Methods

The Health Department of each of the states that make up the United States, as well as the District of Columbia, was contacted with a request to provide basic natality information for the years 2007–2013, inclusive. Specifically, each Health Department was asked to provide, for each calendar year separately, the number of live births that occurred within each week of gestational age at and beyond 37 weeks, and the number of stillbirths that occurred within each week of gestational age at and beyond 37 weeks. Health Departments that either did not provide any requested information, or that provided incomplete information, were recontacted via a series of e-mails and/or telephone calls. Because this investigation did not seek patient-level information, it was exempted from review by the Hershey Medical Center institutional research board. For this study, the early-term period of gestation was defined as 37 weeks, 0 days to 38 weeks, 6 days, the full-term period was defined as 39 weeks 0 days to 40 weeks 6 days, the late-term period was defined as 41 weeks 0 days to 41 weeks 6 days, and the post-term period was defined as equal to or greater than 42 weeks 0 days of gestation.

Datasets from states were considered usable for this study if they included, at a minimum, live birth and stillbirth data for at least 1 year before and 1 year after 2010. Because of privacy issues related to low counts, some of the smaller states were unwilling to provide term stillbirth data for any given week of gestation for any given year if the count was less than 6. Several states provided total term stillbirth data by year only.

Collected data were entered into a Microsoft Excel 2010 spreadsheet (Microsoft Corp, Redmond, WA). The total number of term births and term stillbirths that occurred in each calendar year in each of the states, and for the United States as a whole, was determined. The timing of term childbirth and the timing of term stillbirth, as a function of week of gestation, were determined for each state and for the United States as a whole. Annual overall rates of term stillbirth were calculated for each state and for the United States as a whole. The denominator for each calculation was composed of the sum of live births and stillbirths for any given year and geographic entity. Finally, the rates of term stillbirth for each state, and for the United States as a whole, were calculated for the composite 3-year period before 2010 and for the composite 3-year period after 2010. These 2 composite rates were then compared to generate a relative risk statistic. As a simplified reporting metric, a composite state rate after 2010 (2011−2013) was considered to be “higher” if it increased by 5% or more, or “lower” if it decreased by 5% or more, compared with that state’s rate before 2010 (2007−2009). A composite state rate after 2010 was considered to be “unchanged” if it was neither “higher” nor “lower” compared with that state’s rate before 2010.

The pattern of term stillbirth, as a function of both calendar year and gestational age, was determined for each state and for the United States as a whole. Patterns were based on the proportion of all term stillbirths that occurred in any given week of term gestation in any given year for each geographic entity multiplied by the overall term stillbirth rate for that geographical entity in that year (ie, “rate-proportion”, or the burden of stillbirth activity by week of gestational age for each geographical entity = [# stillbirths in any given week/total number of term stillbirths for that year] * that year’s overall term stillbirth rate for the geographical entity). Similarly, patterns of the rate of term stillbirth also were determined for each state and for the United States as a whole for each “phase” of pregnancy (ie, “early-term,” “full-term,” and “post-term”) by identifying the proportion of term stillbirth rate occurring in each phase of pregnancy for each year for each geographical entity.

During the analysis of the timing of term childbirth in the United States between 2007 and 2013, continuous changes were seen in the proportion of deliveries that occurred at different gestational ages. In addition, the change in the rate of term stillbirth in the United States appeared to be continuous rather than dichotomous. Accordingly, linear regression was used to estimate the rate of this continuous change in national term stillbirth rate over time. The estimated rate of change in term stillbirth over time was used to estimate the change in the incidence of term stillbirth in the United States in 2013 compared with 2007.

Sensitivity analyses were performed to evaluate the potential impact of missing data on both the timing of term delivery and the estimation of changes in the rate of term stillbirth over the 7-year period. Finally, using the change in the proportion of deliveries that occurred before the 39th week of gestation in 2013 compared with 2007, and using the estimated change in the incidence of term stillbirth that occurred in 2013 compared with 2007, we estimated the number needed to treat (NNT). This NNT represents the number of women who would need to be delivered as they would have been in 2007 (ie, early-term delivery without an accepted indication) to prevent 1 term stillbirth.

Results

Queries to all 50 states, and the District of Columbia, resulted in 46 usable datasets ( Table 1 ). A total of 33 states and the District of Columbia provided full datasets; 12 states provided partial datasets. Usable data could not be obtained from 5 states. One state was unable to provide any data beyond 2010, 2 states were not able to produce any of the requested data, and 2 states had administrative requirements for data release that could not be met. Four states suppressed data for any given week of gestational age in any given year if the count was less than 6, and 4 states provided term stillbirth data as a total figure for each year. Approximately 86% of US term birth data and stillbirth data were obtained for analysis (calculations not shown).

In all states that provided data, and in the United States as a whole, there was a clear decrease during the 7-year period in the proportion of births in the 38th week, and a clear increase in the proportion of births in the 39th week of gestation ( Figure 1 for US data; Internet link for state data ). There was a decrease in the proportion of deliveries that occurred in the 40th week of gestation. Less-dramatic changes were seen in the proportion of childbirth that occurred in the other weeks of gestation. Using linear regression methodology, we determined that the changes that occurred during the 7-year period in both the 38th and 39th weeks of gestation were statistically significant in most states and the United States as a whole (data not shown). When we compared the patterns of the timing of term childbirth for 2007 and 2013, the proportion of term births that occurred in the early term period were 32.3% and 27.9%, respectively. Approximately 6.4% of the delivering population in 2013 might have been delivered in the early-term period rather than the full-term period if they had received their care in 2007.

Live term births by year and by week of gestation

Nicholson et al. The 39-week rule and higher rates of term stillbirth. Am J Obstet Gynecol 2016 .

In addition to changes in the timing of term childbirth, the annual rate of term stillbirth during the 7-year period increased in most states and in the United States as a whole ( Table 1 ). When we compared 2007−2009 with 2011−2013, the rate of term stillbirth increased in 24 states (median increase 16%), decreased in 5 states (median decrease 9%), and was unchanged in 17 states. The relative risk of term stillbirth in each state, along with the respective 95% confidence interval (95% CI), was calculated ( Table 1 ) and graphed ( Figure 2 ). For the United States as a whole, the overall rate of term stillbirth increased from 1.103/1000 [2007−2009] to 1.177/1000 [2011−2013], relative risk 1.067, 95% CI (1.038–1.096).

Forest plot of relative risk of term stillbirth

CI , confidence interval; M-H , Mantel-Haenszel test statistic.

Nicholson et al. The 39-week rule and higher rates of term stillbirth. Am J Obstet Gynecol 2016 .

Considering the United States as a whole, changes in the timing of term childbirth involving the 38th and 39th weeks of gestation were continuous during the 7-year period ( Figure 1 ). Changes in the annual rate of US term stillbirth also appeared to be continuous ( Figure 3 ). Linear regression revealed an estimated slope of 0.0186/1000/year (95% CI 0.002–0.035; P = .034) for the continuous increase in the US term stillbirth rate. An analysis of the changes in the patterns of term stillbirth revealed that the increases were greatest in the 38th and 39th weeks of gestation ( Figure 4 ) and that these increases during these weeks were continuous in nature. States with the fastest rate of reduction in the proportion of term deliveries occurring in the 38th week of gestation had the fastest increase in overall rate of term stillbirth (data not shown). Sensitivity analyses revealed that the findings of this study would not have been significantly affected if several of the smaller states had provided complete data. Finally, the NNT to prevent one term stillbirth was estimated to be 1:449.

Change in the rate of term stillbirth, 2007−2013

Nicholson et al. The 39-week rule and higher rates of term stillbirth. Am J Obstet Gynecol 2016 .

Stillbirth rate proportions∗ by week of gestational age and year

∗ Rate-Proportions (RP) defined by the following formula: RP = (number of stillbirths in that week of gestation/total number of term stillbirths for that year) x overall term stillbirth rate for that year.

Nicholson et al. The 39-week rule and higher rates of term stillbirth. Am J Obstet Gynecol 2016 .

Results

Queries to all 50 states, and the District of Columbia, resulted in 46 usable datasets ( Table 1 ). A total of 33 states and the District of Columbia provided full datasets; 12 states provided partial datasets. Usable data could not be obtained from 5 states. One state was unable to provide any data beyond 2010, 2 states were not able to produce any of the requested data, and 2 states had administrative requirements for data release that could not be met. Four states suppressed data for any given week of gestational age in any given year if the count was less than 6, and 4 states provided term stillbirth data as a total figure for each year. Approximately 86% of US term birth data and stillbirth data were obtained for analysis (calculations not shown).

Table 1

State information

State name	Percent of US population (for 2010)	Study data status	Years reported	Change in proportion of early-term delivery	Change in rate of term stillbirth (07−09 vs 11−13)	Relative risk of term stillbirth (with 95% confidence interval) (07−09 vs 11−13)
Alabama	1.63%	Complete	2007–2013	Decreased	Increased	1.16 (0.97–1.38)
Alaska	0.32%	Complete	2007–2013	Decreased	Increased	1.26 (0.77–2.04)
Arizona	2.45%	Complete	2007–2013	Decreased	Increased	1.05 (0.89–1.24)
Arkansas	1.06%	Complete a	2007–2013	Decreased	–No change–	0.97 (0.76−1.22)
California	14.53%	Five years	2007–2011	Decreased	Increased	1.11 (1.00−1.23)
Colorado	1.88%	Complete	2007–2013	Decreased	–No change–	1.02 (0.86–1.22)
Connecticut	1.04%	Six years	2007–2012	Decreased	–No change–	0.84 (0.64−1.10)
Delaware	0.31%	Five years	2007–2011	Decreased	Decreased	0.94 (0.48−1.85)
District of Columbia	0.25%	Complete	2007–2013	Decreased	Decreased	0.76 (0.47–1.22)
Florida	5.98%	Complete	2007–2013	Decreased	–No change–	0.96 (0.86–1.07)
Georgia	3.58%	Six years	2007–2012	Decreased	Increased	1.19 (1.04–1.36) b
Hawaii	0.53%	Six years c	2007–2012	Decreased	Increased	1.37 (0.85–2.21)
Idaho	0.65%	Complete	2007–2013	Decreased	Increased	1.27 (0.93–1.72)
Illinois	4.57%	Complete	2007–2013	Decreased	Increased	1.08 (0.95–1.23)
Indiana	–	Not usable	None	Unknown	Unknown	Unknown
Iowa	1.08%	Complete	2007–2013	Decreased	–no change–	1.01 (0.78–1.31)
Kansas	1.15%	Complete	2007–2013	Decreased	Increased	1.28 (1.00–1.63) b
Kentucky	1.47%	Complete	2007–2013	Decreased	Increased	1.10 (0.89–1.36)
Louisiana	1.66%	Four years a	2008–2011	Decreased	Decreased	0.88 (0.68–1.15)
Maine	0.37%	Complete	2007–2013	Decreased	Increased	1.24 (0.78–1.96)
Maryland	2.05%	Complete	2007–2013	Decreased	–No change–	1.02 (0.85–1.22)
Massachusetts	2.07%	Complete a	2007–2013	Decreased	–No change–	0.96 (0.80–1.16)
Michigan	3.12%	Complete	2007–2013	Decreased	–No change–	0.99 (0.85–1.16)
Minnesota	1.94%	Complete	2007–2013	Decreased	Increased	1.17 (0.98-1.41)
Mississippi	1.07%	Complete	2007–2013	Decreased	–No change–	1.03 (0.80–1.32)
Missouri	–	Unobtained	None	Unknown	Unknown	Unknown
Montana	0.34%	Five years c	2008–2012	Decreased	Increased	1.40 (0.78–2.54)
Nebraska	0.73%	Complete	2007–2013	Decreased	Decreased	0.74 (0.54–1.01)
Nevada	0.97%	Complete	2007–2013	Decreased	Increased	1.30 (1.01–1.68) b
New Hampshire	0.37%	Complete	2007–2013	Decreased	–No change–	1.03 (0.64−0 1.65)
New Jersey	–	Not usable	None	Unknown	Unknown	Unknown
New Mexico	0.76%	Complete c	2007–2013	Decreased	–No change–	0.98 (0.71–1.35)
New York	7.00%	Six years	2007–2012	Decreased	–No change–	0.96 (0.86–1.08)
North Carolina	3.32%	Complete	2007–2013	Decreased	Increased	1.05 (0.93–1.20)
North Dakota	0.26%	Six years c	2008–2013	Decreased	Increased	1.33 (0.79–2.23)
Ohio	3.87%	Complete	2007–2013	Decreased	Increased	1.21 (1.06–1.38) d
Oklahoma	1.47%	Complete a	2007–2013	Decreased	–No change–	1.04 (0.83–1.30)
Oregon	–	Unobtained	None	Unknown	Unknown	Unknown
Pennsylvania	3.96%	Six years	2007–2012	Decreased	Increased	1.12 (0.97–1.28)
Rhode Island	0.31%	Complete	2007–2013	Decreased	Increased	1.48 (0.91–2.41)
South Carolina	1.61%	Complete	2007–2013	Decreased	Increased	1.07 (0.88–1.29)
South Dakota	0.34%	Complete	2007–2013	Decreased	Increased	1.06 (0.71–1.58)
Tennessee	2.17%	Complete	2007–2013	Decreased	–No change–	1.02 (0.87–1.20)
Texas	10.23%	Complete	2007–2013	Decreased	–No change–	0.71 (0.65–0.78) e
Utah	1.47%	Complete	2007–2013	Decreased	–No change–	0.96 (0.76–1.20)
Vermont	−	Not usable	None	unknown	Unknown	unknown
Virginia	2.88%	Complete	2007–2013	Decreased	–No change–	1.01 (0.86–1.19)
Washington	2.41%	Complete	2007–2013	Decreased	Increased	1.12 (0.94−1.33)
West Virginia	0.57%	Complete c	2007–2013	Decreased	Increased	1.16 (0.78–1.73)
Wisconsin	1.90%	Five years	2008–2012	Decreased	Increased	1.68 (1.34–2.09) e
Wyoming	0.22%	Four years	2008-2009, 2011-2012	Decreased	Decreased	0.92 (0.44–1.94)

Only gold members can continue reading. Log In or Register to continue

Share this:

• X
• Facebook

Related

Related posts:

1. Prior uterine evacuation of pregnancy as independent risk factor for preterm birth: a systematic review and metaanalysis
2. Development and validation of a spontaneous preterm delivery predictor in asymptomatic women
3. Exercise during pregnancy protects against hypertension and macrosomia: randomized clinical trial
4. Maternal and fetal morbidity associated with uterine rupture of the unscarred uterus

Stay updated, free articles. Join our Telegram channel

Join