What the researchers studied
Knight and colleagues evaluated Flowlytics, an Invistics system, using historical records from 10 hospitals in four health systems and 22 known diversion cases.1
In blinded retrospective testing, cases were detected 160 days earlier on average, with a median improvement of 74 days, compared with existing methods.1
How to read those findings
The study focused on known cases and nursing personnel. Its small case sample and incomplete assessment of false-positive performance in the historical analysis limit generalization. WasteWitness was not evaluated.1
Ask vendors to show the events behind an alert and explain why it was prioritized. Consider the effort needed to review a concern, as well as the ability to find it.
Connect patterns with evidence from the waste
Analytics finds patterns in medication activity. Chemical waste verification adds evidence about what was actually discarded. PharmID sees value in using both.
Records can balance even when the discarded drug or concentration differs. WasteWitness assesses identity and concentration within supported criteria and measures volume separately, adding evidence for review.
Evaluate how these findings fit alongside your existing analytics and support follow-up decisions.
NIDA funded the research; Knight and May disclosed employment with Invistics. The detection gains cannot be attributed to WasteWitness.1
Read the original sources
- Detecting drug diversion in health-system data using machine learning and advanced analytics
Knight T, et al. American Journal of Health-System Pharmacy. 2022;79(16):1345 - 1354. DOI: 10.1093/ajhp/zxac035.
Read the full research article (opens in a new tab)
