Suicide and Substance Use Prevention Using Community Health Informatics (C.H.I): Leveraging DHIS2 for Early Detection and Intervention
Adekola George Adepoju *
Department of Health Informatics, Indiana University Indianapolis, Indiana, USA.
Daniel Adeyemi Adepoju
Department of Health Informatics, Indiana University Indianapolis, Indiana, USA.
Daniel K. Cheruiyot
Department of Health Informatics, Indiana University Indianapolis, Indiana, USA.
Zeyana Hamid
Department of Health Informatics, Indiana University Indianapolis, Indiana, USA, Fairbanks School of Public Health, Indiana University Indianapolis, Indiana, USA and Luddy School of Informatics and Computing, Indiana University Indianapolis, Indiana, USA.
*Author to whom correspondence should be addressed.
Abstract
Background: Suicide remains a significant global public health issue, with substance abuse as a key risk factor. Current prevention strategies often lack integration between risk assessment tools and health information systems, hindering early intervention. This study bridges this gap by leveraging Community Health Informatics (C.H.I) and the District Health Information Software 2 (DHIS2) to enhance suicide prevention through systematic data integration.
Methods: The study employed the Drug Abuse Screening Test (DAST) and Columbia-Suicide Severity Rating Scale (C-SSRS) within the DHIS2 platform. Data from the Randolph County Caring Community Partnership (RCCCP) were analysed using R Studio for statistical modelling and Power BI for visualisation. Logistic regression, Pearson correlation, and ROC curve analyses were conducted to evaluate risk prediction accuracy. Technical implementation involved configuring DHIS2 with Apache Tomcat, PostgreSQL, and Open JDK.
Results: The analysis revealed a weak positive correlation (r=0.265) between DAST and C-SSRS scores. The logistic regression model achieved 93% accuracy in identifying high-risk individuals, with an AUC of 0.988 confirming excellent predictive capability. Power BI dashboards effectively visualised risk trends by demographic factors, though full DHIS2 integration required further development.
Conclusion: Integrating DAST and C-SSRS into DHIS2 enables scalable, early-risk detection. Future work should address technical barriers (e.g., real-time data synchronization) and expand implementation to diverse settings.
Keywords: Risk stratification, predictive modeling, digital health surveillance, risk assessment, substance abuse, mental health, public health surveillance