The program was planned and executed over 20 weeks (Fig. 1). Program leaders pre-selected three topics, based on pressing public health issues: (1) disrupting current opioid use disorder (OUD) treatment patterns; (2) leveraging ChatGPT to promote patient engagement among underserved populations; and (3) strategies to slow diabetes progression. Students were asked to complete an application for the program, indicating demographic characteristics and educational background. Applicants were asked to select one topic, providing a 300-word narrative justifying their selection. A final prompt asked applicants to indicate their experience and views on interprofessional teams. Flyers to advertise the program were distributed via social media and emails to student advisors across colleges and student-led organizations.

Completed applications were blinded, scored, and ranked by faculty, using a rubric. Of the 46 students who applied, 12 were selected. The small number was due to limited funding to test proof of concept. The final teams consisted of three groups of four students (see Fig. 2 for team composition). Selected students represented the Colleges of Medicine, Pharmacy, Business, and Communications, ensuring diverse perspectives in the creation of realistic and well-rounded solutions. Due to scheduling constraints, all medical students were between their first and second years of training. Graduate students from other colleges were a mix of master’s and doctoral students.

For each topic, secondary data sources were pre-identified for teams to use, and students were added to the institutional review board (IRB) protocol for their respective studies. All IRB submissions were completed by program leaders, and approvals obtained before the start of the program. After IRB approval, program leaders identified faculty mentors with expertise in data analytics, community-based participatory research, epidemiology, health services research, and public health to serve as mentors to the student teams. Each topic had two mentors: one had subject-matter expertise while the other had data analysis expertise. The entire administrative planning process lasted 20 weeks and was completed prior to program initiation. Each student received a $1000 stipend for their participation in the program and competition.

The program occurred primarily virtually over 2 weeks in the summer. Each day commenced with an hour-long didactic session led by industry experts. Topics included “Technology and Health Disparities,” “Caring for the Underserved,” and more. In the second half of the morning, designated as the “Data-Thon” sessions, students engaged with their mentors to delve into their datasets. During the data-thon sessions, students engaged in applied data analyses, covering topics such as “PICO Research Questions,” “Univariate Analysis,” and “Adjusted Models.” These active data sessions exposed students to Excel, R, ArcGIS, and Stata statistical software. Students also met with a community health worker team to ensure their solutions were culturally appropriate to the communities they targeted.

At the end of the virtual 2 weeks, teams pitched their evidence-based solution to an in-person expert panel using a “Shark Tank” format. Shark selection was strongly based on the three topics. Faculty identified five non-university affiliated health systems leaders with respective expertise on substance use (Meninger Clinic), health technology (Microsoft ChatGPT and Care Message), and diabetes care (Humana, Veterans Administration). Invitations were sent 3 months before the start of the program. Sharks were not compensated monetarily, and they did not provide any financial incentives for the solutions proposed by the students.

After the presentation, each team had a public dialogue with the Shark panel where they fielded probing questions into the feasibility, viability, and impact of their proposed solutions. To remove bias from the judging, student presentations were judged using a rubric weighing the quality of the presentation, scientific content, innovation, and feasibility, which were averaged together to create a ranked score. In addition to formal recognition, the winning team received travel funds to present their work at a conference.