Risk assessments/profiling based on claims and EHRs data using FHIR

Founding engineering team and F-CTO for a startup.

Challenge

Produce risk profiles for patients using existing health data (claims, patient charts), and integrate them in a clinical review process.

Solution

We built a solution that uses a FHIR server to store the data, custom code to import and process the data from various sources (CMS claims, EHRs through a data broker), and our own bonFHIR library to present the information and orchestrate the workflow. Creating the risk assessments required close collaboration of the clinical and engineering teams. We also use FHIR resources to manage the parameters for the risk assessments (medication groups, condition groups, threshold levels, etc.), enabling the clinical team to be autonomous in varying those parameters without requiring costly engineering work.

Result

A solution that produces risk assessments for patients, highlights specific conditions that would otherwise be lost in the mass of patients that clinicians need to see and creates relevant recommendations vetted by actual clinicians. The review process is fluid and the clinical team can assess the modifications required to the algorithm and autonomously tweak its parameters to refine it and increase productivity for the next round.