The two organizations did a rapid improvement event to
get at the root causes of back orders. They found that certain
factors often lead to “surprise” demand spikes or shortages,
such as:
- A hospital hiring a new surgeon
- Doctors attending a major clinical conference and coming back with requests for new products
- Members of a regional GPO signing a committed-volume contract with a single supplier
The speakers said such changes can and should be better
anticipated, and that expected shifts should be communicated
through the supply chain. Still, stock-outs can’t be
prevented entirely. They recommended that providers collaborate
in advance with their distributor on an approved substitution
list. “You even need to have a sub for the sub,” said
Clouse, because if you move from one product to its sub,
demand for that second item will spike and may exceed
supply.
Collins asked suppliers to proactively identify the best subs
when a product is back-ordered. “We have to take care of
patients, so please tell us what alternative products will work,
even if they’re from your competitor.”
Build Clinical Support And Trust
Russell Royer, Director of Heart and Vascular Services for
Loma Linda University Medical Center, agreed that while
there is a mystique around the unpredictability of healthcare
demand, it’s really not that much of a surprise: “We can
predict the length of a procedure, the supplies that are
needed. Almost everything is algorithmic – if not A, then B.”
The challenge, said Royer, is convincing clinicians to trust
the system. If they don’t have faith that the right products will be available when needed, they will work around the
system through behaviors like over-ordering and hoarding.
Make Demand More “Forecastable”
The more volatile your demand is, the harder it is to forecast
for and prepare for, according to forecasting expert Mike
Gilliland. The “surest strategy for improvement is to make
demand more forecastable by reducing volatility.” He noted
that while some demand spikes are unpredictable, a considerable
among of demand variation is “artificial volatility”
caused by business practices.
To make demand more forecastable, he recommended:
- Rethinking practices like end-of-quarter promotions that lead to a spike in buying followed by a big drop
- Avoiding SKU proliferation, because the fewer stock-keeping units you have, the less forecasting you have to do
- Forecasting and preparing for a demand range rather than a single point of expected demand
What’s Next
Participants at the Supply Chain Visibility Conference
were enthusiastic about addressing supply and demand issues
through an ongoing industry effort. Priorities they identified
included:
- Turning up the volume on this issue: define the pain, and establish a “maturity path” to help organizations improve their practices
- Creating event response guides built around specific scenarios such as hurricanes or epidemics
- Agreeing on standard definitions for common metrics such as rolling back order or fill rate
- Developing standard substitution lists