Background
Planned home births in the United States are associated with fewer interventions but with increased adverse neonatal outcomes such as perinatal and neonatal deaths, neonatal seizures or serious neurologic dysfunction, and low 5-minute Apgar scores. The American College of Obstetricians and Gynecologists’ Committee on Obstetric Practice states that, to reduce perinatal death and to improve outcomes at planned home births, strict criteria are necessary to guide the selection of appropriate candidates for planned home birth. The committee lists 3 absolute contraindications for a planned home birth: fetal malpresentation, multiple gestations, and a history of cesarean delivery.
Objective
The aim of this study was to evaluate whether there are risk factors that should be considered contraindications to planned home births in addition to the 3 that are listed by the American College of Obstetricians and Gynecologists.
Study Design
We conducted a population-based, retrospective cohort study of all term (≥37 weeks gestation), normal weight (≥2500 grams), singleton, nonanomalous births from 2009–2013 using the Centers for Disease Control and Prevention’s period-linked birth-infant death files that allowed for identification of intended and unintended home births. We examined neonatal deaths (days 0–27 after birth) across 3 groups (hospital-attended births by certified nurse midwives, hospital-attended births by physicians, and planned home births) for 5 risk factors: 2 of the 3 absolute contraindications to home birth listed by the American College of Obstetricians and Gynecologists (breech presentation and previous cesarean delivery) and 3 additional risk factors (parity [nulliparous and multiparous], maternal age [women <35 and ≥35 years old], and gestational age at delivery [37–40 and ≥41 weeks]).
Results
The overall risk of neonatal death was significantly higher in planned home births (12.1 neonatal death/10,000 deliveries; P <.001) compared with hospital births by certified nurse midwives (3.08 neonatal death/10,000 deliveries) or physicians (5.09 neonatal death/10,000 deliveries). Neonatal mortality rates were increased significantly at planned home births, with the following individual risk factors: breech presentation (neonatal mortality rate, 127.52/10.000 births), nulliparous pregnant women (neonatal mortality rate, 22.5/10,000), previous cesarean delivery (18.91/10,000 births), and a gestational age ≥41 weeks (neonatal mortality rate, 17.17/10,000 births). Planned home births with ≥1 of the 5 risk factors had significantly higher neonatal death risks compared with deliveries with none of the risks. Neonatal death risk was further increased when a woman’s age of ≥35 years was combined with either a first-time birth or a gestational age of ≥41 weeks.
Conclusions
In this study, we show 2 risk factors with significantly increased neonatal mortality rates at planned home births in addition to the 3 factors that are listed by the American College of Obstetricians and Gynecologists. These additional risks factors have neonatal mortality rates that are approaching or exceeding those for planned home birth after cesarean delivery: first-time births and a gestational age of ≥ 41 weeks. Therefore, 2 additional risk factors (first-time births and a gestational age of ≥41 weeks) should be added to the 3 absolute contraindications of planned home births that are listed by the American College of Obstetricians and Gynecologists (previous cesarean delivery, malpresentation, multiple gestations) for a total of 5 contraindications for planned home births.
Home births in the United States have increased over the last decade. The 37,551 home births in the United States in 2014 (0.94% of all US births) are now the highest in absolute numbers of all industrialized countries. Planned home births in the United States are associated with fewer interventions but with an increased risk of perinatal and neonatal death ; a 3-fold increased risk of neonatal seizures or serious neurologic dysfunction ; an increased risk of 5-minute Apgar score of 0, <7, and <4 ; an increased risk for neonatal death in a breech presenting fetus, and in women with previous cesarean births.
The American College of Obstetricians and Gynecologists’ (ACOG) Committee on Obstetric Practice described the safety of planned home births in the United States as controversial. In addition, ACOG states that in order to reduce perinatal mortality at planned home births and achieving favorable home birth outcomes, “…strict criteria are necessary to guide selection of appropriate candidates for planned home birth,” and it lists three absolute contraindications for a planned home birth: fetal malpresentation, multiple gestations, and a history of previous cesarean delivery.
The aim of this study was to analyze the association of neonatal death with additional risk factors such as nulliparity, a gestational age ≥ 41 weeks, and women ≥ 35 years of age, and to evaluate whether these risk factors should be considered additional contraindications to planned home birth.
Materials and Methods
Study population
This is a population-based, retrospective cohort study of all term (≥37 weeks gestation), normal weight (≥2500 g), singleton, nonanomalous births from 2009–2013, the last 5 years of the available data, in states that used the 2003 revised birth certificate and the Centers for Disease Control and Prevention’s period-linked birth-infant deaths files that allowed for identification of intended and unintended home births. We excluded births if they met any of the following criteria: birthplace outside the hospital or home; unintended home births; gestational age <37 weeks or not recorded; neonatal birthweight <2500 g or not recorded; multiple gestations; any congenital anomaly, Down syndrome or other chromosomal disorder confirmed or pending; and residents of a foreign country. Multiples births were excluded from the data evaluation because there were too few multiples among planned home births to create meaningful data. The data included the location of deliveries (home vs hospital), the attendant at the delivery, and, for deliveries that occurred at home, whether it was intended or unintended. This analysis included only intended home births; for hospital births, the analysis included those who attended the delivery (physician or midwife). Most deaths are linked to their corresponding birth certificates (approximately 99%); however, the deaths are weighted with the use of the weights that were assigned by the Centers for Disease Control and Prevention to account for unlinked deaths.
Statistical analysis
The dataset was analyzed to examine total neonatal deaths (death of a live-born neonate between days 0–27 of life) across 3 groups: hospital-attended births by certified nurse midwives, hospital-attended births by physicians, and planned home births.
Descriptive statistics for births and neonatal deaths were calculated separately for midwife-attended and physician-attended hospital births and compared with intended home births with the use of chi-square and Fisher’s exact tests with significance set at a probability value of <.05.
We calculated the risk of neonatal death per 10,000 deliveries and the standardized mortality ratio (SMR) for planned home births and physician-attended hospital births vs midwife-attended hospital births using the indirect method that accounted for parity (nulliparous vs multiparous), maternal age (women <35 vs ≥35 years old), and gestational age at delivery (37-40 vs ≥41 weeks gestation).
After restricting the sample to a relatively low-risk set of singleton births that delivered at ≥37 weeks gestation, ≥2500 g, and without congenital anomalies, we selected the most parsimonious set of confounding variables to facilitate the adjusted analysis. We controlled for age, parity, and postterm dates because these were determined a priori to be the strongest predictors for neonatal death.
An SMR >1 indicates that the risk of neonatal death in the study population (eg, intended home births or physician-attended hospital births) is higher than expected if the risk of neonatal death were similar to that experienced among midwife-attended hospital births. We also calculated 95% confidence intervals; if the 95% confidence interval did not cross 1.00, the difference in mortality rate was considered statistically significant. The SMRs for 5 risk factors for neonatal death were evaluated: 2 of the 3 absolute contraindications to home birth that are listed by ACOG (breech presentation and previous cesarean delivery) and 3 additional risk factors (parity, maternal age, and gestational age at delivery), as described earlier. We also calculated the risks and SMR for any and none of the aforementioned risk factors and for combinations of parity, maternal age, and gestational age. All data analysis was completed in SAS software (version 9.4; SAS Institute Inc, Cary, NC).
Because nonidentifiable data from a publicly available dataset were used, our study was not considered human subjects research and did not require review by the Weill Medical College of Cornell University Institutional Review Board.
Results
Characteristics
The study population included a total of 12,953,671 singleton, nonanomalous, term (≥37 weeks) deliveries with infants who weighed ≥2500 g who delivered in states that used the 2003 revised birth certificate from 2009–2013 in a hospital or were intended (or planned) home births. Of the whole group, 11,779,659 deliveries (90.9%) were hospital deliveries by physicians; 1,077,197 deliveries (8.3%) were hospital deliveries by certified nurse midwives, and 96,815 deliveries (0.7%) were intended (planned) home births. Of the total of 6467 neonatal deaths, 6015 deaths (93.0%) were hospital deliveries by physicians; 334 deaths (5.2%) were hospital deliveries by certified nurse midwives, and 118 deaths (1.8%) were intended (planned) home births.
Table 1 shows the characteristics of the 3 subgroups for all deliveries and neonatal deaths. Women with planned home births were significantly more likely to be non-Hispanic white, older, parous, self-payers for delivery, or to deliver infants at ≥4000 g and ≥41 weeks of gestational age (postterm). Approximately 1 in 168 planned home births was a breech delivery, and approximately 1 in 23 planned home births was a vaginal birth after cesarean delivery.
| Variable | Deliveries (n=12,953,671) | P value b | Neonatal deaths (n=6467) | P value b | ||||
|---|---|---|---|---|---|---|---|---|
| Hospital midwife (n=1,077,197), n (%) | Hospital physician (n=11,779,659), n (%) | Intended home birth (n=96,815), n (%) | Hospital midwife (n=334), n (%) | Hospital physician (n=6015), n (%) | Intended home births (n=118), n (%) | |||
| Maternal ethnicity | <.001 | <.001 | ||||||
| Non-Hispanic | ||||||||
| White | 577,665 (53.6) | 6,276,662 (53.3) | 87,253 (90.1) | 188 (56.3) | 3176 (52.8) | 110 (93.2) | ||
| Black | 137,484 (12.8) | 1,539,889 (13.1) | 1,890 (2.0) | 54 (16.2) | 1054 (17.5) | 0 | ||
| Hispanic | 283,687 (26.3) | 3,042,950 (25.8) | 4,643 (4.8) | 65 (19.5) | 1362 (22.6) | 4 (3.4) | ||
| Other | 71,335 (6.6) | 828,184 (7.0) | 1,614 (1.7) | 22 (6.6) | 332 (5.5) | 1 (0.8) | ||
| Unknown | 7,026 (0.7) | 91,974 (0.8) | 1,415 (1.5) | 5 (1.5) | 90 (1.5) | 3 (2.5) | ||
| Maternal age, y | <.001 | <.001 | ||||||
| <25 | 390,660 (36.3) | 3,745,494 (31.8) | 15,294 (15.8) | 134 (40.1) | 2311 (38.4) | 26 (22.0) | ||
| 25–34 | 567,481 (52.7) | 6,336,325 (53.8) | 61,409 (63.4) | 168 (50.3) | 2906 (48.3) | 64 (54.2) | ||
| ≥35 | 119,030 (11.0) | 1,697,022 (14.4) | 20,106 (20.8) | 32 (9.6) | 797 (13.3) | 27 (22.9) | ||
| Unknown | 26 (0.0) | 818 (0.0) | 6 (0.0) | 0 | 1 (0.0) | 0 | ||
| Maternal education, y | <.001 | .37 | ||||||
| <13 | 496,538 (46.1) | 4,998,057 (42.4) | 38,443 (39.7) | 173 (51.8) | 3122 (51.9) | 56 (47.5) | ||
| ≥13 | 566,134 (52.6) | 6,642,060 (56.4) | 57,729 (59.6) | 153 (45.8) | 2722 (45.3) | 60 (50.8) | ||
| Unknown | 14,525 (1.3) | 139,542 (1.2) | 643 (0.7) | 9 (2.7) | 172 (2.9) | 1 (0.8) | ||
| Prenatal visits, n | <.001 | <.001 | ||||||
| 0 | 8,839 (0.8) | 145,689 (1.2) | 2,652 (2.7) | 10 (3.0) | 228 (3.8) | 13 (11.0) | ||
| 1–5 | 49,372 (5.6) | 519,504 (4.4) | 13,255 (13.7) | 20 (6.0) | 482 (8.0) | 24 (20.3) | ||
| ≥6 | 973,834 (90.4) | 10,679,376 (90.7) | 79,765 (82.4) | 281 (84.1) | 4934 (82.0) | 78 (66.1) | ||
| Unknown | 45,152 (4.2) | 435,090 (3.7) | 1,143 (1.2) | 23 (7.9) | 371 (6.2) | 3 (2.5) | ||
| Insurance c | <.001 | <.001 | ||||||
| Private | 294,262 (27.3) | 3,464,544 (29.4) | 12,174 (12.6) | 71 (21.3) | 1346 (22.4) | 11 (9.3) | ||
| Government | 319,590 (29.7) | 3,337,667 (28.3) | 6,145 (6.3) | 118 (35.3) | 2096 (34.8) | 4 (3.4) | ||
| Self-pay/other | 47,071 (4.4) | 423,746 (3.6) | 42,808 (44.2) | 18 (5.4) | 257 (4.3) | 67 (56.8) | ||
| Unknown | 12,945 (1.2) | 89,663 (0.8) | 3,055 (3.2) | 4 (1.2) | 55 (0.9) | 4 (3.4) | ||
| Not reported | 403,329 (37.4) | 4,464,039 (37.9) | 32,633 (33.7) | 123 (36.8) | 2262 (37.6) | 31 (26.3) | ||
| Parity | <.001 | .25 | ||||||
| Nulliparous | 424,060 (39.4) | 4,756,609 (40.4) | 20,125 (20.8) | 157 (47.0) | 2482 (41.3) | 45 (38.1) | ||
| Parous | 641,625 (59.6) | 6,952,531 (59.0) | 75,809 (78.3) | 171 (51.2) | 3475 (57.8) | 70 (59.3) | ||
| Unknown | 11,512 (1.1) | 70,519 (0.6) | 881 (0.9) | 6 (1.8) | 56 (0.9) | 2 (1.7) | ||
| Previous cesarean delivery | <.001 | <.01 | ||||||
| No | 1,048,436 (97.3) | 9,961,948 (84.6) | 92,199 (95.2) | 323 (96.7) | 5108 (84.9) | 106 (89.8) | ||
| Yes | 22,176 (2.1) | 1,782,055 (15.1) | 4,273 (4.4) | 5 (1.5) | 867 (14.4) | 8 (6.8) | ||
| Unknown | 6,585 (0.6) | 35,656 (0.3) | 343 (0.4) | 6 (1.8) | 40 (0.7) | 4 (3.4) | ||
| Newborn weight, g | <.001 | <.001 | ||||||
| 2500–3999 | 982,994 (91.3) | 10,744,142 (92.2) | 76,428 (78.9) | 318 (95.2) | 5560 (92.4) | 98 (83.1) | ||
| ≥4000 | 94,203 (8.7) | 1,035,517 (8.8) | 20,387 (21.1) | 16 (4.8) | 455 (7.6) | 20 (16.9) | ||
| Gestational age, wk | <.001 | <.001 | ||||||
| 37-38 | 256,151 (23.8) | 3,341,327 (28.4) | 14,205 (14.7) | 93 (27.8) | 2261 (37.6) | 18 (15.3) | ||
| 39-40 | 606,165 (56.3) | 6,645,173 (56.4) | 54,232 (56.0) | 164 (49.1) | 2824 (46.9) | 50 (42.4) | ||
| ≥41 | 214,881 (19.9) | 1,793,159 (15.2) | 28,378 (29.3) | 78 (23.4) | 930 (15.5) | 49 (41.5) | ||
| Presentation | <.001 | <.01 | ||||||
| Cephalic | 1,036,683 (96.2) | 10,977,624 (93.2) | 93,462 (96.5) | 321 (96.1) | 5325 (88.5) | 105 (89.0) | ||
| Breech | 1,921 (0.2) | 300,204 (2.5) | 553 (0.6) | 3 (0.9) | 358 (6.0) | 7 (5.9) | ||
| Other | 11,189 (1.0) | 259,162 (2.2) | 470 (0.5) | 2 (0.6) | 170 (2.8) | 1 (0.8) | ||
| Unknown | 27,404 (2.5) | 242,669 (2.1) | 2,330 (2.4) | 8 (2.4) | 162 (2.7) | 5 (4.2) | ||
| Risk composite d | <.001 | .21 | ||||||
| No risk present | 414,744 (38.5) | 3,464,701 (29.4) | 37,286 (38.5) | 108 (32.3) | 1689 (28.1) | 28 (23.7) | ||
| Any risk present | 637,530 (59.2) | 8,124,803 (69.0) | 57,831 (59.7) | 218 (65.3) | 4185 (69.6) | 87 (73.7) | ||
| Unknown | 24,923 (2.3) | 190,155 (1.6) | 1,698 (1.8) | 8 (2.4) | 141 (2.3) | 3 (2.5) | ||
a At ≥37 weeks gestation and ≥2500 g by place of delivery and attendant; US national data (among states using the 2003 revised birth certificate), 2009–2013, total births: n=12,953,671; neonatal deaths: n=6494
b Probability values were calculated with the use of the Chi square test for deliveries and Fisher’s exact test for neonatal deaths, which compared planned home births/deaths with hospital midwife-attended births/deaths
d Risk composite (age ≥35 years and/or nulliparous and/or postterm, previous cesarean delivery or breech) vs no risk composite (age <35 years, parous, term, no previous cesarean delivery, and cephalic).
Neonatal death
The risk of neonatal death was significantly higher in planned home births (12.1 neonatal deaths/10,000 births; P <.001) compared with hospital births by certified nurse midwives (3.08 neonatal deaths/10,000 births) or physicians (5.09 neonatal deaths/10,000 deliveries; Table 2 ). Women with the highest increased individual risk for neonatal death at planned home births were those with breech presentation (127.52 neonatal death/10.000 births or 1 in 78 breech births), followed by nulliparous women (22.5 neonatal deaths/10,000 births or 1 in 444 first-time births), those with previous cesarean delivery (18.91 neonatal deaths/10,000 births or 1 in 529 births), pregnancies with a gestational age ≥41 weeks (17.17 neonatal deaths/10,000 births or 1 in 582 births with a gestational age ≥41 weeks), and women who were ≥35 years old (13.61 neonatal deaths or 1 in 735 births of women ≥35 years of age; Table 3 ). The difference in risks of neonatal death between women <35 and >35 years old was not so large (11.66 vs 13.61 neonatal deaths per 10,000 births). For those who were >35 years old, the neonatal death risk was below the risk of those with previous cesarean delivery (13.61 neonatal deaths/10,000 births among those ≥35 years of age vs 18.91 neonatal deaths/10,000 births for those with previous cesarean deliveries). Physician-attended hospital births had a higher neonatal mortality rate when compared with midwife-attended hospital births.
| Variable | Risk of neonatal death (deaths per 10,000 births) | Standardized mortality ratio a (95% confidence interval) |
|---|---|---|
| Hospital midwife births | 3.08 | 1.00 (reference) |
| Intended home births | 12.1 | 4.13 (3.38-4.88) |
| Hospital physician births | 5.09 | 1.66 (1.62-1.71) |
a Indirectly standardized with the use of maternal age (<35 vs ≥35 years), parity (nulliparous vs parous), and gestational age (37-40 vs ≥ 41 weeks).
| Variable | Risk in midwife hospital births neonatal deaths per 10,000 births (95% confidence interval) | Risk in intended home births neonatal deaths per 10,000 births (95% confidence interval) | Risk in physician hospital births neonatal deaths per 10,000 births (95% confidence interval) | Standardized mortality ratio (95% confidence interval) | |
|---|---|---|---|---|---|
| Planned home birth vs midwife hospital births | Physician vs midwife hospital births | ||||
| Parity | |||||
| Parous | 2.66 (2.26–3.06) | 9.29 (7.12–11.46) | 5.00 (4.83–5.16) | 3.49 (2.68–4.30) | 1.88 (1.82–1.94) |
| Nulliparous | 3.71 (3.13–4.29) | 22.50 (15.95–29.04) | 5.22 (5.01–5.42) | 6.06 (4.30–7.83) | 1.41 (1.35–1.46) |
| Gestational age, wk | |||||
| 37–40 | 2.95 (2.59–3.32) | 9.94 (7.57–12.32) | 5.07 (4.93–5.21) | 3.37 (2.57–4.17) | 1.72 (1.67–1.77) |
| ≥41 | 3.59 (2.79–4.40) | 17.17 (12.33–22.0) | 5.18 (4.85–5.52) | 4.78 (3.43–6.12) | 1.44 (1.35–1.54) |
| Maternal age, y | |||||
| <35 | 3.13 (2.78–3.49) | 11.66 (9.23–14.08) | 5.16 (5.02–5.30) | 3.72 (2.95–4.50) | 1.65 (1.60–1.69) |
| ≥35 | 2.67 (1.73–3.60) | 13.61 (8.49–18.74) | 4.65 (4.32–4.97) | 5.11 (3.19–7.03) | 1.74 (1.62–1.87) |
| Fetal presentation | |||||
| Cephalic | 3.10 (2.76–3.44) | 11.19 (9.05–13.34) | 4.85 (4.72–4.98) | 3.61 (2.92–4.31) | 1.57 (1.52–1.61) |
| Breech | 15.66 (0.00–33.35) | 127.52 (34.00–221.04) | 11.93 (10.69–13.16) | 8.14 (2.17–14.11) | 0.76 (0.68–0.84) |
| Previous cesarean delivery | |||||
| No | 3.08 (2.75–3.42) | 11.46 (9.27–13.64) | 5.13 (4.99–5.27) | 3.72 (3.01–4.43) | 1.66 (1.62–1.71) |
| Yes | 2.27 (0.29–4.25) | 18.91 (5.88–31.93) | 4.86 (4.54–5.19) | 8.33 (2.59–14.07) | 2.14 (2.00–2.29) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
