Integrating Data into Meaningful HIV Indicators Using Business Intelligence

BACKGROUND: Business Intelligence (BI) technology can facilitate access to health data to support public health interventions with a goal of improved population-level outcomes. By utilizing this technology, health departments and clinicians can intervene in a more directed manner with the HIV-infected community along their continuum of care, beginning with diagnosis through linkage to care, retention in care, and viral load suppression. Targeted measurements can evaluate where each HIV-infected individual is along this “cascade” and then be aggregated to describe a public health indicator. An HIV BI tool provides a reporting system for population indicators and thus is an effective tool to monitor and further impact the epidemic. 

METHODS: Using a standard data model, demographic, geographic and HIV-specific data (including clinical visits, viral load and CD4 counts) were integrated from multiple public and personal healthcare sources to define HIV care indicators. Data were assessed for cleanliness, accuracy and where necessary surrogates developed (e.g., use of AIDS diagnosis date when the HIV diagnosis was missing) before storing in the BI data model. Additionally, interfaces were created through a visualization process to promote comprehension of meaningful data.

RESULTS: Joint development sessions allowed HIV indicators to be developed consistent with the quality and accuracy of the data. To describe engagement in and outcomes of HIV care, each stage of the HIV continuum of care cascade was specifically defined and a corresponding indicator identified (i.e., estimated number of HIV infected, number diagnosed, number linked to care, number retained in care, and number with a suppressed viral load). The tool allows users to drill down to patient level data to generate lists for clinical and public health outreach purposes.

CONCLUSIONS: HIV continuum of care population indicators were developed and calculated for one jurisdiction using data from a variety of public and personal healthcare sources.  The HIV BI tool increased disease surveillance capacity, incorporated novel indicators, and may now support and supplement clinical decision making.  In the next phase of development, the tool will be expanded to include more metro-wide data sources for cross-county comparisons.  The geographic component allows public health officials to better understand disparities in the continuum of HIV care.