The preterm birth rate in the United States declined to 11.4% in 2013, the lowest level since 1997. Although the United States has one of the highest preterm birth rates in the developed world, we are improving this outcome and therefore improving the lives of thousands of infants. Demographic changes that may be responsible include a reduced teenage birth rate and fewer higher-order multiple births. Additionally, a public policy shift to prevent nonmedically indicated births at <39 weeks’ gestation and smoking bans in several states have been associated with the reduced rate of preterm births. Last, interventions such as 17 hydroxyprogesterone caproate, vaginal progesterone, and the use of cerclage in selected populations probably are contributing to the reduction in preterm deliveries. However, a large portion of these births could still be prevented with greater access and implementation of our current interventions, the reduction of modifiable risk factors for preterm birth, and expanded reporting of outcomes and risk factors to facilitate research for both prevention and treatment.
The rate of preterm birth (PTB) <37 weeks’ gestation in the United States was 11.4% in 2013, which was the lowest since 1997 ( Figure ). This rate steadily increased from 9.4% in 1981 until it peaked in 2006 at 12.8%. Over the last 7 years from 2006-2013, the PTB rate has declined 11%. During this same time span, early (<34 weeks’ gestation) and late (34-36 6/7 weeks’ gestation), PTBs decreased by 7.1% and 12.7%, respectively. These reductions equate to approximately 92,500 fewer PTBs in 2013, compared with 2006. Although the United States has reduced the PTB rate, our incidence of PTB still remains among the highest in the world, especially compared with other developed countries. An important approach to greater PTB prevention is understanding the reasons behind the recent decrease in PTB rates.
Reasons that the PTB rate declined
The reasons behind the decrease in the rate of PTB in the United States in the last 7 years may include changes in characteristics and risk factors of the US obstetric population, implementation of specific evidence-based guidelines, novel interventions in women with identifiable risk factors, and public health policies and regulations. A selective group of risk factors are presented in Table 1 .
| Risk factor | Year | Trend 2006-2013 | Approximate impact per risk factor a | ||
|---|---|---|---|---|---|
| 2006 | 2012 | 2013 b | |||
| Teen births (15-19 y), % c | 10.2 | 7.7 | 6.9 | Significant decrease | 28,100 fewer preterm births |
| Age, % c | |||||
| 35-39 y | 11.7 | 11.9 | 12.3 | Significant increase | 7800 fewer preterm births |
| 40-44 y | 2.5 | 2.8 | 2.8 | Significant increase | 200 more preterm births |
| Single marital status, % | 38.5 | 40.7 | 40.6 | Significant increase | 31,000 fewer preterm births f |
| Body mass index >30 kg/m 2 , % d , e | 29.1 | 30.3 | N/A | Trend analysis not appropriate | |
| Assisted reproductive technology (total deliveries), n e | 41,343 | 51,267 | N/A | Trend analysis not appropriate | |
| Multiple births, % | |||||
| Twins | 32.1 | 33.1 | 33.7 | Significant increase | 7500 fewer preterm births f |
| Triplets+ (per 100,000 births) | 153.3 | 124.4 | 119.5 | Significant decrease | 1600 fewer preterm births |
| Infection (per 100,000 women ), n e | Trend analysis not appropriate | ||||
| Chlamydia | 515.8 | 643.3 | N/A | ||
| Gonorrhea | 124.3 | 108.7 | N/A | ||
| Early term deliveries (all, 37-38 wks’ gestation, singleton birth), % | 28.91 | 24.7 | 24.5 | Significant decrease | N/A |
a Final data for 2013; does not control for other factors or interventions that may account for change within the group
b Personal communication from Joyce A. Martin, Centers for Disease Control and Prevention
c Percentage of births per total births or rate per 1000 women
e Trend analysis not available or appropriate because of inconsistent reporting between states and current data not being representative of the national trends
f Births to unmarried women and twins increased, although with fewer absolute preterm births in these groups, which is likely because of the interaction of other factors.
Characteristics of the obstetric population
Age
Teenagers and women of advanced maternal age (≥35 years old) both have an increased risk of PTB, with a dose-response relationship (the more extreme the age, the higher the rate of PTB). For teenagers, the total birth rate decreased by 6% in 2012, compared with 2011, and has been trending down since 1991. In 2013, the teen birth rate in those between 15-17 and 17-19 years old fell by an additional 13% and 8%, respectively. Although this reduction in teen births has lowered the overall rate of PTB, this effect may be diminished by a corresponding increase in PTB that is the result of more women delaying childbearing for personal reasons and/or career opportunities. This delay is most evident in the 35-39 and 40-44 year age groups, with an increase of 3% and 1% in birth rate, respectively, from 2012-2013. Births to women >45 years old also rose 7% over the past year from 0.7-0.8 births per 1000 women. However, when the rate of PTB in each respective age group from the Vital Statistics 2006 report and the absolute number of teen births and advanced maternal age births in 2013 is taken into account, the magnitude of decline in the teenage birth rate outweighs the increased PTB rate in the advanced maternal age population and may account for a small part of the decline in PTBs. From 2006-2012, there were approximately 23,000 fewer preterm neonates born to teen mothers. In contrast, there were only 9000 additional PTBs for women 35-44 years old.
Multiple gestations
Multiple gestations have long been associated with increased risk for poor obstetric outcomes, including PTB. From 1980-2009, the twin live birth rate rose 76%, primarily related to the increased use of assisted reproductive technology (ART), rather than to the increase in pregnancies in women ≥35 years old. The twin live birth rate has increased from 3.2% to 3.4% since 2002. Interestingly, in 2013, there were fewer total PTBs in twins compared with 2006, with 7521 fewer deliveries at <37 weeks’ gestation, likely the result of other demographic changes. By contrast, triplet deliveries decreased for all races since 1998 and have continued to decline in 2012 in all racial/ethnic groups except non-Hispanic black births. The reduction in higher order multiple births is certainly influenced by the revised guidelines released by the American Society for Reproductive Medicine in 2006. These guidelines were further refined to recommend single embryo transfer for women who were <35 years old with a favorable prognosis. Although reduction in higher order multiple births has been successful, they accounted for only 120 per 100,000 births in 2013, with a negligible effect on the total PTB rate (4387 of 542,893 total PTBs in 2013). In absolute numbers, there were approximately 450 fewer ART-associated triplet PTBs in 2010, based on a 97% rate of PTB in these pregnancies. However, reducing the amount of twin births as a result of ART has a great potential to impact the PTB rate in the United States. ART accounted for 19% of twin pregnancies in 2010, with a 65% PTB rate. Focusing specifically on these issues will have a much larger effect on the PTB rate.
Evidence-based guidelines
Adherence to guidelines derived from high-quality evidence can help unify the standard of care and reduce morbidity across the spectrum of pregnancy care from conception through delivery ( Table 2 ).
| Intervention | No. of preterm births prevented annually in the United States | Estimated reduction in preterm births, % |
|---|---|---|
| 17P with singleton births and previous spontaneous preterm birth | 10,000 | 0.02 |
| Screening asymptomatic singleton births without previous spontaneous preterm birth with the use of transvaginal ultrasound scanned cervical length and treating with vaginal progesterone of those with cervical length ≤20 mm before 25 weeks | 9500 a | |
| Screening asymptomatic singleton births with previous spontaneous preterm birth with transvaginal ultrasound scanned cervical length and performing cerclage in those with cervical length ≤25 mm before 24 weeks | 23,000 | 0.22 |
In 1999 and again in 2009, the American College of Obstetricians and Gynecologists (ACOG) recommended that births at <39 weeks’ gestation be scheduled only for valid medical and obstetric indications. Later in 2013, ACOG and the Society for Maternal Fetal Medicine (SMFM) recommended the cessation of nonindicated deliveries at <39 weeks’ gestation. It is possible that the effort to prevent nonmedically indicated deliveries at <39 weeks’ gestation reduced the number of women who otherwise would have delivered in the late preterm period (34 0/7 -36 6/7 weeks’ gestation) as physicians go through a paradigm shift in defining what a term birth is. From 2006-2012, the US rates of induction of labor declined 3% at 35 weeks’ gestation and 7% at 36 weeks’ gestation. Observance of these guidelines by practitioners and acceptance by the lay public has been aided by national campaigns that have been conducted by the March of Dimes, ACOG, the SMFM, and individual state-sponsored birth quality initiatives. For example, in South Carolina, there was a 4.7% decrease in late PTBs from 2011-2012 after the South Carolina Birth Outcomes Initiative instituted a hard-stop policy to nonmedically indicated early term deliveries. This collaboration between state health departments, hospitals, insurers, and others led to a 50% decline in nonmedically indicated early-term deliveries in South Carolina. The frequency of scheduled births at 36-38 completed weeks’ gestation increased substantially from 1990-2006 in Ohio. In 2007, the Ohio Perinatal Quality Collaborative (OPQC) was founded to pursue a mission of the use of collaborative improvement science methods to reduce PTBs and improve outcomes of preterm newborn infants in Ohio as rapidly as possible. The OPQC 39-week scheduled birth initiative reduced scheduled births between 36 0/7 and 38 6/7 weeks’ gestation that lacked a medical indication by 70%. Coincident with the implementation of this OPQC 39-week project, the rates of PTBs of 34-35 weeks’ gestation in Ohio declined to levels below the national average (unpublished data). This is remarkable because late PTB rates in Ohio have equaled or exceeded the national average since 1999. Although a causal relationship cannot be inferred from these data, this decline suggests that increased attention to the consequences of near-term birth may have influenced care for pregnancies at 34-36 weeks’ gestation. This is supported by recent findings of the decline in not only sPTB but also in indicated PTB (17.2% decline from 2005-2012). The potential for such an initiative to reduce prematurity and its associated sequelae is provocative and deserves further investigation in states with similar programs. Reductions in early inductions and cesarean deliveries have the potential to decrease the PTB rate in the United States by 0.61%, according to Chang et al and the Born Too Soon preterm prevention analysis group.
Interventions for high-risk women
Progesterone for singleton births with previous spontaneous PTB
In 2003, the Eunice Kennedy Shriver National Institutes of Child Health and Human Development Network of Maternal Fetal Medicine Units published results of a randomized placebo-controlled trial that demonstrated a significant reduction in recurrent PTB in women with a previous spontaneous PTB (sPTB) that had been treated with 17 α-hydroxyprogesterone caproate (17P). Use of 17P increased as prenatal care providers became more familiar and accepting of its use in women who have had a previous sPTB. Surveys of subspecialists and obstetrics/gynecology providers demonstrated a progressive increase in 17P use from 35-67% from 2004-2005 and up to 74% in 2007. The effect of this type of progesterone supplementation in practice on population-based statistics is estimated to be approximately 10,000 fewer PTB annually in the United States.
Cerclage for previous sPTB and short cervical length
Although cerclage was described decades ago, randomized trials that detail the indications for which it would be most effective (ie, ultrasound-indicated) only became available in the last 14 years. Metaanalyses have shown that cerclage decreases PTB in singleton births with a previous sPTB and short cervical length (CL) <25 mm at <24 weeks’ gestation. Approximately 40% of women with previous sPTB will experience a CL of <25 mm before 24 weeks. A policy of screening singleton pregnancies with previous sPTB with CL and performing cerclage in those who experience a CL of <25 mm at <24 weeks’ gestation is estimated to prevent >23,000 PTBs annually in the United States. If fully implemented, cerclage could reduce the PTB rate by 0.22%.
Vaginal progesterone with short CL
The use of 17P and cerclage will reduce the PTB rate in those with previous sPTB, but the greatest decrease will occur only when we can prevent sPTB in those who have not yet experienced one. In randomized controlled trials published in 2007 and 2009 and in a metaanalysis of patient-level data published in 2012, vaginal progesterone treatment for short cervix in conjunction with universal transvaginal ultrasound CL screening has been shown to reduce PTB in women with singleton births without a previous PTB and a short CL ≤20 mm before 25 weeks. Furthermore, cost-effectiveness studies have estimated that, if universal CL screening was used, 248 births at <34 weeks’ gestation would be prevented per 100,000 deliveries. This would translate to the prevention of approximately 9500 singleton deliveries at <34 weeks’ gestation annually in the United States, based on 2012 national vital statistics data. As we continue to answer questions regarding the optimal progesterone route and dosing and the populations to be targeted, progesterone therapy may reduce the overall PTB rate by 0.02%.
Public health policies and regulations
Smoking
Significant reductions in the rate of PTB have been observed after adoption of smoking bans in public places. Ordinances that provide for 100% smoke-free laws in the workplace (including restaurants and bars) have grown exponentially since 1993 when the first laws were passed. In 2006, when the PTB rate was at its peak, there were 181 ordinances against smoking in the workplace; currently, there are >628 ordinances in the United States. The recent metaanalysis by Been et al demonstrated a statistically significant 10.45% reduction in PTB in states where a smoking ban was in place compared with states without a ban. Individual studies examined this effect over a range of 1–5 years, depending on the study. Modeling based on smoking cessation programs in pregnancy estimate a 0.02% decline in the PTB rate. This rate is based on the fact that not all women will be successful in smoking cessation during that pregnancy. It has been estimated that, in the United States, the PTB rate would decrease by approximately 15% if smoking in pregnancy stopped. This information should lead public health policy toward increasing these bans and to further research worldwide in developing nations where these data are currently not available.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
