It’s road construction season across the country and crews are working to smooth out the potholes left behind by another season of freezing and thawing. And, while these efforts will improve driving conditions for a few months, we all know the fresh blacktop won’t last forever. Traffic and weather will eventually wear the new road surface down again and the potholes will reappear, maybe even increasing in number and severity.

Patched blacktop might not seem like a natural analogy for hospital supply chain, but there’s a similar phenomenon at play when it comes to physician preference items (PPI). 

Physician preference cards are meant to detail the preferred items a physician uses for a given procedure. However, it doesn’t take long for gaps and inconsistencies to surface in this process as physician preferences or the realities of surgical situations change, calling for a different set of supplies. 

When the supplies that were originally called for on the preference card can no longer be used or returned to stock, this leads to hospital supply waste. In fact, it’s estimated that inefficient preference card management in the United States is a major contributor to roughly $765 billion in waste each year — or about a quarter of all healthcare spending. That’s a significant hit to hospital revenue.

The good news is that advanced artificial intelligence (AI) and machine learning can be used to automate the preference card process and deliver real-time, shared data about actual supply utilization. Effective hospital supply chain software will use this data to select items for procedures based on actual consumption by a specific doctor for a specific procedure, with adjustments made for the demographics of the patient. This reduces the chance of items coming into the OR only to be thrown away, and can save hospitals money by reducing waste. 

And unlike the highways and byways of America, the same old supply chain potholes won’t keep popping up because the system of managing materials on a preference card is now automated. The “highway” never develops a “pothole” because the computer stops them before they materialize. At the end of the day, this makes managing PPI feel like smooth sailing.