In-Hospital Birth Center with the Same Medical Guidelines as Standard Care: A Comparative Study of Obstetric Interventions and Outcomes

BIRTH 1 In-Hospital Birth Center with the Same Medical Guidelines as Standard Care: A Comparative Study of Obstetric Interventions and Outcomes Karin Gottvall, RN, RM, MSc, PhD, Ulla Waldenström, RN, RM, BA, PhD, Charlotta Tingstig, RN, RM, BSc, and Charlotta Grunewald, MD, PhD ABSTRACT: Background: A challenge of obstetric care is to optimize maternal and infant health outcomes and the mother s experience of childbirth with the least possible intervention in the normal process. The aim of this study was to investigate the effects of modified birth center care on obstetric procedures during delivery and on maternal and neonatal outcomes. Methods: In a cohort study 2,555 women who signed in for birth center care during pregnancy were compared with all 9,382 low-risk women who gave birth in the standard delivery ward in the same hospital from March 2004 to July 2008. Odds ratios (OR) were calculated with 95% confidence interval (CI) and adjusted for maternal background characteristics, elective cesarean section, and gestational age. Results: The modified birth center group included fewer emergency cesarean sections (primiparas: OR: 0.69, : 0.58 0.83; multiparas: OR: 0.34, : 0.23 0.51), and in multiparas the vacuum extraction rate was reduced (OR: 0.42, : 0.26 0.67). In addition, epidural analgesia was used less frequently (primiparas: OR: 0.47, : 0.41 0.53; multiparas: OR: 0.25, : 0.20 0.32). Fetal distress was less frequently diagnosed in the modified birth center group (primiparas: OR: 0.72, : 0.59 0.87; multiparas: OR: 0.45, : 0.29 0.69), but no statistically significant differences were found in neonatal hypoxia, low Apgar score less than 7 at 5 minutes, or proportion of perinatal deaths (OR: 0.40, : 0.14 1.13). Anal sphincter tears were reduced (primiparas: OR: 0.73, : 0.55 0.98; multiparas: OR: 0.41, : 0.20 0.83). Conclusion: Midwife-led comprehensive care with the same medical guidelines as in standard care reduced medical interventions without jeopardizing maternal and infant health. (BIRTH 38:2 June 2011) Key words: birth center, midwife-led care, midwifery, obstetric outcome, perinatal health A challenge of obstetric care is to optimize maternal and infant health outcomes and the mother s experience of childbirth with least possible intervention in the normal process (1). During the 1970s and 1980s concerns were raised about what was perceived to be an unnecessary Karin Gottvall is a Senior Researcher at the Karolinska Institutet, Department of Public Health Sciences, Division of Global Health (IHCAR); Ulla Waldenström is a Professor at the Karolinska Institutet, Department of Women s and Children s Health, Division of Reproductive and Perinatal Health; Charlotta Tingstig is a Midwife at South General Hospital (Södersjukhuset, Södra BB); and Charlotta Grunewald is an Associate Professor and Senior Obstetrician at the Karolinska Institutet, Department of Clinical Science and Education, Division of Obstetrics and Gynecology, Södersjukhuset, Stockholm, Sweden. increase in medical interventions. New models of maternity care focusing on continuity of care and limited use of medical technology, such as freestanding and inhospital birth centers, opened in the United States, Australia, and some European countries (2,3). Evaluations This study was supported by research grants from the Stockholm County Council, Stockholm, Sweden. Address correspondence to Karin Gottvall, RN, RM, MSc, PhD, Radhusallén 9, SE-167 54 Bromma, Sweden. Accepted October 4, 2010 Ó 2011, Copyright the Authors Journal compilation Ó 2011, Wiley Periodicals, Inc.

2 BIRTH of such models of care consistently showed lower intervention rates (2,4 8) and higher rates of satisfaction with care among mothers (2,5 9). In 1997, Waldenström et al published a randomized controlled trial of 1,860 women who gave birth in an in-hospital alternative birth center in Stockholm, Sweden. The perinatal mortality rate was four times higher, but not statistically significant, in first-born infants in the alternative birth center group compared with the standard care group (6). This worrisome finding resulted in the conduction of a register study by Gottvall et al, in which the outcomes of 3,256 birth center confinements were compared with 180,380 in standard care (10). This study confirmed an increase in perinatal mortality in first-born babies (OR: 2.2, : 1.3 3.9). Infants born postterm were overrepresented, suggesting that the birth center applied too liberal criteria when allowing women with prolonged pregnancy into the center without fetal surveillance by electronic fetal monitoring. Based on these results the alternative birth center was closed and replaced by a modified version of a birth center, with medical technology available on site. Like the alternative birth center, the modified new model was an option for women at low medical risk and an important feature was comprehensive care with antenatal, intrapartum, and postpartum care located on the same premises, with the same team of midwives and no ancillary caregivers involved in patient care. All women in active labor had an allocated personal midwife who provided care for only one woman in labor at a time. Midwives working at the modified birth center shared a vision to encourage normal birth without unnecessary interventions. The major change was in the medical guidelines during labor, with the introduction of electronic fetal monitoring on admission and intermittently during labor. To minimize the number of transferal during labor, augmentation of labor, epidural analgesia, and vacuum extraction could be handled on site. If severe maternal or fetal complications occurred during pregnancy (e.g., preeclampsia, antepartum hemorrhage, preterm birth, or intrauterine growth restriction), women gave birth in the standard care ward. In standard care, midwives assisted by ancillary caregivers generally provided care for more than one woman in active labor at the same time. Like the alternative birth center, the new model operated within the public sector and was tax funded in the same way as intrapartum, postpartum, and most antenatal care in standard care. Further details of the modified birth center care model and standard care are presented in Fig. 1. The challenge of the new model of care was to improve safety and at the same time maintain the positive results of birth center evaluations in terms of fewer medical interventions and increased patient satisfaction. The aim of the present study was to investigate the effects of modified birth center care on obstetric procedures during delivery and on maternal and neonatal outcomes. Methods Study Population Modified birth center care is defined as comprehensive care from early pregnancy to postnatal discharge; hence, the study group included all strictly selected low-risk pregnant women admitted to the modified birth center for antenatal care during pregnancy. These women were compared with a control group in standard care in the same medical low-risk category in early pregnancy. The exclusion criteria were: diabetes, hypertension, epilepsy, obesity (body mass index >29), cesarean section prior to Modified Birth Center Care () Standard Care (SC) Continuity of care with the same team of midwives providing antenatal, intrapartum, and postpartum care on the same premises Antenatal, intrapartum, and postpartum care in different settings and with different midwifery staff Intrapartum: Midwives as birth attendants and obstetrician from standard delivery ward present at midwives request Intrapartum: Midwives as birth attendants and obstetrician present on the delivery ward 24 hours/day Medical technology concealed in the furnishings in the birthing room Medical technology visible Homelike environment, with two beds, one for each parent and en suite bathroom Clinical hospital environment with one delivery bed, one chair for the partner, and no en suite bathroom Small scale (two wings in one ward with about 300 births/year, in total about 600 births/year) Large scale (about 6,000 births/year) Fig. 1. Characteristics of modified birth center care () and standard care (SC).

BIRTH 3 the present pregnancy, history of perinatal mortality, multiple pregnancy, maternal age over 40 years if nulliparous, and smoking. All women admitted to the modified birth center from its opening in March 2004 to July 2008 and who gave birth at Södersjukhuset (South General Hospital, Stockholm, Sweden), either in the modified birth center or in the standard delivery ward, were included in the study. The standard care comparison group comprised all lowrisk women with a personal identification number and a singleton pregnancy who gave birth in the standard delivery ward during the same time period with accessible antenatal and intrapartum data files. A total of 2,555 pregnant women were identified in the modified birth center group, 28 percent of whom had been transferred during pregnancy because of maternal or fetal complications, and had given birth in the standard delivery ward. Of the total of 21,518 births in the standard delivery unit during the corresponding observation period, modified birth center group excluded, 12,446 women had complete data, which included antenatal care. Of these, 9,382 women met the same criteria as in the modified birth center group, and thus constituted the final standard care group. Data Collection Data were retrieved from medical records in the hospital s database, Obstetrix, into which information about pregnancy, labor, and the postpartum period was continuously entered. Data on civil status, country of birth, level of education, and income were retrieved from Statistics Sweden by linking the personal identification number of the participants with the Swedish Register of Population and Population Changes. As with the intervention group, women in the control group could have given birth more than once during the observation period. Each pregnancy is in this context regarded as one woman. Maternal and infant diagnoses were classified according to the Swedish version of the International Classification of Diseases, 10th revision (ICD10). Maternal age was defined at the time of birth, civil status by married or unmarried, and country of birth by born in Sweden or elsewhere. Educational level was dichotomized as high (>12 yr and including college university) and low ( 12 yr). Income was dichotomized as higher as or lower than 20,000 SEK (approximately U.S.$2,880) per month). Emergency cesarean section was defined as nonelective and performed within a time frame of 6 hours from the decision. Uterine dystocia was defined according to the ICD10 system as: primary inadequate contractions (O62.0), secondary uterine inertia (O62.1), other specified uterine inertia and unspecified dystocia (O62.8-9), and prolonged labor (O63.0-1, 8-9). Anal sphincter tears included third and fourth degree lacerations (O70.2-3). Labor and delivery complicated by fetal stress (distress) was defined by the code O68. The infant diagnoses were classified according to the Swedish version of the ICD10. Respiratory problems were defined as P22, P24-28 (P23 excluded); intracranial hemorrhage and central nervous system lesion as P100-109, P523-529; cerebral symptoms including seizures as P90-91; peripheral nerve lesion including brachial plexus as P110-119, P140-149; infections including pneumonia as P23, P027, P35-39; immunization and hyperbilirubinemia as P550, P599; hypoglycemia as P703-709; fractures, including clavicle as P130-139; hypoxia asphyxia as P20-21; labor and delivery complicated by fetal stress (distress) as O68; and small-forgestational age as P050-051. Newborn babies were classified as being small-for-gestational age if a birthweight was two standard deviations (SD) below the mean weight for the gestational age according to the Swedish standard (11). This classification of diagnoses of infant morbidity was likewise used in a previous publication (12). Sample Size Power calculation was based on an estimated reduction of the emergency cesarean section rate from 8 percent in the control group to 5 percent in the modified birth center group (80% power; p < 0.05), and an estimated reduction of the epidural rate from 25 percent to 18 percent. The estimated figures in the control group were based on previous findings from a Swedish low-risk population (10), which would require 1,125 and 568 women per group, respectively. For the analysis of perinatal mortality, we were primarily interested in infants of first-time mothers, because this group was exposed to a higher risk in the previous birth center study (10). According to the Swedish Medical Birth Register, the national perinatal mortality rate was 4.8 per 1,000 live births in 2008 (13). To detect a deviation of ±0.1 percent (to 0.4% or 0.6%), an observation period of at least 10 years would be necessary. Such a long study period was not possible within the framework of this study, and perinatal mortality was therefore defined as a secondary outcome. However, we doubled the sample size to stratify by parity. Statistical Analyses Analyses followed the principle of intention-to-treat, where all women booked at the modified birth center were included in the modified birth center group,

4 BIRTH regardless of whether they were transferred to standard care during pregnancy, in labor, or postnatally. Outcomes in the two groups were compared using descriptive statistics and the chi-square analysis. All outcomes were analyzed separately for primiparas and multiparas. By means of logistic regression analyses we adjusted for maternal age, country of birth, education, income, smoking before pregnancy, gestational age, and elective cesarean section. We refrained from adjusting for body mass index because of the high level of values missing in both groups. Results are presented as odds ratios (OR) with 95% confidence interval (CI) comparing the modified birth center group with the standard care group as reference. The study was approved by the Research Ethics Committee at Karolinska Institutet, Stockholm, Sweden (2007 360-31). Results Baseline Characteristics The number of women (pregnancies) was 2,555 in the modified birth center group (49.4% primiparas) and 9,382 in the standard care group (52.0% primiparas; p < 0.02). The proportion of primiparas was higher than the overall proportion of primiparas in Sweden, which was 44 percent in 2008 (13). Women in the modified birth center group had a higher mean age than in the standard care group (primiparas 30.8 vs 29.6 years; multiparas 33.3 vs 32.7 years; p < 0.001). The national figure for maternal mean age was 28.4 years for primiparas and 31.9 years for multiparas. The proportion of women born in Sweden was 90.9 percent in the modified birth center group and 75.5 percent in the standard care group (p < 0.001) in comparison with national figures of 80 percent in 2004 and 77 percent in 2008 (13). Higher education (74.9 vs 53.9%; p < 0.001) and higher monthly income (42.3 vs 37.4%; p < 0.001) were also overrepresented in the modified birth center group; however, no statistical difference was found in marital status (61.3 vs 61.5%; p=0.846). Women in the modified birth center group were less often smokers before pregnancy (4.5 vs 5.7%; p < 0.02). In addition to the similarities in maternal health as a consequence of the study exclusion criteria, the women in the two groups did not differ with respect to history of infertility (>1 yr before the current pregnancy), number of previous miscarriages, and extrauterine pregnancies. We chose to treat preterm birth during the current pregnancy as a background variable, because we believe Table 1. Obstetric Procedures During Birth in the Modified Birth Center () Group and the Standard Care Group () in Primiparas and Multiparas: Number, Proportions, and Odds Ratio (OR) with 95% Confidence Interval (CI) Primiparas Multiparas Obstetric Procedures (n = 1,263) (n = 4,878) OR * (n = 1,292) (n = 4,504) OR * Induction of labor 210 (16.6) 944 (19.4) 0.69 0.58 0.83 116 (9.0) 619 (13.7) 0.52 0.42 0.65 Pain relief Epidural analgesia 475 (37.6) 2,623 (53.8) 0.47 0.41 0.53 99 (7.7) 1,022 (22.7) 0.25 0.20 0.32 Nitrous oxide 830 (65.7) 3,460 (70.9) 0.68 0.59 0.78 668 (51.7) 2,869 (63.7) 0.45 0.40 0.52 Opioids 18 (1.4) 93 (1.9) 0.77 0.46 1.30 2 (0.2) 25 (0.6) 0.28 0.06 1.20 Acupuncture 422 (33.4) 646 (13.2) 3.00 2.59 3.47 252 (19.5) 242 (5.4) 3.70 3.04 4.51 Papules 166 (13.1) 165 (3.4) 4.12 3.27 5.20 57 (4.4) 69 (1.5) 2.59 1.78 3.76 TNS 157 (12.4) 411 (8.4) 1.33 1.09 1.62 140 (10.8) 181 (4.0) 2.09 1.65 2.65 No pain relief 61 (4.8) 63 (1.3) 3.89 2.69 5.64 245 (19.0) 264 (5.9) 3.79 3.10 4.63 Mode of delivery CS, all 239 (18.9) 1,249 (25.6) 0.61 0.52 0.72 43 (3.3) 672 (14.9) 0.19 0.14 0.26 Elective CS 46 (3.6) 355 (7.3) 0.44 0.32 0.61 15 (1.2) 411 (9.1) 0.11 0.07 0.19 Emergency CS 191 (15.1) 887 (18.2) 0.69 0.58 0.83 28 (2.2) 261 (5.8) 0.34 0.23 0.51 Elective CS, indication 6 (0.5) 123 (2.5) 0.27 0.11 0.63 3 (0.2) 219 (4.9) 0.12 0.04 0.38 of fear of childbirth Instrumental delivery 194 (15.4) 818 (16.8) 0.86 0.72 1.02 21 (1.6) 146 (3.2) 0.42 0.26 0.67 Spontaneous vaginal 830 (65.7) 2,811 (57.6) 1.40 1.22 1.61 1,228 (95.0) 3,686 (81.8) 2.80 2.06 3.82 delivery Episiotomy (CS excluded) 52 (5.1) 322 (8.8) 0.57 0.42 0.78 12 (1.0) 75 (2.0) 0.43 0.23 0.83 * for maternal age, country of birth, education, income, smoking before pregnancy, elective cesarean section, and gestational age. CS = cesarean section; TNS = transcutaneous nerve stimulation; papules indicate intracutaneous or subcutaneous injection of sterile water or saline.

BIRTH 5 Table 2. Maternal Outcomes in the Modified Birth Center () Group and the Standard Care Group () in Primiparas and Multiparas: Numbers, Proportions, and Odds Ratio (OR) with 95% Confidence Interval (CI) Primiparas Multiparas Maternal Outcomes (n = 1,264) (n = 4,878) OR a (n = 1,292) (n = 4,606) OR a Prelabor Preeclampsia 44 (3.5) 190 (3.9) 0.81 0.57 1.13 6 (0.5) 57 (1.3) 0.42 0.18 0.98 Amnionitis 3 (0.2) 12 (0.2) 0.96 0.26 3.47 1 (0.1) 2 (0.0) 2.02 0.15 27.68 Hepatosis 5 (0.4) 29 (0.6) 0.58 0.22 1.53 8 (0.6) 54 (1.2) 0.47 0.22 1.01 Placental abruption 2 (0.2) 11 (0.2) 0.70 0.15 3.21 3 (0.2) 24 (0.5) 0.58 0.17 1.98 Maternal distress b 60 (4.7) 245 (5.0) 0.92 0.69 1.24 5 (0.4) 30 (0.7) 0.58 0.22 1.52 Intrapartum Dystocia c 401 (31.7) 1,574 (32.3) 0.90 0.78 1.03 41 (3.2) 289 (6.4) 0.48 0.34 0.67 Anal sphincter tear d 62 (6.0) 265 (7.3) 0.73 0.55 0.98 9 (0.7) 70 (1.8) 0.41 0.20 0.83 Length of second stage of labor d 60 min 567 (80.5) 1,901 (89.2) 1.0 814 (97.6) 2,220 (98.3) 1.0 >60 min 137 (19.5) 229 (10.8) 1.97 1.55 2.49 20 (2.4) 39 (1.7) 1.40 0.81 2.44 Total length of labor e 12 hr 746 (61.4) 2,486 (55.0) 1.0 1,197 (93.9) 3,601 (88.1) 1.0 >12 hr 469 (38.6) 2,037 (45.0) 0.75 0.65 0.85 78 (6.1) 485 (11.9) 0.53 0.41 0.68 Postpartum Partus cum 226 (21.2) 810 (20.1) 1.03 0.87 1.23 181 (15.1) 604 (15.5) 1.03 0.85 1.24 hemorrhagia >600 ml Urine retention 30 (2.4) 171 (3.5) 0.64 0.43 0.95 6 (0.5) 54 (1.2) 0.34 0.14 0.80 Anemia 111 (8.8) 414 (2.5) 0.99 0.79 1.24 41 (3.2) 245 (5.4) 0.63 0.45 0.90 Length of hospital stay <3 days 269 (21.5) 443 (9.2) 1.0 993 (77.5) 2,366 (53.1) 0.1 3 days 982 (78.4) 4,388 (90.8) 0.33 0.28 0.40 288 (22.5) 2,096 (47.0) 0.40 0.34 0.46 a for maternal age, country of birth, education, income, smoking before pregnancy, elective cesarean section, and gestational age; b maternal distress during labor and delivery, ICD10, O75.0; c dystocia, ICD10, O62 0-2.8-9, O63 0-1.8-9; d all cesarean sections excluded; e elective cesarean sections excluded. this outcome was more related to the selection of women into the two groups rather than to the model of care. We found that the modified birth center group included a smaller proportion of preterm births: 2.5 percent in the modified birth center group versus 5.1 percent in the standard care group (p < 0.001). In the statistical analyses we therefore adjusted for gestational age, in addition to the sociodemographic variables. We also treated elective cesarean section as a background variable, as it was obvious that the decision about this mode of delivery in some cases must have been made before the decision about model of care. We found that more women in the control group had an elective operative delivery for psychosocial reasons, such as fear of childbirth, and these women would probably not opt for modified birth center care. Obstetric Procedures The proportion of spontaneous vaginal deliveries was 80.6 percent in the modified birth center group and 69.3 percent in the standard care group (p < 0.001), and the corresponding figures in primiparas were 65.7 and 57.6 percent (p < 0.001), and in multiparas 95.0 and 81.8 percent (p < 0.001), respectively. Table 1 shows the obstetric procedures in the two groups, stratified by parity. Fewer women in the modified birth center group used epidural analgesia and nitrous oxide, but acupuncture and other nonpharmacological pain relief was used approximately three times as often compared with the standard care group. The proportion of women with no pain relief was almost four times higher in the modified birth center group. The proportion of cesarean sections, both elective and emergency, was significantly lower in the modified birth center group compared with the standard care group, especially in multiparas. Episiotomy was less frequently performed in the modified birth center group, both in primiparas and in multiparas. The only procedure that did not differ between the groups was instrumental vaginal delivery in primiparas.

6 BIRTH Maternal Outcomes Table 2 presents antenatal, intrapartum, and postnatal outcomes. Preeclampsia and dystocia were less common in the modified birth center group in multiparas but not in primiparas. No statistical differences between the groups were found in amnionitis, hepatosis, and placental abruption. Anal sphincter tears were less frequent in multiparas in the modified birth center group compared with the standard care group, even when adding instrumental delivery as a confounder (OR: 0.45, : 0.22 0.92) (not in table). The second stage of labor in vaginal births was longer in primiparas in the modified birth center group: mean 45 minutes versus 35 minutes in the standard care group (p < 0.001), but not in multiparas. The total duration of labor estimated from admission to the delivery ward to delivery, elective cesarean section excluded, was shorter in the modified birth center group compared with the standard care group: in primiparas a mean of 11.0 hours versus 12.7 hours (p < 0.001) and in multiparas the mean was 4.3 hours versus 6.3 hours (p < 0.001). In the postnatal period urinary retention and anemia differed between the groups, both in primiparas and multiparas. The women in the modified birth center group had a shorter length of stay after birth than the women in the standard care group. In addition, when excluding all cesarean births from the analysis, the women in the modified birth center group stayed a shorter time at the hospital after childbirth (adjusted OR: 0.33, : 0.28 0.41 for primiparas; adjusted OR: 0.42, : 0.36 0.49 for multiparas). Infant Outcomes Table 3 shows infant outcomes. Respiratory distress, infections, rhesus immunization, or hyperbilirubinemia were less common in infants of multiparas in the modified birth center group than in those of the standard care Table 3. Neonatal Outcomes in the Modified Birth Center () Group and in the Standard Care (SC) Group, in Primiparas and Multiparas: Numbers, Proportions, and Odds Ratios (OR) with 95% Confidence Interval (CI) Primiparas Multiparas Neonatal Outcomes OR a OR a Gestational age at birth (wk) b <37 29 (2.3) 196 (4.0) 0.54 0.39 0.75 21 (1.6) 189 (4.2) 0.37 0.24 0.59 37 41 1,204 (95.4) 4,575 (93.9) 1.0 1,231 (95.4) 4,130 (91.7) 1.0 42 29 (2.3) 101 (2.1) 0.86 0.68 1.08 39 (3.0) 183 (4.1) 0.71 0.50 1.02 Birthweight (g) b <2,500 29 (2.3) 196 (4.0) 0.56 0.38 0.83 11 (0.9) 110 (2.4) 0.35 0.19 0.64 2,500 4,500 1,204 (95.4) 4,575 (93.9) 1.00 1,213 (94.0) 4,194 (93.2) 1.00 >4,500 29 (2.3) 101 (2.1) 1.09 0.72 1.66 67 (5.2) 194 (4.3) 1.19 0.90 1.59 Morbidity Respiratory problems c 45 (3.6) 196 (4.0) 0.81 0.58 1.13 15 (1.2) 103 (2.3) 0.52 0.30 0.92 Cerebral symptoms (seizures 8 (0.6) 16 (0.3) 1.79 0.75 4.24 1 (0.1) 4 (0.1) 0.78 0.08 7.41 included) d Peripheral nerve lesion (brachial 1 (0.1) 8 (0.2) 0.57 0.07 4.62 2 (0.2) 5 (0.1) 2.49 0.40 15.38 plexus included) e Infection (pneumonia included) f 9 (0.7) 57 (1.2) 0.62 0.30 1.27 3 (0.2) 31 (0.7) 0.28 0.08 0.93 Immunization 36 (2.8) 137 (2.8) 1.03 0.70 1.51 11 (0.9) 70 (1.6) 0.51 0.26 0.98 hyperbilirubinemia g Hypoglycemia h 33 (2.6) 143 (2.9) 0.90 0.61 1.34 11 (0.9) 62 (1.4) 0.56 0.29 1.08 Fractures (includes clavicle) i 1 (0.1) 10 (0.2) 0.46 0.06 3.70 3 (0.2) 11 (0.2) 0.94 0.25 3.56 Small-for-gestational age j 18 (1.4) 88 (1.8) 0.77 0.46 1.30 12 (0.9) 37 (0.8) 1.17 0.59 2.32 Labor and delivery complicated 151 (11.9) 710 (14.6) 0.72 0.59 0.87 26 (2.0) 181 (4.0) 0.45 0.29 0.69 by fetal stress (distress) k Hypoxia or asphyxia l 8 (0.6) 45 (0.9) 0.30 1.41 2 (0.2) 18 (0.4) 0.40 0.09 1.77 Apgar score <7 at 5 min 9 (0.7) 61 (1.3) 0.56 0.28 1.14 6 (0.5) 40 (0.9) 0.61 0.3 1.5 Meconium-stained amniotic fluid 95 (7.5) 353 (7.2) 0.96 0.76 1.22 113 (8.7) 343 (7.6) 1.05 0.84 1.33 a for maternal age, country of birth, education, income, smoking before pregnancy, elective cesarean section, and gestational age; b crude odds ratio; c l based on International Classification of Diseases, ICD10: c ICD10: P22, P24-28 (P23 excluded); d ICD10: P90-91; e ICD10: P110-119, P140-149; f ICD10: P23, P027, P35-39; g ICD10: P550, P599; h ICD10: P703-709; i ICD10: P130-139; j ICD10: P050-051; k ICD10: O68; l ICD10: P20-21.

BIRTH 7 group, but no statistical differences were found in infants of primiparas. Imminent or manifest fetal asphyxia was less frequently diagnosed in the modified birth center group compared with the standard care group. No statistical difference in signs of neonatal asphyxia as estimated by Apgar score less than 7 at 5 minutes or by the diagnosis of neonatal hypoxia was found. Four perinatal deaths occurred in the modified birth center group, all stillbirths (0.16%), and 42 deaths (0.45%) in the standard care group (37 stillbirths, 5 deaths 0 6 days postpartum), a nonsignificant difference (adjusted OR: 0.40, : 0.14 1.13). Discussion A common view held by advocates of alternative birth options for low-risk women is that easy access to medical equipment can lead to unnecessary interventions, which explains why birth centers usually do not have electronic fetal monitors and devices for stimulation of labor, epidural analgesia, or vacuum extraction on site. However, we found that a modified version of birth center care, where the medical guidelines were the same as those in standard intrapartum care with medical equipment located within the unit but concealed in the furnishings, reduced the number of interventions without jeopardizing maternal or infant health in comparison with a control group in standard care. The study did not have sufficient power to assess perinatal mortality, but no negative trends were found in perinatal deaths or infant morbidity, such as neonatal distress, or in maternal morbidity, such as fewer perineal tears. On the contrary, outcomes pointed in favor of the modified birth center group. Our findings with respect to medical interventions are similar to those reported in a Cochrane review of midwife-led care versus standard care with less use of regional analgesia and instrumental deliveries (14); however, in contrast to the review, fewer cesarean sections occurred in the modified birth center group in our study. The worldwide increase in cesarean birth rates, in Sweden from 5.3 percent in 1973 to 17.2 percent in 2008 (13) with the corresponding short- and long-term risks for both mother and child, is a concern. It was therefore comforting to observe that fewer women had a cesarean (in total 11.0% in the modified birth center group vs 20.5% in the control group). The explicit focus of the modified birth center on women at low risk and on normal childbirth may have affected the behavior of both staff and women with respect to thinking twice before intervening in the process of labor. One would expect that the midwifery staff would be more likely to demonstrate positive attitudes to natural childbirth compared with the midwives who attended the births on the standard delivery ward. The pregnant women in the two study groups may also have differed in attitudes to childbirth (15), but it was not possible to adjust for such a difference in the statistical analyses. The small-scale format and strict selection of women at low medical risk in the modified birth center may have reduced the risk of suboptimal care. In a recent study including all infants delivered during the period from 2004 to 2006 in the entire Stockholm County, 313 infants with low Apgar score were compared with healthy controls, and it was found that 62 percent were subjected to some form of substandard care during labor compared with 36 percent of the controls (16). The main reasons for the nonoptimal care were related to misinterpretation of findings from electronic fetal monitoring, not acting sufficiently timely on such findings, and incautious use of oxytocin. We speculate on the influence of significantly higher volumes of nonselected parturitions on the standard delivery wards, and also on the risk of acting too quickly, for example, on suspicion of fetal distress, as a consequence of more stressful environments. The individual support by the midwife in the modified birth center, in contrast to that of standard care where midwives are responsible for more than one delivery at a time, may have been important, as supportive nurse-midwife care has been associated with a reduced incidence of cesarean section (17). Current knowledge endorses the taking of specific measures to make deliveries safer. A step-wise intervention including standardized guidelines for improved patient safety at a university clinic in the United States resulted in a significant reduction in the number of serious incidents (18). Standardized educational efforts in obstetric emergency skills training and in fetal surveillance have proved beneficial to an improved neonatal outcome (19 21). In the United Kingdom a risk scoring system, a so-called Maternity Dashboard, has been introduced to increase awareness and activity in highrisk situations (22). The opening of the modified birth center was preceded by educational efforts and emergency skills training, which, together with the previously reported worrisome findings about infant safety in the birth center (10), may have contributed to increased alertness and improvement in the midwives obstetric skills, resulting in the current positive findings. Epidural analgesia during labor has been associated not only with the woman s choice but also with cultural practice within the delivery unit (23), and the substantially lower rate in the modified birth center group could therefore be an effect of both selection bias and model of care. Women interested in natural childbirth without pharmacological pain relief may have been overrepresented in the modified birth center group at baseline, but the small-scale format and continuity of care from early pregnancy to postnatal discharge may also have contrib-

8 BIRTH uted to confidence in giving birth without epidural analgesia. The women s experiences of the two models of care will be reported in a subsequent publication (Tingstig C, Gottvall K, Waldenström U, Grunewald C, unpublished manuscript). The lower epidural rate may have shortened the length of labor (24), reduced the number of instrumental vaginal deliveries (25), and possibly also reduced the number of women with urinary retention (26). Although the total duration of labor was shorter in the modified birth center group, the second stage in primiparas was longer than in the control group, a finding contrasting with a study reporting that a long second stage (>2 hr) was associated with the use of oxytocin and epidural analgesia (27). Our findings may be explained by a greater acceptance of a longer second stage among midwives in the modified birth center. The lower rate of anal sphincter tears in the modified birth center group may be related to less use of vacuum extraction (28,29), but as the skills of the birth attendant is another factor affecting the incidence of anal sphincter tears (30), it cannot be ruled out that midwives in the modified birth center were more experienced in this regard. The incidence of anal sphincter tears in this study corresponded to the Swedish national figures when all cesarean sections were excluded from the analysis (13). As stated previously, it was not possible to evaluate the effect of modified birth center care by a randomized controlled trial, and hence we cannot exclude the possibility that the women in the modified birth center group might have been of lower risk and focused on a natural childbirth at baseline compared with the women in the standard care group. However, we controlled for as many medical and sociodemographic factors as possible, and this process was even more rigorous than that in our previous study, where we reported an increased risk of perinatal deaths in first-born babies in the birth center (10), as we were able to link our data to the Swedish Register of Population and Population Change. Conclusion A modified version of comprehensive birth center care for women at low medical risk, with the same medical guidelines as those in standard care and with medical technology available on site, may reduce interventions during labor and birth without jeopardizing maternal or infant health; it may possibly improve them. Acknowledgments The authors would like to acknowledge Christine Sjöö, and all women and midwives at Södra BB and Södersjukhuset who contributed to this study. References 1. World Health Organization, Maternal and Newborn Health Care Safe Motherhood Unit. Care in Normal Birth: A Practical Guide, WHO Reference Number: WHO FRH MSM 96.24, Geneva, Switzerland: Author. 1996:4 5. 2. Flint C, Poulengeris P, Grant A. The Know Your Midwife scheme A randomised trial of continuity of care by a team of midwives. Midwifery 1989;5(1):11 16. 3. Rooks JP, Weatherby NL, Ernst EK. The National Birth Center Study. Part III Intrapartum and immediate postpartum and neonatal complications and transfers, postpartum and neonatal care, outcomes, and client satisfaction. J Nurse Midwifery 1992;37(6): 361 397. 4. Harvey S, Jarrell J, Brant R, et al. A randomized, controlled trial of nurse-midwifery care. Birth 1996;23(3):128 135. 5. Turnbull D, Holmes A, Shields N, et al. Randomised, controlled trial of efficacy of midwife-managed care. Lancet 1996;348 (9022):213 218. 6. Waldenstrom U, Nilsson CA, Winbladh B. The Stockholm birth centre trial: Maternal and infant outcome. Br J Obstet Gynaecol 1997;104(4):410 418. 7. Byrne JP, Crowther CA, Moss JR. A randomised controlled trial comparing birthing centre care with delivery suite care in Adelaide, Australia. Aust N Z J Obstet Gynaecol 2000;40(3):268 274. 8. Hodnett ED, Downe S, Edwards N, Walsh D. Home-like versus conventional institutional settings for birth. Cochrane Database Syst Rev 2005;4:CD000012. 9. Waldenstrom U, McLachlan H, Forster D, et al. Team midwife care: Maternal and infant outcomes. Aust N Z J Obstet Gynaecol 2001;41(3):257 264. 10. Gottvall K, Grunewald C, Waldenstrom U. Safety of birth centre care: Perinatal mortality over a 10-year period. BJOG 2004;111 (1):71 78. 11. Marsal K, Persson PH, Larsen T, et al. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr 1996;85(7):843 848. 12. Gottvall K, Winbladh B, Cnattingius S, Waldenstrom U. Birth centre care over a 10-year period: Infant morbidity during the first month after birth. Acta Paediatr 2005;94(9):1253 1260. 13. National Board of Health and Welfare. Pregnancy, Delivery and Newborn Infants The Swedish Medical Birth Register 1973 2008, Stockholm, Sweden: Author, ISBN 978-91-86301-73-6. In Swedish. Available at: http://www.socialstyrelsen.se/publikationer 2009/2009-12-11. Accessed December 13, 2010. 14. Hatem M, Sandall J, Devane D, et al. Midwife-led versus other models of care for childbearing women. Cochrane Database Syst Rev 2008;4:CD004667. 15. Waldenstrom U, Nilsson CA. Characteristics of women choosing birth center care. Acta Obstet Gynecol Scand 1993;72(3):181 188. 16. Berglund S, Grunewald C, Pettersson H, Cnattingius S. Risk factors for asphyxia associated with substandard care during labor. Acta Obstet Gynecol Scand 2010;89(1):39 48. 17. Butler J, Abrams B, Parker J, et al. Supportive nurse-midwife care is associated with a reduced incidence of cesarean section. Am J Obstet Gynecol 1993;168(5):1407 1413. 18. Pettker CM, Thung SF, Norwitz ER, et al. Impact of a comprehensive patient safety strategy on obstetric adverse events. Am J Obstet Gynecol 2009;200(5):492, e1 e8. 19. Blix E, Oian Pl. Interobserver agreements in assessing 549 labor admission tests after a standardized training program. Acta Obstet Gynecol Scand 2005;84(11):1087 1092. 20. Draycott T, Sibanda T, Owen L, et al. Does training in obstetric emergencies improve neonatal outcome? BJOG 2006;113(2): 177 182.

BIRTH 9 21. Beckley S, Stenhouse E, Greene K. The development and evaluation of a computer-assisted teaching programme for intrapartum fetal monitoring. BJOG 2000;107(9):1138 1144. 22. Chandraharan E, Sutton J, Beattie J, Arulkumaran S. The role of the Maternity Dashboard in improving patient care: 2 year experience at a tertiary centre. Int J Gynecol Obstet 2009;107(Suppl 2):S139 S164. 23. Schytt E, Waldenstrom U. Epidural analgesia for labor pain: Whose choice? Acta Obstet Gynecol Scand 2010;89(2):238 242. 24. Kjaergaard H, Olsen J, Ottesen B, et al. Obstetric risk indicators for labour dystocia in nulliparous women: A multi-centre cohort study. BMC Pregnancy Childbirth 2008;8:45. 25. Sharma V, Colleran G, Dineen B, et al. Factors influencing delivery mode for nulliparous women with a singleton pregnancy and cephalic presentation during a 17-year period. Eur J Obstet Gynecol Reprod Biol 2009;147(2):173 177. 26. Musselwhite KL, Faris P, Moore K, et al. Use of epidural anesthesia and the risk of acute postpartum urinary retention. Am J Obstet Gynecol 2007;196(5):472, e1 e5. 27. O Connell MP, Hussain J, Maclennan FA, Lindow SW. Factors associated with a prolonged second state of labour A casecontrolled study of 364 nulliparous labours. J Obstet Gynaecol 2003;23(3):255 257. 28. Ekéus C, Nilsson E, Gottvall K. Increasing incidence of anal sphincter tears among primiparas in Sweden: A population-based register study. Acta Obstet Gynecol Scand 2008;87(5):564 573. 29. Andrews V, Sultan AH, Thakar R, Jones PW. Risk factors for obstetric anal sphincter injury: A prospective study. Birth 2006;33(2):117 122. 30. Laine K, Gissler M, Pirhonen J. Changing incidence of anal sphincter tears in four Nordic countries through the last decades. Eur J Obstet Gynecol Reprod Biol 2009;146(1):71 75.