Computer simulation modeling
has evolved over the past twenty years into a powerful tool for understanding operational
issues in health care delivery and the relationship between operational performance and
building design. As pressures to achieve efficiencies in operation continue decisions will
increasingly focus on understanding both "first" dollar and life cycle costs
associated with capital plans. Identification of the space necessary to meet the
anticipated demand and to implement new organizational plans on staff efficiency are the
type of questions which must be addressed in the planning.
A key characteristic of simulation modeling is the
ability to incorporate the behavior of key system elements into a computer
"mock-up" of the area under investigation. In an emergency room, for example,
the arrival of patients into the ER, their diagnoses, use of diagnostic services and
length of stay in the facility vary significantly. To plan based on the
"average" patient can lead to erroneous estimates of space requirements and
staffing needs. Simulation models can take the statistical patterns for arrivals, length
of stay, and other variables and incorporate them into the model of an existing or
proposed service. This allows the decision makers to have detailed estimates of the
performance of a service during peak periods of demand and other time periods. Room and
staff utilization can be estimated along with information regarding queues that may
develop for resources and information regarding waiting room capacity.
The development and implementation of a simulation model typically
moves through three phases:
- Model definition and data collection
- Model validation
- Simulation of proposed plans and outcome analysis
Our typical process is to work with clients to
identify key issues and goals of the modeling. Once the outcome goals are identified, we
will design a plan to collect required data. It is not uncommon to require some
prospective time sampling of activities in the service that are not maintained in existing
records and databases. Collection of this data is usually undertaken by staff from the
client. We will frequently participate in limited observations to become familiar with the
operations of the service and to help in the analysis of the results. Based on this
information we design and implement a model tailored to the specific characteristics of
the hospital's service.
Model validation is a cross-check to assure that the
logic and data used in the model can replicate existing conditions. Once we are
comfortable that the model is working correctly the final stage of analysis is initiated.
Simulations of all proposed
organizational and workload options are run to estimate the behavior of the proposed
system. Outputs from these models are analyzed and presented in graphic and tabular form,
along with specific recommendations.
We have recently started using a new simulation
package designed specifically for medical applications. MedModelŠ , developed by ProModel
Corporation, allows the testing of both operational and facilities attributes through a
powerful animation component. We can, for example, take a proposed plan drawn in AutoCAD
and, based on anticipated workload patterns and operational policies, animate the use of
the space along with providing estimated performance statistics. The animation
characteristics provide a valuable tool to communicate the model logic to staff and
administration, along with checking the circulation flow of the proposed design.