Predictive accuracy of perceived baby birth size for birth weight: a cross-sectional study from the 2016 Ethiopian Demographic and Health Survey

OBJECTIVES: The study was designed to evaluate the accuracy of maternally perceived baby birth size assessments as a measure of birth weight and examine factors influencing the accuracy of maternal size assessments. STUDY DESIGN: Cross-sectional study. SETTING: The study is based on national data from the 2016 Ethiopian Demographic and Health Survey. PARTICIPANTS: We included 1455 children who had both birth size and birth weight data. MAIN OUTCOME MEASURES: Predictive accuracy of baby birth size for low birth weight. Level of discordance between maternally perceived birth size and birth weight including factors influencing discordance. RESULTS: Mother-reported baby birth size had low sensitivity (57%) and positive predictive value (41%) to indicate low birth weight but had high specificity (89%) and negative predictive values (94%). The per cent of agreement between birth weight (<2500 g vs ≥2500 g) and maternally perceived birth size (small size vs average or above) was 86% and kappa statistics indicated a moderate level of agreement (kappa=0.41, p<0.001). Maternal age, wealth index quintile, marital status and maternal education were significant predictors of the discordance between birth size and birth weight. CONCLUSIONS: Maternal assessment of baby size at birth is an inaccurate proxy indicator of low birth weight in Ethiopia. Therefore, a mother’s recall of birth size should be used as a proxy indicator for low birth weight with caution and should take maternal characteristics into consideration.