For the August 2023 issue of CJEM, we collaborated with their team to present “Machine learning to identify attributes that predict patients who leave without being seen in a pediatric emergency department”1 in a visual abstract format.
A mismatch in patient load and ED resources that allow for timely care have led to increased numbers of patients who want to leave without being assessed by a physician or advanced care provider (LWBS). Reported proportions are highly variable, however some studies report this occurs up to 17% of the time2. Machine learning applications in healthcare have been expanding at a tremendous rate, with applications in the USA now able to explore and predict LWBS trends. The two strongest predictors, insurance type and status, are much less applicable to a Canadian context.
This study, by Sarty et al., created machine learning algorithms to create effective models to predict LWBS using historical patient data at a Canadian pediatric ED. These models can be applied in real-time to predict LWBS dispositions, creating an opportunity for patient-level interventions. For a summary of what the authors found, including the five most influential attributes in a Canadian machine learning model, a .pdf version of a visual abstract on the topic can be found below:
- 1.Sarty J, Fitzpatrick EA, Taghavi M, T. VanBerkel P, Hurley KF. Machine learning to identify attributes that predict patients who leave without being seen in a pediatric emergency department. Can J Emerg Med. Published online July 28, 2023:689-694. doi:10.1007/s43678-023-00545-8
- 2.Gaucher N, Bailey B, Gravel J. Who Are the Children Leaving the Emergency Department Without Being Seen by a Physician? Academic Emergency Medicine. Published online February 2011:152-157. doi:10.1111/j.1553-2712.2010.00989.x