Risk Prediction Tool for Medical Appointment Attendance Among HIV-Infected Persons with Unsuppressed Viremia

Area of Science:

• Infectious Diseases
• Public Health
• Health Informatics

Background:

• Successful HIV treatment relies on consistent clinical follow-up and patient engagement.
• A previously validated risk-prediction tool (RPT) uses electronic health record (EHR) data to predict 1-year virologic failure.
• Predicting immediate outcomes like appointment attendance could enable timely interventions.

Purpose of the Study:

• To evaluate if a modified RPT, using EHR data, can predict non-attendance at HIV medical appointments.
• To identify high-risk patients for targeted interventions to improve appointment adherence.

Main Methods:

• An observational cohort study was conducted at the Vanderbilt Comprehensive Care Clinic.
• Patients with routine appointments and HIV-1 viral load >200 copies/mL were included.
• A modified RPT was used to calculate risk scores, and multivariable logistic regression estimated odds ratios for missing appointments.

Main Results:

• Among 510 participants, higher RPT risk scores were significantly associated with missing the next appointment.
• Medium (OR 3.95) and high (OR 9.55) RPT risk scores independently predicted non-attendance.
• The RPT effectively stratified patients based on their risk of appointment non-attendance.

Conclusions:

• Risk-prediction tool scores derived from EHR data can identify patients at high risk for HIV medical appointment non-attendance.
• This risk stratification can guide targeted interventions to improve adherence.
• Utilizing readily available EHR data for RPTs supports resource allocation for improved patient outcomes in HIV care.