In the US, it is a well-known problem that the cost of healthcare is increasing at a much higher rate than the quality of care. This is partially due to the existing incentives in the system being based on quantity of services provided, rather than quality. The Accountable Care Act shifted the focus of reimbursement toward population health to reduce costs and incentivize quality for payers and providers. Our client, a large health insurance company, was interested in piloting an Accountable Care Organization, one example of a population health model. However, they had very little experience in population health, and their business operated in an entirely different way.
My team was tasked with designing and building a system and process that allowed the client to manage their ACOs, which required accessing huge amounts of data.
We developed and formalized business processes that surround the system in order to maintain ACOs.
Data analysis algorithms
The most complex parts of the system are the algorithms that perform large-scale data analysis required for ACOs to function. One example is the Attribution and Assignment algorithm, involved in the first step of the high-level business process. A diagram like this was used to communicate how the algorithm works at a high level.
The system included several components to support the algorithm functionality and meet other user/client needs:
1. Front-end screens in a web-based application to manage data and kick off algorithms
2. Reports of algorithm results to communicate with provider partners
The business processes were developed alongside, while all elements of the system were built using the iterative agile approach shown here. My role, including management and delivery responsibilities of functional team members, is called out.
The project was carried from an early assessment phase through implementation of production system and process.