Conference Coverage

N.J. law, EMR alerts appear effective at reducing opioid prescriptions



Implementing an alert on electronic medical records to reinforce a change in public policy appears to be having a positive effect on opioid prescriptions.

Dr. Margaret Lowenstein Gregory Twachtman/MDedge News

Dr. Margaret Lowenstein

Researchers looked at prescribing patterns of doctors in the Penn Medicine health system, which straddles both the Philadelphia area and southern New Jersey, following the implementation of prescribing limits in New Jersey.

The law in question is a 5-day limit on new opioid prescriptions, which was passed in February 2017 and implemented in May 2017. Penn Medicine implemented an EMR alert in their New Jersey locations to alert physicians within the Penn Medicine system of the change in their state law 2 months after the law went into effect. Researchers looked at prescribing patterns before passage, during the transition between passage and the implementation of the EMR alert and following implementation of the EMR alert, as well as secondary outcomes such as rate of refills, telephone calls, and utilization.

“The implementation of the prescribing limit and EMR alert was associated with a decrease in the volume of opioids prescribed in acute prescriptions without changes in the rates of refills, telephone calls or utilization,” Margaret Lowenstein, MD, of the University of Pennsylvania, Philadelphia, said at the annual meeting of the Society of General Internal Medicine.

“This combination of the policy and the EMR alert may be an effective strategy to influence prescriber behavior,” she added.

Researchers compared outcomes before and after the implementation of the law in New Jersey, using prescribing patterns in Pennsylvania as the control. The cohort of patients was those with a new opioid prescription within Penn Medicine ambulatory nonteaching practices. It excluded specialties not represented in both states as well as patients with cancer, those in hospice and palliative care and those in treatment for opioid use disorder, since the law does not apply to those groups.

In New Jersey, there were 434 patients receiving new prescriptions in the 12 months prior to the implementation of the law, with 234 patients receiving new prescriptions in the 9 months after the EMR alert was implemented in New Jersey. In Pennsylvania, the cohort included 2,961 patients prior to the law going into effect and 1,677 after the EMR intervention went live in New Jersey.

For New Jersey, the morphine milligram equivalent (MME) per prescription was steady at about 350 during the period prior to the law’s implementation, but dropped to nearly 250 by the end of the postintervention period examined. In Pennsylvania, the prelaw implementation period had an MME per prescription a little higher than 200, which leveled off at around 200 during the postintervention period.

“In New Jersey, there is a significantly higher MME than in Pennsylvania and this difference persists in the transition period but what you see in the post period is a significantly greater decline in the MME per prescription in New Jersey as compared to the rate of change in Pa.,” Dr. Lowenstein said. “That difference was statistically significant.”

She said similar results were seen regarding the quantity of tablets prescribed. In New Jersey before the law’s passage, the number of tablets per prescription was close to 50, dropping down to about 35 post period. Pennsylvania saw a slight decrease from about 35 pills per prescription to about 33 during the same period.

No significant changes occurred in the other outcomes measured following implementation of the EMR alert.

Dr. Lowenstein noted that, because the transition period between the law going into effect and the implementation of the EMR alert was so short, whether the greater decreases in opioid prescriptions in New Jersey relative to Pennsylvania was because of the law alone, the EMR alert alone, or both changes is unclear.

Based on the limited amount of change in prescribing patterns during the transition period, it appears that the EMR intervention may be driving the change, “but we weren’t powered to make that determination,” she added.

Dr. Lowenstein and her colleagues reported no disclosures.

Next Article: