• birth certificate data;
  • elective delivery;
  • hospital discharge data;
  • preterm birth



We propose a methodology for identifying and analysing ‘elective’ preterm births (PTBs) using administrative data, and apply this methodology to California data with the objective of providing a framework to further explore the potential rationales for early delivery.


Using the California linked birth cohorts for 1999, 2002 and 2005, singleton PTBs were identified using birth certificate gestational age ≥ 24 and <37 weeks. Through a hierarchical scheme that first removed cases with standard or ‘hard’ indications for early delivery (e.g. severe preeclampsia, placenta previa), cases of ‘elective’ PTB were identified with coding for medical intervention, that is, elective caesarean or labour induction. We calculated rates of elective PTB, with subanalyses of early (<34 weeks of gestational age) and late PTB (34 to <37 weeks of gestational age) using hierarchical logistic regression models.


Of 1 387 565 singleton deliveries, 99 614 (7.2%) were preterm. Elective PTBs increased 27.7% over the 6-year study period, with nearly all cases confined to the late PTB stratum; elective late PTB rates rose from 10.5% to 13.5% of all late PTBs (P < 0.0001). Indications for delivery in this Elective Group (‘soft indications’) included prior pelvic floor repair, mental health conditions, fetal anomalies, malpresentation and oligohydramnios. Six per cent of patients with a late PTB had a medical intervention with no hard or soft indication for delivery.


Using administrative data, we developed a method for identifying and trending the proportion of PTBs that is ‘elective’. This method can be used to explore and monitor potential strategies for the prevention of elective PTB.