Med students used donuts to improve opioid prescribing

Students led the way in an effort to reduce opioid overuse.


Although COVID-19 is the top priority for most U.S. hospitals today, the opioid epidemic remains an important concern. At the end of 2017, hospitalist and clinical informaticist Surafel Tsega, MD, decided to help improve clinicians' opioid-prescribing practices at discharge by enlisting first-year medical students enrolled in a high-value care elective—plus the power of donuts and coffee.

“It seemed like, at the time, the interest in the opioid epidemic was really at its peak,” he said. “I knew that there would be enthusiasm from the med students to put together a project.”

After Dr. Tsega presented the idea, three students began assessing prescribing patterns by looking at documentation in the state prescription drug monitoring database, duration of opioid prescriptions, and number of follow-up appointments, said Gabriela Hernandez-Meza, PhD, who is now a fourth-year med student at the Icahn School of Medicine at Mount Sinai. “Once we knew where our baseline was, we started thinking about ways in which we could improve all of those items,” she said.

How it works

To increase awareness about opioid-prescribing issues, the students presented an educational session in the resident lounge called I-STOP for Donuts, named after New York's prescription database, the Internet System for Tracking Over-Prescribing. The sessions entailed distribution of donuts, coffee, and cards with best practices for opioid prescribing, as well as conversation, said Dr. Hernandez-Meza. “When we first started talking to the residents, we started finding some challenges that we were going to have, for example, that they hadn't really all been trained how to access the database and how to document.”

The students also gave the residents feedback on their documentation for patients discharged with opioids, she said. “It had up arrows or down arrows comparing to the week before that allowed them to see whether they were improving at it or not. I think having that visual cue was key in reminding them that they needed to do it and that previous residents were doing it—that also gave them the initiative to continue to do it,” said Dr. Hernandez-Meza.

To target the other metrics, the students held structured educational sessions on best practices for residents, hospitalists, and advanced practitioners, as well as presented goals of the project on interdisciplinary rounds. In addition, a utilization management coordinator was hired to help make follow-up appointments for patients within a week of discharge, said Dr. Hernandez-Meza.


A similar number of patients were discharged with opioids in the baseline and intervention periods (12.3% and 11.4%, respectively, with about 60% being new prescriptions). The intervention was associated with a decrease in the percentage of patients discharged with opioid prescriptions longer than seven days, from 45.2% to 39.5%, as well as an increase in I-STOP documentation, from 32.5% to 39.7%, according to results published in the March Joint Commission Journal on Quality and Patient Safety. Most notably, the percentage of patients with follow-up appointments within seven days of discharge jumped from 38.6% to 65.9%.

The slight improvement in I-STOP documentation wasn't as satisfying as the other results, said Dr. Tsega, lead author of the study and an assistant professor of medicine at Mount Sinai. “It showed that there are a lot of systems factors at play that make it difficult for people to comply with certain regulations,” he said.


While some residents had access to the I-STOP database, some did not, so obtaining access for them was one of the first challenges, said Dr. Hernandez-Meza. “If they didn't have access to it, then they're relying on somebody else to access it,” she said. “So the chain of command of who's in charge of documenting it gets lost.”

Data gathering at baseline was another challenge. “We knew anecdotally that there was overprescription going on, and there had been several cases where we had seen patients being discharged with excessively long prescription lengths, but we had to really get that data, and that slowed things down a little bit,” said Dr. Tsega. Without an automated auditing system in the electronic health record, the students had to manually retrieve the data from the charts of discharged patients, said Dr. Hernandez-Meza.

Lessons learned

First-year medical students were proactive and enthusiastic about the effort, so “We should consider engaging them in these types of quality improvement projects,” Dr. Tsega said.

Next steps

While the educational campaign was effective in starting culture change around opioid prescribing, technological solutions may ultimately be more effective, said Dr. Tsega. “One of the next steps, that has been paused in the midst of COVID, is how best we can create IT solutions that will help nudge individuals towards making the appropriate decision,” such as making the default opioid prescription duration five to seven days, he said.