Machine Learning for Development of an Expert System to Support Nurses’ Assessment of Preterm Birth Risk

Abstract

Early and accurate detection and treatment of preterm labor can prolong gestation with improved outcomes for both the infant and family. Review of the literature provided both theoretical and empirical support for research and development of an expert system to provide decision support for nurses’ assessment of preterm labor risk. Determining preterm labor risk in pregnant women and making decisions about interventions remain problematic in the clinical setting [1]. The problems related to preterm labor risk assessment include a poorly defined and complex knowledge base. The plethora of information about preterm labor risks remains disorganized and of little guidance to patients and providers of prenatal care. Review of the literature found no conceptual or theoretical models of preterm labor risk, which may account for poor reliability and validity of existing manual screening techniques [2]. Existing preterm labor risk screening instruments include factors that are not valid predictors of preterm labor and delivery risk, and fail to include factors reported in the literature that may be valid predictors of preterm labor [3,4,5]. Although existing instruments are only about 44% accurate, and are not adequately predictive of preterm delivery, current preterm labor prevention programs use these invalid, unreliable tools to intervene with pregnant women on a daily basis.