The client is a leading American healthcare company and national hospital-based physician group
Business Problem
Client’s system supporting over 10,000+ clinicians faced the following issues:
Demand forecasting was inaccurate and inconsistent across all facilities, hindering planning and availability.
Forecasting was high level and lacked hourly granularity.
IT systems were ad-hoc, error prone, labor intensive and expensive to maintain.
Inconsistent system integration was hampering ability to scale, and the client lacked visibility of operations.
The Solutions
The solution was rolled out for Radiology and Emergency Medicine service lines.
Machine Learning (ML) algorithms were designed to accurately predict future volume requirements by the hour for the specific provider service lines and facilities.
Custom analysis tools were developed to use the output from the ML algorithms to predict future resource requirements, and thus optimize staffing levels for the facility.
Legacy enterprise systems were then subsequently integrated to allow the access of a broad set of historical data.
The system could quickly adapt to the recent demand changes, looking at the past short & long term fluctuations.
Value Delivered
Streamlined disparate processes into an agile, consistent, accurate and user-friendly process.
Removed inefficiencies in the system that were due to manual, laborious, one off, inaccurate departmental efforts.
Created easy to use solution interfaces that resulted in supporting a quick onboarding process for the Client’s team and removed the need for elaborate training.
System showed future capacity taking into account each clinician's individual productivity. This helped in identified arriving at the optimal mix of physicians, APPs, NPs, and scribes (in EM) to meet the future demand.
Accurately predict for fluctuating demand scenarios. The system could provide insights on reduced volumes in real-time during the peak Covid times.
Improved reporting using better visual and graphical formats to capture gaps in future planning
