BACKGROUND: The Massachusetts Department of Public Health (MDPH) utilizes open-source software called EMR Support for Public Health (ESP) to automatically report selected notifiable diseases. Data are automatically extracted from participating medical practices, mapped to standardized tables, and analyzed for events of interest. Cases are electronically transmitted to MDPH’s electronic disease surveillance and case management system (MAVEN). ESP has successfully been implemented as a cross-sectional reporting system for incident cases of selected diseases and was recently expanded to enable longitudinal surveillance for patients with HIV and chronic hepatitis C virus infection. This abstract describes HIV continuum of care reporting using ESP and provides provisional, preliminary aggregate-level data from several community health centers in Massachusetts.
METHODS: ESP detects HIV cases using laboratory tests, antiretroviral prescriptions, and diagnosis codes. Cases are transmitted to MAVEN via HL7 messages. Case reports are currently being enhanced to include risk factors, treatments, HIV viral load and CD4 count test results, opportunistic infections, and medical visits, with selected cases validated by chart review. These additional data are being used to create new workflows in MAVEN to identify and triage, for example, patients who appear to be out of care. ESP can also provide population-level aggregate summaries; this capability was utilized to identify the count of new HIV patients at 17 health centers in 2015 and, among them, the following were separately assessed: proportion on treatment, virally suppressed (last viral load <200 copies/mL), and with ≥1 HIV-related follow-up encounter.
RESULTS: The HIV detection algorithm has been validated at 1 multi-site practice and is under validation at 5 individual health centers. Current challenges include determination of adequate chart confirmation of new data elements, ongoing lab mapping maintenance, and identification of indicators of “in care” status. Longitudinal HIV case reporting is expected to begin in early 2017. Using the aggregate querying function of ESP, 325 incident HIV cases were identified in 2015 from 17 health centers. Of those, 95.4% started on an HIV medication and 86.5% had at least one follow-up encounter. Among those with a viral load documented, irrespective of treatment, 65.7% had a suppressed viral load.
CONCLUSIONS: Groundwork has been laid for longitudinal case reporting of HIV detection and care using a system that automates data collection, reducing the burden of reporting by clinical sites, and improving MDPH’s ability to monitor HIV continuum of care. Lessons learned from the HIV implementation will be applied to other notifiable chronic diseases.