Finding undiagnosed patients with hepatitis C infection: an application of artificial intelligence to patient claims data.

摘要: Hepatitis C virus (HCV) remains a significant public health challenge with approximately half of the infected population untreated and undiagnosed. In this retrospective study, predictive models were developed to identify undiagnosed HCV patients using longitudinal medical claims linked prescription data from ten million in United States (US) between 2010 2016. Features capturing information on demographics, risk factors, symptoms, treatments procedures relevant extracted patients' history. Predictive algorithms based logistic regression, random forests, gradient boosted trees stacked ensemble. Descriptive analysis indicated that exhibited known symptoms average 2-3 years prior their diagnosis. The precision was at least 95% for all low levels recall (10%). For >50%, ensemble performed best 97% compared 87% just 31% regression. context, Center Disease Control recommends screening an at-risk sub-population estimated prevalence 2.23%. artificial intelligence (AI) algorithm presented here has which is substantially higher than rates associated recommended clinical guidelines, suggesting AI have potential provide step change effectiveness screening.