Real-Time Prescription Monitoring and Alerting

How can I visualize and analyze all of my HL7 data in one place and in real-time?


The Problem

Through new payment models and electronic health records (EHR), a healthcare provider had access to significant amounts of clinical data. They recognized the opportunity to use this data for healthcare analytics but the traditional approach of business intelligence requiring extraction, transformation, and load (ETL) from many different databases, staging in a temporary data store for schema validation, and then placing in a data warehouse and running BI tools on top was not only slow, complex but very expensive. They were looking for a better alternative that would be inexpensive, simple, and fast. Doing so would help to improve care, save lives, and lower costs.

An area of particular interest involved minimizing fraudulent activity surrounding drug prescriptions. This interest was due in part to DEA auditing of medical providers. In 2014 more than 16 million Americans reported using a prescription drug for nonmedical reasons, numbers that caught the DEA's attention.

As a DEA registrant, the healthcare provider knew that any violations uncovered in an audit could result in revocation of DEA registration, in civil fines, and even in criminal conviction. They realized that their HL7 messages were a rich data source containing all the information surrounding prescriptions, but they lacked a simple way to mine the information.

Desired Outcome

  • A simple way to extract, analyze, and report on HL7 message data with no modifications to their EHR or other applications.
  • Mine and report HL7 messages in real-time for data involving drug prescriptions.
  • Trend and alert information based on drug type, facility, doctor, time, and frequency of prescription.

As a DEA registrant, the healthcare provider knew that any violations uncovered in an audit could result in revocation of DEA registration, in civil fines, and even in criminal conviction. They realized that their HL7 messages were a rich data source containing all the information surrounding prescriptions, but they lacked a simple way to mine the information.


The Solution

The healthcare provider's interface engineers wanted to track prescriptions written for the top 10 most commonly-abused prescription drugs. Prescriptions typically are found within either ORM or RDE HL7 message types, or in this case, within the RDE message. ExtraHop Application Inspection (AI) Triggers for HL7 messages provide not only the full message but also break out each segment and even the individual fields within the segment.

Once they had a network SPAN in place feeding ExtraHop a copy of all application traffic, they wrote a simple AI Trigger that would execute whenever an HL7 RDE message event was seen by ExtraHop, extract data from the RXE or pharmacy order segment and the drug identification field, and perform a simple comparison with the watch list of abused drugs. A trend-based alert fired automatically whenever the prescription counts deviated from normal numbers.

Within a month after deploying the ExtraHop platform, the alert caught a serious abnormality. Prescriptions for OxyContin began to suddenly rise for one doctor. Once informed, the clinic discovered that the doctor was missing a prescription pad, and prescriptions were being forged in his name.

User Impact

By using the ExtraHop platform, the healthcare provider not only has a platform for real-time analysis of prescriptions but is extending it to incorporate other analytics around admission, discharges, and transfers which should take less than a few hours to implement and plans to rapidly expand to other real-time analytic scenarios at no additional cost. They also are applying the platform's native capabilities to monitor the performance of their EHR delivered over Citrix to automatically discover, map, and show all dependencies and performance bottlenecks across their application delivery chain—something they've been blind to and wholly dependent upon their application vendor.

With the drug prescription analysis alone they estimated a $100,000 savings in annual audit costs and anticipate reducing their EHR consulting costs by 50 percent because the majority of that cost had been centered around performance tuning and troubleshooting. Ultimately the Director of Applications and the Director of Informatics said that ExtraHop puts them in the position of greater control of their environment and the ability to create new patient care insights much more cost effectively, faster, and simpler than previously possible.

Contact us Try our free online demo