Development and initial validation of a simplified gestational age score in low birth weight newborns in India

Abstract

Background: Assessment of gestational age (GA) based on the last date of menstrual period is often inaccurate. Scores to predict GA of newborns are difficult to use particularly by health workers. An accurate, easy to use the method of assessing GA, particularly in low birth weight (LBW) babies, is needed to detect and link premature newborns to the appropriate level of care. Objectives: The objectives were (1) to develop a “simplified GA score” (SGAS) by selecting items from 3 standard GA scores that significantly predict the GA as measured by the best obstetric estimate (BOE), (2) to validate the accuracy of SGAS against the BOE, (3) to compare the accuracy of SGAS to assess the GA as compared to the accuracy of GA assessment by the New Ballard Score (NBS), and (4) to assess inter rater agreement of SGAS as compared to that of NBS. Materials and Methods: Both the development and validation studies were cross-sectional studies. In the development study, two neonatology residents trained in the use of the 3 scores assessed LBW (<2,500 g) newborns within 24 h of birth. The residents were blinded to each other’s assessment and the BOE (GA obtained from last menstrual period [LMP] and confirmed by ultrasound (USG) to be within 2 weeks of the GA ascertained by LMP). Items significantly predictive of GA in multiple regressions were included in the SGAS. In the validation study, two different neonatology residents trained in the use of SGAS assessed the same LBW newborns within 24 h of birth, blinded to each other’s assessment and the BOE. Results: In 171 LBW newborns enrolled in the development study, 4 items (Skin, breast, genitals, and posture) were selected for the SGAS. The prevalence of very preterm (<32 weeks) was similar to the actual prevalence. Agreement between the two ratters for SGAS (Cohen’s kappa 0.825) was better than that for the NBS (Cohen’s kappa 0.709). SGAS had higher positive predictive value for <32 weeks and for ?32 weeks to ?35 weeks as compared to the NBS. Conclusions: SGAS is a promising scale for assessment of GA. It needs further validation by public health nurses and community health workers of low resource settings.