Confidential client, name withheld under NDA

Predictive Readmission and Population Health Analytics for a US Regional Health System

Quick answer: A Snowflake data warehouse can unify EHR, claims and operational data, score readmission risk before discharge and feed same-day population health dashboards. Bacancy Technology built this for a US multi-hospital regional health system.

Outcomes deliveredDelivered March 2025 to September 2025 by a team of 5
5 to 7

Days before discharge risk is flagged

Not after an emergency department return

20%

Fewer 30-day readmissions

Across the flagged high-risk segment

Same day

Population health reporting

Down from multi-day manual pulls

1

Patient profile from 3 to 4 records

EHR, claims and operational data

Client name withheld under NDA. To honour our non-disclosure agreement and compliance obligations, we describe this client by industry and region only. The project scope, technology and outcomes are shared without details that could identify them.

About the Client

The client is a US multi-hospital regional health system offering acute, outpatient and specialty care. Its care and finance teams needed one view of each patient and current reporting.

EHR, claims and operational data sat in separate systems. Readmission risk surfaced only after discharge, often after a patient returned to the emergency department. Population health reports took days of manual pulls, and staff could see more patient data than their roles required.

A 5-person Bacancy Technology team unified the data in a governed Snowflake warehouse and added a predictive readmission model. Role-based access and dynamic masking protect PHI. Care teams now see high-risk patients 5 to 7 days before discharge. Finance gets near real-time cost-of-care reporting.

Client type
Multi-hospital regional health system
Industry
Healthcare
Location
USA
Team
5 team members
Timeline
March 2025 to September 2025
Project size
Non-disclosable
Engagement
Data analytics and data science

Challenges & Solutions

Our data engineers and data scientists solved each problem on one governed Snowflake platform.

Challenge 1: Unifying EHR, claims and operational data

Coordinators had to query 3 systems and reconcile records by hand to see a full patient picture.

Solution: Our data analytics services team loaded EHR, claims and operational data into one Snowflake warehouse with Fivetran. They modeled conformed patient-level tables, so a full record comes from 1 query instead of 3.

Challenge 2: Flagging readmission risk before discharge

Readmission risk had no early warning, so care teams found out only after an emergency department return.

Solution: Our data science and predictive analytics team trained a scikit-learn model on clinical history, prior admissions and comorbidities. It scores every patient ahead of planned discharge, and care teams get a ranked, explainable risk list days in advance.

Challenge 3: Replacing multi-day manual reporting

Population health and cost-of-care reports took several days to compile and were outdated on arrival.

Solution: Our team modeled the reporting layer in dbt and connected it to Power BI dashboards that refresh on current warehouse data. A dedicated Snowflake developer maintains it, and reports now arrive the same day.

Challenge 4: Scoping access to patient data

Staff saw more patient data than their roles required, and no log showed who accessed which record.

Solution: Our healthcare IT team set up role-based access and dynamic data masking at the schema level in Snowflake, with PHI fields masked by default. Every access event is logged, giving the client a queryable audit trail to support HIPAA reviews.

How the Platform Works

The client readmission and population health dashboard

Core Features

  • Unified patient data warehouseEHR, claims and operational records in one Snowflake profile per patient.
  • Readmission risk scoringRanked, explainable risk list generated ahead of each planned discharge.
  • Population health dashboardsPower BI views of clinical, population and cost-of-care metrics, refreshed same day.
  • Role-based access and maskingEach user sees only permitted fields; PHI masked by default and access logged.

The stack behind the analytics

Cloud data warehouse
Snowflake
Ingestion and transformation
Fivetran | dbt
Orchestration
Apache Airflow
Machine learning
Python | scikit-learn | MLflow
BI and dashboards
Power BI
Security and governance
Snowflake RBAC | Dynamic data masking | Access logging
CI/CD
Git | GitHub Actions

Ready to start?

2500+ projects delivered by Bacancy Technology

Get access to an experienced team of developers and engineers from Bacancy Technology, handpicked to ace your goals. Kickstart within 48 hours, no-risk trial.

Book A 30 Min Call
14+Years of business experience
1458+Happy customers
30+Countries with happy customers
1050+Agile-enabled employees

More Case Studies

View All Case Studies