Comparison of birth certificate and hospital discharge data to identify early elective deliveries

Background:  Early elective deliveries (EED) are associated with infant morbidity and mortality and increase costs by $1 billion annually.  To aid quality improvement initiatives, Illinois released EED reports using birth certificate data and an algorithm based on the Joint Commission (JC) perinatal core measure.  Birth certificates likely overestimate EED rates since they do not have all 73 JC indications for early delivery available.  The extent of this overestimation, however, is unknown.

Methods:  We compared the 2012 birth certificate and 2012 hospital discharge files on the number of singleton live births with any reported JC indications and used the observed hospital discharge indication rates to quantify potential overestimation when using birth certificates for EED studies.

Results:  From the 2012 birth certificates, there were 36,402 singleton early term live births, and 3,682 (10.1%) had at least one of 43 JC indications available on the birth certificate.  The hospital discharge data captured all JC indications, but did not have gestational age available.  Of the 128,885 discharges of singleton live births at ≤40 weeks gestation, 27,047 (21.0%) had at least one of the indications available on the birth certificate.  Of the remaining 101,838 births, an additional 23,972 (23.5%) had indications not available on the birth certificate.  If the hospital discharge indication rates were applied to birth certificate data, 14,402 (29.4%) early term births would have had indications; therefore approximately 29% of early term births may be misclassified as having no indications for early delivery. 

Conclusions:  The birth certificate overestimates the EED rate because of underreporting of medical conditions and missing indications.  Improvements in data systems, including linkage of discharge and birth certificates, could provide more accurate data for EED quality improvement initiatives.