Automated calls added to SMS reminders reduce missed appointments among high-risk patients

Missed outpatient appointments (or no-shows) pose a significant challenge to health systems by reducing access to care, wasting resources, and negatively impacting patient outcomes. Traditional interventions — such as overbooking, transportation services, phone calls, and financial incentives — have yielded mixed results and face limitations in terms of scalability. Predictive modeling offers a promising tool for identifying patients at high risk of missing their appointment, allowing for targeted outreach. However, targeted outreach typically needs to be paired with manual outreach efforts to ensure patients make it to their appointments. To address these challenges, Penn Medicine’s Patient Access team partnered with the Center for Health Care Transformation and Innovation (CHTI) to better understand the reasons for appointment no-shows and design and implement a scalable intervention. A text message survey of 186 patients who missed appointments revealed the top reason was “I did not know I had an appointment” (22% of patients). The team developed an automated targeted outreach program for high-risk patients using an interactive voice response (IVR) system that complemented the existing text message appointment reminder program. Leveraging the system’s electronic health record no-show predictive model and CHTI’s telecommunications platform, an automated IVR outreach campaign was developed to remind these patients of their upcoming appointments. The IVR call offered patients options to confirm, cancel, or reschedule their appointments, and represented a way to reduce no-show rates with minimal effort. Using these methods jointly to reduce no-show rates — while not overburdening patients at low risk of no-show — became the center of the design and intention. A rapid randomized trial (ClinicalTrials.gov number NCT06767423) was conducted with 59,994 patients at high risk of missing their appointment over a 4-week period in 2024. The no-show rate was 1.7 percentage points lower (down to 9.6% from 11.3%), and the net number of appointment completions was 1.9 percentage points higher (up to 77.8% from 75.9%) in the group of patients who got an IVR call in addition to text message reminders than in the group of patients who got text message reminders alone. The intervention had the greatest effect on reducing no-shows among patients in the highest risk quartile. It also increased appointment completion rates the most among Black patients, helping to reduce a preexisting equity gap. Overall, the results imply that 19,000 additional appointments could be completed per million new appointment slots. Based on the results of the trial, the intervention was immediately implemented at scale. Follow-up data from over 244,000 high-risk patients over a 6-month period demonstrate that the improved appointment completion rates were sustained. This case study demonstrates that integrating predictive modeling with IVR outreach can serve as a low-cost, scalable solution for improving appointment completion, increasing clinic efficiency, enhancing health equity, and generating revenue for health systems.

• Adding interactive voice response (IVR) to standard text message reminders can reduce appointment no-shows and increase appointment completion in patients at increased risk of missing their appointment.
• Conducting a rapid randomized trial is a highly effective testing method to rigorously assess the impact of automated interventions and return on investment and to guide enterprise-wide implementation decisions.
• Having a multidisciplinary team, including operational leaders and experts in informatics, health care innovation, and pragmatic evaluations, can facilitate rapid cycle learning and health system transformation.
• Even small percentage-point improvements in high-volume events can have a significant impact when scaled across health systems.
• Delivering these IVR reminders only for appointments at high risk of no-show accomplishes the goal of increasing appointment utilization while not burdening low-risk patients with an extra reminder call.

Laura Kavanaugh, Shane Thomas, Sean Lynch, Damien Leri, Ruiying A Xiong, Lin Xu, Christopher K Snider, Jeffrey P Ebert, Michael O Harhay, Yingying Lu, Nathan Orwig, Noelle Desir, Iain Noel Encarnacion, Nicholas Mollanazar, Rachel Thomas, Charles Bae, Yevgeniy Gitelman, Lauren Hahn, Kathleen Lee, Shivan J Mehta, Raina M Merchant, M Kit Delgado