Reducing non-attendance in outpatient appointments: predictive model development, validation, and clinical assessment

Area of Science:

• Healthcare Management
• Predictive Analytics
• Patient Engagement

Background:

• Hospital outpatient appointment non-attendance negatively impacts healthcare resource planning and patient care quality.
• Delayed assessments and increased waiting lists are consequences of missed appointments.
• A predictive model and intervention were developed to address patient no-shows.

Purpose of the Study:

• To develop and validate a model for predicting patient non-attendance at hospital outpatient appointments.
• To assess the effectiveness of a targeted intervention for reducing non-attendance based on the predictive model.

Main Methods:

• Retrospective data from dermatology and pneumology outpatient services were used to build predictive models (decision trees).
• Model predictive capacity was prospectively validated.
• A pilot study assessed the effectiveness of phone call reminders for high-risk patients identified by the model.

Main Results:

• The developed models demonstrated good predictive accuracy for non-attendance in both services.
• Prospective validation confirmed the model's predictive performance.
• Targeted phone call reminders significantly reduced non-attendance rates by 50.61% in dermatology and 39.33% in pneumology.

Conclusions:

• Patient information in medical records can effectively estimate non-attendance risk.
• Stratifying patients by non-attendance risk enables targeted interventions, like phone reminders, to reduce no-show rates and improve service delivery.