Implementing an Automated Prediction Model to Improve Prescribing of HIV Preexposure Prophylaxis

Area of Science:

• Public Health
• Health Informatics
• Clinical Decision Support

Background:

• Antiretroviral pre-exposure prophylaxis (PrEP) is highly effective for HIV prevention but underutilized in key populations.
• Primary care providers require effective tools to identify and prescribe PrEP to at-risk individuals.

Purpose of the Study:

• To develop and validate an automated electronic health record (EHR) decision support tool with interactive alerts.
• To assess the feasibility, clinician acceptance, and preliminary impact of the EHR tool on PrEP care in primary care settings.

Main Methods:

• A pilot study was conducted in three federally qualified health centers.
• The EHR tool generated alerts for providers, and its impact was compared to matched control clinics.
• Qualitative interviews were conducted with providers to assess tool acceptance and perceived impact.

Main Results:

• The tool was feasible, with providers receiving alerts for 2.2% of patients.
• New PrEP prescriptions were 4.5 times higher in pilot clinics compared to control clinics.
• Providers reported the tool facilitated sensitive PrEP discussions, overcoming patient stigma.

Conclusions:

• Automated EHR decision support tools can increase PrEP discussions and prescribing in primary care.
• Successful implementation hinges on provider collaboration, trust, and integrating data-driven insights with clinical judgment.
• This approach holds promise for implementing predictive models across various medical domains.