Leveraging Electronic Health Record Data to Improve HIV Surveillance Registry

Monday, June 5, 2017: 2:00 PM
410B, Boise Centre
Bridget J Anderson , New York State Department of Health, Albany, NY
Jennifer Cukrovany , New York State Department of Health, Albany, NY
Brenda Moncur , New York State Department of Health, Albany, NY
Joseph Kerwin , New York State Department of Health, Albany, NY
Travis O'Donnell , New York State Department of Health, Albany, NY
Charles Fisher , New York State Department of Health, Albany, NY
Monish Madhani , New York State Department of Health, Albany, NY

BACKGROUND:  The National HIV/AIDS Strategy encourages states to use electronic health record (EHR) data to enhance public health data systems. The New York State Department of Health (NYSDOH) partnered with Healthix, the largest Regional Health Information Organization in New York State, to receive Continuity of Care Documents (CCDs) for patients with an HIV diagnosis. The collaboration allows the NYSDOH HIV surveillance program to access additional information on diagnosed persons for a more comprehensive assessment of health care measures important to the epidemic, including measures of HIV testing, linkage to care, retention in care, antiretroviral therapy, and viral suppression.

METHODS:  NYSDOH and Healthix developed an algorithm based on ICD-9-CM, ICD-10-CM codes and text-based information to identify HIV-diagnosed patients. An initial CCD is sent for each patient identified using the aforementioned algorithm. NYSDOH receives updated EHR information from Healthix via HL7 version 2.5 messaging. In addition, Healthix sends a monthly patient expiration file containing previously identified patients who are now flagged as deceased by a facility reporting to Healthix. The Healthix data is parsed and matched against the NYS HIV registry.

RESULTS:  57,606 Healthix patients have been identified, with 80% matching to the HIV registry. Of the matches, 6.5% match to an unconfirmed or incomplete case in the registry. Modification of the algorithm and resolution of aberrant data are ongoing. An initial bolus of 175 historical patient expiration records were received; 40% matched to the registry. Since then an additional 72 patient expiration records were transmitted, yielding a 74% match rate, with most of the fact of death information (85%) being new to the registry.

CONCLUSIONS:  Automated use of EHR data has great potential to augment HIV surveillance and data to care activities. In addition to case ascertainment and near real-time death ascertainment, a major benefit includes leveraging up-to-date locating information for persons with diagnosed infection. The ability of a surveillance system to track individuals over time and facilitate re-engagement in care is dependent upon accurately matching received data to existing case records. Additionally, automated use of electronic data from the CCD affords the NYSDOH efficiencies over the traditional surveillance strategy of individual medical record review. However, the extent to which the EHR data can be leveraged may be limited by data quality and completeness and the general absence of utilization or reliable adherence to national coding standards.