The use of data-mining to support maternal and child health management

Abstract

Introduction:
Data mining (DM) is part of KDD (Knowledge Discovery in Database), which is a computational process focused on discovering new, valid and useful knowledge in databases in order to substantiate decision-making processes. Objective: The aim of the current integrative review is to investigate the potential use of DM to support maternal and child health management processes. Methods: Search process was adapted from the PRISMA method and applied to BVS, PubMed, Scopus (Elsevier) and IEEE Xplore repositories based on MeSH® terms such as “data mining” AND health AND child* OR maternal OR pregnant*. Results: Twenty-nine documents were included in the study: nine of them confirmed previous findings, ten focused on investigating models’ accuracy (lack of discussion with previous studies) and ten reported understandable, valid, new and useful results – associations between new chemical elements and ozone on asthma, between specific Australian ethnicity and stillbirth, between Indian American race and prematurity, between vaccination and prematurity, between geriatric pregnancy and low prenatal adherence, among others. Conclusion: Results have indicated DM’s potential to support maternal and child health management processes, mainly in interdisciplinary domains. A gap was identified between DM use and the effective application of its results.