Outcomes Associated with Planned Home and Planned Hospital Births in Low-Risk Women Attended by Midwives in Ontario, Canada, 2003-2006: A Retrospective Cohort Study

Abstract

Midwives in Ontario, Canada, provide care in the home and hospital and are required to submit data for all births to the Ontario Ministry of Health database. The purpose of this study was to compare maternal and perinatal/neonatal mortality and morbidity and intrapartum intervention rates for women attended by Ontario midwives who planned a home birth compared with similar low-risk women who planned a hospital birth between 2003 and 2006.
The database provided outcomes for all women planning a home birth at the onset of labor (n = 6,692) and for a cohort, stratified by parity, of similar low-risk women planning a hospital birth.
The rate of perinatal and neonatal mortality was very low (1/1,000) for both groups, and no difference was shown between groups in perinatal and neonatal mortality or serious morbidity (2.4% vs 2.8%; relative risk [RR], 95% confidence intervals [CI]: 0.84 [0.68-1.03]). No maternal deaths were reported. All measures of serious maternal morbidity were lower in the planned home birth group as were rates for all interventions including cesarean section (5.2% vs 8.1%; RR [95% CI]: 0.64 [0.56, 0.73]). Nulliparas were less likely to deliver at home, and had higher rates of ambulance transport from home to hospital than multiparas planning home birth and had rates of intervention and outcomes similar to, or lower than, nulliparas planning hospital births.
Midwives who were integrated into the health care system with good access to emergency services, consultation, and transfer of care provided care resulting in favorable outcomes for women planning both home or hospital births.

Full text

180 BIRTH 36:3 September 2009
Outcomes Associated with Planned Home and
Planned Hospital Births in Low-Risk Women
Attended by Midwives in Ontario, Canada,
2003–2006: A Retrospective Cohort Study
Eileen K. Hutton, PhD, Angela H. Reitsma, BSc, BHSc(Midwifery), and
Karyn Kaufman, DrPH
ABSTRACT: Background: Midwives in Ontario, Canada, provide care in the home and
hospital and are required to submit data for all births to the Ontario Ministry of Health
database. The purpose of this study was to compare maternal and perinatal/neonatal mortality
and morbidity and intrapartum intervention rates for women attended by Ontario midwives
who planned a home birth compared with similar low-risk women who planned a hospital birth
between 2003 and 2006. Methods: The database provided outcomes for all women planning a
home birth at the onset of labor (n=6,692) and for a cohort, stratified by parity, of similar
low-risk women planning a hospital birth. Results: The rate of perinatal and neonatal mortality
was very low (1/1,000) for both groups, and no difference was shown between groups in
perinatal and neonatal mortality or serious morbidity (2.4% vs 2.8%; relative risk [RR], 95%
confidence intervals [CI]: 0.84 [0.68–1.03]). No maternal deaths were reported. All measures
of serious maternal morbidity were lower in the planned home birth group as were rates for
all interventions including cesarean section (5.2% vs 8.1%; RR [95% CI]: 0.64 [0.56, 0.73]).
Nulliparas were less likely to deliver at home, and had higher rates of ambulance transport
from home to hospital than multiparas planning home birth and had rates of intervention and
outcomes similar to, or lower than, nulliparas planning hospital births. Conclusions: Midwives
who were integrated into the health care system with good access to emergency services,
consultation, and transfer of care provided care resulting in favorable outcomes for women
planning both home or hospital births. (BIRTH 36:3 September 2009)
Key words: birth outcomes, cesarean section rates, home birth, midwifery care
Midwives in Ontario, Canada, provide care to low-risk
women in a model of continuity or case-based care
wherebyawomanisattendedbythesamesmallgroup
of midwives throughout pregnancy, birth, and the post-
partum period. Women who meet selective criteria have
the choice of a hospital birth or a planned home birth.
Midwives are expected to attend both home and hospital
births to the extent that maintenance of registration with
the regulatory college depends on a record of providing
care in both settings.
Midwives are well integrated into the Ontario
health care system; they have admission and discharge
Eileen K. Hutton is an Assistant Dean, Faculty of Health Sciences,
Director Midwifery Education Program and Associate Professor in
the Department of Obstetrics and Gynecology, McMaster University,
Hamilton; Angela H. Reitsma is a Registered Midwife, Practicing
Member at The Hamilton Midwives, Hamilton; and Karyn Kaufman is
a Professor Emeritus Faculty of Health Sciences, McMaster University,
Hamilton, Ontario, Canada.
Address correspondence to Dr. Eileen K. Hutton, McMaster University
Midwifery Program, 1200 Main Street West, MDCL 2210, Hamilton,
Ontario, Canada L8N 3Z5.
Accepted March 23, 2009
©2009, Copyright the Authors
Journal compilation ©2009, Wiley Periodicals, Inc.

BIRTH 36:3 September 2009 181
privileges at their local hospital(s), and access to other
health care providers for consultation or transfer of care
as required. Two midwives are in attendance at births
either in the home or in the hospital. The regulatory
body, the College of Midwives of Ontario, sets out eligi-
bility criteria for a home birth: women with twin, breech,
or medically complicated pregnancies; with more than
one previous cesarean section; or with gestation less
than 37 weeks or more than 43 weeks at labor onset
are ineligible. The college also prescribes the prenatal,
intrapartum, and postpartum conditions for which con-
sultation and transfer of care to a physician are manda-
tory (1,2). Random practice audits are undertaken by the
College of Midwives to ensure that standards and regu-
lations are followed.
We sought to compare the outcomes of home and
hospital births to inform the midwifery profession
and the wider obstetric community, because birth at
home remains controversial (3–9), and practitioners are
subjected to criticism for attending home births (10).
Although a single study cannot provide a definitive con-
clusion about the relative safety of birth setting, we
thought it was important to assess the relatively recent
experience within Ontario.
Evaluating the outcomes of home births poses sev-
eral challenges. Because adverse outcomes are infre-
quent, large samples are needed to detect important
differences; however, the number of births occurring at
home is not large in most Western countries. Studies of
home birth have been limited by sample size (4,11–24);
absence of a comparison group (15,18,20–22,25–30);
comparison with care provided by other practition-
ers (17,23,24,29,31,32); use of self-reporting survey
methods (25,27,30); or inclusion of both planned and
unplanned home births (31,33).
Midwifery care in Ontario is part of publicly funded
health care in the province. All midwives are required
to submit a “client tracking form,” which includes pre-
natal, intrapartum, and postpartum information for each
woman in care. A database of midwifery care has
Table 1. Details of Allocation of 419 Records Inconsistent with Home Birth Criteria
Number of
Record Allocation Record Details Cases
Retained in primary analysis (n=74) Breech delivering at home or transferred in labor 13
Preterm 35 wk 20
Antenatal transfer of care with possible return to midwifery care 41
Removed from any analysis (n=90) Breech with antenatal transfer of care and elective cesarean section 36
Very preterm 28 wk 25
Antenatal transfer of care for conditions judged to be permanent 29
Retained for sensitivity analysis (n=255) Breech with hospital delivery; no known transport in labor 30
Preterm <35 wk; >28 wk 41
Antenatal transfer of care unclear re possible return to midwifery 20
Oxytocin induction 164
been systematically compiled at the Ontario Ministry of
Health. Since 2003, the submission of data has been tied
to reimbursement, which together with ongoing audit and
adjustment for inconsistent information virtually ensures
a complete database.
During the study period (2003–2006), approximately
130,000 births (34) per year were reported in Ontario,
with midwives attending about 8,600 (6.6%) births annu-
ally. Actual home births accounted for 1.6 percent of
the total provincial births and 25 percent of midwife-
attended births (35).
The purpose of this study was to compare maternal
and perinatal/neonatal mortality and morbidity and intra-
partum intervention rates for women attended by Ontario
midwives who planned a home birth, compared with
similar low-risk women who planned a hospital birth
between 2003 and 2006.
Methods
The Ministry of Health database of midwifery care was
used to undertake a retrospective cohort study of 13,384
midwife-attended births that occurred between April 1,
2003, and March 31, 2006, to compare the outcomes
for all women who planned a home birth (n=6,692)
at the outset of labor with a comparable low-risk group
of women who planned a hospital birth at the outset
of labor. All outcomes were analyzed according to the
planned place of birth irrespective of actual birth set-
ting. The groups were matched with respect to parity and
previous lower segment cesarean section. Midwives reg-
istered with and regulated by the College of Midwives in
Ontario attended all births included in the study. Ethics
approval for the project was received from McMaster
University Research Ethics Board.
Selection of Study Groups
Women discuss with their midwives their preference for
place of birth at several points during the pregnancy, and

182 BIRTH 36:3 September 2009
a Evidence from other data fields that care was organized for a hospital birth.
b Planned home birth seemed unlikely but could not be ruled out by information available; all records indicating oxytocin induction.
c Home birth occurred.
All midwife
births, 2003-
2006
(n=25,720)
Indicated home,
other, or undecided
(n=7,037)
Indicated planned
hospital birth
(n=18,683)
Logic check
indicates possible
contraindication to
home birth
(n=419)
Excluded due to:
>1 previous cesarean section
Breech presentation
Multiple pregnancy
Preterm delivery
Medical induction
Any antenatal transfer of care
(n=5,259)
Incompatible
with home
birtha —
excluded
from data set
(n=90)
Retain for
sensitivity
analysisb
(n=255)
Return to
home birth
data setc
(n=74)
Eligible for
matching
(n=13,424)
Primary analysis
(n=6,692)
Randomly selected
and matched on:
Parity
1 previous
cesarean section
(n=6,692)
•
•
•
•
•
•
•
•
Sensitivity
analysis
(n=6,947)
Fig. 1. Selection of included records.
this information will be documented on the client medi-
cal record. The data form provides information about the
planned location of birth when labor begins, since inter-
vening situations and changes of preference can modify
the initial plan.
For this study, the planned home birth group included
all client records where “planned place of birth at the
outset of labor” was “home,” “other out-of-hospital loca-
tion” (Ontario offers no formal out-of-hospital alter-
native settings to home birth, such as birth centers),
or “undecided,” because in any of these situations the
occurrence of the birth outside the hospital was a pos-
sibility. To detect possible coding errors for the planned
place of birth, we carried out logic checks to identify
the records with contraindications to planned home birth
or records with interventions that were inconsistent with
home birth (oxytocin induction of labor) or in which an
antenatal transfer of care to a physician was documented.
Two experienced midwives independently used an algo-
rithm, developed a priori, to decide whether to include
the record in the main analysis, permanently exclude the
record, or reserve the record for the sensitivity analysis
(Fig.1,Table1).
The comparison group was derived from the remain-
ing records in the data set from the same time period,
all of which indicated that the hospital was the planned
place of birth at the outset of labor. To ensure a low-risk
cohort comparable to the planned home birth cohort, we
removed all the records in which a home birth would
have been contraindicated (had it been planned), or in
which a prelabor intervention had occurred that was
inconsistent with, or unlikely at, a home birth. Thus,
from the planned hospital group, we excluded the records
with more than one previous cesarean section, breech
presentation, multiple pregnancies, preterm delivery
prior to 37 weeks’ gestation, oxytocin induction of labor,
prostaglandin cervical ripening, or any antenatal transfer
of care to a physician.
To minimize the risk of bias, we stratified our cur-
rent low-risk hospital records on parity (0 or 1) and on
previous cesarean section (none or one), and from each

BIRTH 36:3 September 2009 183
strata selected a random sample of records matching the
groups on parity and previous cesarean section (36).
Outcomes
Our primary outcome was a composite measure of peri-
natal and neonatal mortality or serious morbidity, defined
as the presence of one or more of the following: death
(stillbirth or neonatal death 0–27 days, excluding lethal
anomalies and fetal demise before the onset of labor);
Apgar score of less than 4 at 5 minutes of age; neonatal
resuscitation requiring both positive pressure ventilations
and cardiac compressions; admission to a neonatal or
pediatric intensive care unit with a length of stay greater
than 4 days; or birthweight less than 2,500 g. We note
that although birthweight is not an outcome that can
be a result of birth place, failure to screen adequately
for extremes of weight can result in planning for an
inappropriate place of birth. Thus, low birthweight was
included as part of the composite. Perinatal or neonatal
mortality cases that listed a lethal anomaly were inde-
pendently reviewed by two authors blinded to group, and
Table 2. Baseline Characteristics of Women Planning
Home and Hospital Birth
Planned Planned
Home Hospital
(n=6,692) (n=6,692)
Characteristic No. (%) No. (%)
Age (yr)
<25 729 (10.9)844 (12.6)
25–34 4,428 (66.1)4,630 (69.2)
35 1,503 (22.5)1,199 (17.9)
Missing 32 (0.5)19 (0.3)
Parity
02,293 (34.3)2,298 (34.3)
1–4 4,172 (62.3)4,289 (64.1)
>4 221 (3.3)105 (1.6)
Missing 6 (0.1)–
Geographical location
South rural 1,022 (15.3)901 (13.5)
South urban 5,305 (79.3)4,937 (73.8)
North rural 91 (1.4)192 (2.9)
North urban 271 (4.0)661 (9.9)
Missing 3 (0.0)1(0.0)
Repeat Ontario midwifery client
Yes 3 ,044 (45.5)2,331 (34.8)
No 3,642 (54.4)4,357 (65.1)
Missing 6 (0.1)4(0.1)
Previous cesarean section
06,479 (96.8)6,485 (96.9)
1 200 (3.0)207 (3.1)
>16(0.1)–
Missing 7 (0.1)–
Median gestation at booking 11.011.0
Median gestation at birth 40.040.0
a consensus decision was made with respect to exclusion
from the composite outcome.
Other outcomes included maternal mortality, signifi-
cant morbidity, rates of intrapartum interventions, and
breastfeeding. We included any maternal death from
a direct obstetrical cause (as determined by a provin-
cial coroner’s review of all maternal deaths) occurring
between the onset of labor and 6 weeks postpartum. Sig-
nificant maternal morbidity included blood loss greater
than 1,000 mL or bleeding requiring a consultation with
a physician; any infection after onset of labor requir-
ing a consultation with a physician; any third or fourth
degree laceration; and any postpartum transfer of care to
a physician. Any situation requiring a consultation or a
transfer of care was deemed to be a significant outcome
because the indications for consultation and/or transfer
of care are well defined by the College of Midwives of
Ontario. So, whereas practitioners might typically under-
estimate blood loss greater than 1,000 mL, any symp-
tomatic blood loss or transfusion would be captured in
our analysis by including the consultation or transfer of
care in the postpartum period.
Intrapartum interventions included rates of labor aug-
mentation, use of any pharmaceutical pain relief, epi-
siotomy and perineal trauma, assisted vaginal delivery,
and cesarean section. Rates of infant feeding with for-
mula supplementing breastfeeding or exclusive formula
feeding were compared at 1 and 6 weeks of age.
Data Analysis
All the analyses were conducted using SPSS 15.0 (37).
We used descriptive statistics to report baseline char-
acteristics. Comparative analyses used chi-square and
relative risk (RR) and 95 percent confidence inter-
vals (CI). We removed the possible misclassified home
birth records from the primary analysis and retained them
for a sensitivity analysis of the primary outcome. We also
conducted a subgroup analysis of nulliparas and multi-
paras using descriptive statistics.
Results
Data Cleaning
We found 7,037 records indicating that at the outset
of labor, birth was intended to take place at home. Of
these, we identified 419 records in which an intervention
inconsistent with home birth was undertaken (induction
with oxytocin [n=164]) or a contraindication to home
birth existed (breech delivery [n=79], preterm labor
[n=86]), or an antenatal transfer of care had occurred
(n=90). These records were analyzed as described in
Fig. 1 and Table 1, resulting in 6,692 records for the
primary analysis, and 6,947 for the sensitivity analysis.

184 BIRTH 36:3 September 2009
Table 3. Comparison of Variables in Planned Home and Hospital Groups with Relative Risk Presented for Selected
Outcomes
Planned Home Planned Hospital
(n=6,692) (n=6,692)
Variable No. (%) No. (%) pRR [95% CI]
Birth
Actual place of birth
Home 5,259 (78.6)208 (3.1)
Hospital 1,371 (20.5)6,467 (96.6)
Other location 62 (0.9)17 (0.3)
Ambulance transport from home
during or immediately after birth
Yes 361 (5.4)44 (0.7)
No 6,307 (94.2)6,544 (97.8)
Missing 24 (0.4)104 (1.5)
Estimated intrapartum blood loss
<500 mL 6,048 (90.4)5,909 (88.3)
500–1000 mL 568 (8.5)678 (10.1)
>1000 mL 56 (0.8)82 (1.2)0.026 0.68 [0.49–0.96]
Missing 20 (0.3)23 (0.3)
Consultation or transfer of care for
bleeding
79 (1.2)106 (1.6)0.046 0.75 [0.56–1.00]
Laceration
Any laceration 3,612 (54.0)4,081 (61.0)
1st degree perineal 1,109 (16.6)1,186 (17.7)
2nd degree perineal 1,695 (25.3)1,939 (29.0)
3rd degree perineal 78 (1.2)123 (1.8)
4th degree perineal 21 (0.3)22 (0.3)
Labial 413 (6.2)381 (5.7)
Vaginal 474 (7.1)542 (8.1)
Any 2nd–4th degree perineal, labial
or vaginal tear, or episiotomy
2,589 (38.7)2,979 (44.5)0.000 0.87 [0.83–0.90]
Intrapartum transfer of care 837 (12.5)1,270 (19.0)0.000 0.66 [0.61–0.71]
Postpartum transfer of care 119 (1.8)104 (1.6)0.311 1.14 [0.88–1.49]
Intrapartum intervention
Labor induction∗
None 6,586 (98.4)6,524 (97.5)
ARM before labor 73 (1.1)136 (2.0)
Prostaglandin 27 (0.4)—
Unknown/missing 15 (0.2)21 (0.3)
Labor augmentation
None 4,797 (71.7)4,203 (62.8)
Any augmentation 1,852 (27.7)2,426 (36.3)0.000 0.76 [0.72–0.80]
ARM 1,496 (22.4)1,889 (28.2)
Oxytocin 551 (8.2)878 (13.1)
Unknown/missing 43 (0.6)63 (0.9)
Pharmaceutical pain relief
None 5,570 (83.2)3,667 (54.8)
Any pharmaceutical pain relief 1,122 (16.8)3,025 (45.2)0.000 0.37 [0.35–0.39]
Nitrous oxide 221 (3.3)1,206 (18.0)
Narcotic analgesia 117 (1.7)423 (6.3)
Regional anesthesia
Epidural 655 (9.8)1405 (21.0)
Spinal 114 (1.7)199 (3.0)
General anesthesia 26 (0.4)48 (0.7)
Local anesthesia 191 (2.9)352 (5.3)
Episiotomy 286 (4.3)393 (5.9)0.000 0.73 [0.63–0.84]
Mode of delivery
Spontaneous vaginal 6,146 (91.8)5,852 (87.4)
Assisted vaginal 195 (2.9)293 (4.4)0.000 0.67 [0.56–0.80]
Forceps 81 (1.2)141 (2.1)
Vacuum 124 (1.9)168 (2.5)
Cesarean section 348 (5.2)544 (8.1)0.000 0.64 [0.56–0.73]
*Oxytocin induction is not undertaken at home and was an exclusion for the hospital birth cohort.
ARM =artificial rupture of membranes.

BIRTH 36:3 September 2009 185
Table 4. Comparison of Neonatal Variables in Planned Home and Hospital Groups with Relative Risk Presented for
Selected Outcomes
Planned Home Planned Hospital
(n=6,692) (n=6,692)
Variable No. (%) No. (%) pRR [95% CI]
Breech presentation 12 (0.2)—
Multiple birth — —
Gestational age (wk)
<37 17 (0.3)—
37–41, 6 days 6,555 (98.0)6,638 (99.2)
>41, 6 days 117 (1.7)54 (0.8)
Missing 3 (0.0)—
Birthweight (g)
<2,500 38 (0.6)56 (0.8)
2,500–4,000 5,364 (80.2)5,410 (80.8)
>4,000 1,279 (19.1)1,220 (18.2)
Missing 11 (0.2)6(0.1)
Apgar scores
<4at1min 89(1.3)107 (1.6)
Missing 49 (0.7)36 (0.5)
<7at5min 47(0.7)58 (0.9)
Missing 40 (0.6)34 (0.5)
Infant resuscitation
PPV 432 (6.5)382 (5.7)
PPV and chest compressions 21 (0.3)23 (0.3)
Significant congenital anomalies
None 6,515 (97.4)6,513 (97.3)
Major 28 (0.4)24 (0.4)
Minor 140 (2.1)144 (2.2)
Life threatening 5 (0.1)4(0.1)
Unknown/missing 4 (0.1)7(0.1)
Infant feeding at 1 wk
Exclusively breastfed 6,122 (91.5)5,638 (84.2)
Breastfed with supplement, or exclusively formula fed 544 (8.1)1,016 (15.2)0.000 0.53 [0.48–0.59]
Unknown 26 (0.4)38 (0.6)
Infant feeding at 6 wk
Exclusively breastfed 5,853 (87.5)5,140 (76.8)
Breastfed with supplement, or exclusively formula fed 620 (9.3)1,260 (18.8)0.000 0.49 [0.44–0.53]
Unknown/missing 219 (3.3)292 (4.4)
Any mortality 9 (0.1)9(0.1)
Stillbirth∗3(0.0)4(0.1)
Neonatal mortality†6(0.1)4(0.1)
Infant death 28–42 days — 1 (0.0)
*No congenital anomalies noted in either group.
†Includes 2 infants with a major congenital anomaly in the planned hospital group (1 brain tumor, 1 liver cirrhosis).
PPV =positive pressure ventilation.
Participants
Once the records were classified, the planned home
birth group comprised 6,692 women. A sample of 6,692
women was then randomly selected and stratified for par-
ity and one previous cesarean section from the low-risk
planned hospital group for the analysis. The groups were
similar on baseline characteristics (Table 2). Most women
were between 25 and 34 years of age, multiparous, and
lived in the southern urban part of the province. More
women planning home births (45.5%) had received care
from a midwife in a previous pregnancy compared with
those in the hospital birth group (34.8%).
Outcomes
Of 6,692 women in each group, 5,259 (78.6%) of those
in the planned home birth group gave birth at home
and 6,467 (96.6%) of those in the planned hospital
birth group gave birth in hospital (Table 3). Overall,
98.6 percent of newborns were born between 37 and 41
completed weeks of pregnancy (Table 4). The planned
home birth group had more babies with a gestational age
greater than 42 weeks (1.7% vs 0.8%).
Table 5 displays the perinatal/neonatal composite out-
come (2.4% vs 2.8% RR [95% CI]: 0.84 [0.68, 1.03]).
No differences were reported in the rates of perinatal or

186 BIRTH 36:3 September 2009
Table 5. Primary Outcome Including Primary and Sensitivity Analyses
Planned Home Planned Hospital
(n=6,692) (n=6,690)
Outcome No. (%) No. (%) pRR [95% CI]
Primary outcome: Perinatal/neonatal
morbidity or mortality
Composite outcome in primary analysis∗159 (2.4)190 (2.8)0.092 0.84 [0.68–1.03]
Composite outcome in sensitivity analysis
(planned home birth, n=6,947)
180 (2.6)190 (2.8)0.371 0.91 [0.75–1.12]
Composite component outcomes
(primary analysis)
Perinatal/neonatal mortality 9 (0.1)6(0.1)0.600 1.50 [0.53–4.21]
Stillbirth (after onset of labor) 3 (0.0)4(0.1)
Neonatal death <28 days 6 (0.1)2(0.0)
Neonatal morbidity 155 (2.3)189 (2.8)0.063 0.82 [0.66–1.01]
Birthweight <2,500 g 38 (0.6)56 (0.8)
Apgar <4at5min 10(0.1)8(0.1)
PPV and cardiac compressions 21 (0.3)23 (0.3)
NICU >4 days 102 (1.5)115 (1.7)
∗Two infants with a major congenital anomaly excluded from planned hospital group.
PPV =positive pressure ventilation; NICU =neonatal intensive care unit.
neonatal mortality between the groups, or in the rates
of serious morbidity (2.3% vs 2.8% RR [95% CI]: 0.82
[0.66, 1.01]). The rate of perinatal and neonatal mortal-
ity was 1 death per 1,000 live births for both groups,
with 9 in the planned home birth group and 8 in the
planned hospital birth group (of which 2 had significant
congenital conditions and were removed from the com-
posite outcome). One additional death occurred between
28 and 42 days of age in the planned hospital group.
With respect to infant feeding, infants born to women
planning to birth at home were one half as likely to use
supplemental formula feeding or be exclusively formula
fed at 1 week (8.1% vs 15.2%, RR [95% CI]: 0.53 [0.48,
0.59]) and at 6 weeks of age (9.3% vs 18.8%, RR [95%
CI]: 0.49 [0.44, 0.53]).
No maternal deaths occurred in either group. Women
in the planned home birth group experienced fewer
intrapartum interventions for each specific intervention
studied (induction, augmentation, pharmaceutical pain
relief, episiotomy, assisted delivery), including an abso-
lute decrease of 2.9 percent in the rate of cesarean
section (5.2% vs 8.1%, RR [95% CI]: 0.64 [0.56, 0.73]).
Women in this group also had less perineal trauma and
reduced incidence of blood loss greater than 1,000 mL.
Intrapartum transfers of care from midwives to another
practitioner (typically an obstetrician) were significantly
fewer in the planned home birth group (12.5% vs 19.0%,
RR [95% CI]: 0.66 [0.61, 0.71]) (Table 3).
Sensitivity Analysis
As described earlier, after completing the logic checks,
we excluded from our primary analysis 255 records that
were identified as women who were planning a home
birth. Although we judged it unlikely that these records
were truly planned home births at the outset of labor, we
undertook a sensitivity analysis and added them to the
planned home birth group, and reanalyzed our primary
outcome. Inclusion of these additional records did not
change the findings (Table 5).
Subgroup Analysis
The results of the subgroup analysis are reported in
Table 6. In our sample, nulliparas were less likely to
deliver at home, and had higher rates of ambulance trans-
port from home to hospital compared with multiparas
planning a home birth. Compared with multiparas, nul-
liparas in both the planned home and hospital groups
were more likely to experience outcomes such as blood
loss greater than 1,000 mL; experience 2nd, 3rd, or 4th
degree tears; require consultation or transfer of care;
and have interventions such as augmentation, pain relief,
assisted vaginal delivery, and cesarean section. The rates
of intervention and outcomes among nulliparas plan-
ning home birth were similar to, or lower than, among
those planning hospital births. Neonatal morbidity and
mortality was similar among all subgroups. Multiparas
planning home birth were most likely to be exclusively
breastfeeding at 6 weeks, followed by nulliparas plan-
ning birth at home. Nulliparas planning hospital birth
had the lowest rate of breastfeeding.
Discussion
As expected in a low-risk population, overall rates of
intervention, maternal morbidity, and perinatal/neonatal
morbidity and mortality were low for both groups. With

BIRTH 36:3 September 2009 187
Table 6. Subgroup Analysis of Selected Variables for Nulliparas and Multiparas by Planned Home and Hospital Groups
Planned Home Planned Hospital
Nulliparas (n=2,293) Nulliparas (n=2,298)
Multiparas (n=4,393) Multiparas (n=4,394)
Variable No.a(%) No.a(%)
Actual place of birth—home
Nulliparas 1,364 (59.5)76 (3.3)
Multiparas 3,891 (88.6)132 (3.0)
Ambulance transport from home during or
immediately after birth
Nulliparasb188 (8.2)14 (0.6)
Multiparasc173 (3.9)30 (0.7)
Estimated intrapartum blood loss >1,000 mL
Nulliparasd29 (1.3)31 (1.3)
Multiparase27 (0.6)51 (1.2)
Consultation or transfer of care for bleeding
Nulliparas 32 (1.4)39 (1.7)
Multiparas 47 (1.1)67 (1.5)
Laceration–any 2nd–4th degree perineal, labial,
or vaginal tear, or episiotomy
Nulliparas 1,406 (61.3)1,382 (60.1)
Multiparas 1,182 (26.9)1,597 (36.3)
Intrapartum transfer of care
Nulliparas 638 (27.8)798 (34.7)
Multiparas 197 (4.5)472 (10.7)
Postpartum transfer of care
Nulliparas 66 (2.9)49 (2.1)
Multiparas 53 (1.2)55 (1.3)
Any labor augmentation
Nulliparasf817 (35.8)1,038 (45.5)
Multiparasg1,032 (23.7)1,388 (31.9)
Any pharmaceutical pain relief
Nulliparas 782 (34.1)1,434 (62.4)
Multiparas 338 (7.7)1,591 (36.2)
Episiotomy
Nulliparas 229 (10.0)277 (12.1)
Multiparas 57 (1.3)116 (2.6)
Assisted vaginal delivery
Nulliparas 166 (7.2)221 (9.6)
Multiparas 28 (0.6)72 (1.6)
Cesarean Section
Nulliparas 276 (12.0)365 (15.9)
Multiparas 71 (1.6)179 (4.1)
Infant feeding at 1 wk—exclusively breastfed
Nulliparash1,996 (87.4)1,825 (79.9)
Multiparasi4,122 (94.2)3,813 (87.2)
Infant feeding at 6 wk—exclusively breastfed
Nulliparasj1,962 (89.1)1,680 (76.7)
Multiparask3,887 (91.1)3,460 (82.2)
Composite perinatal neonatal morbidity/mortalityl
Nulliparas 80 (3.5)85 (3.7)
Multiparas 79 (1.8)105 (2.4)
Perinatal/neonatal mortalityl
Nulliparas 5 (0.2)4(0.2)
Multiparas 4 (0.1)2(0.1)
Neonatal morbidity
Nulliparas 78 (3.4)84 (3.7)
Multiparas 77 (1.8)105 (2.4)
aDenominator has been adjusted for missing data with missing data in home birth group and hospital birth group as follows: b8, 35; c16, 68; d6, 6;
e14, 15; f12, 15; g31, 48; h10, 15; i16, 23; j91, 107; k127, 185; l2 infants with a major congenital anomaly excluded from planned hospital group.

188 BIRTH 36:3 September 2009
an overall rate of maternal mortality in Canada approx-
imately 7 per 100,000 live births, it is clear that catas-
trophic events in the low-risk population are very rare,
and that this study cannot address this issue (38). Peri-
natal mortality rates in our study were 1 death per 1,000
live births in both groups. Statistics Canada reports a
perinatal mortality rate of 6.7 per 1,000 in Ontario in
2003, and 6.3 in all of Canada (39). This rate, however,
includes outcomes for pregnancies of all risk categories,
and is not directly comparable with the low-risk term
population that is reported in our study. Our findings
were comparable to, or lower than, those reported in
2002 of home birth in British Columbia, Canada, where
the perinatal mortality rate reported for home birth group
was 3.5 per 1,000 and for a similar group of physician
attended hospital births 1.3 per 1,000 (14).
We found no increase in risk of adverse perinatal and
neonatal outcomes in morbidity or mortality between
groups, but we found significantly lower rates of intra-
partum intervention and of serious maternal morbidity in
the planned home birth group compared with the planned
hospital group. With respect to the cases included in the
sensitivity analysis, it is possible that desire for home
birth at the outset of labor (despite a contraindication)
is construed at times to be a “plan” for a home birth.
We judged that for some of the 419 cases, where the
data file indicated that home birth was planned at the
outset of labor in the presence of a clear contraindi-
cation, some midwives might have erroneously entered
responses to indicate that a home birth had at some point
in the pregnancy been planned or desired. We believe
that our careful categorization of the two study groups,
our approach to these cases, and the sensitivity analysis
minimize the overestimation or underestimation of the
outcome rates for the planned home birth cohort.
Although the planned hospital birth group experienced
somewhat higher rates of labor augmentation, epidural
use, assisted vaginal birth, and an increase in cesarean
section, the women’s outcomes are reassuring. In con-
sidering explanations for the very favorable outcomes of
the planned home birth group, we believe it is unlikely
that variations in midwifery practice are responsible. The
same midwives cared for both groups of women. When
midwives provide care in the home, they may use differ-
ent skills to enhance normal labor and birth, but it seems
unlikely that such differences would extend systematically
across the number of midwives who provided care to the
study women. We cannot discount the possibility of biased
reporting of outcomes for the planned home birth group,
but audits of midwifery data forms compared with hospi-
tal records have not shown systematic underreporting of
problems for women transported from home to hospital.
In addition, one or more midwives are present for each
birth, and typically a student is in attendance, so several
individuals contribute to completing a client record.
It is far more likely that differences are due to the
women themselves, since they self-select the planned
place of birth. We noted no major confounders between
the two groups; however, more subtle but important dif-
ferences likely exist between the women planning home
and those planning hospital births. In Ontario, midwives
are already caring for a self-selected subgroup of the
population who has selected midwifery care over the
more typical obstetrical model. Women who plan a home
birth are often more motivated to avoid interventions
such as epidural analgesia, which reduces the potential
for other interventions compared with other women (16).
A finding that supports the hypothesis of differences
between the groups is the decreased rate of supple-
mented or formula fed infants at 6 weeks of age in
the planned home birth group. It is also possible that
multiparas in our sample who planned a home birth
did so because they had an uncomplicated first birth,
whereas those who experienced problems with previous
births were more likely to choose a hospital setting. We
were unable to verify this hypothesis, since we did not
have access to actual medical records with detailed client
histories.
It is also plausible that the setting itself influences the
outcomes. Being in one’s own home environment may
go beyond mere comfort and enhance the very process of
labor and birth. We found the 40 percent decrease in the
rate of cesarean section, an absolute difference of nearly
3 percent, surprising and speculated about the possible
influence of the hospital setting. Several hospitals in the
province require that midwives transfer care to a physi-
cian according to institutional criteria, the result being
more transfers than if criteria of the College of Midwives
were followed. This factor is borne out with our finding
of an increased rate of intrapartum transfer of care in the
planned hospital group, and leads to a greater number of
women having care in labor managed by hospital staff.
Distress at losing a known midwife as the attendant, a
more interventionist style of care, or both may contribute
to an increase in assisted or operative births.
Conclusions
Whatever their relative balance, we conclude that multi-
ple factors contribute to the decreased rate of interven-
tions and of maternal morbidity we found in the group
who planned home birth. Appropriate self-selection by
the women themselves and good screening by mid-
wives likely contribute to the observed results. A well-
integrated midwifery care system is known to have a
positive impact on outcomes (40), and it is likely that
the Ontario context of self-regulation, good integration
into the health care system with access to emergency
services, and consultation and transfer of care when nec-
essary also contribute to favorable outcomes.

BIRTH 36:3 September 2009 189
References
1. College of Midwives of Ontario. Statement of Homebirth. Toronto,
Canada: Author, 1994.
2. College of Midwives of Ontario. Indications for Mandatory Dis-
cussion, Consultation and Transfer of Care. (updated June 15).
Toronto, Canada: Author, 2000.
3. Royal College of Obstetricians and Gynaecologists/Royal Col-
lege of Midwives. Home Births. Joint statement No. 2. London:
Authors, 2007.
4. Wiegers TA, Keirse MJNC, van der Zee J, Berghs GAH. Out-
comes of planned home and planned hospital births in low risk
pregnancies: Prospective study in midwifery practices in the
Netherlands. BMJ 1996;313:1309–1313.
5. Campbell R, Macfarlane A. Where to Be Born: The Debate and
the Evidence. Oxford, England: National Perinatal Epidemiology
Unit, 1987.
6. Tew M. Safer Childbirth? A Critical History of Maternity Care.
2nd ed. London: Chapman and Hall, 1998.
7. Olsen O. Meta-analysis of the safety of home birth. Birth
1997;24(1):4–13.
8. Olsen O, Jewell D. Home versus hospital births. Cochrane
Database Syst Rev. 1998, Issue 3. Art. No.: CD000352. DOI:
10.1002/14651858.CD000352.
9. Amelink-Verburg MP, Verloove-Vanhorick SP, Hakkenberg RMA,
et al. Evaluation of 280,000 cases in Dutch midwifery practices:
A descriptive study. BJOG 2008;115:570–578.
10. American College of Obstetrics and Gynecology. American Col-
lege of Obstetrics and Gynecology (ACOG) Statement on Home
Births. News release February 6, 2008. Accessed March 11,
2009. Available at: http://www.acog.org/from_home/publications/
press_releases/nr02-06-08-2.cfm.
11. Berghs G, Spnjaards E, Driessen L, et al. Neonatal neurological
outcome after low-risk pregnancies. Eur J Obstet Gynecol Reprod
Biol 1995;62:167–171.
12. Shearer JML. Five year prospective survey of risk of booking for
a home birth in Essex. Br Med J 1985;291:1478–1480.
13. Mehl LE, Peterson GH, Whitt M, et al. Outcomes of elec-
tive home births: A series of 1,146 cases. J Reprod Med
1977;19:281–290.
14. Janssen PA, Lee SK, Ryan EM, et al. Outcomes of planned home
births versus planned hospital births after regulation of midwifery
in British Columbia. CMAJ 2002;166(3):315–323.
15. Crotty M, Ramsay AT. Smart R, Chan A. Planned home births in
South Australia 1976–1987. Med J Aust 1990;153:664–671.
16. Ackermann-Liebrich U, Voegeli T, Gunter-Witt K, et al. Zurich
Study Team. Home versus hospital deliveries: Follow up
study of matched pairs for procedures and outcome. BMJ
1996;313:1313–1318.
17. McKenna P, Mathews T. Safety of home delivery compared with
hospital delivery in the Eastern Region Health Authority in Ireland
in the years 1999–2002. Ir Med J 2003;96(7):198–200.
18. Murphy PA, Fullerton J. Outcomes of planned home birth in
nurse-midwifery practice: A prospective descriptive study. Obstet
Gynecol 1998;92:461–470.
19. Hinds MW, Bergersen GH, Allen DT. Neonatal outcomes
planned versus unplanned out-of-hospital births in Kentucky.
JAMA 1985;253:1578–1582.
20. Davies J, Hey E, Reid W, Young G, Home Birth Study Steering
Group. Prospective regional study of planned home birthday. BMJ
1996;313:1302–1306.
21. Ford C, Iliffe S, Franklin O. Outcome of planned home birth in
an inner city practice. BMJ 1991;303:1517–1519.
22. Schneider G, Soderstrom B. Analysis of 275 planned and
10 unplanned home births. Can Fam Physician 1987;33:
1163–1171.
23. Woodcock HC, Read AW, Bower C, Stanley FJ, et al. A matched
cohort study of planned home and hospital births in Western Aus-
tralia 1981–1987. Midwifery 1994;10:125–135.
24. Durand AM. The safety of home birth: The farm study. Am J
Public Health 1992;82(3):450–452.
25. Bastian H, Keirse M, Lancaster P. Perinatal death associated with
planned home birth in Australia: Population based study. BMJ
1998;317:384–388.
26. Aikens Murphy P, Fullerton J. Outcomes of intended home births
in nurse midwifery practice: A prospective descriptive study.
Obstet Gynecol 1998;92(3):461–469.
27. Johnson KC, Davis BA. Outcomes of planned home birthday with
certified professional midwives: Large prospective study in North
America. BMJ 2005;330:1416–1422.
28. Tyson H. Outcomes of 1,001 midwife-attended home birthday in
Toronto, 1983–1988. Birth 1991;18(1):14–19.
29. Woodcock HC, Read AE, Moore DJ, et al. Planned homebirths
in Western Australia 1981–1987: A descriptive study. Med J Aust
1990;153:672–678.
30. Anderson RE, Aikins Murphy P. Outcomes of 11,788 planned
home births attended by certified nurse-midwives: A ret-
rospective descriptive study. J Nurse Midwifery 1995;40(6):
483–492.
31. Pang JW, Heffelfinger JD, Huang GJ, et al. Outcomes of planned
home birth in Washington State: 1989–1996. Obstet Gynecol
2002;100(2):253–259.
32. Schramm WF, Barnes DE, Bakewell JM. Neonatal mortality
in Missouri home births, 1978–84. Am J Public Health
1987;77(8):930–935.
33. Mori R, Dougherty M, Whittle M. An estimation of intrapartum-
related perinatal mortality rates for booked home births in
England and Wales between 1994 and 2003. BJOG 2008;115:
554–559.
34. Statistics Canada. Births 2005 . Statistics Canada Catalogue
number 84F0210X1E (1): Table 1, Births by geography,
Canada, 2004 and 2005;Accessed March 11, 2009. Available at:
http://www.statcan.ca/english/freepub/84F0210XIE/84F0210XIE
2005001.pdf.
35. Perinatal Partnership Program of Eastern and Southeastern
Ontario, Provincial Perinatal Surveillance System Committee.
2006 Provincial Perinatal Report. Ottawa, Canada: Author,
August 5, 2008; Accessed March 11, 2009. Available at
http://www.pppeso.on.ca/en/pppeso/NIDAY_Perinatal_Database_
p484.html.
36. College of Midwives of Ontario. Statement on VBAC and Choice
of Birthplace. Toronto, Canada: Author, 2002.
37. SPSS Inc. SPPS Version 15.0. Chicago, Illinois: Author, 2006.
38. World Health Organisation. Maternal Mortality in 2005 –Es-
timates Developed by WHO, NICEF, UNFPA and The World
Bank. World Health Organisation 2007;23, Annex 3: Estimates
of number of maternal deaths, lifetime risk, MMR and range
of uncertainty (2005). Accessed March 11, 2009. Available at
www.who.int/making_pregnancy_safer/documents.
39. Statistics Canada. Deaths 2003 . StatsCan Catalogue No.
84F0211XIE, Table 7: Perinatal mortality and components by
geography. 2005;12:47. Accessed March 11, 2009. Available at
www.statcan.gc.ca.
40. Weigers TA. Home or hospital birth: A prospective study of mid-
wifery care in the Netherlands. Eur J Obstet Gynecol Reprod Biol
1998;79:139–141.