Division of Reproductive Health, Centers for Disease Control and Prevention, 4770 Buford Highway NE, MS k34, Atlanta, GA 30341-3717, USA.
Matern Child Health J. 2012 Nov;16(8):1703-8. doi: 10.1007/s10995-011-0877-7.
To assess the validity of probabilistic linkage (PL) in combining national surveillance data on assisted reproductive technology (ART) with Massachusetts birth and infant death data, for the purpose of monitoring maternal and child health outcomes of ART. A study conducted in 2006 utilized direct identifiers to match Massachusetts birth records with records on ART procedures performed to Massachusetts residents in fertility clinics located in Massachusetts and Rhode Island, achieving a linkage rate of 87.5%. The present study employed PL using the program Link Plus, without access to direct identifiers. The primary linking variables were maternal and infant dates of birth, and plurality. Ancillary variables such as maternal ZIP code and gravidity helped resolve duplicate matches and capture additional matches. PL linked 5,390 (87.8%) of 6,139 deliveries, correctly identifying 96.4% of the matches previously obtained using deterministic linkage methods. PL yielded a high linkage rate with satisfactory validity; this method may be applied in other states to help monitor the maternal and child health outcomes of ART.
为了评估概率链接 (PL) 在将全国辅助生殖技术 (ART) 监测数据与马萨诸塞州出生和婴儿死亡数据相结合,以监测 ART 的母婴健康结果的有效性。2006 年进行的一项研究利用直接标识符将马萨诸塞州的出生记录与在马萨诸塞州和罗得岛的生育诊所为马萨诸塞州居民进行的 ART 程序记录进行匹配,链接率达到 87.5%。本研究使用 Link Plus 程序进行了 PL 分析,而无需直接标识符。主要的链接变量是母婴的出生日期和多胎妊娠。辅助变量如产妇的邮政编码和孕次有助于解决重复匹配并捕获其他匹配。PL 链接了 6139 次分娩中的 5390 次(87.8%),正确识别了之前使用确定性链接方法获得的 96.4%的匹配。PL 具有较高的链接率和令人满意的有效性;这种方法可以在其他州应用,以帮助监测 ART 的母婴健康结果。