Diverters of controlled substances are getting more savvy about finding ways to get around documentation safeguards and protocols. Using machine learning, Bluesight for Controlled Substances is able to recognize patterns learned over time and apply them to client data sets to expose bad behavior.
For example: the perfect documentation fallacy.
When we compare what was dispensed, wasted, and returned from the OR’s automated dispensing cabinet against what was administered in the EHR, everything should add up:
Dispense – Admin – Waste – Return = 0
In this particular hospital, Bluesight for Controlled Substances noted considerable lags from the time some OR patients’ dispense records were opened to the time the administration records were completed. At first glance, everything looked okay, since the medications being checked out during the time the cases were open were also recorded as being administered to patients. Since the “math” was in order, these transactions wouldn’t be noticed in a manual audit or in standard variance reporting.
But even if the documentation looks perfect, a seemingly innocuous element—like time—can make a significant difference.
With Bluesight for Controlled Substances, a new picture emerged, hidden in the perfect documentation: two providers were routinely taking 10 hours or more to complete their controlled substance documentation.
This particular “learned” pattern can now be applied to all client data sets to reveal the same behavior, allowing them to investigate, and truly close the loop. The Bluesight for Controlled Substances prescriptive intelligence platform analyzes usage data to focus on important variances—making the invisible visible.