Features of serial cranial ultrasound detected neuropathology in very preterm infants

Abstract

BACKGROUND:

The purpose of this study was to identify the pattern and factors associated with changes in cranial ultrasound (CUS) -detected findings in infants born at or less than 28 weeks of gestation.

METHODS:

We compared readings of CUS performed at the end of the first week of life and at 4–5 weeks of age. Alteration of CUS findings was classified as: (i) unchanged, if no deviation was detected (Group 1); (ii) worsening, if there were new findings (Group 2); and (iii) improvement, if there was normalization or reduction in severity (Group 3). Descriptive statistics, multivariate controlled logistic regression, and kappa (k) statistics with 95% Confidence Interval (95% CI) were reported.

RESULTS:

Among 510 studied infants, 82.3% (95% CI 78.8–85.4) were in Group 1, 10.0% (95% CI 7.7–12.9) in Group 2, and 7.7% (95% CI 5.7–10.3) in Group 3. Overall agreement between the two scans was moderate (k 0.62; 95% CI 0.55–0.69). Worsening of CUS findings was associated with neonatal morbidities independently from gestational age and birth weight. The probability for worsening of CUS findings was higher in infants with an initial diagnosis of intraventricular hemorrhage (IVH) grade 2, than in those reported as no pathology/IVH grade 1 (Odds Ratio 5.79; 95% CI 2.42–13.91) or IVH grade 3–4 (Odds Ratio 3.81; 95% CI 1.10–13.21).

CONCLUSIONS:

In very preterm born infants, the initial CUS findings in combination with neonatal morbidities can help predict the brain lesions that are seen at the end of the first month of life and could be useful in their clinical management.