Flexible budgeting: a tool for better forecasting.
Change abounds. Patient census has begun to drop in hospitals across the country with the onset of prospective payment. There are drives to cut laboratory utilization for inpatient cases and the laboratory budget along with it. At the same time, many laboratories are seeking a new workload balance through outside marketing efforts.
Amid so much flux, it becomes difficult to anticipate what the laboratory will need to perform its job. Sometimes how much the laboratory needs will in fact determine whether a change is implemented. How can we project budget requirements as accurately as possible and thereby make better-informed decisions?
A flexible budget helped us show administrators in our 379-bed university hospital that lower volume in the clinical laboratory would not warrant a proportionate budget cut, just as expanded volume would not call for a commensurate increase in spending. These outcomes, involving the difference between fixed and variable costs, are well known in a general way. But we were able to furnish fairly exact numbers because we had found a means of pinpointing variable costs.
This article describes the process we used. We will define a simplified flexible budgeting method that realistically relates costs to volume, then apply the method in a typical laboratory setting and suggest actions based on the results.
When analyzing the effects of changing test volume on lab expenses, we must consider fixed and variable costs. Fixed costs--those that remain constant regardless of workload--generally include such items as equipment depreciation, maintenance costs, and supervisory salaries. Variable costs cover supplies and reagents, among other items that fluctuate in a direct relationship to the volume of work performed.
Figure I presents a very simplified view of fixed and variable costs as a function of workload. The variable component increases along with workload, while the fixed component is constant. Let's dissect this relationship to reveal the variable expenses.
Financial experts use two main ways to sort expenses into fixed and variable components. The first, which we call the synthetic approach, starts with a determined level of fixed costs and adds on variable costs. This approach presumes that the business--in this case, a laboratory--exists initially with a baseline workload of zero. All necessary expenses independent of volume are considered fixed; the remainder are classed as variable.
We tried this approach, but found it very unrealistic in practice. Costs extrapolated from a workload of zero have no meaning. How would a laboratory be staffed if no testing were being performed? We would need instruments to perform the first test, but which ones? The shortcomings of the synthetic method become clear in any calculations outside the realm of theory.
The second approach is called the analytical method. It simply allocates current workload expenses into fixed and variable categories--general costs versus those directly related to the assay in question. Essentially, each test is cost-accounted to the nearest cent.
This method also has its short-comings. Where are third-shift salaries allocated, for example? How much labor cost is fixed, and how much is variable? If test ordering goes down, how will costs in both categories be affected?
Taking our analysis beyond a strict examination of current expenses, we created a budget that would help forecast lab expenditures at various volume levels. We called our method "the 50 per cent solution," since it assumes 50 per cent of present test volume as baseline, determines baseline expenses at the 50 per cent volume, and computes the variable component.
Why use 50 per cent of current workload as a basis for calculation? For one thing, it's a convenient figure. We could also safely assume that workload would be unlikely to drop below this level in the foreseeable future without dramatic consequences for the laboratory and the hospital.
Test volume will actually fall somewhere between 50 per cent and 100 per cent in coming years due to decreasing hospital census and lower test utilization--unless the lab's market expands. Assuming a linear relationship between expenses and workload in that range, we could move up from the baseline and add the appropriate amount of variable costs to reach any particular test volume point.
We first tried this simplified approach in the chemistry section. Figure II presents the five major components of the clinical chemistry budget: depreciation, maintenance, supplies, quality control, and labor. Note that the total section budget is $1,120,000 annually at the current workload. Now let's analyze various ways to pin down baseline expenses at 50 per cent of that workload.
We can assume that depreciation and maintenance costs will remain constant at all workload values between 50 per cent and 100 per cent of current volume. The depreciation of $100,000 shown in Figure II is based upon previous capital expenditures. Maintenance for various instruments, amounting to $40,000 a year, is also constant regardless of instrument usage.
Annual supply costs of $300,000 a year consist mainly of reagents for the laboratory's current annual workload of 1 million tests. Of this amount, $240,000 buys reagents and disposables consumed in test performance. The remaining $60,000 buys supplies needed regardless of test volume. Thus, projected supply costs at half the current test volume are $180,000 a year (half of $240,000 plus $60,000)--assuming that present reagent discounts would continue at a lower volume. (In reality, reagent discounts would probably be somewhat smaller, but for simplicity we left this factor out.)
We estimated quality control at $30,000 for materials and $50,000 for reagents consumed in QC-related tests. To estimate fixed and variable QC costs at 50 per cent of current workload, we surveyed the clinical chemistry director and supervisors.
Based on their input, we decided that 75 per cent of the existing controls would still be necessary at 50 per cent volume, giving us residual quality control expenses of $60,000.
Labor was the hardest factor to project at the baseline volume. Again, we postulated that half as many specimens would be drawn per day, batches would be 50 per cent of current size, and the total number of tests per day and per shift would be half that now performed. And we made one further assumption: that the quality of service would not be compromised, even at half our present workload. Turnaround time, frequency of test performance, and lab responsiveness were not to suffer.
Figure III shows the chemistry section's staffing needs in full-time equivalents (FTEs) at 50 per cent workload versus current workload. We now have 27.9 FTEs in the chemistry lab. Even at half our volume, the director, supervisor, four section supervisors, and the research and development technologist would all still be necessary. The number of technologists who could be saved with a 50 per cent in workload is estimated at 4.2 FTEs. This includes 2.2 daytime and weekend FTEs, 1 of 4 second-shift FTEs, and 1 of 2 third-shift FTEs.
We used an analogous approach for the other sections of the laboratory. Figure IV compares FTEs needed for current and 50 per cent workload in chemistry, blood bank, hematology, microbiology, and shared lab services. The total number of employees decreases from 103.15 to 88.85--an effective drop of 2.9 FTEs per 10 per cent drop in workload. Note that shared lab services are least sensitive to volume changes, while microbiology is most sensitive to them.
Figure V summarizes how various chemistry costs compare at both workload levels. At the lower volume, total expenses decrease some 20 per cent from $1,120,000 to $900,000 annually, producing savings of $220,000 in chemistry.
In theory, chemistry department costs rise in a linear relationship to test volume. In practice, the relationship is somewhat more variable, but we will assume a linear correlation. Figure VI shows this relationship, assuming that workload cuts are randomly distributed among all types of testing.
Annual performance of 500,000 chemistry tests would cost $900,000; the current workload of 1 million tests now costs $1,120,000. Divide the cost difference of $220,000 by 500,000 tests, and you obtain the variable component per unit test: 44 cents.
What does this variable component tell us about the financial implications of a campaign to decrease overutilization? It's the amount of money that would be saved for each test not performed between 50 and 100 per cent of present workload. If the laboratory were successful in cutting test volume by one-fifth--that is, by 100,000 tests a year--we would save a total of $44,000 (i.e. 100,000 test X 44 cents per test result).
We have avoided referring to fixed costs, for a good reason. Fixed costs, we believe, conform more closely to theory than to reality. We can estimate them, however, if we extrapolate the curve in Figure VI to zero testing annually. If the cost-volume relationship remained linear, fixed costs would equal $680,000 a year, or some 60 per cent of total chemistry expenses. From a purely academic perspective, we can state that clinical chemistry costs are probably 60 per cent fixed and 40 per cent variable.
After this foray into flexible budgeting, we applied the same analysis to the laboratory as a whole, excluding costly variable factors like send-out tests and procured blood bank products. The inclusion of shared laboratory services--computer, phlebotomy, managerial and secretarial staff--adds another important fixed cost at 50 per cent of current workload.
The bottom line, we found, was that only one dollar in six would be saved if the average workload were cut in half throughout the laboratory!
Obviously, we can view the hospital laboratory as a business with a relatively small component of variable costs. Even a reduction of 50 per cent in total volume will yield only a 20 per cent reduction in total expenses. (Our $220,000 savings from halving the workload equals 20 per cent of the chemistry section's $1,120,000 in current expenses.)
These very high fixed costs present the clinical laboratory with a dilemma. Management can grab the dilemma by either of two horns: the compression strategy or the expansion strategy.
* Compression. This approach assumes that the laboratory's overriding goal is to decrease utilization and minimize in-house testing. As instruments outlive their usefulness, they can be replaced by smaller, less expensive ones. Laboratory service may suffer in terms of turnaround time and frequency of testing. For example, a Stat electrolyte determination may take an hour rather than 30 minutes, or performance of thyroxine tests may be cut back from three days a week to a weekly schedule. When appropriate, one technologist may handle several work stations, and an evening or night technologist may perform tasks in various sections. The lab's space may even be reduced to cut expenses.
These compression tactics will yield some savings, and many laboratories around the country are practicing them. Their usefulness is limited, though. In most labs, expenses will never penetrate below a certain fixed floor.
* Expansion. This alternative takes the opposite tack. It aims to increase the workload and do more tests in-house. Let's say the lab performs 2 million chemistry tests a year instead of 1 million. Naturally, it would be more expensive. But it would not be twice as expensive. Given our variable cost of 44 cents per test result, we can project that the additional million tests would cost $440,000 more than the current budget. Total annual expenses would increase to $1.56 million--admittedly a conservative estimate, since a twofold rise in volume contains other hidden expenses like marketing and delivery costs and additional instrumentation.
The expansion option won't work for the majority of laboratories. There just aren't enough test requests generated to warrant expansion in every lab across the country. And those that do try to expand face stiff competition from referral facilities, other hospital laboratories, and physicians' office labs.
For some laboratories, however, successful marketing efforts and higher volume will maximize returns on those high fixed costs. When we showed our administrators that each additional test beyond our current workload would cost only 44 cents to perform, they agreed that the expansion strategy was worth pursuing.
Flexible budgeting provided us with other strategic information--facts and figures that can protect the laboratory from indiscriminate budget-cutting. Administration pointed out that the hospital's census was dropping. One entire nursing ward, in fact, had recently closed. We agreed to investigate how our staffing needs would respond to a cut in volume, and presented the figures described above.
At first, our administrators were a little incredulous that we could cut so little in staff and other costs, even at half our workload. As we walked them through each stage of the analysis, however, they acknowledged that the only way to achieve significantly greater savings was to cut service. Right now, that's not an option we are ready to consider. And, as we mentioned, they gave the nod to a more aggressive policy of marketing and expanding laboratory services.
To review our conclusions: The clinical laboratory must now be run as a business, and one key element of a businesslike approach is the flexible budget. The goal of flexible budgeting is to reveal the variable cost component in testing. With this information, we can relate costs to actual workload. It's the nature of our particular business, however, that our expenses are fairly inelastic in relation to workload.
To put it another way: The flexible budget is a valuable management tool--but in a laboratory setting, it's not all that flexible!