Você está na página 1de 7

Obstetrics: Original Research

Neonatal and Maternal Adverse Outcomes


Among Low-Risk Parous Women at 39–41
Weeks of Gestation
Han-Yang Chen, PhD, William A. Grobman, MD, MBA, Sean C. Blackwell, MD,
and Suneet P. Chauhan, MD, Hon DSc

OBJECTIVE: To compare the composite neonatal or live births. The risk of the composite neonatal adverse
maternal adverse outcome among low-risk, parous outcome was higher for those delivered at 40 (aRR 1.18;
women at 39–41 weeks of gestation. 95% CI 1.15–1.22) and 41 (aRR 1.59; 95% CI 1.53–1.65)
METHODS: This was a retrospective cohort study using weeks of gestation when compared with 39 weeks. The
the U.S. vital statistics data sets (2012–2016). We evalu- overall rate of the composite maternal adverse outcome
ated low-risk parous women with nonanomalous single- was 2.31 per 1,000 live births. The risk of the composite
ton gestations who delivered at 39, 40, or 41 weeks of maternal adverse outcome was also significantly higher with
gestation (as reported in completed weeks, eg, 39 weeks delivery at 40 (aRR 1.15; 95% CI 1.11–1.19) and 41 weeks of
includes 39 0/7–39 6/7 weeks of gestation). The primary gestation (aRR 1.50; 95% CI 1.42–1.58) than at 39 weeks.
outcome, the composite neonatal adverse outcome, CONCLUSION: Though only modestly, the rates of the
included any of the following: Apgar score less than 5 composite neonatal and maternal adverse outcomes
at 5 minutes, assisted ventilation for longer than 6 hours, increase, from 39 through 41 weeks of gestation, among
neonatal seizure, or neonatal mortality. The secondary low-risk parous women.
outcome, the composite maternal adverse outcome, (Obstet Gynecol 2019;134:288–94)
included any of the following: intensive care unit admis- DOI: 10.1097/AOG.0000000000003372
sion, blood transfusion, uterine rupture, or unplanned

E
hysterectomy. We used multivariable Poisson regression vidence from randomized trials and population-
analyses to estimate the association between gestational based reports indicates that, among low-risk nul-
age and adverse outcome (using adjusted relative risks liparous women, neonatal and maternal morbidities
[aRR] and 95% CI).
are significantly more common at 40 and at 41 weeks
RESULTS: Of 19.9 million live births during the study of gestation than they are at 39 weeks. Such neonatal
period, 5.4 million (27.1%) met inclusion criteria. Among complications include Apgar score less than 5 at 5 mi-
them, 54.4% delivered at 39 weeks of gestation, 35.7% at 40 nutes, respiratory morbidity, ventilation for 6 hours or
weeks, and 9.9% at 41 weeks. The overall rate of the
longer, neonatal seizure and mortality. The maternal
composite neonatal adverse outcome was 4.86 per 1,000
complications more common at 40 and 41 weeks of
gestation than at 39 weeks include hypertensive disor-
From the Department of Obstetrics, Gynecology, and Reproductive Sciences, der of pregnancy, cesarean delivery, admission to the
McGovern Medical School, the University of Texas Health Science Center at
Houston, Houston, Texas; and the Department of Obstetrics and Gynecology, intensive care unit, transfusion, uterine rupture, and
Northwestern University, Chicago, Illinois. unplanned hysterectomy.1,2
Each author has confirmed compliance with the journal’s requirements for Of the 2.39 million parous deliveries in 2017,
authorship. more than half were low-risk.3,4 Yet, in contradistinc-
Corresponding author: Han-Yang Chen, PhD, Department of Obstetrics, tion to uncomplicated nulliparous women,1,2,5–11
Gynecology, and Reproductive Sciences, University of Texas Health Science there is a paucity of published data about whether
Center at Houston, Houston, TX; email: han-yang.chen@uth.tmc.edu.
adverse outcomes increase in frequency with advanc-
Financial Disclosure
The authors did not report any potential conflicts of interest. ing gestational age among low-risk parous women
© 2019 by the American College of Obstetricians and Gynecologists. Published
after 39 weeks of gestation.
by Wolters Kluwer Health, Inc. All rights reserved. Thus, the primary objective of this analysis was to
ISSN: 0029-7844/19 compare adverse outcomes among a large sample of

288 VOL. 134, NO. 2, AUGUST 2019 OBSTETRICS & GYNECOLOGY

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
low-risk parous women who delivered at 39, 40 and to the obstetric estimate as the standard for estimating
41 weeks of gestation. The primary outcome was the the gestational age of a newborn.14
composite neonatal adverse outcome, and the sec- The main exposure variable for our analysis was
ondary outcome was the composite maternal adverse gestational age at delivery (ie, 39, 40, 41 weeks of
outcome. gestation). The primary outcome was the composite
neonatal adverse outcome, which included any of the
following: Apgar score less than 5 at 5 minutes,15 as-
METHODS
sisted ventilation for longer than 6 hours, neonatal
This was a population-based retrospective cohort
seizure, or neonatal mortality (defined as death within
study using the Period Linked Birth-Infant Death
27 days). The determination of death within 27 days
Data Files of U.S. vital statistics data from 2012 to was made by linking birth and death certificates
2016. These data, including all live births and linked (which has the variable “Age at Death in Days”).
to infant deaths within the first year, were assembled The secondary outcome was the composite maternal
by the National Center for Health Statistics and adverse outcome, which included any of the follow-
reported annually by the Centers for Disease Control ing: admission to the intensive care unit, maternal
and Prevention. Our study sample was restricted to blood transfusion, uterine rupture, or unplanned hys-
parous women who delivered between 2012 and 2016 terectomy. For determining the frequency of occur-
with singleton, nonanomalous gestations who did not rence for both composites, newborns or women with
have hypertensive disorders, pregestational or gesta- more than one outcome were counted only once.
tional diabetes, who experienced labor and delivered Differences in the maternal characteristics strati-
from 39 through 41 weeks of gestation, had cephalic fied by gestational age at delivery were examined
presentations, and had birth data recorded using the using chi-square tests for categorical variables. The
2003 revised birth certificate. rates of the composite neonatal and maternal adverse
Since 2003, the updated 2003 revision of birth outcomes, along with the individual components of
certificates has been incorporated gradually on a state- these composite measures, were reported as the
wide basis. Compared with the 1989 birth certificate number of cases per 1,000 live births. We used
version, the 2003 version contains more detailed multivariable Poisson regression models with robust
obstetric, medical, and demographic data.12 The error variance to examine the association between
revised birth certificate was used by 38 states and gestational age at delivery (using 39 weeks of gestation
Washington, DC, in 2012, by 41 states and Washing- as the reference) and the risks of composite and
ton, DC, in 2013, by 47 states and Washington, DC, individual neonatal and maternal adverse outcomes,
in 2014, by 48 states and Washington, DC, in 2015, while adjusting for maternal age, race and ethnicity,
and by all 50 states and Washington, DC, in 2016, education, marital status, total births, prepregnancy
which represented 86%, 90%, 96%, 97%, and 100% body mass index (BMI, calculated as weight in kilo-
of live births in the United States, respectively. grams divided by height in meters squared), prenatal
Because the data are publicly available and do not care, smoking during pregnancy, neonatal sex, and
contain direct personal identifiers, this study was con- delivery year. The results are presented as adjusted
sidered exempt from review by the institutional relative risk (aRR) with 95% CI. Those missing data
review board at the McGovern Medical School at for maternal race and ethnicity, maternal education,
University of Texas Health Science Center at prepregnancy BMI, prenatal care, and smoking dur-
Houston. ing pregnancy were categorized and analyzed as an
The 2003 revision of the birth certificate replaced “unknown” group.
the “clinical estimate of gestation” with the “obstetric We performed two sensitivity analyses. The first
estimate of gestation.” Detailed information of the sensitivity analysis was to ascertain whether the
methods for this obstetric estimate of gestation can associations of adverse maternal outcomes with ges-
be found elsewhere.13 The obstetric estimate of gesta- tational age persisted if the composite did not include
tion is reported in completed weeks (ie, 39 weeks in- maternal transfusion, which may not be reliably
cludes deliveries from 39 0/7–39 6/7 weeks of recorded or represent severe morbidity.16–18 The sec-
gestation). The reason for selecting obstetric estimate, ond sensitivity analysis was to ascertain whether the
over clinical estimate, is increasing evidence of the associations of adverse outcomes with gestational age
greater validity of the obstetric estimate compared persists among women without a prior cesarean.19 All
with the LMP-based measure. Additionally, in 2014, statistical analyses were conducted using SAS 9.4 and
the National Center for Health Statistics transitioned STATA 15.

VOL. 134, NO. 2, AUGUST 2019 Chen et al Adverse Outcomes in Low-Risk Parous Women at 39–41 Weeks 289

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
RESULTS The multivariable adjusted analysis showed that
From 2012 to 2016, of the 19,858,574 live births in the the risk of the composite neonatal adverse outcome
U.S. national database, 18,804,092 (94.7%) used the was higher at 40 (aRR 1.18, 95% CI 1.15–1.22) and 41
2003 revised birth certificate. After exclusion criteria weeks of gestation (aRR 1.59, 95% CI 1.53–1.65) than
were applied, 5,385,728 women remained for our at 39 weeks. The risks for three individual neonatal
analysis, of whom 54.4% (n52,928,605) delivered at morbidities—Apgar score less than 5 at 5 minutes, as-
39 weeks of gestation, 35.7% (n51,924,925) at 40 sisted ventilation for longer than 6 hours, and neona-
weeks, and 9.9% (n5532,198) at 41 weeks (Fig. 1). tal seizure—also were significantly higher at 40 and 41
Women delivered at 39 weeks of gestation were weeks of gestation than at 39 weeks; the risk of neo-
more likely to be younger (less than 35 years of age), natal death, however, was not significantly different at
of minority race or ethnicity, have lower education, different weeks of gestation (Table 2).
and have used cigarette during pregnancy, but were The overall rate of the composite maternal
less likely to be married (Table 1). The overall rate of adverse outcome was 2.31 per 1,000 live births
cesarean was 7.0% (n5204,317) at 39 weeks of gesta- (12,421/5,380,863). At 39 weeks of gestation, the
tion, 7.4% (n5141,941) at 40 weeks, and 11.4% overall rate was 2.10 per 1,000 live births (6,159/
(n560,929) at 41 weeks (P,.001). 2,926,240); at 40 weeks 2.40 per 1,000 live births
The overall rate of the composite neonatal (4,607/1,923,050); at 41 weeks 3.11 per 1,000 live
adverse outcome was 4.86 per 1,000 live births births (1,655/531,573). The proportion of missing
(26,078/5,361,898). At 39 weeks of gestation, the data was 0.09% (n54,865) for all individual compo-
composite neonatal adverse outcome was 4.33 per nents and 0.09% (n54,865) for the composite mater-
1,000 live births (12,627/2,917,124); at 40 weeks 5.10 nal adverse outcome. There was no meaningful
per 1,000 live births (9,760/1,915,405); at 41 weeks difference in the proportion of missing data between
6.96 per 1,000 live births (3,687/,529,369). The 39, 40, 41 gestational weeks of gestation.
missing data ranged 0–0.36% (n519,546) for individ- The multivariable adjusted analysis showed that
ual components, and 0.44% (n523,830) for the com- the risk of the composite maternal adverse outcome
posite neonatal adverse outcome. There was no was higher at 40 (aRR 1.15; 95% CI 1.11–1.19) and 41
meaningful difference in the proportion of missing weeks of gestation (aRR 1.50; 95% CI 1.42–1.58) as
data between 39, 40, 41 gestational weeks of gestation. compared with that at 39 weeks. The risk of two

Fig. 1. Flow chart of live births in


the United States (2012–2016),
eligibility and sample size. *Items
not mutually exclusive.
Chen. Adverse outcomes in Low-Risk
Parous Women at 39–41 Weeks. Ob-
stet Gynecol 2019.

290 Chen et al Adverse Outcomes in Low-Risk Parous Women at 39–41 Weeks OBSTETRICS & GYNECOLOGY

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
Table 1. Maternal Characteristics

Gestational Age (wk)


Characteristic Total (N55,385,728) 39 (n52,928,605) 40 (n51,924,925) 41 (n5532,198) P

Maternal age (y)


Younger than 20 131,906 (2.4) 73,706 (2.5) 46,396 (2.4) 11,804 (2.2) ,.001
20–34 4,364,404 (81.0) 2,376,433 (81.1) 1,558,688 (81.0) 429,283 (80.7)
35 or older 889,418 (16.5) 478,466 (16.3) 319,841 (16.6) 91,111 (17.1)
Maternal race and ethnicity
Non-Hispanic white 2,981,427 (55.4) 1,598,134 (54.6) 1,069,269 (55.5) 314,024 (59.0) ,.001
Non-Hispanic black 690,199 (12.8) 386,795 (13.2) 238,791 (12.4) 64,613 (12.1)
Hispanic 1,269,726 (23.6) 697,594 (23.8) 458,786 (23.8) 113,346 (21.3)
Non-Hispanic other 402,655 (7.5) 225,379 (7.7) 142,430 (7.4) 34,846 (6.5)
Unknown 41,721 (0.8) 20,703 (0.7) 15,649 (0.8) 5,369 (1.0)
Maternal education
Less than high school 845,076 (15.7) 451,290 (15.4) 308,060 (16.0) 85,726 (16.1) ,.001
High school 1,325,484 (24.6) 743,554 (25.4) 459,445 (23.9) 122,485 (23.0)
More than high school 3,156,637 (58.6) 1,703,746 (58.2) 1,135,579 (59.0) 317,312 (59.6)
Unknown 58,531 (1.1) 30,015 (1.0) 21,841 (1.1) 6,675 (1.3)
Marital status
Not married 1,949,550 (36.2) 1,076,199 (36.7) 686,585 (35.7) 186,766 (35.1) ,.001
Married 3,436,178 (63.8 1,852,406 (63.3) 1,238,340 (64.3) 345,432 (64.9)
Total birth order
2 2,377,591 (44.1) 1,251,772 (42.7) 878,106 (45.6) 247,713 (46.5) ,.001
3 1,448,538 (26.9) 801,631 (27.4) 510,605 (26.5) 136,302 (25.6)
4 772,546 (14.3) 435,853 (14.9) 265,890 (13.8) 70,803 (13.3)
5 or greater 787,053 (14.6) 439,349 (15.0) 270,324 (14.0) 77,380 (14.5)
Prenatal care
No 63,494 (1.2) 32,243 (1.1) 24,373 (1.3) 6,878 (1.3) ,.001
Yes 5,184,883 (96.3) 2,820,904 (96.3) 1,851,522 (96.2) 512,457 (96.3)
Unknown 137,351 (2.6) 75,458 (2.6) 49,030 (2.5) 12,863 (2.4)
Smoking during pregnancy
Yes 430,280 (8.0) 250,938 (8.6) 140,317 (7.3) 39,025 (7.3) ,.001
No 4,842,218 (89.9) 2,619,451 (89.4) 1,741,898 (90.5) 480,869 (90.4)
Unknown 113,230 (2.1) 58,216 (2.0) 42,710 (2.2) 12,304 (2.3)
Prepregnancy BMI (kg/m2)
Underweight (less than 18.5) 170,793 (3.2) 101,057 (3.5) 56,645 (2.9) 13,091 (2.5) ,.001
Normal weight (18.5–24.9) 2,506,014 (46.5) 1,364,873 (46.6) 900,924 (46.8) 240,217 (45.1)
Overweight (25–29.9) 1,404,482 (26.1) 755,647 (25.8) 505,925 (26.3) 142,910 (26.9)
Obesity class I (30.0–34.9) 690,652 (12.8) 374,349 (12.8) 244,710 (12.7) 71,593 (13.5)
Obesity class II (35.0–39.9) 290,507 (5.4) 158,698 (5.4) 101,250 (5.3) 30,559 (5.7)
Obesity class III (40.0 or more) 162,487 (3.0) 90,336 (3.1) 55,297 (2.9) 16,854 (3.2)
Unknown 160,793 (3.0) 83,645 (2.9) 60,174 (3.1) 16,974 (3.2)
Delivery route
Vaginal 4,978,541 (92.4) 2,724,288 (93.0) 1,782,984 (92.6) 471,269 (88.6) ,.001
Cesarean 407,187 (7.6) 204,317 (7.0) 141,941 (7.4) 60,929 (11.4)
Neonatal sex
Female 2,676,730 (49.7) 1,446,378 (49.4) 966,530 (50.2) 263,822 (49.6) ,.001
Male 2,708,998 (50.3) 1,482,227 (50.6) 958,395 (49.8) 268,376 (50.4)
Year
2012 976,114 (18.1) 531,258 (18.1) 351,903 (18.3) 92,953 (17.5) ,.001
2013 1,006,247 (18.7) 546,036 (18.6) 361,808 (18.8) 98,403 (18.5)
2014 1,095,282 (20.3) 594,096 (20.3) 391,026 (20.3) 110,160 (20.7)
2015 1,142,305 (21.2) 619,929 (21.2) 407,851 (21.2) 114,525 (21.5)
2016 1,165,780 (21.6) 637,286 (21.8) 412,337 (21.4) 116,157 (21.8)
BMI, body mass index.
Data are n (%) unless otherwise specified.

individual maternal morbidities—blood transfusion of admission to the intensive care unit and unplanned
and uterine rupture—also were significantly higher at hysterectomy were significantly higher only at 41
40 and 41 weeks of gestation than at 39 weeks; the risk weeks of gestation compared with 39 weeks (Table 3).

VOL. 134, NO. 2, AUGUST 2019 Chen et al Adverse Outcomes in Low-Risk Parous Women at 39–41 Weeks 291

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
Table 2. Composite and Individual Neonatal Adverse Outcome

Gestational Total Live Rate/1,000 Live Unadjusted RR Adjusted RR


Outcome Age (wk) Births n Births (95% CI) (95% CI)

Composite neonatal Total 5,361,898 26,078 4.86 (4.80–4.92)


adverse outcome
39 2,917,124 12,627 4.33 (4.25–4.40) 1.00 1.00
40 1,915,405 9,764 5.10 (5.00–5.20) 1.18 (1.15–1.21) 1.18 (1.15–1.22)
41 529,369 3,687 6.96 (6.74–7.19) 1.61 (1.55–1.67) 1.59 (1.53–1.65)
Apgar score less than 5 at Total 5,366,182 15,193 2.83 (2.79–2.88)
5 min
39 2,919,399 7,093 2.43 (2.37–2.49) 1.00 1.00
40 1,916,951 5,819 3.04 (2.96–3.11) 1.25 (1.21–1.29) 1.25 (1.21–1.30)
41 529,832 2,281 4.31 (4.13–4.48) 1.77 (1.69–1.86) 1.74 (1.66–1.83)
Assisted ventilation longer Total 5,381,281 10,760 2.00 (1.96–2.04)
than 6 h
39 2,926,267 5,338 1.82 (1.78–1.87) 1.00 1.00
40 1,923,300 3,939 2.05 (1.98–2.11) 1.12 (1.08–1.17) 1.13 (1.09–1.18)
41 531,714 1,483 2.79 (2.65–2.93) 1.53 (1.44–1.62) 1.51 (1.42–1.60)
Neonatal seizure Total 5,381,281 963 0.18 (0.17–0.19)
39 2,926,267 432 0.15 (0.13–0.16) 1.00 1.00
40 1,923,300 372 0.19 (0.17–0.21) 1.31 (1.14–1.50) 1.31 (1.14–1.50)
41 531,714 159 0.30 (0.25–0.35) 2.03 (1.69–2.43) 1.96 (1.63–2.35)
Neonatal death Total 5,385,728 1,297 0.24 (0.23–0.25)
39 2,928,605 699 0.24 (0.22–0.26) 1.00 1.00
40 1,924,925 459 0.24 (0.22–0.26) 1.00 (0.89–1.12) 1.02 (0.91–1.15)
41 532,198 139 0.26 (0.22–0.30) 1.09 (0.91–1.31) 1.09 (0.91–1.31)
RR, relative risk.
The composite neonatal adverse outcome includes Apgar score less than 5 at 5 minutes, assisted ventilation longer than 6 hours, neonatal
seizure, or neonatal mortality. Adjusted for maternal age, race and ethnicity, education, marital status, total births, prepregnancy body
mass index, prenatal care, smoking during pregnancy, neonatal sex, year. Bold indicates statistically significant results.

In the first sensitivity analysis which excluded and maternal adverse outcomes were still more com-
maternal transfusion, results were similar to the mon among parous women who labor at 40 and 41
primary analysis, with a higher risk of the composite weeks of gestation than 39 weeks.
maternal adverse outcome at 40 (aRR 1.10; 95% CI As expected among low-risk parous women, the
1.03–1.17) and 41 (aRR 1.53; 95% CI 1.39–1.68) absolute rate of any individual adverse outcome was
weeks of gestation than at 39 weeks (Appendix 1, uncommon, with rates of the composite neonatal
available online at http://links.lww.com/AOG/ adverse outcome being about 5 per 1,000 live births
B452). The second sensitivity analysis of women with and the composite maternal adverse outcome being 2
no prior cesarean also showed that both composite per 1,000 live births. Nevertheless, more than 1.2
neonatal and maternal adverse outcomes were signif- million low-risk parous women3,4 deliver in the
icantly higher at 40 and 41 weeks of gestation than at United States annually, and therefore even low-
39 weeks (Appendix 2, available online at http://links. frequency complications can result in a substantial
lww.com/AOG/B452). absolute number of women-newborn dyads being
affected by adverse outcomes. Because randomized
DISCUSSION trials,1,6 population-based studies,2 and retrospective
The results of this population-based study indicate analysis7,8 often focus on nulliparous women, these
that, among low-risk parous women who delivered at data help to fill a knowledge gap about low-risk par-
39–41 weeks of gestation, almost half (45.6%) were at ous women.10
40 or 41 weeks. The risk for overall composite neo- The strengths of the present study include its
natal and maternal adverse outcomes, as well as for population-based design with a large sample of live
many individual components of the composite out- births over 5 years, which account for 86–100% of live
comes, were significantly higher at 40 and 41 weeks births in the United States between 2012 and 2016.
of gestation than at 39 weeks. Our sensitivity analyses The data, which are recent, provide contemporary
showed consistent results that the composite neonatal determination of the rates of the composite neonatal

292 Chen et al Adverse Outcomes in Low-Risk Parous Women at 39–41 Weeks OBSTETRICS & GYNECOLOGY

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
Table 3. Composite and Individual Maternal Adverse Outcome

Gestational Total Live Rate/1,000 Live Unadjusted RR Adjusted RR


Outcome Age (wk) Births n Births (95% CI) (95% CI)

Composite maternal Total 5,380,863 12,421 2.31 (2.27–2.35)


adverse outcome
39 2,926,240 6,159 2.10 (2.05–2.16) 1.00 1.00
40 1,923,050 4,607 2.40 (2.33–2.46) 1.14 (1.10–1.18) 1.15 (1.11–1.19)
41 531,573 1,655 3.11 (2.96–3.26) 1.48 (1.40–1.56) 1.50 (1.42–1.58)
Admission to ICU Total 5,380,863 2,693 0.50 (0.48–0.52)
39 2,926,240 1,380 0.47 (0.45–0.50) 1.00 1.00
40 1,923,050 953 0.50 (0.46–0.53) 1.05 (0.97–1.14) 1.06 (0.98–1.15)
41 531,573 360 0.68 (0.61–0.75) 1.44 (1.28–1.61) 1.46 (1.30–1.64)
Maternal transfusion Total 5,380,863 9,780 1.82 (1.78–1.85)
39 2,926,240 4,813 1.64 (1.60–1.69) 1.00 1.00
40 1,923,050 3,664 1.91 (1.84–1.97) 1.16 (1.11–1.21) 1.17 (1.12–1.22)
41 531,573 1,303 2.45 (2.32–2.58) 1.49 (1.40–1.58) 1.51 (1.42–1.61)
Uterine rupture Total 5,380,863 1,079 0.20 (0.19–0.21)
39 2,926,240 491 0.17 (0.15–0.18) 1.00 1.00
40 1,923,050 424 0.22 (0.20–0.24) 1.31 (1.15–1.50) 1.31 (1.15–1.49)
41 531,573 164 0.31 (0.26–0.36) 1.84 (1.54–2.19) 1.81 (1.52–2.17)
Unplanned hysterectomy Total 5,380,863 1,089 0.20 (0.19–0.21)
39 2,926,240 581 0.20 (0.18–0.21) 1.00 1.00
40 1,923,050 358 0.19 (0.17–0.21) 0.94 (0.82–1.07) 0.95 (0.83–1.08)
41 531,573 150 0.28 (0.24–0.33) 1.42 (1.19–1.70) 1.45 (1.21–1.73)
RR, relative risk; ICU, intensive care unit.
The composite maternal adverse outcome includes admission to ICU, maternal transfusion, uterine rupture, or unplanned hysterectomy.
Adjusted for maternal age, race and ethnicity, education, marital status, total births, prepregnancy body mass index, prenatal care,
smoking during pregnancy, neonatal sex, year. Bold indicates statistically significant results.

and maternal adverse outcomes among low-risk par- the national rate of adverse outcomes. However, our
ous women at 39–41 weeks of gestation. In addition, findings of increased risks of composite maternal and
results from both sensitivity analyses were consistent neonatal adverse outcomes at 40 and 41 weeks of
with our original analyses: the composite adverse out- gestation had similar directionality as that among
come in low-risk parous women increases with low-risk nulliparous women.2 We excluded women
advancing gestational age between 39 and 41 weeks with hypertensive diseases, which may develop after
of gestation. 39 weeks of gestation,1 to make the groups maximally
We do acknowledge several limitations. Because comparable. However, this decision should bias
our analysis used data from the U.S. vital statistics toward the null; had we included women with hyper-
data sets, a variety of clinical information (eg, post- tensive disorders after 39 weeks of gestation, the
partum hemorrhage) was unavailable for analysis and adverse outcomes at these later gestational ages would
there is potential for unmeasured confounding. be expected to be worse. Finally, because some health
Although the data are stratified by gestational age, and medical items are likely underreported, research
the actual comparison of outcomes was not related to based on birth certificates has been criticized21; how-
specific management strategies because coding of ever, studies evaluating birth certificate data have con-
approaches to labor on the birth certificate, particu- sistently shown that the demographic and selected
larly for labor induction, has been reported to lack medical and health items (ie, method of delivery, birth
high accuracy.20 We used the 2003 revised birth cer- weight, and plurality) are collected with a high degree
tificate; the revised birth certificate represented 86%, of completeness and accuracy.22 Although maternal
90%, 96%, 97%, and 100% of live births in the United morbidity measures have been added to the birth data
States from 2012 to 2016, respectively. Therefore, our files since 2011, there is a paucity of research assessing
findings may not be generalizable to the whole U.S. the validity of these measures.23
population. Also, our results may not be applicable to In conclusion, among low-risk parous women
high-risk pregnancies among parous women. Thus, who delivered from 39 through 41 weeks of gestation,
the rate of composite maternal and neonatal adverse the risks of neonatal and maternal adverse outcomes,
outcomes described here should not be generalized to although uncommon, increase as gestational age

VOL. 134, NO. 2, AUGUST 2019 Chen et al Adverse Outcomes in Low-Risk Parous Women at 39–41 Weeks 293

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
advances. Our findings extend the previous study 13. National Center for Health Statistics. Guide to completing the
facility worksheets for the certificate of live birth and report of
examining low-risk nulliparous women and may be fetal death (2003 revision). Available at: https://www.cdc.gov/
useful for counseling women, and be nidus for nchs/data/dvs/GuidetoCompleteFacilityWks.pdf. Retrieved
interventional trials, with a primary outcome similar March 18, 2019.
to that reported by Grobman et al.1 14. Martin JA, Osterman MJK, Kirmeyer SE, Gregory ECW. Mea-
suring gestational age in vital statistics data: transitioning to the
obstetric estimate. National Vital Statistics Reports. Vol 64.
REFERENCES Hyattsville (MD): National Center for Health Statistics; 2015.
1. Grobman WA, Rice MM, Reddy UM, Tita ATN, Silver RM, 15. American College of Obstetricians and Gynecologists, Task
Mallett G, et al. Labor induction versus expectant management in Force on Neonatal Encephalopathy; American Academy of
low-risk nulliparous women. N Engl J Med 2018;379:513–23. Pediatrics. Neonatal encephalopathy and neurologic outcome.
2. Chen HY, Grobman WA, Blackwell SC, Chauhan SP. Neona- 2nd ed. Washington, DC: American College of Obstetricians
tal and maternal morbidity among low-risk nulliparous women and Gynecologists; 2014.
at 39–41 weeks of gestation. Obstet Gynecol 2019;133:729–37. 16. Main EK, Abreo A, McNulty J, Gilbert W, McNally C, Poeltler
3. Martin JA, Hamilton BE, Osterman MJK, Driscoll AK, Drake P. D, et al. Measuring severe maternal morbidity: validation of
Births: final data for 2017. National Vital Statistics Reports. Vol potential measures. Am J Obstet Gynecol 2016;214:643.e1–10.
67. Hyattsville (MD): National Center for Health Statistics; 2018. 17. Callaghan WM, Grobman WA, Kilpatrick SJ, Main EK, D’Al-
4. Chauhan SP, Rice MM, Grobman WA, Bailit J, Reddy UM, ton M. Facility-based identification of women with severe
Wapner RJ, et al. Neonatal morbidity of small- and large-for- maternal morbidity: it is time to start. Obstet Gynecol 2014;
gestational-age neonates born at term in uncomplicated preg- 123:978–81.
nancies. Obstet Gynecol 2017;130:511–9. 18. Kilpatrick SJ, Berg C, Bernstein P, Bingham D, Delgado A,
5. Gibbs Pickens CM, Kramer MR, Howards PP, Badell ML, Callaghan WM, et al. Standardized severe maternal morbidity
Caughey AB, Hogue CJ. Term elective induction of labor review: rationale and process. Obstet Gynecol 2014;124:361–6.
and pregnancy outcomes among obese women and their off- 19. Vaginal birth after cesarean delivery. ACOG Practice Bulletin
spring. Obstet Gynecol 2018;131:12–22. No. 205. American College of Obstetricians and Gynecologists.
6. Miller NR, Cypher RL, Foglia LM, Pates JA, Nielsen PE. Elec- Obstet Gynecol 2019;133:e110–27.
tive induction of labor compared with expectant management 20. Martin JA, Wilson EC, Osterman MJK, Saadi EW, Sutton SR,
of nulliparous women at 39 weeks of gestation: a randomized Hamilton BE. Assessing the quality of medical and health data
controlled trial. Obstet Gynecol 2015;126:1258–64. from the 2003 birth certificate revision: results from two states.
7. Osmundson S, Ou-Yang RJ, Grobman WA. Elective induction National Vital Statistics Reports. Vol 62. Hyattsville (MD):
compared with expectant management in nulliparous women National Center for Health Statistics; 2013.
with an unfavorable cervix. Obstet Gynecol 2011;117:583–7. 21. Schoendorf KC, Branum AM. The use of United States vital
8. Osmundson SS, Ou-Yang RJ, Grobman WA. Elective induc- statistics in perinatal and obstetric research. Am J Obstet Gy-
tion compared with expectant management in nulliparous necol 2006;194:911–15.
women with a favorable cervix. Obstet Gynecol 2010;116: 22. Martin JA, Wilson EC, Osterman MJK, Saadi EW, Sutton SR,
601–5. Hamilton BE. Assessing the quality of medical and health data
9. Alexander JM, McIntire DD, Leveno KJ. Forty weeks and from the 2003 birth certificate revision: results from two states.
beyond: pregnancy outcomes by week of gestation. Obstet Gy- Natl Vital Stat Rep 2013;62:1–19.
necol 2000;96:291–4. 23. Curtin SC, Gregory KD, Korst LM, Uddin SFG. Maternal
10. Verhoeven CJ, van Uytrecht CT, Porath MM, Mol BW. Risk morbidity for vaginal and cesarean deliveries, according to pre-
factors for cesarean delivery following labor induction in mul- vious cesarean history: new data from the birth certificate, 2013.
tiparous women. J Pregnancy 2013;2013:820892. National Vital Statistics Reports. Vol 64. Hyattsville (MD):
11. Carlson NS, Neal JL, Tilden EL, Smith DC, Breman RB, Lowe National Center for Health Statistics; 2015.
NK, et al. Influence of midwifery presence in United States cen-
ters on labor care and outcomes of low-risk parous women: a Con-
sortium on Safe Labor study. Birth 2018 [Epub ahead of print]. PEER REVIEW HISTORY
12. Osterman MJ, Martin JA, Mathews TJ, Hamilton BE. Received April 1, 2019. Received in revised form May 6, 2019.
Expanded data from the new birth certificate, 2008. Natl Vital Accepted May 16, 2019. Peer reviews and author correspondence
Stat Rep 2011;59:1–28. are available at http://links.lww.com/AOG/B453.

294 Chen et al Adverse Outcomes in Low-Risk Parous Women at 39–41 Weeks OBSTETRICS & GYNECOLOGY

© 2019 by the American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.