Evaluating Facility Type for HIV-Related Laboratory Tests – New York City, 2008-2013

BACKGROUND:  The New York City (NYC) Bureau of HIV/AIDS Prevention and Control’s (BHIV) surveillance program uses electronic laboratory reporting to monitor the clinical status of persons living with HIV/AIDS (PLWHA). BHIV’s surveillance unit assigns a category for the type of medical facility associated with the ordering provider for each laboratory test. In 2013, there were 2,832 new HIV diagnoses and 117,618 PLWHA in NYC. Overall, 63% of PLWHA received HIV medical care, using viral load (VL) and CD4 tests as a proxy for care. BHIV’s work to monitor and improve care engagement among PLWHA could be enhanced by specific knowledge of where and how PLWHA access the healthcare system. This evaluation sought to assess the completeness and quality of facility type data within the surveillance registry.

METHODS:  Electronic laboratory report data, contained within the HIV surveillance registry, were analyzed for NYC-based medical facilities for HIV-related tests conducted between January 1, 2008 and December 31, 2013. These data included VL and CD4 tests, positive HIV diagnostic tests, and viral genotype tests. Completeness, level of detail, and type distribution were assessed for facility type data.

RESULTS:  From 2008-2013, 101,656 individuals received 4,437,540 HIV-related tests in NYC. Most common ordering facility types were inpatient hospital (55.3%), outpatient facility (18.0%), and private physician office (17.3%). Of all reported tests, 6,248 (0.1%) were either missing facility type data or classified as unknown. An additional 18.1% of tests were classified in a general facility type category but were missing specific information about the facility (for example, ‘outpatient’ versus ‘outpatient - adult HIV specialty clinic’). For the 2,832 new diagnostic events in 2013, most common ordering facility types were inpatient hospital (39.6%), outpatient facility (17.8%), and private physician office (17.2%); 4.3% were missing facility type data. Among tests ordered by hospital systems, none could be classified as having been ordered in the emergency department.

CONCLUSIONS:  While the majority of HIV-related laboratory tests include some information about the type of facility where tests were performed, a substantial proportion are lacking in detail, presenting a barrier to using these data to improve BHIV’s understanding of care patterns among NYC PLWHA. Given that over half of laboratory tests are conducted by hospitals, more specific information about ordering department within hospitals would be valuable. The next step will be to perform a data linkage between the HIV surveillance registry and a statewide hospital discharge dataset to augment the facility type data.