Machine Learning in the Maternity Ward
The Challenge: Outsmarting the Odds
More than 8% of babies in the US are born at low birthweight each year, resulting in approximately $13.4B in annual neonatal intensive care costs and often-significant impacts to quality of life and long-term costs of care. Though a number of risk factors have been identified as contributors to low birthweight, prenatal care is known to have a big impact on both the cause and solution to the issue. With this data point in mind, one of the nation’s largest managed care providers came to Amitech for an intelligent Machine Learning technology solution capable of improving outcomes and cutting costs for this high-risk population.
The Solution: Building a Better Model
Based on a historical analysis of the existing program data and the inclusion of more than 100 additional data points about each pregnancy already available to the organization but ignored by the existing risk model, Amitech built, trained and tested a model able to effectively tease out the most accurate prediction possible while also identifying the factors with the most influence on the probability of a low-birthweight delivery.
Want to know how it worked out? Check out the full case study below for additional details and specific results!