State Opioid Database Links with EHR in Pilot Program
This article originally appeared in HealthLeaders Media March 14, 2017.
Despite the growing number of online prescription drug databases that aim to counter the misuse of opioids and other controlled substances, many physicians don’t use them.
In an effort to encourage physicians to use the databases, a pilot program at the University of Colorado Anschutz Medical Campus has linked the Colorado Prescription Drug Monitoring Program (PDMP) with the University of Colorado Hospital’s EHR/EMR.
PDMPs are state-run databases to track information related to a patient’s controlled substance prescription history and are used to monitor suspected abuse or diversion, according to the CDC.
“Prior to now, people have tried to gauge the value of PDMPs based on statewide trends,” says researcher Jason Hoppe, DO, an associate professor of emergency medicine at the University of Colorado School of Medicine.
His team looked at provider decision-making and followed patients to see the impact of those decisions on patients’ lives.
The project, funded through a federal grant through the Bureau of Justice Assistance, was rolled out several months ago at the school of medicine’s emergency medicine department after roughly a year-long building phase, which linked the PDMP with the hospital’s Epic EHR/EMR.
Before the two systems were linked, it could take up to three minutes for clinicians to make their way through the PDMP.
Barriers to Clinician Adoption
They would have to leave the EHR/EMR, visit a separate website, sign into the PDMP using different credentials, navigate several links to get to the correct page, enter patient information, and sort through a list of possible matches before finally seeing the patient’s aggregated prescription history.
Hoppe says providers used the system only about 30% of the time when prescribing a controlled substance.
Some states require providers to consult the PDMP before prescribing, but Colorado is not one of them.
The CDC touts the requirement as key to the success of the PDMPs. In 2012, New York began requiring prescribers to check the state’s PDMP before prescribing opioids and in 2013 it saw a 75% drop in patients’ seeking the same drugs from multiple providers.
The University of Colorado pilot allows providers to search with a single click and from within the EMR itself. “That entire process takes less than two seconds,” Hoppe says.
Rolling Out the Program in Four Phases
Of course, it’s not only accessibility that stops clinicians from using the PDMP. Hoppe says other factors include being unsure what to do with the database information once they get it, and being unaware that a patient is at-risk for opioid abuse.
The pilot program uses a “stepped wedge” research approach that rolls out the program in four phases:
- Evaluating the baseline
- Turning on the system and giving clinicians the information in raw data
- Giving the clinicians data using a risk-stratification tool
- Mandating the tool’s use
-
Each phase takes about three months and will be staggered with the rollout to each research site, allowing the sites to be compared to each other, as well as to themselves.
The school of medicine’s ED is the first site, and Hoppe says the second set of sites will get their EHR/EMR-linked system turned on in about two weeks. When they do, the first site will move onto the next phase.
Within the next three months the system will be active in five EDs with a combined total of 270,000 visits per year, or about one-quarter of hospital admissions through the state.
No official training is provided because researchers want the data to be generalizable.
Researchers want to know:
- Whether provider use of the PDMP increases
- Whether the PDMP makes a difference for high-risk patients
- What happened to patients a year before, during, and a year after their ED visits using de-identified patient data
It’s too early for the pilot to have any data yet, but Hoppe says that anecdotally, provider satisfaction is high, and the time-saving aspect of the tool is popular.