Predicting No-shows at a Student-Run Comprehensive Primary Care Clinic

Area of Science:

• Health Services Research
• Predictive Analytics in Healthcare
• Outpatient Clinic Operations

Background:

• Missed appointments are a significant challenge in outpatient settings, leading to inefficient resource use and loss of patient follow-up.
• The Shade Tree Clinic (STC), a student-run primary care facility, faces high no-show rates impacting its service delivery.
• This study aimed to leverage existing data to predict patient no-shows and enhance clinic operations.

Purpose of the Study:

• To develop a predictive model for patient no-shows at the Shade Tree Clinic.
• To identify key factors associated with appointment no-shows.
• To improve clinic efficiency and quality of care by reducing missed appointments.

Main Methods:

• Utilized appointment data from STC spanning January 2010 to December 2015.
• Integrated external weather data corresponding to each appointment date.
• Employed multivariable logistic regression to construct predictive models based on historical data.

Main Results:

• The analysis included 13,499 appointments with a 69.2% show rate.
• Significant predictors of no-shows included previous show rate, day of the week, automated reminders, snowfall, and high ambient temperature.
• The predictive model achieved a negative predictive value of 61.0% at the 25th percentile cutoff.

Conclusions:

• Readily available data and a novel framework can identify patients at high risk of no-shows.
• Targeted interventions for high-risk patients can improve clinic efficiency and patient outcomes.
• The developed analysis methodology is replicable in other clinics utilizing electronic medical records.