![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: Reducing elective delivery (ED) is one nationally recognized strategy to improve birth outcomes. The Joint Commission ED performance measure is the proportion of singleton infants delivered at 37-38 weeks gestation via induction or Cesarean section without spontaneous labor, after excluding medically indicated deliveries. Although The Joint Commission recommends medical record review, HRSA’s Collaborative Improvement and Innovation Network (CoIIN) leadership developed a method to approximate ED prevalence using US standard birth certificate (BC) data. Louisiana expanded its birth certificate in March 2012 to include nine additional Joint Commission medical indications for delivery not available on the BC, along with fields for “spontaneous labor,” “other” and “no medical reason.” The extent to which ED prevalence differs between BC and the Louisiana expanded birth record (LER) is unknown. This analysis investigates how the prevalence of ED differs between BC and LER and assesses the agreement between sources.
METHODS: Louisiana resident birth certificates of in-state, hospital-based, singleton births occurring between 37-38 completed weeks gestation from April through December 2012 were used for analyses. Differences between BC and LER were assessed in SAS v9.2 using counts and percents, with statistical significance indicated by non-overlapping 95% confidence intervals (95%CI); ED agreement was assessed using Cohen’s Kappa. Limitations include using birth certificate data without medical record validation and classifying 608 Louisiana birth records (4.4%) as elective because the medical indication specified in the “other” field was not included in The Joint Commission medical indications list.
RESULTS: Of 13,337 Louisiana singleton births, 37-38 weeks gestation, there were 2,392 inductions (17.9%), 4,338 Cesarean sections (32.5%), and 6,607 non-induced vaginal deliveries (49.6%). Medical indications for delivery and spontaneous labor were identified for 3,440 and 5,822 BC compared to 4,433 and 6,893 LER records, respectively. Louisiana ED prevalence was 33.6% (95%CI= 32.7%-34.6%) using BC compared to 22.6% (95%CI= 21.7%-23.5%) using LER. Kappa for ED identification between BC and LER was 0.24 (95%CI= 0.22, 0.26).
CONCLUSIONS: Indications for delivery not available on the BC accounted for over 10% of medically indicated deliveries using LER. The BC less frequently identified both medical indications for delivery and spontaneous labor, resulting in a significantly higher estimate of ED compared to LER. ED prevalence agreement between BC and LER was in the fair range for Kappa. Birth certificate data used to identify ED should be validated using claims data or medical record review.