The Client
Leading US-based most comprehensive chain of hospitals offering a complete range of medical services, innovative research and life-enhancing care through a team of 7,000 doctors and 500+ patient care locations.
The Challenge
The client wanted analytical sanity in its operations, tying financial performance with patient outcomes. It worked with an on-premises, self-managed data center to source critical patient care data from 18 network hospitals. This would fall short of supporting a high-performing analytics platform that could help the client factor in a wider data pool comprised metrics currently used by independent healthcare auditing firms such as the Center for Medicare and Medicaid Services (CMS) or the U.S. News & World Report. These healthcare indices are critical for helping patients and doctors select the right hospital for quality care. They feed into the value-based payment model under Medicaid, where providers are reimbursed based on their performance ratings.
The Solution
Persistent was commissioned to create a data analytics layer for a quality reporting dashboard to glean actionable insights to help meet patient care quality metrics used in national and state healthcare indices. We helped migrate the client’s on-premises data center to Google Cloud Platform’s (GCP) BigQuery for scalability, performance, and reliability.
While building the analytical backend for the quality reporting dashboard, we onboarded multiple data sources to Big Query, such as Change Data Capture-based replicas of CLARITY (EPIC) data. We created custom dataflow jobs for multi-format data ingestion, transformations, and analytics. This was done in a completely orchestrated environment along with continuous integration and deployment model to ensure minimal latency in data updates. Data Marts on BigQuery supported patient encounter-based analysis.
Healthcare performance parameters are non-trivial, such as readmission rates, mortality rates, bed utilization, and length of stay. These parameters can also be those used in auditing exercises that factor in patient satisfaction scores around cleanliness and care quality. An in-depth analysis requires attributing proper weightage to each standard and unique parameter. Persistent created a GCP-native data analytics layer that helped the client analyze performance based on hundreds of dimensional analyses and aggregation.
We helped the client qualify this data with multiple dimensions to glean insights into highly specialized services. We built functionality to allow the client’s business users to correlate performance with patient cohorts based on disease and previous treatments. This was complex since this patient data exists as disease and diagnostic codes, the definitions of which are routinely updated by external auditing agencies. Working with the client’s business users, we encoded rules to identify these codes, working with a provision to accommodate dynamic changes in definitions, offering the user a comparative snapshot of a customer cohort changes with new definitions.
The Outcome
Persistent’s GCP-backed analytical backend fed into multiple quality benchmarking dashboards that helped the client make business-critical calls on physician performance, patient outcomes, and hospital care quality. This helped the client:
Outperform leading healthcare indices at the state level.
Build a forward-looking strategy for business growth.
Identify gaps around care demand and physician availability.
Improve care outcomes with a sustainable boost in top-line numbers.
