Making IDD Visible

PROGRAM DESCRIPTION (in development)

Making IDD Visible is a multi-year project to deploy machine learning/artificial intelligence to identify people with IDD in clinical and claim data, address disparities, and build a business case for better care.

The goal is to facilitate the comprehensive identification of people with IDD through development of validated, automated algorithms using machine learning applied to a large dataset of electronic medical record, billing, and non-healthcare data.

A future phase will pilot the use of the algorithms in specific communities to address priorities such as timely diagnosis, reducing disparities in care, and producing actionable business cases for financing and payment.

BACKGROUND

Currently, many people with IDD are poorly identified – not “visible” –  in typical healthcare data such as insurance claims or electronic medical records and charts. This is likely due to a combination of factors (under-screening and diagnosis, lack of self-disclosure among adults, and loss of clinical history during care transitions, such as adolescence into adulthood.)

 These data gaps pose a foundational barrier to (1) addressing disparities in access and quality of care; (2) measuring and improving quality of care; (3) understanding true utilization, cost, and outcome patterns among people with IDD;(4) effectively targeting services and supports (some unlabeled people with IDD hence currently manifest as “high-cost, high-need” patients because their underlying needs aren’t being met); and (5) creating credible business cases for smarter investments in IDD care.