OBJECTIVE: The New Ballard Score (NBS) has been evaluated only until 96 h of age. We studied the validity and reliability of NBS for gestational age (GA) assessment on days 1, 5 and 7 of postnatal age (PNA). DESIGN AND SETTING: This prospective, analytical study was conducted in a level III neonatal unit. PATIENTS: Neonates born at a GA of 29-35 weeks (based on accurate last menstrual period (LMP)) were eligible. Encephalopathy, malformations, and unstable vitals were exclusion criteria. LMP-based GA was the gold standard. NBS was assessed within 24 h of birth by one rater, and two raters assessed NBS on days 5 and 7. All were blinded to LMP and one anothers' ratings. Recruitment continued until >100 subjects were enrolled with > or =25 in each LMP-based GA group: 29-30 weeks, 31-32 weeks, 33-34 weeks and 35 weeks. MAIN OUTCOME: Correlation of GA assessed on day 7 with gold standard. RESULTS: 129 neonates were studied. NBS-based GA on days 5 or 7 did not differ from the gold standard GA by more than 2 weeks in any subject. On day 7, NBS overestimated GA in 26.7% and underestimated GA in 19.8% cases; all discrepancies were < or =2 wks. Compared to gold standard GA, the intra-class correlations (ICCs) of the gold standard GA and the NBS-based GA of the day 1 rater, day 5 rater and day 7 rater were 0.94, 0.94 and 0.92, respectively. ICCs for inter-rater reliability on day 5 and day 7 were 0.97 and 0.96, respectively. Compared to the day 1 rater's raw NBS, the ICCs of day 5 and day 7 raters' total scores were 0.98 and 0.97, respectively; of day 5 and day 7 raters' neurological scores were 0.98 and 0.97; and of day 5 and day 7 raters' physical scores were 0.92 and 0.88. All ICCs mentioned above had p values <0.001. CONCLUSIONS: NBS is a valid and reliable clinical tool for GA assessment until day 7. It slightly overestimates the GA with increasing PNA. Neurological signs are more reliable than physical ones.
