Discovering Key Topics in Emergency Medical Dispatch from Free Text Dispatcher Observations

Abstract

The objective of this work was to discover key topics latent in free text dispatcher observations registered during emergency medical calls. We used a total of 1374931 independent retrospective cases from the Valencian emergency medical dispatch service in Spain, from 2014 to 2019. Text fields were preprocessed to reduce vocabulary size and filter noise, removing accent and punctuation marks, along with uninformative and infrequent words. Key topics were inferred from the multinomial probabilities over words conditioned on each topic from a Latent Dirichlet Allocation model, trained following an online mini-batch variational approach. The optimal number of topics was set analyzing the values of a topic coherence measure, based on the normalized pointwise mutual information, across multiple validation K-folds. Our results support the presence of 15 key topics latent in free text dispatcher observations, related with: ambulance request; chest pain and heart attack; respiratory distress; head falls and blows; fever, chills, vomiting and diarrhea; heart failure; syncope; limb injuries; public service body request; thoracic and abdominal pain; stroke and blood pressure abnormalities; pill intake; diabetes; bleeding; consciousness. The discovery of these topics implies the automatic characterization of a huge volume of complex unstructured data containing relevant information linked to emergency medical call incidents. Hence, results from this work could lead to the update of structured emergency triage algorithms to directly include this latent information in the triage process, resulting in a positive impact in patient wellbeing and health services sustainability.

Deeep Continual Multimodal Multitask Modells for Out-of-Hospital Emergency Medical Call Incidents Triage Support in the Presence of Dataset Shifts

“…The contents of this chapter were published in the conference paper (Ferri et al, 2022a)-thesis contributions C2 and P3.…”

“…Consequently, the latent information contained within these data, including hidden statistical patterns, is not considered to improve triage protocols. Moreover, a substantial portion of this data is in the form of unstructured information, which cannot be automatically processed by current triage protocols (Ferri et al, 2022a;Tollinton et al, 2020). Therefore, an alternative approach is needed to complement the limitations of existing triage protocols and enhance EMD processes.…”

“…Building upon the significance of text features highlighted in previous studies (Ferri et al, 2021(Ferri et al, , 2022a, our study has focused on developing Continual Learning pipelines to ensure consistent decision support for the prediction of life-threatening levels using free text dispatcher observations. The primary objective of this work is to investigate the extent to which various Continual Learning strategies enable lifelong adaptation of deep triage models over time.…”