Skip to main content
. Author manuscript; available in PMC: 2014 Jul 1.
Published in final edited form as: Pharmacoepidemiol Drug Saf. 2013 Apr 18;22(7):776–782. doi: 10.1002/pds.3443

Table 2.

Comparison of women linked to infants using different methods at two sites which augmented linkage using birth certificate data*

SITE 1 SITE 2
Links from birth certificates
(60% of pairs)
Links from health plan records alone
(40% of pairs)
Links from birth certificates
(13% of pairs)
Links from health plan records alone
(87% of pairs)
Hispanic (%) 64.3 41.5 7.4 5.6
Race (%)
 White 91.7 92.1 72.3 81.1
 Black 2.3 1.2 11.0 4.6
 Asian American 0.1 0.5 10.6 9.1
 Pacific Islander 1.1 3.2 2.3 2.6
 Native American 3.9 3.0 2.0 1.1
Education:
< high school graduate (%)
30.3 4.1 16.5 3.7
Age (%)
 <18 8.1 0.7 4.0 1.0
 18–39 91.0 95.4 83.3 76.8
 40+ 0.9 3.9 2.3 4.1
Medicaid insurance (%) 73.8 6.6 32.0 6.0
Smoking status (%)
 Smoker 12.5 4.1 4.0 1.0
 Nonsmoker 85.9 94.2 18.0 20.0
 Unknown 1.6 1.7 78.0 79.0
Nulliparous (%) 71.4 73.9 48.0 42.0
*

Limited to the two sites which identified additional linked mother-infant pairs from the public record portion of state birth certificate files. Characteristics included in the table are derived from birth certificate data. All comparisons between groups linked using different methods within each site are statistically significant (p<0.01).