Revising the Geographic Distribution of People Living with Diagnosed HIV Infection in New York State Using Routinely Collected Laboratory Data

BACKGROUND: Accurate counts of persons living with diagnosed HIV infection (PLWDHI) are required to estimate service needs at state and sub-state levels. New York State (NYS) Department of Health (DOH) routinely reports only persons diagnosed in NYS or a foreign country, by county of residence at HIV/AIDS diagnosis. PLWDHI ever identified as prisoners of state correctional facilities are excluded from sub-state counts outside of New York City (NYC). PLWDHI diagnosed in NYS are presumed living in-state unless known to have died. To improve estimates of PLWDHI, NYSDOH’s alternate method excludes individuals (“lost” cases) for whom no information was received by the surveillance system for several years, includes those diagnosed outside NYS but now residing in the State, and counts former prisoners by county of current residence.

METHODS:    Addresses from HIV-related laboratory test reports 2008-2013 were matched to cases living as of 12/2012 in the 3/2014 analysis file. The file combines information from separate NYS and NYC eHARS registries. Addresses of health care providers, laboratories and non-residential institutions were discarded. Residential addresses were retained if reported at least twice, separated by ≥1 month. Correctional institutionaddresses were flagged. For each case, the most recent address before 12/2012 was selected. AIDS cases without lab reports after 2007 or HIV cases after 2005 were labelled “lost.” Prisoner cases whose updated residential address wasn’t a correctional facility were considered “non-prisoner” in 2012. Updated residential addresses were applied to cases. Next, “lost” cases were dropped, cases with current residence in state but diagnosed out of state were included, and prisoner status was revised.

RESULTS:   Updated addresses were obtained for 81,000 (62%) of the 132,700 NYS standard count. After excluding prisoner cases, use of updated addresses markedly altered county-level estimates of PLWDHI, ranging from a 33% reduction to a 171% increase (-15% to +80% for counties with >50 PLWDHI).   Removing “lost” cases reduced estimated PLWDHI to 108,000 (-19%). Including out-of-state diagnosed cases increased that to 114,700 (+6%).  After excluding prisoner cases, use of updated addresses changed county estimates (-47% to +73%; -36% to +73% for counties with >50 PLWDHI). Individual counties experienced both in- and out-migration. In NYC’s 5 boroughs, outflow and inflow averaged 21% and 22%, respectively. Statewide, 21% of cases were outside the county of diagnosis.

CONCLUSIONS:   Updated addresses from routinely-collected surveillance data can greatly change the estimated PLWDHI within counties. HIV surveillance jurisdictions should routinely use updated address information for local planning purposes.