Evaluation of Perinatal Exposure to HIV Reporting in Pennsylvania

BACKGROUND: Pennsylvania (PA) has required perinatal exposure to HIV reporting since October 2002. With changes such as electronic laboratory reporting (ELR) and the integration of HIV data into PA National Electronic Disease Surveillance System (PA-NEDSS), we recognized that over time the number of cases had declined. In addition, an evaluation conducted in 2011 revealed a completeness rate of 48%. The objective of this project was to evaluate reporting lag for cases identified by different sources.

METHODS: We focused on 15 of PA’s largest birthing hospitals. Letters were sent to medical record departments requesting a list of HIV positive women (code V08 or 042) delivering a live infant during 2012. In addition, a letter was sent to the infection control nurse requesting a pharmacy list of newborns that received AZT. All 15 facilities complied with both requests. All data received were matched to the HIV surveillance database. We evaluated timeliness by date of birth through date of report into the HIV surveillance system.

RESULTS: A total of 30 files (medical record and pharmacy lists from each facility) were received. Fifty-seven mother/infant pairs were identified through medical records. Forty-nine were reported cases. Of the remaining 8, six of the mothers were HIV negative, one child died at birth and one case was new. Fifty-one mother infant/pairs were identified through pharmacy records. Forty-eight of these cases were reported. Of the remaining three, 2 of the mothers were HIV negative and the third case was new. For lag in reporting, average number of days to report was: 270 days (range 172-414) from a pediatric HIV clinic, 219 days (one case) from the birth registry match, 42 days (range 19-91) from Philadelphia pediatric HIV clinics, 15 days (range 7-42) from ELR, and 13 days (range 4-39) from pharmacy.  Our final cohort of 51 cases came from the following reporting sources: 18 pharmacies, 16 Philadelphia pediatric HIV clinics, 13 ELR, 3 pediatric HIV clinic and 1 birth registry.

CONCLUSIONS: Timeliness for both pharmacy reporting and ELR identified cases was similar. The data submitted from medical records was similar to the pharmacy records. Medical records data was less specific and more labor intensive to gather and review.  This project demonstrated the value of engaging more birthing hospitals to adopt a pharmacy triggered reporting system for identifying cases of perinatal exposure to HIV.