A commonly used benchmark for maintenance activities is the ratio of maintenance cost to asset value (MC/AV). ^{1, 2} Some medical equipment maintenance professionals also claim MC/AV to be the best or most reliable benchmark.^{3, 4} Maintenance cost includes both scheduled (often erroneously called preventive maintenance) and nonscheduled (i.e., repairs) maintenance costs, whereas asset value has various interpretations. This article challenges the validity of MC/AV as a benchmark or even a key performance indicator for maintenance activities in most, if not all, industries.

To be clear, the usefulness of the ratio of repair cost to asset replacement value is not being questioned. Whenever a significant or catastrophic failure happens to a piece of equipment and the related repair cost exceeds a significant fraction—as defined by each organization—of the replacement cost (minus resale or salvage value), maintenance leaders should evaluate the wisdom of repair and discuss it with the executive team. Caution should be taken in using replacement cost and not the depreciated book value, as the latter is nothing but an accounting exercise.

On the other hand, it is unclear why one should use MC/AV for benchmarking maintenance. Implicit in this ratio is the assumption that an asset with higher value will demand a higher maintenance expenditure. While it is intuitive that the MC for a moped is lower than that of an automobile, which in turn has a lower MC than an aircraft, the assumption that the MC/AV ratio is constant for transportation equipment doesn’t appear reasonable under scrutiny.

As an example, let’s examine a single class of transportation equipment: automobiles. In order to have significantly different asset values, let’s consider automobiles in three different categories: standard, luxury and ultra-luxury. Figure 1 shows the MC/AV for three such vehicles computed with data collected from Edmunds.com. ⁵ While MC does increase with the car’s manufacturer’s suggested retail price (MSRP)—most likely due to higher costs of replacement parts for the more luxurious vehicles produced at lower volume—MC/AV actually decreases as MSRP increases. Therefore, the maintenance team of a fleet of luxury vehicles cannot claim to be better performers than other teams that maintain standard vehicles. Hopefully, you are not rushing out to buy ultra-luxury cars because of their lower MC/AV!

Figure 1: Average annual MC and ratio of MC to respective MSRP for three gasoline-fueled automobiles of different brands and models and an electric vehicle; MC was obtained by averaging the sum of maintenance and repair costs for the first 5 years after acquisition (Data collected from Edmunds.com)

To make the automobile example more interesting, a purely electric vehicle was added to Figure 1. The annual MC of this vehicle is obviously lower than those with an internal combustion engine. Furthermore, its MC/AV is also lower than the others, except for the ultra-luxury car, although this ratio is likely to increase with future reductions of battery cost and, thus, MSRP. The lesson here is that it is not enough to consider AV; MC is also significantly impacted by technology evolution. This has been most evident in the last half century with the transition of vacuum tubes to transistors, integrated circuits and microprocessors.

Aside from the conceptual challenge, MC/AV also has significant practical challenges with the determination of AV summarized as:

1. Purchase Price as AV: Unlike the MSRP for automobiles, most industrial and professional assets do not have list prices. Even when a list price exists, discounts are typically offered by vendors to buyers with large purchasing power (i.e., bargaining power) or high prestige (i.e., for future sales reference). In this case, MC/AV would favor organizations that pay more for the same asset without any relationship to performance. Furthermore, few organizations keep accurate records of purchase price, with details on optional accessories, extended warranty, freight and installation costs.
2. Replacement Cost as AV: Like purchase price, the replacement cost varies from one organization to another due to bargaining power and marketing considerations. While easier to obtain compared to original purchase price, it is laborious to keep updating this cost every year and often difficult to find exact replacement for older equipment, as technology is evolving rapidly and new features are constantly being introduced.
3. Standardized Price at Purchase or Replacement as AV: In lieu of list price, some proponents of MC/AV have suggested adopting standardized prices at purchase or replacement. While this is an improvement over individual purchase or replacement cost, it is challenging to establish industry-wide standardization. Until standardized prices are widely available and adopted, this interpretation is not likely to be useful.

To illustrate how MC/AV can be misleading, let’s consider the cost to maintain medical equipment in acute care hospitals. For Figures 2-6, data were collected in 2013 from about 300 American hospitals of different sizes and teaching characteristics (major >400 beds, minor, nonteaching and unknown) by Truven Health Analytics through its ActionOI® program. ⁶ Since not every hospital provided data on all the metrics, each graph has a different number of data points. Note the logarithmic scales on both axis. The dashed lines in the graphs are visual guides for readers based on the author’s experience instead of statistical fits due to the high number of obvious outliers. Verification of data accuracy is not possible because hospital identities are protected by Truven.

Figure 2 shows MC/AV (with purchase price as AV) for hospitals of different sizes and teaching characteristics. It appears that major teaching hospitals are just as cost-effective as smaller and nonteaching hospitals, except perhaps for a few very large ones.

However, when an output metric is used as the denominator, one can see that most larger and teaching hospitals are actually not as cost-effective as the rest. One output often used for healthcare is adjusted patient discharges, ⁷ where the actual patient discharge is adjusted by the ratio between total revenue and inpatient revenue in order to account for outpatient care. Figure 3 shows most major teaching hospitals generally spend more on MC than the rest.

Figure 2: Relationship between medical equipment MC and its acquisition cost

Figure 3: Relationship between medical equipment MC and adjusted patient discharges

It is not difficult to understand why major teaching hospitals have higher MC. These organizations typically have more equipment—in quantity, variety and sophistication—than the rest due to their educational needs (see Figure 4). This leads to higher equipment MC. The fact that they enjoy discounts on a few high-end, “showcase” equipment does not significantly reduce their total equipment purchase cost (see Figure 5), as they still have to pay near normal price for the remainder of equipment. In addition, most major teaching hospitals offer higher employee compensation than others, especially those located in expensive major metropolitan areas on both coasts.

Figure 4: Relationship between the number of medical equipment and the number of beds actually in use by each hospital (known as operating beds); note that major teaching hospitals are better fitted with 30 capital devices (equipment) per bed versus 18 for the others

Figure 5: Relationship between medical equipment purchase cost and the number of operating beds

Therefore, it is not surprising that major teaching hospitals have MC/AV comparable to smaller and nonteaching hospitals. Their higher MC is compensated by higher AV. However, this does not justify using MC/AV as a benchmark. Like the classical statistics lesson, “correlation does not imply causation,” one should not take the fortuitous coincidence of comparable MC/AV among hospitals of different sizes and teaching characteristics as proof of it being a good benchmark. To blindly follow MC/AV would lead hospitals to spend more in MC than truly necessary for good patient care.

MC/AV could be a reasonable benchmark if the organizations being compared have exactly the same assets (e.g., brands and models of equipment), with only difference in quantities. This scenario, however, is very unlikely. Even car rental companies are likely to have vehicles of different brands, models, sizes and luxury levels.

From this analysis, it is not surprising that some industries have developed their own financial benchmarks. For example, facility maintenance professionals use square feet as the denominator, ⁸ while commercial airlines use flight hours.⁹

For those who prefer to benchmark in the form of a ratio or percentage, the cost of goods (or services) sold (COGS) seems to be a reasonable alternative denominator, as long as the benchmarking is made within the same industry and includes cost of living adjustments for regional salary differences. Figure 6 shows MC as a function of a hospital’s total operating expense. Most hospitals, including the major teaching ones, have MC/COGS in the range of 0.5 to 2 percent. Notice in Figure 6 that some hospitals have much higher operating expenses than others (i.e., data points shifted to the right). These are typically prestigious, major teaching hospitals that can afford to have consistently negative annual operating incomes due to significant donations and endowments they receive.

Figure 6: Relationship between medical equipment MC and a hospital’s total operating expense

As MC/COGS shows, every benchmark has its own imperfections. Therefore, it is recommended that benchmarking attempts be made with as many benchmarks as possible, including dissecting MC into internal labor, replacement parts, external labor and material, service contracts, etc. This would be analogous to improving location accuracy of a global positioning system (GPS) with more satellites. In fact, a multidimensional model has been proven to work fairly well in comparing or predicting hospital equipment MC. ¹⁰

In essence, maintenance professionals should consider abandoning MC/AV or at least supplement it with additional benchmarks. There is no irrefutable reason that MC should be related to AV, be it purchasing or replacement cost. Experience has demonstrated that MC depends primarily on the main building components (e.g., mechanical, electrical, electronic, etc.) of deployed assets, as well as how often these assets are used and how well they are cared for by the users. Adopting MC/AV as a financial benchmark or key performance indicator for measuring and comparing cost-effectiveness of maintenance teams—especially if used alone or as the primary indicator—would discourage improvements in financial performance.

REFERENCES

1. Levitt, Joel. Handbook of Maintenance Management (2nd Edition). Norwalk: Industrial Press, 2009.
2. BSI. Maintenance. Maintenance key performance indicators, Standard Number: BS EN 15341:2007. London: BSI Group, 2007.
3. Cohen, Theodore. “Validating medical equipment repair and maintenance metrics, Part II: Results of the 1997 Survey.” Biomedical Instrumentation & Technology, 1998: Volume 32, pp 136-144.
4. Lynch, Patrick K. “Three indicators of an excellent HTM department.” Biomedical Instrumentation & Technology, 2013: Volume 47, pp 512-513.
5. Edmunds. True Cost to Own (TCO). https://www.edmunds.com/tco.html.
6. IBM Watson Health. “How Do I Deliver Top-Quality Care While Keeping Costs Under Control?” Truven Health ActionOI®. hhttps://truvenhealth.com/markets/provider/products/ops-financial-performance-improvement/actionoi.
7. Chansky, Brian, Garner, Corby A., and Raichoudhary Ronjoy. “New measure of labor productivity for private community hospitals: 1993–2012.” Monthly Labor Review, U.S. Bureau of Labor Statistics, October 2015. https://www.bls.gov/opub/mlr/2015/article/new-measure-of-labor-productivity-for-private-community-hospitals-1993-2012.htm.
8. Kimmel, Peter S. “Benchmarking for Facilities Professionals.” IFMA Foundation. https://foundation.ifma.org/docs/default-source/Whitepapers/benchmarking-for-facility-professionals-ifma-foundation-whitepaper-small.pdf?sfvrsn=4.
9. Buyers, Tom. “Optimizing Airplane Maintenance Economics.” AERO Magazine, QTR 01-2010. http://www.boeing.com/commercial/aeromagazine/articles/qtr_01_10/2/.
10. Wang, Binseng. Medical Equipment Maintenance: Management and Oversight. Princeton: Morgan & Claypool Publishers, 2012.