Internship project: Identifying optimal measures for hospital comparison

Opportunity

This project’s goal is to understand the best measures for comparison of hospitals. This is important because achieving fair comparisons across hospitals isn’t straightforward, as different healthcare providers have diverse contexts and deal with a wide range of patients of varying complexities. The project will identify the best measures for hospital comparison through a review of the literature and a consultation process with subject matter experts. The identified measures will then be built and validated using an available clinical data set. Unsupervised learning/clustering models will be applied to group hospitals into peer groups. Groupings will be labelled in collaboration with subject matter experts. There may be an opportunity to test the groupings with client facing teams of the industry partner and develop ‘nearest neighbour’ measures for each hospital using victor distance/magnitude measures.

Project Objectives

• Understand the best measures for comparison of hospitals by triangulating subject matter expert opinion with literature.
• Identify comparison measures using an available clinical data set.
• Validation comparison measures using an available clinical data set.