Leveraging Natural Language Processing to Identify Veterans Who Inject Drugs to Assess Preexposure Prophylaxis and Sexually Transmitted Infection Testing Services at the Veterans Health Administration - Scorecard - MDSpire
Advertisement
Leveraging Natural Language Processing to Identify Veterans Who Inject Drugs to Assess Preexposure Prophylaxis and Sexually Transmitted Infection Testing Services at the Veterans Health Administration
Clinical Scorecard: Utilizing Natural Language Processing to Detect Veterans Engaged in Injection Drug Use for Evaluating Preexposure Prophylaxis and STI Testing Services within the Veterans Health Administration
At a Glance
Category
Detail
Condition
Injection drug use among veterans with associated infectious disease risks
Key Mechanisms
Natural language processing (NLP) applied to electronic health records to identify injection drug use behaviors not captured by ICD codes
Target Population
Veterans engaged in injection drug use within the Veterans Health Administration system
Care Setting
Veterans Health Administration facilities across multiple southeastern US locations
Key Highlights
NLP dashboard identified 507 potential PWID among 502,075 veterans; 15% confirmed by chart review.
Confirmed PWID had high prevalence of HIV (6%), hepatitis C antibody positivity (45%), and hepatitis B exposure (13%).
Significant gaps in preventive care: 88% had not received syringes, 74% lacked recent gonorrhea/chlamydia screening, and only one received HIV preexposure prophylaxis.
Guideline-Based Recommendations
Diagnosis
Use NLP tools to identify injection drug use from unstructured clinical notes due to lack of specific ICD codes for IDU.
Confirm NLP-identified cases with targeted chart review for accuracy.
Management
Engage PWID in comprehensive harm reduction services including syringe service programs, STI screening, and preexposure prophylaxis (PrEP).
Leverage existing mental health and social work engagements to improve infectious disease specialist referrals.
Monitoring & Follow-up
Regularly assess HIV, hepatitis B and C status, and STI screening among PWID.
Monitor uptake of harm reduction interventions including syringe distribution and PrEP usage.
Risks
High risk of infectious diseases including HIV, hepatitis B and C, and STIs among PWID.
Underdiagnosis due to lack of specific coding and insufficient preventive care delivery.
Patient & Prescribing Data
Veterans identified as people who inject drugs within the VA healthcare system
Despite high engagement with mental health (94%) and social work (82%), only 29% saw infectious disease specialists and PrEP usage was extremely low (1 patient), indicating underutilization of preventive pharmacotherapies.
Clinical Best Practices
Implement NLP-based screening tools to efficiently identify PWID in large healthcare systems.
Integrate infectious disease services with mental health and social work to close gaps in preventive care.
Prioritize STI and HIV screening and PrEP provision for PWID to reduce transmission risks.
Use multidisciplinary teams including infectious disease physicians and pharmacists for comprehensive care.
Leverage existing therapeutic relationships to enhance harm reduction service delivery.
