OBJECTIVE: To evaluate the accuracy of assisted reproductive technology (ART) reporting on the Massachusetts birth certificate and to explore the individual and hospital-level characteristics associated with ART reporting.

DESIGN: Validity analysis of population-based data.

SETTING: Live-birth deliveries by Massachusetts-resident mothers during 1997-2000.

PATIENT(S): Live births and delivery mothers.

INTERVENTION(S): The ART data maintained by the Centers for Disease Control and Prevention were linked with the live birth-infant death records in Massachusetts. Successfully linked records were used as the gold standard for ART-related deliveries in evaluating the validity of the ART information reported on the Massachusetts birth certificate.

MAIN OUTCOME MEASURE(S): Sensitivity and specificity.

RESULT(S): The sensitivity of ART reporting on the birth certificate was 27% and the specificity >99%. Sensitivity of ART reporting was higher among women with multiple deliveries (twins: 32%; triplets+: 43%) and preterm deliveries (36%).

CONCLUSION(S): During the period evaluated, reporting of ART information on the birth certificate was incomplete, and ART births identified through the birth certificate were a biased sample of the population of ART births. Using delivery hospital data as the sole source of ART information for the standard birth certificate may yield inaccurate information.