Paper Title
Review Of Low Birth Weight Prediction Models In Indonesia

Abstract
Low birth weight is one of the most significant contributors to the neonatal mortality rate. It has been an important public health issue in many developing countries. Several prediction models have been developed to forecast newborn delivery weight using a wide range of regression analysis and maternal and neonatal data from Indonesia. However, there has not been any comparison study to investigate the most efficient models or availability of data required by these models. This paper reviews the last five years research carried out on the prediction of delivery weight, particularly prediction for low birth weight. We also investigate whether the significant characteristics are accessible to be measured in rural settings. The results indicate that majority of the prediction models were developed based on maternal characteristics rather than fetal characteristics or the combinations of both. These prediction models were also designed based on data accessible at delivery time rather than before birth due to the lack of maternal and fetal data during the process of antenatal care. Improvement on health information systems, particularly on the quality of complete and accurate statistics on maternal and fetal data throughout pregnancy is urgently required. This can be done by regular update of these estimates to continuously improve the rate of low birth weight delivery. Index Terms�Coefficient Of Determination, Correlation Coefficient, Mean Prediction Error, Regression Analysis.