Mortality and two year outcome of infants of birthweight 500-1500 g: relationship with neonatal cerebral ultrasound data

Cranial ultrasounds were performed on 218 (96%) of 227 liveborn infants of birthweight 500-1500 g delivered in the Royal Women's Hospital, Melbourne, Australia, in an 18-month period concluding in March 1982. Seventy-two (31.7%) of the children died; 28 children (38.9%) had cerebroventricular haemorrhage, 35 (48.6%) showed no bleeding and there were nine (12.5%) with no data. Paired necropsy and ultrasound data were congruent in 22 (88%) of 25 children. One hundred and forty-eight (95.5%) of 155 survivors were seen at 2 years of age. Forty-one (28%) had cerebroventricular haemorrhage; nine children (6%) had both ventricular dilatation and haemorrhage and two had ventricular dilatation alone. Apart from a marginal advance in gestation and higher number of immigrant and less educated mothers in children without cerebroventricular haemorrhage, all other perinatal, biographical and social variables between those with haemorrhage and those without were similar. The major handicap rate overall was 14.2% (21 patients). The children with cerebroventricular haemorrhage had a trend for greater prevalence of handicap and lower mean Bayley psychological scores. This was even more evident with ventricular dilatation being present. Of children with major handicap 57.1% (12/21) had normal serial ultrasound findings during their primary hospitalization. Major handicap occurred in 15% (3/20) of children with grade 1 haemorrhage, 23.5% (4/17) with grade 2 or 3 bleeds and 25% (1/4) of those with grade 4 haemorrhage. Laterality of cerebral palsy did not correlate with ultrasound findings. Ultrasound findings did not improve statistical prediction of deaths or major handicap.     

Neonatal cranial ultrasonography as predictor of 2 year outcome of very low birthweight infants

Abstract Real time ultrasound scans using an ATL 300C sector scanner with 5–7.5 MHz transducer were performed on days 1, 4, 7 and thereafter as clinically necessary on 153 consecutively discharged very low birthweight (VLBW) infants. One hundred and forty-six long-term survivors were assessed fully at 2 years. The prevalence of cerebroventricular haemorrhage (CVH) in these survivors was 34.2% (grade 1—21.2%; grade 2—4.8%; grade 3—3.4%; grade 4—4.8%), ventricular dilatation 19.9% (including 4.1% with ventriculoperitoneal shunt), and ischaemia 9%. Impairments at 2 years were classified as nil, mild, moderate, severe or multiply severe, based on the criteria of Kitchen et al. Overall, 120 infants (82.2%) were unimpaired and 6.2% had mild, 3.4% had moderate, 4.1% had severe and 4.1% had multiply severe impairment. The major factors associated with impairment were gestational age < 28 weeks, birthweight < 1000 g, vaginal delivery, respiratory distress syndrome, mechanical ventilation, pulmonary air leaks and CVH. When these factors were reanalysed in a logistic regression model for odds ratios, only CVH (P < 0.005) and birth by spontaneous vaginal delivery (P < 0.05) were significant. The prevalence of impairment was 11.4% with no CVH, 6.5% grade 1, 71% grade 2, 20.0% grade 3 and 100.0% grade 4 CVH. The sensitivity of CVH of grade 2 or greater as a screening test was 64.7% for impairment, 78.6% for cerebral palsy and 70% for severe intellectual handicap. The mean general quotient (GQ) (Griffiths) at 2 years for infants with CVH was 89.1, and 97.5 for those without CVH (P < 0.001). Although infants with ventricular dilatation had an average GQ that was 13.1 units less than those with normal ventricles, the difference was only significant in those with a CVH of grade 2–4. The study shows the sonographic diagnosis of CVH and ventricular dilatation, but not ischaemia, to be a useful adjunct in predicting impairment and intellectual performance in VLBW infants but does not replace the need for multidisciplinary follow-up.     

Neonatal cranial ultrasonography as predictor of 2 year outcome of very low birthweight infants

Real time ultrasound scans using an ATL 300C sector scanner with 5-7.5 MHz transducer were performed on days 1, 4, 7 and thereafter as clinically necessary on 153 consecutively discharged very low birthweight (VLBW) infants. One hundred and forty-six long-term survivors were assessed fully at 2 years. The prevalence of cerebroventricular haemorrhage (CVH) in these survivors was 34.2% (grade 1-21.2%; grade 2-4.8%; grade 3-3.4%; grade 4-4.8%), ventricular dilatation 19.9% (including 4.1% with ventriculoperitoneal shunt), and ischaemia 9%. Impairments at 2 years were classified as nil, mild, moderate, severe or multiply severe, based on the criteria of Kitchen et al. Overall, 120 infants (82.2%) were unimpaired and 6.2% had mild, 3.4% had moderate, 4.1% had severe and 4.1% had multiply severe impairment. The major factors associated with impairment were gestational age less than 28 weeks, birthweight less than 1000 g, vaginal delivery, respiratory distress syndrome, mechanical ventilation, pulmonary air leaks and CVH. When these factors were reanalysed in a logistic regression model for odds ratios, only CVH (P less than 0.005) and birth by spontaneous vaginal delivery (P less than 0.05) were significant. The prevalence of impairment was 11.4% with no CVH, 6.5% grade 1, 71% grade 2, 20.0% grade 3 and 100.0% grade 4 CVH. The sensitivity of CVH of grade 2 or greater as a screening test was 64.7% for impairment, 78.6% for cerebral palsy and 70% for severe intellectual handicap. The mean general quotient (GQ) (Griffiths) at 2 years for infants with CVH was 89.1, and 97.5 for those without CVH (P less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)     

Status at Two Years in 121 Very Low Birth Weight Survivors Related to Neonatal Intraventricular Haemorrhage and Mode of Delivery

ABSTRACT. All 121 infants with birthweight 1500 g (VLBW) discharged from our department through an 18-month period were followed up at two years of age. All but 10 infants were examined by cranial ultrasound scanning in the neonatal period. Six had died after discharge; and of these two had severe brain damage following neonatal intraventricular haemorrhage. Twelve children had definite neuromotor abnormality; of these, the nine children with spastic types of cerebral palsy had all been delivered vaginally. Of the remaining children, one half had at least one neurodevelopmental symptom suggesting perinatal brain damage. Outcome was associated to neonatal intraventricular/subependymal haemorrhage when complicated by ventricular dilatation. Outcome in children with simple haemorrhage was similar to that in children without haemorrhage. The association between outcome and haemorrhage was considerably reduced by simultaneously considering the associations of gestational age at birth and the use of mechanical ventilation in the neonatal period. Head circumference at follow-up was greatest in those children with haemorrhage complicated by ventricular dilatation, when adjusted for actual body weight, birth weight, and gestational age at birth.     

Status at two years in 121 very low birth weight survivors related to neonatal intraventricular haemorrhage and mode of delivery

All 121 infants with birthweight less than or equal to 1500 g (VLBW) discharged from our department through an 18-month period were followed up at two years of age. All but 10 infants were examined by cranial ultrasound scanning in the neonatal period. Six had died after discharge; and of these two had severe brain damage following neonatal intraventricular haemorrhage. Twelve children had definite neuromotor abnormality; of these, the nine children with spastic types of cerebral palsy had all been delivered vaginally. Of the remaining children, one half had at least one neurodevelopmental symptom suggesting perinatal brain damage. Outcome was associated to neonatal intraventricular/subependymal haemorrhage when complicated by ventricular dilatation. Outcome in children with simple haemorrhage was similar to that in children without haemorrhage. The association between outcome and haemorrhage was considerably reduced by simultaneously considering the associations of gestational age at birth and the use of mechanical ventilation in the neonatal period. Head circumference at follow-up was greatest in those children with haemorrhage complicated by ventricular dilatation, when adjusted for actual body weight, birth weight, and gestational age at birth.     

Neurodevelopmental outcome at 12 months of age related to cerebral ultrasound appearances of high risk preterm infants

A prospective neurological and developmental assessment at 12 months of age corrected for prematurity was performed on 54 surviving preterm infants of 34 weeks' gestation or less. The babies were allocated into three groups according to their ultrasound (US) appearances: Group I (n = 29), normal scan; Group II (n = 10), isolated periventricular-intraventricular haemorrhage (PVH); Group III (n = 15), association of PVH, periventricular leukomalacia (PVL) and ventricular dilatation. The developmental outcome evaluated with the Griffiths' development quotient (DQ) was good and similar in Groups I and II, while it was worse and variable in Group III. There was also a higher incidence of neurological abnormalities in Group III, as 47% of children only were found to be normal compared to 86% and 80% in Groups I and II, respectively. A major handicap was diagnosed in 5 children of Group III. Infants with small lesions of PVH or PVL or with ventricular dilatation developed as well as children with normal US scan, whereas more diffuse or extensive US changes of PVL had a poorer prognosis. The outcome of a cerebral injury seems to depend on the type, the size and localisation of the lesion, and to some extent, on the neuroplasticity of the developing brain.     

Periventricular leucomalacia and intraventricular haemorrhage in the preterm neonate.

Two hundred very low birthweight infants were prospectively scanned to ascertain the incidence of periventricular leucomalacia (PVL) and haemorrhage. Before collection of data, clear definitions of ultrasound abnormalities believed to represent PVL and intraventricular haemorrhage were described. These referred to small and moderate intraventricular haemorrhage, paenchymal haemorrhage, and PVL, including prolonged flare (echoes in the periventricular region lasting for two weeks or more and not becoming cystic). Sixty nine infants (34%) had no abnormality on ultrasound scans. Intraventricular haemorrhage occurred in 107 babies (37 grade I and 62 grade II), and only eight infants were thought to have true parenchymal haemorrhage. Ultrasound appearances of PVL were seen in 27 infants, 19 of whom developed cysts and eight died in the precystic stage. Prolonged flare occurred in another 25 babies. Unilateral parenchymal haemorrhage occurred in four infants who subsequently developed cystic PVL in the contralateral hemisphere. Twenty one infants developed ventricular dilatation, 12 of whom had associated parenchymal lesions. Haemorrhage, PVL, and flare occurred commonly in infants of 30 weeks'' gestation and below and became markedly less common in more mature infants. We believe prolonged flare represents a form of PVL, and in this study a total of 52 (26%) infants had an ultrasound appearance of periventricular leucomalacia, an incidence considerably higher than previously reported.     

A longitudinal study of post-haemorrhagic ventricular dilatation in the newborn.

Two hundred and two consecutive admissions to a regional neonatal unit were scanned by real-time ultrasound. Sixty-eight (34%) infants had intracranial haemorrhage, 39 (57%) of whom were scanned repeatedly until they were at least 30 days old. Fifteen infants showed some degree of ventricular dilatation. Four had transient dilatation with complete recovery without any form of treatment (group 1), 7 showed persistent but non-progressive dilatation with no treatment (group 2), 3 had rapidly progressive hydrocephalus (group 3), and 1 had cerebral atrophy (group 4). Occipitofrontal head circumference was also followed sequentially from birth and was not abnormal in groups 1 and 2, but abnormal rates of head growth were seen in groups 3 and 4. It is concluded that after intracranial haemorrhage only a small proportion of infants develop frank hydrocephalus, but ventricular dilatation of some degree is common and may require no treatment.     

A study of periventricular haemorrhage, post-haemorrhagic ventricular dilatation and periventricular leucomalacia in Chinese preterm infants

Serial cranial ultrasound scans were performed in 178 preterm Chinese infants (gestation less than 35 weeks, birthweight less than 2000 g) to study the incidence, age of onset and associating risk factors of periventricular haemorrhage (PVH), and also the occurrence of post-haemorrhagic ventricular dilatation and periventricular leucomalacia (PVL). Sixty-four infants developed haemorrhage, giving an incidence of 36%. Among infants of birthweight less than 1500 and less than 1000 g the respective incidence was 52 and 69%. Seventy-two per cent (46 of 64) of haemorrhages were initially detected within the first 3 days of life, but delayed haemorrhage occurring after 1 week of age occurred in nine infants. In eight of these infants PVH had been shortly preceded by a major clinical disaster. Eleven perinatal factors were found to be significantly associated with PVH but only systemic hypotension showed a significant independent association. Post-haemorrhagic ventricular dilatation developed in 17 (46%) of the 37 infants who survived for more than 1 month after PVH. This was transient in 41%, persistent but stable in 29% and progressive in 29%. PVL was detected in eight infants who survived the initial period following PVH.     

Effect of neonatal periventricular haemorrhage on neurodevelopmental outcome.

All 56 infants born between 23 and 28 weeks'' gestation admitted to this hospital in 1981 were examined for periventricular haemorrhage with cerebral ultrasonography. Haemorrhage was diagnosed in 34 (61%)-12 (22%) had germinal layer haemorrhage, 18 (32%) had intraventricular haemorrhage, and four (7%) had intracerebral haemorrhage. The two year outcome of survivors with and without periventricular haemorrhage was compared to determine the effect on neurodevelopment. Only three (16%) of 19 infants with normal scans or germinal layer haemorrhages had evidence of major disability but nine (75%) of 12 infants with intraventricular or intracerebral haemorrhage had major disability. The mental and psychomotor performance on the Bayley scales of infant development was also significantly worse in the latter group. All three survivors with intracerebral haemorrhage had major disability. The continuation of life support treatment for extremely preterm infants who are at very high risk of severe handicap is a matter of increasing concern in neonatal intensive care. Our results show that if extensive periventricular haemorrhage, in particular intracerebral haemorrhage, occurs in this gestational group, extreme pessimism is warranted.     

Cranial ultrasound screening of preterm and term neonates

A cranial ultrasound examination was performed between 48 and 96 h of age on 580 neonates of 25-42 weeks gestation. The incidence of cerebroventricular haemorrhage (CVH) in infants less than 32 weeks gestation was 37%, compared with an incidence of 2.7% in infants of 32 weeks or more. The incidence of CVH unselected healthy term infants (between 37 and 42 weeks) was only 1.1%. Of the 13 infants of 32 weeks or more who were found to have a haemorrhage, nine had a small (Grade I) haemorrhage (69%) and none of these infants had abnormal neurological signs in the neonatal period. The remaining four infants with Grade II, III or IV haemorrhage developed either seizures or episodes of apnoea. Two of the 13 infants of 32 weeks or more with a haemorrhage died, one during the newborn period and the other at 5.5 months of age. CVH in asymptomatic infants of 32 weeks or more gestation is uncommon and does not justify routine cranial ultrasound scanning.     

Cranial ultrasound screening of preterm and term neonates

Abstract A cranial ultrasound examination was performed between 48 and 96 h of age on 580 neonates of 25–42 weeks gestation. The incidence of cerebroventricular haemorrhage (CVH) in infants less than 32 weeks gestation was 37%, compared with an incidence of 2.7% in infants of 32 weeks or more. The incidence of CVH in unselected healthy term infants (between 37 and 42 weeks) was only 1.1%.
Of the 13 infants of 32 weeks or more who were found to have a haemorrhage, nine had a small (Grade I) haemorrhage (69%) and none of these infants had abnormal neurological signs in the neonatal period. The remaining four infants with Grade II, III or IV haemorrhage developed either seizures or episodes of apnoea. Two of the 13 infants of 32 weeks or more with a haemorrhage died, one during the newborn period and the other at 5.5 months of age.
CVH in asymptomatic infants of 32 weeks or more gestation is uncommon and does not justify routine cranial ultrasound scanning.     

Renal ultrasound findings and vesicoureteral reflux in children hospitalised with urinary tract infection

Aims: To determine the sensitivity, specificity, and predictive values of renal ultrasound findings for vesicoureteral reflux (VUR). Methods: Retrospective review of the ultrasound and voiding cystourethrogram (VCUG) results of 162 children under 5 years of age admitted with their first episode of urinary tract infection (UTI) over a two year period. Ultrasound findings were considered suggestive of VUR if "dilatation of the pelvi-calyces", "dilatation of the ureters", or "dilatation of the collecting system" of one or both kidneys was reported. Results: A total of 162 patients were eligible for inclusion (median age 85 days; 71 (44%) were female). The prevalence of VUR was 22%. Ultrasound findings were positive for VUR in 14 of 35 patients with confirmed VUR on VCUG, and positive in 30 of 127 patients without VUR on VCUG. Of 21 patients who had a normal ultrasound but showed VUR on VCUG, 14 had grade II reflux, five grade III reflux, and two grade IV reflux. The sensitivity and specificity of ultrasound in suggesting VUR were 40% and 76%, respectively. The positive predictive value of ultrasound in suggesting VUR was 32%; the negative predictive value was 82%. Conclusion: Renal ultrasound findings are neither sensitive nor specific for VUR in children with a first UTI.     

Renal ultrasound findings and vesicoureteral reflux in children hospitalised with urinary tract infection.

AIMS: To determine the sensitivity, specificity, and predictive values of renal ultrasound findings for vesicoureteral reflux (VUR). METHODS: Retrospective review of the ultrasound and voiding cystourethrogram (VCUG) results of 162 children under 5 years of age admitted with their first episode of urinary tract infection (UTI) over a two year period. Ultrasound findings were considered suggestive of VUR if "dilatation of the pelvi-calyces", "dilatation of the ureters", or "dilatation of the collecting system" of one or both kidneys was reported. RESULTS: A total of 162 patients were eligible for inclusion (median age 85 days; 71 (44%) were female). The prevalence of VUR was 22%. Ultrasound findings were positive for VUR in 14 of 35 patients with confirmed VUR on VCUG, and positive in 30 of 127 patients without VUR on VCUG. Of 21 patients who had a normal ultrasound but showed VUR on VCUG, 14 had grade II reflux, five grade III reflux, and two grade IV reflux. The sensitivity and specificity of ultrasound in suggesting VUR were 40% and 76%, respectively. The positive predictive value of ultrasound in suggesting VUR was 32%; the negative predictive value was 82%. CONCLUSION: Renal ultrasound findings are neither sensitive nor specific for VUR in children with a first UTI.     

Intraventricular hemorrhage. Incidence and outcome in a population of very-low-birth-weight infants

The incidence, extent, and outcome of germinal matrix hemorrhage-intraventricular hemorrhage (GMH-IVH) were determined with the use of ultrasound and autopsy findings in 100 consecutive infants, with a birth weight of less than 1,500 g. Serial ultrasound examinations once or twice weekly were performed with the use of a portable real-time linear-array scanner. The overall incidence of GMH-IVH was 46%. Twenty infants had grade 1 (GMH), 24 had grade 2 (IVH +/- GMH), and two had grade 3 (IVH +/- GMH with intracerebral hemorrhage) conditions. The mortality in infants with GMH-IVH was 35%, compared with 13% in infants without GMH-IVH. Although 11 (37%) of 30 survivors with GMH-IVH had ventricular dilatation, only two infants required ventriculoperitoneal shunts for progressive hydrocephalus. The incidence of GMH-IVH was increased in outborn infants, in those delivered vaginally, and in those who required mechanical ventilation, bicarbonate therapy, or volume expansion in the first 24 hours. The long-term prognostic significance of the ultrasound findings was unknown and will be determined by follow-up studies.     

Primary vesicoureteric reflux--how useful is postnatal ultrasound?

The presence or absence of pelvicalyceal dilatation on postnatal ultrasound continues to appear within diagnostic algorithms to select patients for micturating cystourethrography (MCU) in the investigation of antenatally diagnosed hydronephrosis. Postnatal ultrasound findings were assessed in a population diagnosed as having antenatal hydronephrosis due solely to primary vesicoureteric reflux (VUR) to see whether this is justified. The postnatal ultrasound and MCU findings in 177 patients with primary VUR detected as antenatal hydronephrosis were reviewed retrospectively. A total of 132 (75%) were boys. Reflux was unilateral in 103 cases and bilateral in 74 (42%). Altogether 37% of boys and 33% of girls with a renal pelvic diameter of < or = 10 mm had grade III-V VUR. Calyceal and/or ureteric dilatation had specificities of 87-96% for grade III-V VUR, but sensitivities of only 37-54%. Fifty eight per cent of male and 75% of female renal units with grade III VUR and 17% of male units with grade IV-V VUR were normal on ultrasound. Approximately 25% of ultrasonically normal renal units had grade III-V VUR on MCU. Postnatal ultrasound criteria correlate poorly with the presence and degree of VUR in children with antenatally diagnosed hydronephrosis and should not be used to direct the use of cystography.     

Short, middle and long-term outcome of major peri-intraventricular hemorrhages

Within 4 years 10 months (1981-1985), the Port-Royal Neonatal Intensive Care Unit admitted 2,400 neonates, one third with a birthweight below 1,501 g; 4,631 cranial ultrasound studies were performed in 1,488 of those neonates, mostly less than 1,501 g, detecting 392 consecutive peri-intraventricular hemorrhages (PIVH), of which 130 were major forms (from unilateral grade III to bilateral grade IV PIVH). Overall survival rates were 91% in grade I, 85% in grade II, 42% in grade III, 26% in grade IV; survival rate was significantly lower in bilateral than in unilateral grade II and III PIVH. In major PIVH, deaths occurred early (58% in the first week after birth). Post-hemorrhagic dilatation was constant but mostly regressive; true active hydrocephalus appeared in 1 unilateral grade III PIVH and 8 bilateral grade III PIVH, with ventriculo-peritoneal shunt in the second month of life in 5 infants (2 died), and 4 deaths (surgery not feasible). The neurological and developmental outcome of 42 of 46 survivors (4 losts to follow-up) was evaluated beyond one year of age in 12 unilateral grade III PIVH (10 normal children, 1 minor sequela, 1 moderate sequela), 16 bilateral grade III PIVH (7 normal children, 3 minor sequelae, 1 moderate sequela, 5 major sequelae), 13 unilateral grade IV PIVH (8 normal children, 1 minor sequela, 3 moderate sequelae, 1 major sequela), 1 bilateral grade IV PIVH (major sequela). A persistent major dilatation after 6-9 months of age bore an ominous prognosis.(ABSTRACT TRUNCATED AT 250 WORDS)     

Real-time ultrasonography in the neonate: a systematic study of a high-risk infants population

During a 12-month period a prospective and systematic study was carried out by means of portable real-time ultrasound (US) scanner in order to detect cerebral lesions in a population of high-risk neonates. Newborn infants were allocated into two groups: group A: all premature infants (n = 83) of less than or equal to 34 weeks' gestation or less and group B: neonates (n = 36) of more than 34 weeks' gestation presenting with abnormal neurological signs. Group A: the overall incidence of periventricular haemorrhage (PVH) was 47% (15 grade I, 16 grade II, 3 grade III, 1 Plexus choroid haemorrhage, 1 isolated intraventricular haemorrhage). Infants of 30 weeks or less were at highest risk to develop a PVH. The degree of severity did not depend on gestational age. Repeated scans accurately timed the onset of PVH; 67% developed a PVH within the first 24 hours of life and 31% within the first 6 hours. A post-haemorrhagic ventricular dilatation was noted in 50% of the 28 infants who survived more than 28 days (4 transient, 7 arrested and 3 rapidly progressive). Group B: 15 of 36 infants had US abnormalities. Cerebral lesions were miscellaneous. Diagnosis of PVH, leukomalacia, agenesis of corpus callosum, calcifications in the basal ganglia, hydranencephaly were made and confirmed at autopsy in 9 fatal cases. US has proved useful for the detection of cerebral lesions among high-risk newborn infants in a Neonatal Unit.     

Ultrasound measurements of the lateral ventricles in neonates: why,how and when? A systematic review

Germinal matrix‐intraventricular haemorrhage and subsequent post‐haemorrhagic ventricular dilatation (PHVD) are frequently encountered complications in preterm neonates. As progressive dilatation of the lateral ventricles may be associated with elevated intracranial pressure, ultrasound measurements of ventricular size play a major role in the evaluation of neonates at risk of ventricular dilatation as well as in assessing the effect of intervention for PHVD. A systematic search was carried out in Medline and Embase to identify neonatal and foetal ultrasound studies on lateral ventricular size. This review presents an overview of the available data concerning neonatal reference values for lateral ventricular size, the influence of gender, ventricular asymmetry and the effect of the mode of delivery on the phenomenon of ventricular reopening following birth. Conclusion: Serial cranial ultrasound measurements of the lateral ventricles play a key role in the early recognition and therapeutic evaluation of post‐haemorrhagic ventricular dilation and can be of prognostic value in neonates with ventricular dilatation.     

Timing and evolution of periventricular haemorrhage in infants weighing 1250 g or less at birth.

The brains of 50 consecutively admitted infants who weighed 1250 g or less at birth were examined with real time ultrasound. Of 30 (60%) who had periventricular haemorrhage (PVH), 19 (63%) bled on the first day and 17 (57%) showed extension of the initial haemorrhage on serial scans. The median age was 16 hours when PVH was first detected and 48 hours when PVH reached its maximum extent. Ventricular size at birth correlated with gestation. Progressive ventricular growth was seen after birth in infants both with and without PVH. Charts of normal ranges of ventricular size and head circumference were drawn up from birth to 10 weeks of age. All infants with PVH showed a transient increase in ventricular size at 2 weeks of age but most returned to normal by 6 weeks of age. Ventricular dilatation after PVH that was greater than the 95th centile for this population developed in 5 (31%) of 16 survivors, four of whom subsequently developed hydrocephalus, although none required ventriculo peritoneal shunting. The optimal timing for diagnosis with ultrasound is at the end of the first week for PVH and the second to third week for ventricular dilatation.