Content area
Traditionally, the finance department has assumed responsibility for assessing process costs in healthcare organizations. To enhance process-improvement efforts, however, many healthcare providers need to include clinical staff in process cost analysis. Although clinical staff often use electronic spreadsheets to model the cost of specific processes. PC-based animated-simulation tools offer 2 major advantages over spreadsheets: they allow clinicians to interact more easily with the costing model so that it more closely represents the process being modeled, and they represent cost output as a cost range rather than as a single cost estimate, thereby providing more useful information for decision making.
Full text
Traditionally, the finance department has assumed responsibility for assessing process costs in healthcare organizations. To enhance process-improvement efforts, however, many healthcare providers need to include clinical staff in process cost analysis. Although clinical staff often use electronic spreadsheets to model the cost of specific processes, PC-based animated-simulation tools offer two major advantages over spreadsheets: they allow clinicians to interact more easily with the costing model so that it more closely represents the process being modeled, and they represent cost output as a cost range rather than as a single cost estimate, thereby providing more useful information for decision making.
Historically, the functions associated with understanding, controlling, and avoiding costs in a healthcare organization were almost solely the responsibility of the finance department. In recent years, however, increasing cost pressures have made it necessary for all of a healthcare organization's staff to play a role in controlling costs. Clinical staff, in particular, often can provide insight into cost-related details of clinical processes that financial managers might easily overlook.
This organizationwide sharing of responsibility for cost control has led to the adoption of new PC-based tools to develop cost information that can be used by both financial and nonfinancial staff. Many healthcare organizations have adopted electronic spreadsheets, and most clinical personnel are familiar with these applications. Until recently, however, healthcare organizations have been slow to adopt sophisticated process-simulation applications, largely because of their complexity. The next generation of simulation software offers an increasingly user-friendly means to perform activity-based costing through animated displays, point-and-click model building. and automated output.
An animated process-simulation application facilitates involvement of clinical professionals in process costing by presenting information in user-friendly ways. For example, the application can provide a representation of the physical layout of the area in which the process is performed (eg, the radiology area of an emergency department). Within this display, the data analyst can create graphic images of the various clinical and support personnel and plot their activities, such as movement from room to room or extended presence in one room to perform various activities (eg, MRI examination).
Output values, such as waiting room contents or minutes wait time, can be displayed on-screen while the model is running. The application depicts the process within a time frame that the cost analyst has entered into the program (eg, 100 simulated "days"). Thus, by running multiple processes, the analyst can simulate actual conditions in the process area.
Animated-simulation tools also differ from electronic spreadsheets in that they are designed to account for variations inherent in the delivery of healthcare services by providing output in the form of cost ranges, or distributions, as opposed to specific costs, or point estimates. The simulation tools therefore provide healthcare service administrators with more realistic information for activities such as contract negotiations, cost estimations, and promotion of cost-reduction initiatives. In addition, animated visualization of the process being modeled allows clinical and other nonfinancial staff to more easily see how the costing model relates to the process being analyzed, thereby helping them to ensure that the model accurately reflects the process.
Animated Simulation Modeling and ABC
The advantages of electronic simulation tools can be demonstrated by comparing an activity-based costing (ABC) analysis of a specific clinical process using a traditional spreadsheet with an ABC analysis of the same process using an animated process-simulation program. The same basic steps can be used to perform ABC with either electronic spreadsheets or animated-simulation models:
* Describe the process being costed;
* Visually represent the process (eg, using a flowchart) to ensure that the model conforms to the actual process;
* Collect primary and secondary data;
* Build and run the model (ie, identify variables, constants, and their relationship and input data);
* Analyze initial outputs (ie, cost estimates);
* Review the process and results to identify possible errors and/or omissions, and revise as necessary; and
* Analyze final cost outputs.
For the purpose of comparison, the two approaches were used to perform an ABC analysis of the costs associated with performing X-ray, computed tomography (CT), and magnetic resonance imaging (MRI) examinations of trauma patients who presented with possible cervical spine (c-spine) injury at the emergency department (ED) of a large, state-run, licensed, academic medical center. The comparison of spreadsheet analysis with an animated-- simulation program was an incidental part of a larger study requested by a physician to compare the cost-effectiveness of CT with X-ray modalities for initial screening of high-risk patients. The fact that the physician could be integrally involved in all phases of the study underscores the usefulness of these analytical methods for cost analysis by nonfinancial personnel.
Resources available for the emergency radiological evaluations included three basic X-ray examination rooms, one CT scanner, and one MRI machine-all in immediate proximity to the ED. Patients were categorized into three acuity levels-low, moderate, and high-reflecting the probability of a c-spine injury. To simplify the discussion of the ABC analysis, examples will focus on the X-ray examination process. Although this process involved basically the same steps for all types of patients, it typically was more complicated for high-acuity patients, requiring both more activities and more images. Process Description and Representation
The c-spine X-ray examination process begins after a physician has triaged and examined a patient and determined that it is necessary to generate a computer order for a radiological work-up. The process consists of the following steps:
The patient is wheeled or escorted to the radiology area by a patient transporter (or sometimes by a nurse).
A clerk processes the physician's order and prepares a film jacket for the radiology technologist.
The radiology technologist carries the film jacket into an examination room and calls for the patient to be brought in by the patient transporter for imaging.
* The technologist takes the images (the volume of images taken and the time spent varies according to the patient's physical characteristics and the severity of injury).
* The technologist enters the processing area to develop the films and brings them into the radiologists' reading room.
* The radiologist on duty determines whether the images are satisfactory. Sometimes the images are not clear enough and must be retaken, or the radiologist may request that additional images be taken from a different angle or that the patient also undergo a CT or MRI examination (requiring additional steps for imaging in the CT or MRI area).
* Once imaging is satisfactory, the patient is escorted back into the main ED area.
For purposes of the ABC analysis, a flowchart of this process should be constructed to ensure that it is properly understood. The flowchart may be used either by finance staff to guide their own spreadsheet analysis of process costs or by simulation model builders to develop the process parameters that form the model's internal logic. In either instance, the flowchart's accuracy should be verified by clinical staff who understand the process under review.
Then, data should be collected on cost factors such as labor and equipment costs and time to perform each step. These data provide the basis for building a spreadsheet or animated simulation model.
Spreadsheet-Based Cost Estimations
Exhibit 1 shows how the academic medical center's data were used to develop a spreadsheet model for estimating the costs associated with performing X-ray examinations of c-spine patients as classified by acuity level.
The example focuses on basic X-ray costs, and therefore does not include costs associated with the radiologist's evaluation of examination results and any CT or MRI. The direct labor cost for a patient is a function of the amount of the labor used, the cost per hour, and the probability that a patient would be at a given acuity level. Similarly, the supply cost is a function of the cost of individual supply items and the volume of supplies used according to the patient's acuity level.
In this particular study, two equipment-cost estimates were made: the cost of equipment at actual usage rates and the cost of equipment at maximum reasonable capacity (based on the top five percentile of daily volume over a year). As the conclusions for the purposes of this discussion are the same whichever of these two estimates is used, Exhibit 1 presents results of the maximum capacity simulation only.
Because indirect costs were not a focus of this study, for illustrative purposes they are estimated to be 32 percent of direct costs (based on national averages).
Thus, the total technical cost for a basic X-ray examination, assuming the organization operates at maximum sustainable capacity, is $33.71, $39.64, and $43.96 per patient for low-, moderate-, and high-- acuity patients, respectively.
Animated-Simulation-Model-Based Cost Estimation
The process of building an animated-simulation model differs significantly from that of building a spreadsheet model. With an animated-simulation model, instead of being presented with blanks on the spreadsheet in which to input data, the model builder creates a structured flow process key and supplies key input parameters. To analyze the costs of performing c-spine X-ray examinations, for example, the parameters would include:
* Patients by category (low, medium, and high probability of having a c-spine injury);
*Personnel (radiologist, radiology technologist, nurses, patient transporters);
Bo Other resources (X-ray machine, CT scanner, MRI machine, film);
*Locations (waiting room, hallway, X-ray room);
*Activities (a transporter helping a patient move from the waiting room to an X-ray room, a radiology technologist taking an X-ray);
*Paths (a transporter leaves one X-ray room, goes to the waiting room, and then escorts a patient to an unoccupied X-ray room); and
*Raw data (number of patients, arrival patterns of patients, patients by injury category, clerical time per patient, patients' time in radiography by patient type, volume of c-spine exposures and films by patient type, need for a CT or MRI scan, cost per hour of a radiology technologist, cost per film of radiography machine).
For many of these parameters, the data analyst inputs distributions rather than the point estimates required for a spreadsheet analysis. Exhibit 2 presents a comparison of data inputs for the spreadsheet model and the animated-simulation model. Consider, for example, the difference in data input for patient transport time. In a spreadsheet model, the cost analyst would enter a specific time that reflects the average time for patient transport based on sample data collected over a given time period. Exhibit I shows this average time to be about 11 minutes. By contrast, in the animated-simulation model, the data analyst would enter a range-say five to 17 minutes-to reflect the normal variation in transport times.
Once the parameters are entered into the animated-- simulation model, the data analyst tells the application how many iterations of the process to run to reach a longterm steady-state solution. The animated-simulation program then automatically combines and processes the various distributions and point estimates to determine a cost range, within a 95 percent confidence interval, for the overall costs of the process. The user has the option of viewing an animated representation of the process being analyzed or simply running the application in the background and waiting for it to produce an alphanumeric output of its estimated overall cost ranges.
Exhibit 3 shows the results of cost analysis of the cspine X-ray examination process performed using the animated-simulation application. The technical costs of the procedure, by risk level, are indicated as cost ranges, compared with the amount the organization ordinarily is paid for the procedure by Medicare.
Advantages of Animated-Simulation Modeling
The animated-simulation model offers two distinct advantages over a spreadsheet approach. First, by graphically depicting the process being analyzed, an animated-simulation model allows clinicians to check the model for face validity (eg, normal movement sequences and queue build-ups for X-ray examinations) If the animated representation indicates, for example, that the process results in longer waiting times for examinations than actually occur in practice, then the original input can be easily accessed on-screen and appropriate time adjustments made.
This ability to fine-tune cost analyses to more closely conform with actual processes significantly enhances clinicians' interest and involvement in the whole modeling process. In addition, the visual interface provides a basis for clearer communication between clinicians and financial planners regarding the factors that contribute to process costs.
The second advantage is in the type of statistics that animated-simulation applications can provide (eg, mean costs, standard deviation, and cost ranges within a 95 percent confidence interval). The ability to analyze costs as distributions rather than point estimates gives a broader perspective on process cost savings than can be achieved through a spreadsheet analysis.
Under the spreadsheet analysis of the c-spine X-ray examination process, for example, the total process cost for a high-risk patient was determined to be $43.96 at maximum capacity, including overhead (see Exhibit 1, page 62). By contrast, the animated-simulation model indicates that, within a 95 percent confidence interval, the actual technical cost is between $35.13 and $54.19, with a mean of $44.66 (see Exhibit 3). Armed with such information during contract negotiations, a financial manager would know with near certainty that his or her organization's costs are at least $35.13, and that therefore, any payment below that amount is not acceptable. Moreover, to be certain that all costs are being covered, the organization would have to be paid at least $54.19.
Conclusion
Although spreadsheet models commonly are used today by both healthcare financial managers and clinicians to perform ABC analysis, animated-simulation models have become increasingly user-friendly and may be preferable to spreadsheets, especially for the nonfinancial cost analyst. Healthcare organizations that adopt animated-simulation tools also will incur some initial training costs, but the benefits in improved financial performance and greater clinical staff participation in controlling organizational costs, will far outweigh this preliminary investment.
Animated-simulation models are likely to appeal to clinicians because they offer information that accounts for normal variations in practice and give clinicians a clear, nontechnical way to communicate with financial management regarding process activities and cost factors. Moreover, because the application software calculates outputs automatically, nonfinancial analysts need only provide easily collected or determined point estimates, distributions, and other data.
ABOUT THE AUTHORS
William N. Zelman, PhD, is professor of healthcare financial management, Department of Health Policy and Administration, University of North Carolina at Chapel Hill, School of Public Health, and a member of HFMA's North Carolina Chapter.
Noah D. Glick, MSE, MHA, is senior management engineer, Department of Management Services, University of North Carolina at Chapel Hill.
C. Craig Blackmore, MD, MPH, is assistant professor of radiology, Harborview Medical Center, University of Washington, Harborview, Washington.
Questions or comments regarding this article may be sent to William Zelman at [email protected].
Copyright Healthcare Financial Management Association Sep 2001