Starting mid-week the week of January 11, FDNY EMS will launch an ambulance system optimization process that will be integrated with existing FDNY Computer Aided Dispatch (CAD) algorithms. Information about the optimization process, how it is anticipated to impact 911-participating facilities, and ongoing monitoring of this new process can be found below.

During March and April of 2020, FDNY Emergency Medical Services (EMS) and hospitals in New York City experienced an unprecedented patient surge that severely strained certain hospitals. This new system optimization initiative’s overarching goal is to prevent similar scenarios where a hospital or group of hospitals is overloaded with 911 transports while neighboring hospitals have capacity.

GNYHA and FDNY will hold an informational webinar about the new optimization process. The details can be found here.

Development of the Optimization Process

Based on lessons learned from the spring 2020 COVID-19 patient surge, FDNY has developed this optimization system to support preemptive patient load balancing across the 911 system. It stems from a research collaboration between FDNY and Columbia University’s Department of Industrial Engineering & Operations Research, and relies on the concept of system optimization.

How the Optimization Process Works

Using FDNY 911 transport data and hospital bed availability data from the New York State Department of Health’s (DOH) daily HERDS survey, the optimization formula assesses gaps between supply (hospital capacity defined as available staffed beds) and demand (based on predicted transports in the vicinity of the receiving hospitals for the next day with the assumption that 40% of ambulance transport will result in an admission).

The optimization is run each evening after receipt of that day’s HERDS data, with the output of the optimization instituted by 8:00 p.m. each evening and in effect for the next 24 hours. The optimization is only run for the Critical Care Category (CCC) General Emergency Department, which accounts for 80% of all transports. No specialty CCC codes are eligible for load balancing in this initiative.

Applying the Optimization to Hospitals and Groups of Hospitals

These data are applied at the FDNY atom level. For the purposes of EMS ambulance dispatch activity, New York City is divided into 2,388 unique geographic areas called atoms. For atoms where demand is predicted to exceed supply, the algorithm suggests small changes within CAD’s suggested hospital formula to allocate the next day’s patient transports to a neighboring hospital with capacity. The optimization system’s proactive approach institutes incremental, atom-level changes in areas where demand is predicted to outpace supply. Individual hospitals in an area where capacity appears to be strained would see a change to net transports in increments.

FDNY EMS staff working on the optimization project have also been given access to GNYHA’s Sit Stat 2.0 Hospital Surge Indicator and Surge Behavior data. These data will be used as a secondary check for hospitals in selected atoms.

How Does this Process Work with Redirection and Diversion

The new optimization process is a preventive measure designed to mitigate ambulance overcrowding at a single hospital or groups of hospitals. If it works as expected, it will minimize the need for hospital redirection (automatically set by FDNY when three or more ambulances exceed a 30-minute turnaround time) or hospital diversion requests. However, both diversion and redirect will continue to be available while the load-balancing function is running.

Launch and Monitoring of New Optimization Process

FDNY has been working on the optimization process for several months, including running the optimization against retrospective data dating back to spring 2020. On average, the optimization process suggests changes in 50 to 450 atoms per day out of a total of 2,388 citywide, with the total number of impacted transports in these atoms between 5% and 20%. Optimization parameters have been set conservatively to achieve project goals while causing minimal disruption to routine transport processes.