Demonstrations of the use of Simulation in veterinary ...



Demonstrations of the use of Simulation in veterinary practice management

Duane Steward, DVM, MSIE, and Charles R. Standridge, Ph D

From the Department of Industrial Engineering, Florida Agricultural and Mechanical University - Florida State University College of Engineering, Tallahassee, FL 32306. Dr. Steward’s present address is Laboratory for Computer Science, NE43-415, 545 Technology Square, Cambridge, MA 02139

Simulation offers an analytical method for solving problems in practice design, planning, and operations. It can be used, for instance, when designing a facility, planning for renovations, planning for changes in patient demand, or evaluating practice system operations. The basic principle behind simulation is the use of specialized computer languages to construct a program that models the behavior of a complex system (eg, a veterinary practice). The program is then used to predict how changes in practice details would affect other details or the practice as a whole. Simulation may be used where other modeling methodologies (eg, multiple regression, linear programming) are inadequate or inappropriate.

Details of a simulation model of a generic veterinary practice have been published.[i] In this report, 2 examples of how simulation can be used to assist in making veterinary management decisions are described.

Use of Simulation for Practice Planning

Consider a fictitious veterinary practice that has 1 veterinarian, 2 examination rooms, 1 receptionist, 2 technicians, a ward of 10 cages in which animals can be hospitalized, 1 surgical suite, and 1 area that is used both for in-hospital treatments and for surgical preparation. The current hospital policy is to disallow the scheduling of any appointment later than 60 minutes prior to closing. However, because of variations in the time it takes to process clients through the system, it often happens that some animals are discharged after closing time. When this occurs, staff members must be paid overtime, and the clients are often inconvenienced or become dissatisfied with the hospital’s performance.

The practice management, therefore, would like to decrease the frequency of after-hours discharges and reduce the time interval between closing and any after-hours discharge. One possible solution would simply be to hire another staff member. However, it is unclear whether the savings in overtime would offset the new staff person’s salary. Providing medical services involves communication, and phone conversations, which are an essential part of the hospital’s services, occupy a large part of the hospital staff’s time that could otherwise be used for in-house activities. One might subjectively wonder if the phones are often busy and if increasing the number of phone lines would allow the current staff to deal with phone tasks more rapidly and reduce patient waiting time in the hospital. This would, after all, cost less than adding another staff member to the payroll. Simpler still, one might argue, would be a change in hospital policy that would disallow the scheduling of any appointment later than 90, or even 120, minutes prior to closing. This, however, might have a detrimental effect, by decreasing the number of clients that can be seen.

Simulation allows one to predict the effects that any combination of these 3 changes (addition of a technician, addition of a phone line, a change in the scheduling policy) would have on the performance of the hospital as a whole. The performance measures of interest (ie, outcome variables) would be total number of patients seen by the practice, number of patients ready for discharge after closing but within 4 hours of closing (because it was assumed that any animal ready for discharge more than 4 hours after closing would be held overnight and discharged the following morning) and time between closing and patient discharge, which was assumed to be an indicator of overhead cost and of client satisfaction. The objective was to minimize the latter 2 outcome variables (number of late discharges and time between closing and discharge) while maximizing, or at least not decreasing, the first (total number of patients seen).

Mean values and 90% confidence intervals of the 3 outcome variables were estimated using 10 replications of the simulation. The initial configuration was 2 technicians, 2 phone lines, and a 60 minute cutoff between the last appointment and closing. The model was then changed to include 3 technicians instead of 2 (number of phone lines and scheduling policy remained unchanged) and values were recalculated. Subsequently, the model was changed to include 2 technicians and 3 phone lines, and, finally to include 3 technicians and 3 phone lines. This grid of four configurations was repeated with a policy of disallowing appointments less than 90 minutes prior to closing, and was again repeated with a policy of disallowing appointments less than 120 minutes prior to closing. Analysis of variance was used to determine differences among output for each of the model configurations. Various standard methods were used to minimize variance.[ii]

Results - A quick survey of the projected total number of clients served for each configuration of the model did not reveal marked differences (Table 1). Thus, none of these changes in policy appear to have a substantial impact on total number of clients. Therefore, effects of changes in the number of resources and in operational policies may be compared with vigor and with an absence of fear that overall performance will suffer.

Inspection of the results suggested that mean number of late discharges was affected by the scheduling policy far more than by changing the number of phone lines or technicians (Table 2). Surprisingly, as the interval between last appointment time and closing increased, so did the interval between closing and discharge. This suggests that changing the scheduling policy will mean that staff members will have to work overtime less frequently but perhaps will have to stay longer whenever they do work overtime. This may well be tolerable, however, because the change represents an overall improvement in performance.

Analysis of variance indicated that changes in the number of technicians and the number of phone lines did not account for the differences in the number of late discharges or time between closing and discharge. Changing the scheduling policy to prohibit appointments within 90 minutes or 120 minutes of closing was found to significantly (p ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download