First of all, Ted Cohen has written a very nice article.1 It provides good, basic instructions for someone not familiar with looking at their operation as a business. I believe there are some areas where additional clarification would help those less familiar with the subject, and some of the definitions in the metrics could be cleaned up a bit. From the description in the article, the metric definitions were provided by AAMI rather than Cohen.
I think that the definitions of parts usage are confusing and overcomplicated related to what is included in in-house maintenance and what is included in external maintenance. The focus for parts related to in-house maintenance appears to be on stocked versus nonstocked parts. As Cohen notes, there is a move to “just-in-time” parts acquisition. Parts, especially for high-end equipment, in many cases are too expensive to keep on the shelf. Whether parts are stocked or not is irrelevant to the metric. If internal staff services the equipment, it's an in-house maintenance cost. Any parts required for that service are included in that cost. If a vendor services the equipment, it's an external maintenance cost. Any parts included in that service are included in that cost. Separating those costs can be done either by using different accounts within the budget or using the computerized maintenance management system. The bottom line is that the definitions related to internal and external parts usage could be renamed and simplified.
The other area I wanted touch on is the cost-of-service ratio (COSR). One of the most important considerations for managers is whether they intend to use the COSR as an internal metric only, trended over time, or if they plan to use it to compare themselves with other organizations or a generally accepted standard. (There is no formal standard for COSR, but many quote a figure of 5% or less as a “good” number.) If you are only trending COSR internally, then it doesn't really matter what you use for acquisition cost or how accurate you are at including only acquisition costs for devices for which you actually bear the service costs. The important thing to keep in mind in this situation is that you need to be consistent year to year. One problem I have seen is that managers manage this metric as an internalonly metric, but if they think it looks good, they then will compare it with the widely quoted standard of 5%, or to other organizations—whether that is to their boss or in articles/presentations. If COSR is going to be used to compare organizations, acquisition costs have to be consistent across organizations. The actual cost is irrelevant. All that matters is that having everyone use a fixed, estimated cost for each type of device makes it more of an apples-to-apples comparison.
And yes, I realize that devices that have a lot of options that affect the acquisition cost (e.g., magnetic resonance imaging and computed tomography scanners) make this a bit more complicated. That doesn't make the creation of a fixed acquisition cost table impossible, just a bit more challenging. Creating this table for everyone around the country to use can and should be done. The other thing that can make COSR inaccurate is the degree of service provided on a device. Cohen is accurate with his comment that “whatever is included in the numerator should be included in the denominator, and vice versa.” But it goes beyond that. Organizations will include the full acquisition cost for devices for which they bear no service cost, such as those provided as part of a reagent rental agreement, or for which the clinical department pays the service contract. Or devices for which they only do planned maintenance inspections and the department pays for any repair services. Or devices that are repaired by the in-house department but the clinical department pays for the parts. All of those approaches make the COSR look better than it really is.
COSR is calculated by summing the total acquisition costs for all devices that your department services and dividing it by total service costs. With that in mind, if you use a standardized acquisition cost for all devices and have the resources and ability to capture all of the service costs, including those not paid by your department, then the COSR would be accurate and have value for comparisons across organizations. If you are unable to capture service costs not paid by your department, the only way for COSR to have value across organizations is if definitions are universally adopted for calculating how much of the acquisition value of a particular device you can use for your COSR calculations, depending on the percentage of service costs you actually bear. An example is provided in Table 1. This doesn't have to be perfect, but creating a simple table that shows the percentage of acquisition costs that can be claimed for the most common types of services provided can be done.