The gentlebirth.org website is provided courtesy of
Ronnie Falcao, LM MS,
a homebirth midwife in Mountain View, CA
The gentlebirth.org website is provided courtesy of
Ronnie Falcao, LM MS,
a homebirth midwife in Mountain View, CA
|
An interactive resource for moms on easy steps they can take to reduce exposure to chemical toxins during pregnancy. Other excellent resources about avoiding toxins during pregnancy These are easy to read and understand and are beautifully presented. |
Physician-
and midwife-attended home births. Effects of breech, twin, and post-dates
outcome data on mortality rates.
Mehl-Madrona L, Madrona MM.
J Nurse Midwifery. 1997 Mar-Apr;42(2):91-8.
PHYSICIAN- AND MIDWIFE-ATTENDED HOME BIRTHS
Effects of Breech, Twin, and Post-Dates Outcome Data on Mortality Rates
Lewis Mehl-Madrona, MD, PhD, and Morgaine Mehl Madrona
©1997 by the American College of Nurse-Midwives.
Many studies have reported positive outcomes for home birth (1-26), including births attended by apprentice- trained midwives1 and certified nurse-midwives in the United States and births attended by direct-entry midwives in Canada and the Netherlands. However, there is an ongoing debate regarding whether it is safe for any provider to attend breech deliveries, twin deliveries, and post-dates pregnancies at home. Although standard textbooks of obstetrics do not support home birth for anyone, their authors especially object to home birth for breech, twin, and post-dates pregnancies. Williams Obstetrics states, "The provider who might naively champion any childbirth outside of a hospital setting is either not aware of the hazards of breech delivery in such a setting or is totally insensitive to the welfare of the fetus and the mother" (27).
One of the authors (LM) has frequently been asked to testify in court proceedings involving bad outcomes of home births, 84% of which have involved one of these three types of deliveries occurring at home. At a Midwifery Today conference held in New York City in 1995, several presenters argued for the acceptability and desirability of midwives' attending breech and twin deliveries at home. Review of recent publications in the midwifery literature shows a continued effort to attain authority to deliver breeches, twins, and post-dates women at home. Recent published articles include one suggesting that post-dates or post-maturity is a myth (28), that twins may be safely delivered at home (29), and that breeches may be delivered at home by all midwives (30,31). A 1995 front page article in the Burlington, Vermont area told the story of a lay midwife-attended breech delivery in which the midwife assured the mother that she could feel comfortable about having a breech birth at home (32).
It is difficult to analyze midwife-attended home birth outcomes for a number of reasons: 1) women tend to self-select home birth, and the act of self-selection or the psychological factors influencing the choice for home birth could influence the outcome; 2) many complications are sufficiently rare that large numbers of births would be needed to compare outcomes between management of the problem at home versus management of the problem in the hospital, and, given the small numbers of home births in the United States, large numbers are hard to generate; 3) planned and unplanned home births are not always easily distinguishable; and 4) even within home birth, the variations in practice protocols and attitudes toward medicine and hospitals among practitioners can be enormous, varying from a close, cooperative relationship to rejecting attitudes and antagonism.
Because of the experiences of one of the authors (LM), especially those
in the courtroom and in medical and nursing board proceedings; the ongoing
controversy over attending breech, twin, and post-dates pregnancies at
home; and the difficulty of collecting additional data, we decided to use
an existing data base, composed of home births attended by midwives from
1970 to 1985, to learn more about the safety and risks of attending births
with any of these complicating factors. Although this was a 10-year-old
data base, much of it had never been used before in published research;
furthermore, the published results from earlier versions of this data base
are still used in support of home births attended by both physicians and
midwives. This data base was thought to be able to provide a reasonable
approximation of the safety and risks of attending twin, breech, and post-dates
deliveries at home during the 1970s and 1980s. Although midwifery practices
may have changed and increased experience may better prepare the contemporary
midwife to attend these types of deliveries safely at home, examination
of practitioners' data from the past can at least provide an estimate of
risks against which current practitioners can compare their own results
and make future practice decisions.
To enrich the available sample of apprentice-trained midwife-attended births, beginning in 1975, data were solicited from a larger cohort of midwives in the far west. The cooperation of midwives was solicited at conferences and by mail. Forms, instructions for use of the forms, and definitions of terms were mailed to midwives who agreed to collect these data; however, site visits were made only to a subset of those whose location was convenient. The births occurred between 1970 and 1985, as some midwives offered retrospectively collected data. The larger portion of the births were collected prospectively.
The data collection form used in the earlier studies (1-
4)
was sent with an attached sheet defining all terms, including fetal distress,
postpartum hemorrhage, first- stage labor dystocia, second-stage labor
dystocia, etc.2 Midwives recorded their own data. They were encouraged
to call the first author with any questions about how to define terms or
how to code events. They were encouraged to provide a narrative description
of complications in the event that they were unsure how to code. This procedure
permitted some room for variability in recording between physicians and
midwives.
|
|
|
|
|
|
| 1970-1974 | Santa Cruz, California | 100% midwives for 287 births; no physicians | Chart review by researchers |
|
| 1969-1975 | Northern California | 31% midwives for 355 total births; 69% physicians for 791 total births | Chart review by researchers |
|
| Added more births for the period of 1970-1976 | Northern California and Wisconsin | 35% midwives for 816 total births; 65% physicians for 1,514 total births | Chart review by researchers |
|
| 1977-1985 | Northern California and Wisconsin | Added 2,593 births; all physicians | Chart review by researchers | Not previously published |
| 1970-1985 | Western United States | Added 3,545 births; all midwives | Completion of data form by midwives | Not previously published |
| Cumulative date: 1969-1985 | Western United States and Wisconsin | 4,361 total births with midwives; 4,107 total births with family physicians | Midwives: both self-report and chart review; physicians: only chart review | Total data available for analysis in this paper |
* Entries are in order of when each study was begun. Dates of birth
may be out of sequence because of retrospective data collection.
Midwives were asked to submit consecutive data of all their cases, including all transfers from home to hospital. Data were to be included on what happened at the hospital for all women transferred if they had begun labor with the intention of delivering at home. There were limited checks on the accuracy of data submitted, which could bias the results in favor of better midwife outcomes. (One would expect that if any reporting bias existed, it would tend to be toward underreporting of adverse outcomes.)
Of the 153 midwives who could be identified and were invited to contribute
data, 30% responded. Twelve percent of these provided the bulk of the data,
contributing more than 50 cases per midwife. The mean number of births
contributed for all midwives was 87 (range nine to 626; median 53). Twenty-four
midwives submitted data for the years 1970-1980 (1,919 births); 11 midwives
submitted data for the years 1975-1985 (1,234 births); another 11 midwives
submitted data in Wisconsin for 1977-1983 (392 births). Far-west midwives
practiced in California, Oregon, Washington, Idaho, Arizona, Nevada, and
Hawaii; because some of them practiced illegally, they were given code
names and no identifying data were received. No identifying data were used
for midwives or their patients. Incomplete forms or forms with logically
impossible data (implying data entry error) were discarded if the midwife
could not be contacted for clarification or could not provide further insight.
Data on an additional 3,545 apprentice-trained midwife-attended cases were
accumulated (see Table 1). These midwives did not
have a policy of excluding twin, breech, or postdates pregnancies from
home birth, although the data from the earlier 816 midwives (collected
by chart review) came from practices in which there was a policy not to
attend breech, twin, or post-dates births at home. Combining these two
data sets gave a total of 4,361 apprentice-trained midwife-attended births
for use in matching.
All of the physicians had a policy against delivering breeches and twins
at home, although this occurred accidentally several times in multiparous
patients. A total of 4,107 physician-attended home births were accumulated
for matching.
FPs family physicians; DEMs = direct-entry midwives; NS = not significant.
*Results expressed as cases per 1,000.
Matching was relatively easy because most women were low risk (Popras score 1), in their early twenties, insured, and nulliparous. Cases were selected as matches at random when the computer identified more than one match from the MD sample for a particular midwife birth.
The steps in the analysis were as follows: 1) calculate intrapartum and neonatal mortality rates and the incidence of neonatal resuscitation in both groups (1,000 matched patients in each group), and compare the results; 2) calculate these rates after excluding infants with lethal congenital anomalies; and 3) calculate these rates after the stepwise elimination of breech presentations, twin births, and post-dates pregnancies.
We used t test procedures of the Systat System for Statistics
to assess the significance of differences between the groups. A subsequent
analysis was performed with McNemar's statistics of discordant pairs, with
no substantial change in the results. Logistic regression as implemented
in Systat was used to calculate the odds ratios of risk for attending breech,
twin, and post-dates pregnancies.
For the two entire samples of physicians and midwives (not the matched
samples), the physicians had significantly more lethal congenital anomalies
(cause unknown), whereas the midwives had significantly more deaths of
the second twin during labor, other intrapartum deaths, deaths of post-dates
fetuses during labor, deaths of breech infants during labor, deaths due
to meconium aspiration of term newborns, and deaths due to meconium aspiration
of post-dates newborns (Table 4).
NS = not significant at P < .05.
Logistic regression was used to determine whether the year of delivery had an effect on mortality; no effect was found. When breeches, twins, and post-dates pregnancies were eliminated from the analysis, no effect of type of practitioner was observed. The odds ratio of death for the infant born at home with one or more of the three conditions under study was 3.1 (95% confidence interval, 2.1-12.3, P = .002). Thus, the high-risk population delivering at home was 3.1 times more likely to experience a mortality event than the low-risk population delivering at home.
This study is limited by nonrandom selection of both midwives and physicians.
The data arose from births that occurred between 1969 and 1985, and there
was no way to assess the accuracy and completeness of the data contributed
by the midwives who reported their own data. The data collection procedures
varied between physicians and midwives. Because of these limitations, the
findings and conclusions cannot be generalized to current apprentice-trained
midwifery practice.
MWs = midwives; NS = not significant.
* Results expressed as cases per 1,000.
Adverse outcomes were more commonly reported by the midwives than were found among the physician-attended births. This statistically significant difference was not found, however, when twins, breech births, and postdates births were removed from the samples. One would assume that bias, if it existed, would be in favor of the midwives because they would, if biased, tend to underreport their adverse outcomes. This strengthens the direction of the conclusions.
Parents decide on home births for many reasons. Some insist on home births even when they know that they have an increased risk of serious complications from factors such as multiple pregnancy, nonvertex presentation, and post-dates pregnancy. Births involving these conditions are associated with higher perinatal mortality regardless of the place of birth. Some apprentice-trained midwives are willing to attend these births at home. Thus, our results address a continuing controversy.
Similarly, midwives decide to attend higher-risk births at home for many reasons. Many, even most, of these births can be achieved successfully at home. Being "high risk" only means that the incidence of untoward outcomes is higher; it does not mean that every high-risk birth will have a poor outcome. Some midwives believe that parents who insist on home birth for a higher-risk pregnancy have the right to professional attendance. It is also true that births involving twins, breech presentation, and post-dates pregnancies are at higher risk regardless of the type of delivery. For example, breech infants have a higher perinatal mortality rate even if delivered by elective cesarean at term, as compared with elective cesarean at term for vertex presentations (27).
If the argument is made that today's midwife is better equipped to handle twins, breech, or post-term home births, or that, with better screening, some of these three types of deliveries could be safely managed outside of the hospital, then we believe that this article establishes a reference point against which these claims could be tested. Until proven otherwise, however, practitioners should strive to avoid such deliveries at home, given the findings of this study and despite the age of the data set. To explore this controversy further, the first step should probably be a randomized trial of home-like midwifery care within a hospital for breech births or twin births, which would, of course, include written informed consent and an analysis based on intent to deliver. The results of this study should not be used to argue that breeches, twins, or post-dates infants should never be delivered outside of a hospital. Rather, because currently available data do not support this practice, the burden of proof would be on a future researcher to show under what circumstances, with what type of care, and with what type of practitioner these deliveries could occur safely in alternative settings. Certainly there is a strong need to humanize all high-risk deliveries and to apply the insights and wisdom of normal birth care to the avoidance of excessive intervention for all pregnant women.
Midwifery care and home births are appropriate choices for normal, low-risk births. However, services that may be needed at higher-risk births are not available as quickly or readily at home, and few midwives have enough experience with complicated births to acquire the necessary competency for their delivery at home. Having experience with a few high-risk births that turn out well can be deceiving and may lead to false confidence. Bad outcomes of high-risk home births hurt babies, parents, individual midwives, and the midwifery and home birth movements. The philosophy of home birth is based on normal births-births that do not have a substantially increased risk of serious complications. Attending high-risk births at home undermines that philosophy.
Similar findings regarding post-dates births were seen in the National Birth Center Study (23,33). Post-term births were the only category of births in which outcomes for infants were not better than low-risk hospital comparison groups. The intrapartum death rate per 1,000 births was 2.3 for post-term births and 0.2 for term births. The neonatal death rate per 1,000 births was 1.5 for post-term births and 0.7 for term births. The combined intrapartum and neonatal death rate was 3.8 for post-term births and 0.9 for term births. The combined intrapartum and neonatal death rate among low-risk patients in hospitals ranged from 0.9 to 4.3 in studies of low-risk in-hospital births used as comparisons in the National Birth Center Study (23,33). These data were based on births during 1969-1985, the period in which the births in this study occurred.
For comparison purposes with this data set, it should be noted that the combined intrapartum and neonatal mortality in the Wormerveer study of 7,980 Dutch home births, including those transferred to the hospital during labor, was 2.3 per 1,000, with a 0.4% cesarean rate (24). The same rate for home births in Cardiff, Wales in 1979 was 4.1 per 1,000 births (25). There were no pen- natal deaths in a 5-year prospective study of home birth in Essex, England from 1978 to 1983(26). The National Birth Center Study found a combined intrapartum and neonatal mortality rate of 1.3 per 1,000 and a neonatal mortality rate of 0.8 per 1,000, with a cesarean rate of 4.4%, among 11,814 women giving birth from mid- 1985 through the end of 1987 (23,33). The physician- attended home births from this study and the apprentice- trained midwife-attended home births (excluding breeches. lethal anomalies, twins, and post-dates pregnancies) compare favorably with these international results from the same time.
Previous studies by the first author helped to increase the support for home birth. A concern remains, however, about an overconfident approach to attending high- risk births at home. Although there are limitations to the data on which this study is based, most of these limitations apply also to the earlier studies (1-4). Historic data, such as these previous studies, are consistently used to "prove" the safety of home births. This analysis adds to the historic data that are available.
The authors gratefully acknowledge the help, patience, and comments of Judith Rooks, CNM, MPH in the preparation of this manuscript.
2 Postpartum hemorrhage was defined as blood loss of at least 1,000 mL resulting in significant maternal symptoms (lightheadedness, dizziness, loss of consciousness, nausea, orthostatic blood pressure changes), preferably confirmed by at least a 20% drop in hemoglobin. Neonatal resuscitation was defined as a minimum of vigorous bagging with an ambu-bag/mask set with 100% oxygen in a baby whose 1- minute Apgar score was 4 or less. Most of these babies would be intubated, at least temporarily, in the hospital setting. Other definitions are available upon request along with the data entry form.
Lewis Mehl-Madrona is Research Assistant Professor at the University of Arizona College of Medicine Native American Research and Training Center. He has published many articles on home birth outcomes in the United States. His research on Native American studies is described in his book, Coyote Medicine. He is board certified in family practice and geriatrics and board eligible in psychiatry and emergency medicine, and has attended home births as part of his practice in California, Wisconsin, and Vermont from 1973 until 1996.
Morgaine Mehl Madrona is a midwife who has practiced in the Catskill Mountain area of Oneonta, New York and in the greater Burlington, Vermont region. She is currently writing a book about the future of midwifery and normal childbirth in the United States. She has assisted in promoting midwifery legislation in Vermont and has organized political programs that advocate for the legalization of one level of professional midwife who is able to practice in any setting, rather than the pervasive two-tiered system of certified nurse-midwives practicing primarily in the hospital and several other types of midwives all practicing in the home.
| About the Midwife Archives / Midwife Archives Disclaimer |