Surveillance protocol for the detection of intracranial abnormalities in premature neonates

Objective: To determine the optimal timing of cranial ultrasound scans (USs ) for identifying preterm neonates weighing less than 1500 g at birth who d evelop intracranial complications of prematurity. Design/Setting: Observational study at an urban county hospital. Methods: Serial USs from neonates with less than 1500-g:birth weight (BW) a dmitted to the neonatal intensive care unit between January 1995 and Decemb er 1996 were reviewed by a pediatric neuroradiologist in a blinded random m anner. Results: Two hundred forty-eight neonates (78%) underwent at least 3 USs, 3 2 (10%) had 2 USs and 37 (12%) only 1 US. The initial US was normal in 156 neonates (49%) and abnormal in 161 (57%). The principal abnormalities inclu ded intraventricular hemorrhage (IVH) (n=74), periventricular echogenicity (PVE) (n=68), ventriculomegaly (n= 7), and solitary cysts (n= 9). Severe IV H (n = 17) occurred in 13 (11.44%) of 114 neonates at less than 1000-g EW a nd 4 (5%) of 79 neonates of BW 1000 to 1250 g. In 11 cases (65%), the sever e IVH was clinically unsuspected For neonates weighing less than 1000 g, IV H was diagnosed by days 3 to 5 in 10 (77%) of 13, by days 10 to 14 in II (8 4%) of 13, and by day 28 in all neonates; for neonates 1001 to 1250 g, IVH was diagnosed in 1 (24%) of 4 by days 3 to 5, 2 (50%) of 4 by days 10 to 14 , and 3 (75%) of 4 by day 28. One infant's condition was diagnosed on routi ne US before discharge from the hospital. Cystic periventricular leukomalac ia (PVL) was noted in neonates; in 4 of the 9 cases. increased PVE was pres ent on the initial US and cyst formation was obvious by the second US. For 4 neonates (3 with BW <1000 g), all routine USs were negative and cystic PV L was noted on the predischarge US in these cases. Nonobstructive ventricul omegaly in the absence of IVH or cystic PVL. was observed in 14 neonates. I n 6, it was noted on the initial screening US; in 4 of the cases, it evolve d after the third screening US. Two hundred fifty-six neonates had a US bef ore discharge from the hospital, 181 (72%) were normal and 75 (28%) abnorma l. Nine significant lesions were identified by the US before discharge from the hospital (ie, severe IVH [n= 1], cystic PVL [n=4],and ventriculomegaly [n=4]). Conclusions: The following screening protocol is recommended: (1) Neonates of less than 1000-g BW: initial US on days 3 to 5 (should identify at least 75% of cases of IVH and same PVE abnormalities); second US on days 10 to 1 4 (should detect at least 84% of IVH and identify early hydrocephalus and e arly cyst formation), third scan on day 28 (should detect all cases of IVH, as well as assess PVE and ventricular size); and final scan before dischar ge from the hospital (should detect approximately 20% of significant late-o nset lesions). (2) Neonates of 1000- to 1250-g BW: initial US at days 3 to 5 (should detect at least 40% of significant abnormalities); a second scan at day 28 (should detect at least 70% of significant abnormalities); and a predischarge scan (should detect all late-onset significant lesions). (3) N eonates of 1251- to 1500-g BW: an initial scan at days 3 to 5; and a second scan before discharge from the hospital if the clinical course is complica ted.