Or could you be setting your organization up for stockouts?
by David Hermann
A
s a supply chain consultant, I often get asked
about national benchmarks for inventory values. This is usually a request
for total inventory value or the value per procedural suite (i.e., operating
room or cath lab). While in most areas benchmarks are quite useful, it is my
experience that managing to arbitrary inventory benchmarks potentially
leaves money on the table or, worse, could put an organization at risk for a
high percentage of stockouts.
Inventory benchmark limitations
Inventory benchmarks are by their nature an abstraction that
usually buries many assumptions into the metric, including economies of
scale, quality and reliability of local distribution, geographic area and
other factors. For example:
• A high performer’s (low percentile) benchmark value could
represent a large operating room or cath lab with significant economies of
scale. The larger the department, the more the variability of demand for any
one product can be reduced (for the same reason diversifying a stock
portfolio reduces risk).
• A high performer’s benchmark value could be a hospital
that is geographically close to the vendor’s distribution point. This could
imply the facility experiences low lead times and potentially low lead time
variability.
• A poor performer’s (high percentile) benchmark value could
be a hospital in a remote area with less-than-ideal distribution. This could
mean the facility experiences both high lead times and high lead time
variability.
If inventory benchmarks may not be useful, what can
materials managers do to quantify inventory opportunity? Simply put, maximum
inventory value should be determined through a combination of the optimal
reorder point (ROP) and the reorder quantity (ROQ).
Calculating reorder points
Reorder points need to be based on four factors: Average
Daily Demand, Variability of Daily Demand (measured as standard deviation),
Average Lead Time (time from order to delivery) and Variability of Lead Time
(again, measured as standard deviation). If one sets the ROP without taking
the above four variables into consideration, the facility or department may
carry too little inventory and suffer constant stockouts or carry too much
inventory and tie up capital in non-productive assets.
"Standard deviation" is a statistical measurement of
variability which can be defined as how wide the bell-shaped curve is. As
the bell gets wider (increases in variability), the standard deviation
numerically increases. While standard deviation requires some sophisticated
math to calculate, every major spreadsheet and many scientific calculators
will calculate it for you.
While a detailed explanation of statistics is out of the
scope of this column, it is important to stress that accounting for
variabilities of demand and lead time are extremely important. The higher
the variability, the more inventory you will need to keep on-hand to ensure
a high availability of product. If you only use averages, your inventory
service level is 50 percent by definition.
Several solutions for ROP have been provided by operations
research studies; however, over the years, I’ve developed a formula that
accounts for variable demand and variable lead time with minimally
complicated math (see figure 1). The benefit of this formula is that it
recognizes both the effects of singular spikes in demand or lead time and
the effects of overall trends. While the formula here yields a service level
of 95 percent, it can easily be adjusted to change the service level up or
down. It also has the benefit of being easy to implement in a spreadsheet,
permitting inventory managers to download usage histories and lead times
from their materials management information system into a spreadsheet, apply
the ROP calculation, then upload the result back into the materials
management information system.
Figure 1 |
ROP = (average daily demand + 2 *
standard deviation daily demand) * (average lead time days + 2 *
standard deviation lead time days)
EOQ = square root ((2 * total annual demand * reorder cost per PO
line) / carrying cost) |
Calculating reorder quantities
While there are a few ways to calculate reorder quantities,
the industry generally uses "Economic Order Quantity" or EOQ (see figure 1).
It requires that the person calculating EOQ have access to the hospital’s
carrying costs and reorder costs per purchase order line. This can be a bit
of a challenge because it is my experience that few hospitals have these
calculations easily (and accurately) available and it is calculated using
estimates based upon industry "benchmarks."
Moving beyond inventory benchmarks
It is important that hospital materials managers move beyond
using arbitrary benchmarks for managing their inventories. By managing
inventory to a maximum value of ROP + ROQ, hospitals can reduce the amount
of capital tied up in non-producing assets without sacrificing service
levels, product availability and patient care. If inventory benchmarks must
be used, it is important to understand all of the assumptions buried in that
benchmark. If you do, you will avoid setting your organization up for
stockouts.
David Hermann is with Aspen Healthcare Metrics, a
MedAssets company, Englewood, CO
(dhermann@aspenhealthcare.com).