Procesamiento de lenguaje natural para texto clínico en español: el caso de las listas de espera en Chile

The waiting lists not covered by the Explicit Health Guarantee Plan for new specialty consultation in Chile increased due to the effects of the SARS-CoV-2 coronavirus (COVID-19) pandemic. This represents a problem derived from the delay in the resolution and prioritization of each case. This paper aims to describe the issue of the waiting lists in the Chilean health system and present an example of the application of Natural Language Processing (NLP). Specifically, a methodology for recognizing key information in medical narratives is described. Currently, we have a set of manually annotated medical referrals in the development of the Chilean Waiting List Corpus, with a fraction of 2,000 referrals in which the annotated medical entities were automatically normalized to the Unified Medical Language System concepts using the lexicon MedLexSp. The clinical NLP Group of the Center for Mathematical Modeling of the University of Chile, and other national NLP groups, are developing several tools and resources in medicine that can be transferred to the Chilean health system to support managing clinical text in Spanish.