The dangers of delivering a newborn before 39 weeks for nonmedical reasons have become increasingly clear.1–3 In response to the evidence, the Ohio Perinatal Quality Collaborative developed a multihospital quality improvement project aimed at decreasing nonmedically indicated scheduled births before 39 weeks. In 2008, the Ohio Perinatal Quality Collaborative brought together a group of 20 Ohio maternity hospitals for the purpose of decreasing the rate of nonmedically indicated scheduled deliveries before 39 weeks. The initiative was successful and demonstrated a decrease from more than 15% to less than 5% (P<.05) from July 2008 through June 2009.4
One possible explanation for the decline in the rate of nonindicated scheduled births was that the Ohio Perinatal Quality Collaborative project produced changes in coding of the reason for delivery rather than an actual change in practice. For example, providers who were scheduling deliveries for valid reasons but not recording the indication might have improved their documentation. There was also concern that providers continued to schedule births before 39 weeks that lacked a genuine medical indication but were recording indications that were not actually present or were stretching the definition of legitimate reasons such as hypertension or fetal compromise. Clearly, if the reason for the success of the program was that the charting changed but actual practice did not, this would undermine the purported success of the collaborative.
The Ohio Perinatal Quality Collaborative abstracted data directly from maternal charts at participating sites as the primary measurement of the project's effect. We also obtained concurrent data from Ohio birth certificates as an independent marker of the effect of the intervention. Both sources of data showed that actual average gestational age at delivery was changing at Ohio Perinatal Quality Collaborative participating hospitals, suggesting that the project was producing a change in practice, not in coding documentation. For example, as the project progressed, there was an increase in births after 39 weeks that paralleled a similar decline in births before 39 weeks at Ohio Perinatal Quality Collaborative hospitals.4 In an effort to understand the changes in clinical practice that led to the decrease in nonmedically indicated scheduled births, we investigated rates of recorded indications for delivery over the course of the project.
MATERIALS AND METHODS
After obtaining Institutional Review Board approval from the sites, Ohio Perinatal Quality Collaborative abstractors reviewed charts and recorded the reason for delivery for all scheduled deliveries at 36 0/7 through 38 6/7 weeks of gestation in member hospitals between July 2008 and December 2009. Data were deidentified and sent to a central location for storage and analysis. The 20 member hospitals represented 47% of all Ohio births. Reasons were listed in predefined check boxes. Abstractors were instructed to check as many boxes as were documented on the chart.4 There were specific boxes for reason not documented and elective. Data on amniocentesis for fetal lung maturity were also collected. Additionally, birth certificate data from these same hospitals were provided to the Ohio Perinatal Quality Collaborative by the state vital statistics bureau.5 Initial review of data revealed that many abstractors added free text because the original list of indications was not complete. In July 2009, the form was revised to capture additional common indications such as maternal seizure disorder, substance abuse, and anticoagulation use. We chose to use data from October 2008 forward because before this date data were not complete for all hospitals.
To account for multiple reasons for delivery, we needed a hierarchy of delivery categories. Because we could find no commonly accepted algorithm for acceptable medical reasons for scheduled delivery at that time, maternal fetal medicine specialists and obstetricians at participating hospitals developed a system that ranked reasons for scheduled birth into three categories: strong medical indications for delivery, intermediate indications, and deliveries that did not require scheduled birth before 39 weeks of gestation (unnecessary). The last category included births that were appropriately scheduled only after 39 weeks of gestation as well as births that had no medical or obstetric reason to be scheduled (Table 1). Births were classified into the highest category for which they qualified. For example, if an abstractor checked diabetes and advanced cervical dilation, the patient would be categorized as having a strong (diabetes) indication for scheduled delivery.
Next, a second hierarchy was created incorporating whether the patient had a mature amniocentesis before delivery (Table 2). There were two categories in the second hierarchy, medically indicated and nonmedically indicated. If a woman had a mature amniocentesis and an indication in the nonmedically indicated column, a check was placed in the medically indicated column. If more than one indication was indicated, a check was placed in the highest category to which the woman qualified.
To account for changes in delivery volume, we reported indication categories as a percentage of all deliveries in that month. To insure complete accounting of delivery volume, we used birth certificate data from the participating hospitals to determine delivery volume. Changes in indication categories were evaluated using a nonparametric test for random order (runtest STATA).
There were 23,022 deliveries at the Ohio Perinatal Quality Collaborative hospitals during the study period. There were 4,997 scheduled deliveries with strong indications, 835 scheduled deliveries with intermediate indications, and 896 scheduled deliveries that were unnecessary at less than 39 weeks of gestation. Scheduled deliveries that were unnecessary at less than 39 weeks or had intermediate indications decreased (P<.03; Fig. 1). At the same time, deliveries with strong medical indications were not significantly different over time (P=.99). The decline in the rates of intermediate and unnecessary births translated to a substantial reduction in the total number of scheduled births: 2,363 in the first 4 months of the project compared with 2,038 in the last 4 months, 145 fewer with intermediate reasons, and 265 fewer that were unnecessary. Strong indications increased slightly by 85.
When amniocentesis data are incorporated, similar findings are seen (Fig. 2). Medically indicated deliveries were not significantly different (P=.8), but nonmedically indicated deliveries decreased over time (P=0). Amniocentesis rates showed a slight trend toward more performed over time (P=.046), suggesting that providers performed more amniocenteses to justify scheduled birth before 39 weeks of gestation in women without a medical reason for earlier birth. However, other decreases in nonmedically indicated deliveries overwhelmed this effect.
The Ohio Perinatal Quality Collaborative efforts to reduce nonmedically indicated deliveries at less than 39 weeks of gestation received accolades when it lowered rates by 60% in 1 year. However, doubts about the results remained because of concern that documentation of delivery practices changed but not the practices themselves. The trends presented here support the conclusion that the Ohio Perinatal Quality Collaborative Scheduled Birth Initiative influenced physician decisions about the appropriate gestational age for scheduled births and that the reduction in unnecessary scheduled births was not caused by “up coding” the reason for delivery. The effect of the Ohio Perinatal Quality Collaborative was not primarily driven by improved documentation, but instead reflected a true change in scheduled delivery practices. This is further supported by the graph of gestational age at delivery in Ohio during the study period that was presented in our initial publication from this project.4Deliveries at more than 39 weeks of gestation increased in parallel with the decline in births at 36 0/7 to 38 6/7 weeks. We estimate that 1,000 expected near-term scheduled births were deferred to more than 39 weeks during the first year of the Ohio Perinatal Quality Collaborative initiative.4
As noted in our original report, Ohio Perinatal Quality Collaborative abstractors initially tracked neonatal intensive care unit admissions within 2 hours of birth in neonates born after a scheduled birth between 36 and 38 weeks without a medical indication. However, because neonates born at 36–38 weeks of gestation are often admitted to special care after being observed for more than 2 hours in transitional or well-baby nurseries, this metric was deemed unreliable by Ohio Perinatal Quality Collaborative. Furthermore, we did not have any primary data for women who did not have a scheduled birth, so we could not track how neonatal intensive care unit admission rates changed for the population. It is important to recall that the Ohio Perinatal Quality Collaborative is a quality improvement initiative conducted by its members, and not a funded research project.
The ideal study to measure changes in provider behavior would monitor medical decisions in real time, including consumer and provider interviews about the decision-making process. Medical chart review provides indirect evidence of provider decisions and is a more practical way to assess the effect of quality improvement interventions. However, medical record review may not reveal the reasons for changes in behavior.
Our study is limited in that our algorithm for medically indicated compared with not indicated was empiric, not a nationally accepted guideline. This was necessary because there was no standard set of guidelines at that time. Spong et al6 have recently reported guidelines for late-term and near-term scheduled births. Their guidelines, like ours, are largely empiric, although from a larger group of experts. The empiric guidelines reflect the absence of hard data in this area.6 As nonmedical indications for delivery undergo more scrutiny, we anticipate that firmer guidelines will be created. We decided that nonmedical indications for delivery, when accompanied by a mature amniocentesis, were acceptable because for the majority of the study time that was allowable by ACOG standards. The goal of the Ohio Perinatal Quality Collaborative was not to exceed current standards but rather to live up to them.
The effort to decrease nonmedically indicated deliveries is not unique to Ohio. The Ohio Perinatal Quality Collaborative project likely benefited from secular trends surrounding this issue. Nevertheless, the Ohio Perinatal Quality Collaborative's project drove substantial change in practice. Further work on the best and most economical methods to change provider practices remains to be performed. Clark et al.7 compared three methods to reduce the rate of unneeded scheduled births before 39 weeks of gestation. A “soft stop” approach led to significantly fewer unnecessary scheduled births compared with education alone, but a “hard stop,” wherein hospital personnel are empowered to stop nonmedically indicated deliveries at less than 39 weeks of gestation, was the most effective way to change practice patterns. Hospitals participating in the Ohio Perinatal Quality Collaborative were free to use any method. Seven of the 20 Ohio Perinatal Quality Collaborative hospitals used a “hard stop” and the rest used something softer.
Nonmedically indicated deliveries should be deferred until after 39 weeks, but the reduction in stillbirth related to timely scheduled birth before 39 weeks for women with clear medical indications should not be forgotten.8 For example, the HYPITAT trial clearly showed that delivery at 37 weeks was better for women and their newborns with mild preeclampsia and gestational hypertension compared with later delivery.9 A change in culture surrounding nonmedically indicated deliveries at less than 39 weeks could have unintended consequences if women with strong medical indications for early delivery are not delivered in a timely fashion. We were not able to ascertain the population denominators (number of all ongoing pregnancies at 36–40 weeks) for each site, so we could not determine site-specific rates of stillbirth. We therefore monitored the overall Ohio stillbirth rate during the course of our initiative and observed no change from previous years. Similarly, Oshiro et al10 found no change in the rate of stillbirth during a similar project in Utah. However, Ehrenthal et al11 were able to determine the number of ongoing pregnancies during a similar project at a single site. The stillbirth rate increased from 2.5 to 9.1 per 10,000 term pregnancies during the course of their successful effort to limit elective deliveries before 39 weeks. This issue clearly deserves continuous attention.
Deliveries with strong medical indications remained stable, whereas women with unnecessary or intermediate indications for delivery at less than 39 weeks of gestation decreased significantly during the Ohio Perinatal Quality Collaborative's effort to decrease nonmedically indicated deliveries before 39 weeks. This finding persisted after accounting for a mature amniocentesis.
1. Tita AT, Landon MB, Spong CY, Lai Y, Leveno KJ, Varner MW, et al.. Timing of elective repeat cesarean delivery at term and neonatal outcomes. N Engl J Med 2009;360:111–20.
2. Hibbard JU, Wilkins I, Sun L, Gregory K, Haberman S, Hoffman M, et al.. Respiratory Morbidity in Late Preterm Births. JAMA 2010;304:419–25.
3. McIntire DD, Leveno KJ. Neonatal mortality and morbidity rates in late preterm births compared with births at term. Obstet Gynecol 2008;111:35–41.
4. The Ohio Perinatal Quality Collaborative Writing Committee. A statewide initiative to reduce inappropriate scheduled births at 36 0/7 to 38 6/7 weeks' gestation. Am J Obstet Gynecol 2010;202:243.e1–8.
5. Bailit JL. Rates of labor induction without medical indication are overestimated when derived from birth certificate data. Am J Obstet Gynecol 2010;203:269.e1–3.
6. Spong CY, Mercer BM, D'Alton M, Kilpatrick S, Blackwell S, Saade G. Timing of indicated late-preterm and early-term birth. Obstet Gynecol 2011;118:323–33.
7. Clark SL, Frye DR, Meyers JA, Belfort MA, Dildy GA, Kofford S, et al.. Reduction in elective delivery at <39 weeks of gestation: comparative effectiveness of 3 approaches to change and the impact on neonatal intensive care admission and stillbirth. Am J Obstet Gynecol 2010;203:449.e1–6.
8. Ananth CV, Gyamfi C, Jain L. Characterizing risk profiles of infants who are delivered at late preterm gestations: does it matter? Am J Obstet Gynecol 2008;199:329–31.
9. Koopmans CM, Bijlenga D, Groen H, Vijgen SMC, Aarnoudse JG, Bekedam DJ, et al.. Induction of labour versus expectant monitoring for gestational hypertension or mild pre-eclampsia after 36 weeks' gestation (HYPITAT): a multicentre, open-label randomised controlled trial. Lancet 2009;374:979–88.
10. Oshiro BT, Henry E, Wilson J, Branch DW, Varner MW, Women and Newborn Clinical Integration Program. Decreasing elective deliveries before 39 weeks of gestation in an integrated health care system. Obstet Gynecol 2009;113:804–11.
11. Ehrenthal DB, Hoffman MK, Jiang X, Ostrum G. Neonatal outcomes after implementation of guidelines limiting elective delivery before 39 weeks of gestation. Obstet Gynecol 2011;118:1047–55.