.
Scoring GM quality at term CA in our study did not add to the predictive power of Prechtl's test in predicting the neurological and developmental outcomes at 2 years CA. Earlier studies have reported that the quality of GM had good predictive value (8-10). However, most of these studies examined more mature infants at birth who were either fullterm or preterm infants > 30 weeks gestational age and included infants selected for high risk. An earlier gestational age and a low percentage (< 10%) of abnormal infants in our population of preterm infants may explain why the quality of GM were less predictive of the abnormal oucome at 2 year CA. Previous studies used repeated GM assessments (weekly) until 6 months CA (10). These investigators emphasized that the quality and pre-sence or abpre-sence of fidgety movements usually seen by 2-4 months of age are the most useful indicators of neurological and development outcome(8 ,12). Since our study focused on evaluations during the neonatal to term PMA there are no data at later ages to compare with reports in the literature. GM scoring is an easy method however, it would not be practical for clinical practice if weekly assessments for several months of age were required.
In order to identify the < 10% of infants who were abnormal at 2 years CA for this study, a large population of infants would need screening. Therefore, the specificity and ease of application of any test are important issues. We found that the Prechtl test and the cranial ultrasound had less sensitivity than the quality of general movement and quality of behavioural state development but higher specificity for both neurological and
developmental outcome at 2 years CA. This means the Prechtl and the CU tests are better for identifying normal infants. The Prechtl test had the best overall positive predictive value (about 30%) for all tests scores at 2 years CA. Negative predictive value on the other hand was high for all four tests (about 90%). These results are consistent and not surprising for there was a low % of abnormal outcome in our population. This issue of the relative importance of a positive predictive value is mathematically discussed by Sanghavi from Harvard Children Hospital in a letter to the editor published in Pediatrics in response to an
Predictive value of preterm infants' neurobehavioural tests 139
article by Morgan and Adley (26) on early identification of cerebral palsy using a profile of abnormal motor patterns. He has elegantly discussed the fact that the positive predictive value exponentially declines as one moves from a highly selected high risk group to a randomly selected (representative) clinical population in the NICU. In clinical practice false positives are very important. This problem was not analyzed or could not be analyzed due to small study populations in earlier studies on GM. Inaccurate predictions of abnormal outcome would present a significant emotional and physical burden and cost for individual parents which should be avoided.
In conclusion, though Prechtl score at term CA and cranial ultrasounds both have high specificity, Prechtl's score has better overall predictive power. Neurobehavioural tests combined with new advancements in quantitative magnetic resonance imaging (27,28) may improve the ability to prognosticate outcome. We are currently testing these methods in a collaborative study between Stanford and Amsterdam Universities.
Acknowledgements
We would like to thank all participating infants and their parents for their cooperation. We are grateful to all medical and nursing staff of our neonatal department for their
contributions to the study, to DR. A. G. van Wassenaer and H. Smolders-De Haas for the Prechtl tests, to Dr. A.G. van Wassenaer and J.M. Briët for providing the data from all 2 years follow-up examinations; to Dr. J.H. Kok, Dr. B.J. Smit and Dr. P. Tamminga for scoring the cranial ultrasounds; to Dr. M. Hadders-Algra for her contribution to the analyses of GM quality and to Ronald L. Ariagno, M.D, Prof, of Pediatrics Stanford University and currently visiting Prof. At University of Amsterdam, Academic Medical Center, Dept. Medical Physics for revising the paper.
Y.G.H. Maas was financially supported by Nutricia, The Netherlands.
This report is part of a study in fulfilment of the Degree in Philosophy in Science for Y.G.H. Maas.
7.6 References
1. Volpe JJ. Neuroloy of the Newborn, 3rd ed. Philadelphia: WB Saunders Company; 1995.
2. Aziz K, Vickar DB, Sauve RS, Etches PC, Pain KS, Robertson CMT. Province-based
study of neurologic disability of children weighing 500 through 1249 grams at birth in relation to neonatal cerebral ultrasound findings. Pediatrics 1995;95:837-44.
3. Pinto-Martin JA, Riolo S, Onaan A, Holzman C, Süsser MW, Paneth N. Cranial ultrasound predition of disabling and nodisabling cerebral-palsy at age 2 in a low-birth-wight population. Pediatrics 1995;95:249-54.
4. Wilson-Costello D, Borawski E, Friedman H, Redline R, Fanaroff AA, Hack M. Perinatal correlates of cerebral palsy and other neurologic impairment among very low birth weight children. Pediatrics 1998;102:315-22.
5. Prechtl H. The neurological examination of the full term newborn infant. 2nd. Ed. Clinics in Developmental Medicine No 63. London, UK: William Heinemann medical books Ltd; 1977.
6. Ferrari F, Cioni G, Prechtl HFR. Qualitative changes of general movements in preterm infants with brain lesions. Early Hum Dev 1990;23:193-231.
7. Hadders-Algra M. The assessment of general movements is a valuable technique for the detection of brain dysfunction in young infants. A review. Acta Paediatr Suppl
1996;416:39-43.
8. Cioni G, Ferrari F, Einspieler C, Paolicelli PB, Barbani MT, Prechtl HFR. Comparison between observation of spontaneous movements and neurological examination in preterm infants. J Pediatr 1997;130:704-11.
9. Hadders-Algra, M, Klip-Van den Nieuwendijk AWJ, Martijn A, Van Eykern LA. Assessment of general movements: towards a better understanding of a sensitive method to evaluate brain function in young infants. Dev Med Child Neurol
1997;39:89-99.
10. Prechtl HFR, Einspieler C, Cioni G, Bos AF, Ferrari F, Sontheimer D. An early marker for neurological deficits after perinatal brain lesions. Lancet 1997;349:1361-3. 11. Prechtl HFR. The behavioral states of the newborn infant (a review). Brain Res
1974;76:185-212.
12. Nijhuis JG, Prechtl HFR, Martin CB, Bots RSGM. Are there behavioral states in the human fetus? Early Hum Dev 1982;6:177-95.
Predictive value of preterm infants' neurobehavioural tests 141
et al. Sleep state organization in premature infants of less than 35 weeks gestational age. Pediatr Res 1993;34:624-8.
14. Curzi Dascalova L, Mirmiran M, (eds). Manual of Methods for Recording and Analyzing Sleep-Wakefulness States in Preterm and Full-term Infants. Paris: INSERM; 1996.
15. Van Wassenaer AG, Kok JH, de Vijlder JJM, Briët JM, Smit BJ, Tamminga P, et al. Effects of thyroxine supplementation on neurological development in infants born at less than 30 weeks' gestation. N Engl J Med 1997;336:21-6.
16. Dubowitz LMS, Dubowitz V, Goldberg C. Clinical assessment of gestational age in newborn infants. J Pediatr 1970;77:1-10.
17. Volpe JJ. Neurology of the newborn. 2nd ed. Philadelphia: WB Saunders Company; 1987: p.331.
18. De Vries LS, Eken P, Dubowitz LMS. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res 1992;49:1-6.
19. Levene MI. Measurements of the lateral ventricles in preterm infants with real-time ultrasound. Arch Dis Child 1981;56:900-4.
20. Touwen BCL, Huisjes HJ, Jurgens-van der Zee AD, Bierman-van Eendenburg MEC, Smrkovsky M, Olinga AA. Obstetrical condition and neonatal neurological morbidity. An analysis with the help of the optimality concept. Early Hum Dev 1980;4/3:207-28. 21. Jurgens-van der Zee AD, Bierman-van Eendenburg MEC, Fidler VJ, Olinga AA,
Visch JH, Touwen BCL, et al. Preterm birth, growth retardation and acidemia in relation to neurological abnormality of the newborn. Early Hum Dev 1979;3/2:141-54.
22. Touwen BCL, Hempel MS, Westra LC. The development of crawling between 18 months and four years. Dev Med Child Neurol 1992;34:410-6.
23. Meulen BF van der, Smrkovsky M. De Bayley Ontwikkelings-schalen (BOS 2-30). Lisse: Swets and Zeitlinger; 1983.
24. Meulen BF van der, Smrkovsky M. De Bayley Ontwikkelings-schalen (BOS 2-30), Niet-Verbale versie. Lisse: Swets and Zeitlinger; 1987.
25. Dixon WJ, ed. BMDP statistical software manual. Berkeley, Los Angeles, Oxford: University of California Press; 1992.
26. Morgan AM, Aldag JC. Early identification of cerebral palsy using a profile of abnormal motor patterns. Pediatrics 1996;96:692-7.
27. Battin MR, Maalouf EF, Counsell SJ, Herlihy AH, Rutherford MA, Azzopardi D, et al. Magnetic resonance imaging of the brain in very preterm infants: visualization of the germinal matrix, early myelination, and cortical folding. Pediatrics
1998;101:957-62.
28. Huppi PS, Warfield S, Kikinis R, Barnes PD, Zientara GP, Jolesz FA, et al.
Quantitative magnetic resonance imaging of brain development in preterm and mature newborns. Ann Neurol 1998;43:224-35.
