AI Models for Patient Volume Prediction in Hospital ADT Units

Executive Summary

Hospitals around the globe grapple with a significant challenge – the persistent hurdle of reactive decision-making. A major hospital in the United States found itself entangled in the same scenario. Delayed information to make decisions was rippling across multiple departments, affecting patient care, operational efficiency, and financial growth. The fluctuating patterns in admissions, discharges, and transfers compounded the problem, causing additional delays that impacted both patient outcomes and the financial health of the institution. The cascading effect led to further inefficiencies in staff scheduling.

In collaboration with Factspan, the hospital sought a holistic solution to overcome the challenge. The team crafted an advanced AI- driven time series forecasting model, leveraging machine learning to streamline operations. The team deployed these models to guarantee precise predictions for admissions, discharges, and transfers. The initiative materialized in a user-friendly Tableau Dashboard, aligning with some of the most advanced healthcare analytic systems available. It resulted in proactive decision-making, optimized staffing, and notable cost reductions for the hospital

About the Client

The client, a faith-based not-for-profit organization, is one of the most trusted healthcare institutions in the United States. With nearly 2 million patient visits per year and with over 4000 physicians, they have earned a reputation for providing exceptional medical care and are recognized as a leader in the healthcare industry.

The organization is known for its commitment to patient-centered care, with a focus on personalized treatment plans and advanced medical technologies.

Business Challenge

As proactive patient care requires accurate demand planning, the lack of predictive analytics led to reactive decision-making in the hospital system. This issue affected patient care, operations, and finances, causing delays and uncertainties. The unpredictable patterns in admissions, discharges, and transfers posed a technical puzzle, requiring precise solutions for resource allocation and patient management.

Inaccuracies in staffing predictions further complicated the technical aspects, disrupting smooth workflows. Deciphering this multifaceted technical challenge was imperative, demanding a solution capable of understanding complex data dynamics and providing tailored insights for enhanced operational strategies at a technical level

Our Solution

To address the challenges, Factspan engineered a unique solution centered on time series forecasting models. The project workflow encompassed understanding data sources, building necessary ETLs, data collection, cleaning, exploratory data analysis, feature engineering, and model development.

The team developed machine learning and classical time series models to ensure accuracy in predicting admissions, discharges, and transfers across Inpatient, OBS, Step-down, ICU, and MedSurge units. The multifaceted approach also included the creation of a Tableau Dashboard, seamlessly integrating advanced healthcare analytics capabilities.

By mitigating operational blockages, reducing costs, and providing a faster turnaround time on discharges, the solution proved instrumental in elevating healthcare operations to new levels of efficiency for the healthcare provider. This strategic intervention not only addressed immediate challenges but also set the stage for sustained efficiency in healthcare operations

Business Impact

• Streamlined staffing response, cutting turn-around time by 30%
• Enhanced workload management efficiency by 25%
• Realized a 40% increase in operational productivity
• Achieved precise volume predictions with 95% accuracy