CreateASoft, Inc - Simulation and Digital Twin software with AI
Search
× Search

Monday, September 29, 2025

Improve Patient Flow, Reduce Length-of-Stay, and Optimize Providers with Digital Twin Studio

Case Study
Improve Patient Flow, Reduce Length-of-Stay, and Optimize Providers with Digital Twin Studio®

Problem statement

A mid-size hospital (211 Beds, 11 OR Rooms, 25 ED Beds) was experiencing significant operational challenges that impacted patient care and resource efficiency. Emergency Department (ED) crowding had intensified, leading to prolonged ED-to-inpatient boarding times and increased risk of patients left without being seen (LWBS). Inpatient units faced highly variable lengths of stay (LOS), which disrupted the scheduling of elective procedures and contributed to under-utilization of operating rooms (ORs). Additionally, poor visibility into downstream bottlenecks—such as delays in imaging, lab processing, and post-acute placement—combined with inefficient staff deployment, further strained hospital operations.

These issues collectively hindered the hospital’s ability to deliver timely, coordinated care and maintain optimal throughput. The overarching goal was to reduce average LOS across both ED and inpatient settings, lower wait times and LWBS rates, and improve frontline provider utilization, all while maintaining high standards for quality and safety.

To address these challenges, the hospital leveraged Digital Twin Studio® to build a validated, data-driven digital twin model capable of testing interventions, forecasting disruptions, and guiding operational improvements.

Digital Twin Model Development

A discrete-event simulation (DES) and hybrid model was developed in Digital Twin Studio® to represent key hospital operations, including the emergency department (ED), imaging services, inpatient units, discharge processing, and patient transfers. This model captures variability in arrival patterns by hour and day, treatment cycle times based on patient acuity, lab and imaging service durations, and transport delays.

To further enhance the model’s capabilities and scalability, the simulation was integrated into Digital Twin Studio’s data analytics and Regenerative AI engine, providing a more comprehensive and dynamic data-driven digital twin environment. Digital Twin Studio® allows for real-time and historical data integration, enabling the model to reflect live operational conditions and respond to changes in patient volume, resource availability, and care protocols. The platform supports advanced analytics, machine learning, and visualization tools, which help stakeholders explore “what-if” scenarios and predict the impact of operational changes before implementation.

Model Validation

The digital twin model was validated using a multi-step process to ensure the simulation accurately reflected real-world operations. The first step was to compare simulated key performance indicators (KPIs)—such as hourly patient arrivals, median length of stay (LOS), and throughput metrics—with historical data extracted from hospital systems. This comparison helped identify discrepancies and guided model refinement.

A crucial component of this validation was the use of historical data replay. By feeding actual patient flow and event data into the simulation, the digital twin was tested under real-world conditions. This replay allowed for a direct comparison between simulated behavior and observed outcomes, providing a robust benchmark for accuracy. It also helped uncover whether the model could replicate not just average performance, but peak loads, bottlenecks, and variability in service delivery.

To further refine the model, AI generated service-time lead times —such as treatment durations, lab and imaging turnaround times, and transport delays—were generated based on different operational conditions. Once validated, the digital twin became a powerful tool for scenario testing, operational planning, and decision support, enabling hospital leaders to explore interventions with confidence.

Experiments and Optimization

Digital Twin Studio was then used for extensive scenario testing to evaluate operational interventions without impacting real patients. Hundreds of simulations and emulations were run across a wide range of configurations to assess the effectiveness of proposed changes in care delivery, resource allocation, and patient flow. These scenarios included:

  • Split-flow and fast-track models in the Emergency Department (ED), such as implementing physician-in-triage protocols for low-acuity patients
  • Dynamic bed allocation strategies, including prioritized patient transfers and predictive bed staging
  • Imaging prioritization, such as reserving dedicated CT slots for ED patients
  • Provider reallocation models, including surge nurse float pools and staggered shift starts
  • Discharge acceleration initiatives, like nurse case-manager rounding and earlier discharge order workflows

Digital Twin Studio’s simulation engine allows these interventions to be tested under various demand conditions, staffing levels, and operational constraints. The platform’s multi-scenario analyzer enabled the identification of the most effective strategies based on KPIs such as wait times, LOS, resource utilization, and patient outcomes.

Optimization and Analysis Results

  • ED LOS reduction:
    • The digital twin model showed that a split-flow/physician-in-triage + fasttrack implementation would reduce the emergency department’s median length-of-stay and time-to-disposition; after deployment, this intervention reduced disposition times by an average of 12.7 minutes.
  • Overall inpatient LOS:
    • Optimizing discharge workflows and prioritizing imaging/transport reduced inpatient delay components; the average patient LOS was reduced by 11.3%, while low acuity patients saw reductions of 19.45%.
  • LWBS and wait times:
    • Modeling showed that small increases in front-end staffing at peak hours (or a fast-track lane) reduced LWBS and decreased wait times. The optimization generated an 83% reduction in LWBS for all shifts.
  • Provider utilization & throughput:
    • Rebalancing shifts to match predicted arrival patterns increased provider productive time and reduced idle/overload periods. On average, provider efficiency increased by 14.2% throughout all departments.

____________________________________________________

Total financial impact

When combined, the modeled improvements delivered:

  • Direct savings + cost avoidance: $11.2M/year.
  • Revenue lift freed capacity & reduced LWBS: $2.5M/year.

Conclusion

This case study illustrates how CreateASoft’s Digital Twin Studio® revolutionized all aspects of hospital optimization and generates detailed analysis to reduce operational cost. By merging simulation, emulation, and AI, the hospital transformed its internal processes—achieving sustainable improvements in productivity, quality, and responsiveness. In addition, the model is still used for operational analysis to provide constant insights into the operation including scheduling, provider allocation, and process control. This success demonstrates the value of digital twins not just as a onetime analysis, but as a core operational intelligence platform for the future of healthcare.

Print
0 Comments
Rate this article:
5.0

Categories: BlogNumber of views: 346

Tags:

Please login or register to post comments.
Terms Of UsePrivacy Statement© 1992-2025 CreateASoft, Inc. All Rights Reserved.
Back To Top