Can maternal recalled birth size be used as a proxy measure of birth weight? An evaluation based on a population health survey in Oman

To evaluate the utility of maternal recalled birth size (BS) as a proxy measure for actual birth weight (BW) when BW data are missing. Data for the study come from the 2000 National Health Survey of Oman. Frequency distribution, sensitivity and specificity analysis, bivariate, and multivariate statistical techniques were used for data analysis. The BW data exhibited a moderate level of heaping on measurements ending in 0 or 5, suggesting that health personnel often round when recording. About 31 % of actual BW data were missing due to non-availability of health cards. Maternal assessment of BS was found to be closely linked to BW on an aggregate level. However, on an individual level, there was misclassification of birth weights across all categories of BS. The overall agreement between recalled BS and recorded BW was moderate (κ = 0.44). Infants with missing BW records were more likely to be low birth weight (LBW). Maternal recalled BS appeared to be a poor proxy for BW. Estimates of LBW based on maternal assessments of BS as small should be considered as an underestimate of its actual prevalence. As infants with missing BW data have different characteristics from those with recorded BW, estimates of LBW depending solely on available BW records will produce a biased prevalence. Health personnel should record actual BW without rounding. They should inform mothers of the birth weight and advise them to retain health cards for future reference.