Homebirth Death Statistics: Studies

Showing posts with label Studies. Show all posts

Friday, July 13, 2012

Uterine rupture: how much time do you have to save the baby?

A new study to be published in the April issue of Obstetrics and Gynecology demonstrates that in the wake of a uterine rupture, providers have no more than 18 minutes to deliver the baby before the baby experiences significant hypoxia, and only 30 minutes until the baby suffers major neurological impairment.

The paper, entitled Uterine Rupture With Attempted Vaginal Birth After Cesarean Delivery: Decision-to-Delivery Time and Neonatal Outcome by Calla Holmgren et al. was undertaken to provide the most accurate information possible about the conditions needed to make attempted vaginal birth after cesarean (VBAC) as safe as possible.

The authors note:

The increasing rate of primary and repeat cesarean delivery in the United States is of concern to physicians and patients, and vaginal birth after cesarean delivery is considered to be one way to lower the overall cesarean delivery rate. Trial of labor after cesarean delivery (TOLAC), which peaked at 31% in 1998, has decreased progressively since (8.5% by 2006), primarily because of issues surrounding uterine rupture. Although rupture of the uterus during TOLAC is rare, it can be devastating for both the mother and neonate when it occurs, and it is a major liability risk for physicians. The American College of Obstetricians and Gynecologists’ guidelines advise physicians that TOLAC is most safely undertaken in hospitals where staff can immediately carry out an emergency cesarean delivery. This view is based on the premise that the ability to rapidly intervene will minimize adverse neonatal outcomes. However, immediate availability is loosely defined, and it is not clear how rapidly the fetus must be delivered after uterine rupture to prevent neonatal death or neurologic sequelae. The purpose of this study was to examine whether an association exists between neonatal outcomes and the time from diagnosis of uterine rupture to delivery of the neonate.

In other words, should hospitals refuse a trial of labor to women with a previous C-section if they cannot guarantee that both an obstetrician and and anesthesiologist are on site to start a C-section with less than a half hour?

What did they look for?

The primary adverse outcome was defined as an abnormal umbilical pH level less than 7.0 or a 5-minute Apgar score of 7 or less. Secondary adverse outcomes included fetal or early neonatal death and neonatal neurologic injury attributed to uterine rupture. Neonatal neurologic injury was defined as otherwise unexplained seizures, clinically obvious cerebral palsy, or developmental delay attributable to hypoxia resulting from the uterine rupture.

Who was included in the study?

Within the 10 hospitals studied, 40,772 women were identified with a prior cesarean delivery between January 1, 2000, through December 31, 2009. Of these, 11,195 women (27.5%) attempted TOLAC, with successful vaginal delivery for 9,419 (84.1%) patients... In total, there were 36 cases of documented uterine rupture (0.32%) during TOLAC.

What did they find?

Of the 36 patients, 13 (36.1%) met our criteria for a primary adverse outcome of umbilical artery pH level less than 7.0 or 5-minute Apgar score less than 7. These patients were compared with the 23 patients without this outcome. Median (range) time to delivery for the primary adverse outcome group (n=13) was 19 (9–40) minutes compared with 14 (0 –38) for the nonadverse outcome group. Results after stratifying the sample by hospital type yield a similar result, with those experiencing the primary outcome having, on average, a 5.5-minute (95% confidence interval [CI] 0.0 –15.0) longer time to delivery than those who did not experience the outcome...

Seventeen neonates (47.2%) were delivered less than 18 minutes after identification of uterine rupture. Of these, two neonates had an abnormal 5-minute Apgar score, but both of these neonates had an umbilical pH level greater than 7.0 and none had neurologic injury. Eighteen patients were delivered more than 18 minutes after suspicion of uterine rupture (50.0%). Of these, 11 met criteria for a primary adverse outcome and three met criteria for an adverse secondary outcome. One patient did not have suspected uterine rupture during labor.

Seventeen neonates (47.2%) were delivered less than 18 minutes after identification of uterine rupture. Of these, two neonates had an abnormal 5-minute Apgar score, but both of these neonates had an umbilical pH level greater than 7.0 and none had neurologic injury. Eighteen patients were delivered more than 18 minutes after suspicion of uterine rupture (50.0%). Of these, 11 met criteria for a primary adverse outcome and three met criteria for an adverse secondary outcome. One patient did not have suspected uterine rupture during labor.

A chart of the data makes the relationship between time to delivery and risk of adverse outcomes quite clear.

More than 83% of babies delivered more than 30 minutes after uterine rupture experienced major neurological complications.

As the authors explain:

Our study focused on serious neurologic morbidity in cases of confirmed uterine rupture during TOLAC. In 36 cases of acute uterine rupture, there were no fetal or neonatal deaths. Delivery within less than 18 minutes was associated with adverse primary outcome in two cases, but this was based on a 5-minute Apgar score less than 7; both neonates had normal umbilical pH levels. Three neonates in our study sustained long-term neurologic damage. These neonates were delivered 31, 40, and 42 minutes after uterine rupture was suspected on clinical grounds. When uterine rupture was identified in a timely fashion and delivery occurred in less than 30 minutes, there was no long-term neonatal morbidity in our study. However, delivery within 30 minutes did not prevent every case of low umbilical cord pH level or low 5-minute Apgar score, so these results should be interpreted with caution...

The authors conclude:

Uterine rupture during TOLAC is a rare but serious complication that requires prompt recognition and delivery of the fetus. The response time necessary to prevent neonatal injury has been uncertain and controversial. In our study, all neonates delivered within 18 minutes from decision to delivery had normal umbilical cord pH levels. Delivery within 30 minutes was associated with good long-term outcomes...

This is an important study that has the potential for wide impact. The study strongly confirms the ACOG recommendation that babies should be delivered as soon as possible in the wake of a uterine rupture. Intervals longer than 18 minutes resulted in demonstrable hypoxia and intervals longer than 30 minutes resulted in major neurological impairment. Studies like these make it extremely unlikely that hospitals and malpractice insurers will liberalize access to VBAC.

This piece first appeared on The Skeptical OB in March 2012.

International comparisons of neonatal and infant mortality are invalid

The annual meeting of the Society for Maternal-Fetal Medicine is currently being held in Dallas. Two different papers will be presented today, both of which highlight the increased risks of homebirth.

The first is Neonatal outcomes associated with intended place of birth: birth centers and home birth compared to hospitals (abstract 65) by Cheng, Snowden and Caughey. According to the authors:

This was a retrospective cohort study of singleton live births that occurred in 2008 in the U.S. that had specified birthing facility information. Deliveries were categorized by location of occurrence: hospitals, birthing centers, or intended home births...

What did they find?

While the risk of cesarean delivery was much lower for women who delivered/or intend to deliver outside of hospitals (0.02-4% vs. 24%, <0.001), the odds of 5-minute Apgar score <7 and neonatal seizure was significantly higher for intended home births compared to hospital birth.

In other words, while homebirth has a dramatically lower C-section rate, the price is double to triple the rate of birth asphyxia. This study actually substantially underestimates the risk of these serious complications at homebirth because it compares homebirth to all risk hospital birth instead of low risk hospital birth.

The second study is entitled Does planned home birth affect neonatal mortality? (abstract 563) by Joel Larma.

This is a population based retrospective cohort study of all births between 37 and 42 weeks gestation using the National Health Center for Vital Statistics 2005 Linked Birth/Infant Death Cohort Data Set. The primary outcome was neonatal mortality and the primary predictor was planned home birth. The referent group for the regression model was births that occurred in a hospital...

The authors did not separate homebirths attended by CNMs from homebirths attended by non-nurse midwives. The analysis showed that homebirth more than doubled the risk of neonatal death. This actually underestimates the risk because it compares homebirth to term hospital birth in women without pre-existing medical conditions, but apparently includes hospital births with pregnancy complications.

The adjusted OR for neonatal mortality among individuals having a planned home birth was 2.32 (95% CI 1.33, 4.06) after controlling for the aforementioned covariates [age, race, marital status, education, prenatal care, tobacco use, composite medical comorbidities].

CONCLUSION: The odds of neonatal mortality are significantly increased among those individuals having a planned home birth compared to those individuals giving birth in the hospital.

The amount of data on planned homebirth is slowly growing and the findings are remarkably robust. These studies, like all the existing scientific evidence, state and national statistics show that homebirth doubles or triples the rate of neonatal death and other adverse outcomes, and that the analyses almost certainly underestimate the real increased risks at homebirth.

addendum: These studies underestimate the risk even further because, as far as I can determine, bad outcomes that occurred after homebirth transfer are included in the hospital group and removed from the homebirth group. The real number of bad outcomes and deaths is higher than what the authors were able to determine.

This piece first appeared on The Skeptical OB in March 2012.

Two new studies show increased risk of death, serious complications at homebirth

This was a retrospective cohort study of singleton live births that occurred in 2008 in the U.S. that had specified birthing facility information. Deliveries were categorized by location of occurrence: hospitals, birthing centers, or intended home births...

What did they find?

While the risk of cesarean delivery was much lower for women who delivered/or intend to deliver outside of hospitals (0.02-4% vs. 24%, <0.001), the odds of 5-minute Apgar score <7 and neonatal seizure was significantly higher for intended home births compared to hospital birth.

This is a population based retrospective cohort study of all births between 37 and 42 weeks gestation using the National Health Center for Vital Statistics 2005 Linked Birth/Infant Death Cohort Data Set. The primary outcome was neonatal mortality and the primary predictor was planned home birth. The referent group for the regression model was births that occurred in a hospital...

The adjusted OR for neonatal mortality among individuals having a planned home birth was 2.32 (95% CI 1.33, 4.06) after controlling for the aforementioned covariates [age, race, marital status, education, prenatal care, tobacco use, composite medical comorbidities].

CONCLUSION: The odds of neonatal mortality are significantly increased among those individuals having a planned home birth compared to those individuals giving birth in the hospital.

Near-miss maternal mortality

Maternal mortality has dropped 99% in the past 100 years. A maternal death is now, fortunately, a rare event. Attention, therefore, is shifting to maternal morbidity, in particular, life-threatening morbidity. The results of a new study are instructive.

Near-Miss Maternal Mortality: Cardiac Dysfunction as the Principal Cause of Obstetric Intensive Care Unit Admissions by Small et al. will be published in the February 2012 issue of Obstetrics and Gynecology. The title gives away the principle finding; heart disease is the most common cause of maternal ICU admission.

The study took place at Duke University from January 2005 to April 2011. There were 19,575 births and 5 maternal deaths for a maternal mortality rate of 25/100,000. That is approximately double the US maternal mortality rate, but that is only to be expected in a tertiary center that receives the most complicated cases from the surrounding area. There causes of the five maternal deaths were: two from metastatic cancer, two secondary to cystic fibrosis, and one the result of sepsis.

The authors then looked at maternal admissions to the intensive care unit:

Ninety-four obstetric patients—five per 1,000 deliveries—were admitted to ICUs. Eight declined participation in the study. Eighty-six patients were included in this analysis.

... African American women comprised the largest population admitted to the ICU (45%). Significant differences were found by race and ethnicity in the following variables: parity, BMI, and marital and insurance status. African American (mean 35) and Hispanic women (mean 36) had significantly higher BMIs than white women (mean 28). African American and Hispanic women were also more likely to have Medicaid or no insurance and were more likely to be unmarried and multiparous.

The following table shows the reasons for ICU admission.

The authors write:

... The leading admission diagnosis for pregnant and postpartum women was maternal cardiac disease (36%). Maternal hemorrhage (both obstetric and nonobstetric) was the second leading reason for admission (29%). Hypertensive disease accounted for 9% of ICU admissions...

The majority of cardiac conditions prompting ICU admission resulted from cardiomyopathy. Acute complications associated with peripartum cardiomyopathy
comprised the majority of this group. Congenital heart disease is the underlying etiology for many of these valvular lesions and cardiomyopathies. Congenitally
acquired conditions were the second leading cause of maternal cardiac ICU admissions...

The findings of this study are notable for the following:

Race is a major risk factor for near-miss maternal mortality.
Obesity (BMI greater than 30) is a major risk factor for near-miss mortality.
The leading cause of near-miss mortality is cardiac disease.
Infection and bleeding account for only one third of the near-miss events.

This paper demonstrates that near miss-maternal mortality, like maternal mortality, is the result of complex medical and non-medical factors. Both race and maternal BMI appear to play important roles. Pre-existing medical conditions account for a substantial proportion of near-miss maternal mortality. The traditional causes of maternal mortality and near-miss mortality have been eclipsed by cardiac complications of pregnancy.

In other words, contrary to the claims of natural childbirth and homebirth advocates, maternal mortality and near-miss maternal mortality are not related to obstetric interventions. The most common risk factor is pre-existing maternal health issues. Women with serious medical problems (including obesity) prior to pregnancy are the ones most likely to develop life threatening medical problems during pregnancy and childbirth.

This piece first appeared on The Skeptical OB in January 2012.

The UK Birthplace Study: homebirth increases the risk of death

The largest, most comprehensive study ever done of homebirth has released its results and there's nothing left to argue about: homebirth increases the risk of perinatal death.

The Birthplace Study, a large multi-year study, was designed to address the safety of place of birth by controlling for the many factors that had not been handled properly in other studies. The study looked at intended place of birth to rule out improperly assigning transferred patients to the hospital group, and included only the lowest possible risk women. The study was conducted by The National Perinatal Epidemiology Unit in the United Kingdom.

The authors found that homebirth increases the risk of death, brain damage and serious neonatal injury.

The authors chose to evaluate the results by creating an index of primary events comprising intrapartum stillbirths, early neonatal deaths, neonatal encephalopathy [brain damage] meconium aspiration syndrome, brachial plexus injury, and fractured humerus or clavicle. Using this measurement:

... [T]here was a significant excess of the primary outcome in births planned at home compared with those planned in obstetric units in the restricted group of women without complicating conditions at the start of care in labour. In the subgroup analysis stratified by parity, there was an increased incidence of the primary outcome for nulliparous women in the planned home birth group (weighted incidence 9.3 per 1000 births, 95% confidence interval 6.5 to 13.1) compared with the obstetric unit group (weighted incidence 5.3, 3.9 to 7.3).

In other words, the risk of death and serious injury was approximately double in the homebirth group and that increase was seen mainly among first time mothers.

The authors did not include the number and distribution of specific primary events within the paper itself, but did publish a 78 page supplementary file including this information. The following tables are adapted from that file. (OU stands for Obstetric unit [hospital], AMU stands for along side maternity unit [in hospital birth center], and FMU for free-standing maternity unit [independent birth center].)

Stillbirths

Early neonatal deaths (to 7 days)

Encephalopathy [brain damage]

The authors put the best possible face on the outcome:

... Adverse perinatal outcomes are uncommon in all settings, while interventions during labour and birth are much less common for births planned in non-obstetric unit settings. For nulliparous women, there is some evidence that planning birth at home is associated with a higher risk of an adverse perinatal outcome...

What can we conclude?

Homebirth increases the risk of perinatal death and brain damage in the lowest risk women receiving care from highly trained midwives (often two) and liberal access to transfer.

Homebirth increases the risk of perinatal death and brain damage even when, at the start of labor, breech, twins, VBAC. positive GBS status, gestational diabetes and obesity were excluded. All routinely occur at homebirths in the US, the UK and Australia.

And how about the purported "risks" of interventions that homebirth advocates are always taking about?

Homebirth increases the risk of perinatal death and brain damage even though the incidence of epidural use was 5 times higher in the hospital group.

Homebirth increases the risk of perinatal death and brain damage even though the incidence of pitocin augmentation was 5 times higher in the hospital group.

Homebirth increases the risk of perinatal death and brain damage even though the incidence of operative vaginal delivery was 3-4 times higher in the hospital group.

Homebirth increases the risk of perinatal death and brain damage even though the C-section rate was 4 times higher in the hospital group.

In other words, any way you choose to look at it, no matter how carefully you slice and dice the data, there is simply no getting around the fact that homebirth increases the risk of perinatal death and brain damage.

This piece first appeared on The Skeptical OB in November 2011.

New Zealand study tries to buy increased homebirth death rate

No wonder Melissa Cheyney refuses to release the death rate of Oregon homebirths.

In August, I wrote:

The minutes of the August 5, 2010 Board meeting reports that the state of Oregon asked for the ability to retrieve information on Oregon midwives from the database:

"Cheyney stated that the MANA board's official policy is to give state-level accounts to professional organizations as a tool to evaluate areas where more training might be needed for the purpose of self regulation, and to not provide the data to regulatory entities."

In other words, the database is only to be used by MANA itself, and not shared with anyone who could potentially identify unqualified midwives and discipline them.

It doesn't take a rocket scientist to speculate that there have been an extraordinary number of deaths. Now comes information from a new source that confirms that suspicion.

A new website, Oregon Homebirth Midwife Info, has compiled a Midwife Directory that makes for stomach-churning reading. The directory lists midwives by name and includes reports of deaths as well as other major morbidity and actions taken against the midwives.

It is an incomplete list; not all Oregon midwives are included and there may have been additional deaths that are not recorded. Nonetheless, the statistics are no less than horrifying.

In the past decade, no less than 19 babies have died at the hands of Oregon homebirth midwives.

To put that in perspective, consider that there are approximately 1000 homebirths per year in Oregon and that the neonatal death rate for low risk women in a hospital setting is 4/10,000 (0.4/1000). That means that you would expect approximately 4 homebirth deaths per decade. Instead there were at least 19 deaths, for a rate more than 4X higher (375%) than expected.

No less than 16 midwives have presided over at least one death. Interestingly, only 2 were unlicensed midwives. The rest were licensed at the time of the death(s) and almost all had complaints filed against them with the Board of Direct Entry Midwifery. In other words, this information is available to Melissa Cheyney in her role as a member of the Board.

Homebirth kills babies. No one knows that better than Melissa Cheyney, who has, until now, successfully hidden the number of homebirth deaths at the hands of Oregon homebirth midwives, and who continues, in her role as Director of Research at MANA (Midwives Alliance of North America), to hide the number of babies who died at the hands of homebirth midwives across the country.

This piece first appeared on The Skeptical OB in November 2011.

Reducing early elective delivery leads to more deaths

You could have seen this coming.

In a flourish of righteous zeal, the March of Dimes went on record strongly opposing early elective delivery before 39 weeks gestation. They railed against the increase in NICU admissions; they railed against the increase in C-sections; and they railed against the increase in costs. What they inexplicably failed to take into account was the inevitable increase in stillbirths.

When the Christiana Care Health System in Delaware implemented the March of Dimes recommendations, NICU admissions decreased, C-section rates decreased and cost decreased. And more babies died.

Neonatal Outcomes After Implementation of Guidelines Limiting Elective Delivery Before 39 Weeks of Gestation by Ehrenthal et al published in the forthcoming issue of the journal Obstetrics and Gynecology looked at neonatal outcomes before and after limiting elective delivery prior to 39 weeks of gestation.

All singleton deliveries 37 or more completed gestational weeks during the periods of interest were included. Any fetal death was considered a stillbirth; all others were considered live births and were analyzed separately. Each stillbirth was verified and cause of death determined by review of the hospital medical record by the study investigators...

We assessed change in obstetric practice by determining the percentage of neonates delivered during the early term if the delivery was at 37 or 38 weeks compared with full term if the delivery was 39 or more completed weeks...

We had three primary neonatal outcomes for this study: admission to the NICU for at least 24 hours, fetal macrosomia, and stillbirth...

What did they find?

The new policy achieved the objective of lowering births prior to 39 completed weeks gestation:

... the overall percentage of deliveries during the early term fell from 33.1% to 26.4% (P<.001) after the guidelines were introduced when compared with before. This changed for the cohort overall and for both cesarean and vaginal deliveries. The magnitude of the change was greater for those women with an induced labor and repeat cesarean delivery; the change was greatest for those undergoing an electively induction of labor...

NICU admissions dropped:

The overall rate of admission to the NICU was significantly different between the two periods; before the intervention, there were 1,116 admissions (9.29% of term live births), whereas after, there were 1,027 (8.55% of term live births) and this difference was significant (P=.044). Multivariable logistic regression revealed a reduced odds of a NICU admission (adjusted OR 0.92, 95% CI 0.84–1.01) after the intervention...

But the stillbirth rate more than tripled:

... The overall rate of stillbirth of nonanomalous fetuses differed between the periods with an overall increased risk of stillbirth after the intervention (relative risk 2.14, 95% CI 0.87–5.26, P=.06); this overall increase was not statistically significant. However, stratification by gestational age group of the stillbirth revealed the increased risk in the after group was limited to stillbirths before 39 weeks, which increased from 2.5 to 9.1 per 10,000 term pregnancies (relative risk 3.67, 95% CI 1.02–13.15, P=.032), whereas there was no change in risk of stillbirth at 39 weeks or more (relative risk 0.91, 95% CI 0.23–3.64, P=.896).

Because this increase in stillbirths is so large, the authors reviewed each stillbirth to be sure that they were not the result of risk factors that would have triggered a medically indicated induction.

The authors carefully reviewed the medical records of each stillbirth to identify cause of death and the presence of a maternal risk factor ... No definitive cause-of-death pattern emerged.

The reduction in early elective delivery achieved the aims for which the March of Dimes advocated. The reduction in early elective delivery reduced NICU admissions, reduced both the induction rate and the C-section rate, and (although the authors did not measure this) presumably reduced costs. However, these benefits were achieved at a very steep price. The stillbirth rate increased from 2.5 to 9.1 per 10,000 term pregnancies. Instead of 3 stillbirths between 37-39 weeks among 12,000 patients, there were 11 stillbirths between 37-39 weeks among a similar number of patients after reduction in early elective deliveries.

This finding is not unexpected. CDC data shows that the stillbirth rate rises from approximately 3 per 10,000 at 37 weeks to 4.5 per 10,000 at 39 weeks. An increase of 6 stillbirths in a population of 12,000 women is almost exactly what you would expect from reducing deliveries between 37-39 weeks.

This brings us to the heart of the matter. We have traditionally approached the inherent dangers of childbirth by attempting to reduce perinatal mortality. Our efforts have been so successful, that we have turned our attention to reducing perinatal morbidity under the assumption that any reduction in morbidity would be added on to the existing reductions in mortality.

That assumption in clearly not justified. That's because low rates of perinatal mortality have been achieved, in part, by exchanging mortality for morbidity. There are fewer deaths when you deliver babies before 37-39 weeks (whether for indicated or elective reasons); those babies who otherwise would not have lived experience relatively mild, self limited problems related to borderline prematurity. Attempts to reduce these morbidities by preventing borderline premature delivery may simply result in the deaths of these babies, not an overall improvement in outcomes. That's certainly what the existing data on stillbirths and gestational age would predict and that's precisely what happened in this study.

This piece first appeared on The Skeptical OB in October 2011.

New Dutch study raises troubling questions about the safety of homebirth

A new Dutch study of homebirth appears in the forthcoming issue of the journal Obstetrics and Gynecology. Planned Home Compared With Planned Hospital Births in The Netherlands by van der Kooy et al. is large, comprehensive and raises troubling questions about the safety of homebirth.

The large amount of data is analyzed in a bewildering number of ways, but the bottom line is that homebirth is safe when nothing goes wrong; in the presence of life threatening problems, homebirth increases the risk of death. Moreover, while homebirth with a Dutch midwife in the absence of complications is nearly as safe as hospital birth with a Dutch midwife, the perinatal mortality rate in both groups is 33% higher than comparable risk women delivered in hospitals by obstetricians just across the border in Flanders.

This study is one of many undertaken in the Netherlands to investigate the high perinatal mortality rate.

The debate on the safety of home births continues in the literature ... In The Netherlands, approximately 50% of women give birth under the supervision of a community midwife. The community midwives are independent health care professionals in The Netherlands operating either solely or in group practices.

The proportion of home birth deliveries in The Netherlands has steadily decreased over the last decade but is currently stable at 25% of all births. Several Anglo-Saxon countries are considering the reintroduction of home births based on recent claims of sufficient safety. The reverse trend is observed in The Netherlands, where the debate has intensified since the national perinatal mortality rate showed it to be one of the highest in Europe.

The authors started with a very large and comprehensive database, and analyzed it in a variety of ways. They started with 679,952 births: all the low risk births attended by midwives from 2000-2007. They looked at the difference in perinatal death rates (defined restrictively as intrapartum deaths and neonatal deaths up to 7 days of age) between home and hospital birth, first by analyzing what actually happened, then by constructing hypothetical groups of patients, all of whom were ideal candidates for homebirth.

As primary analysis, we present the results of the natural prospective approach resembling an intention- to-treat analysis. For comparison, we added a perfect guideline approach resembling a per-protocol analysis. The natural prospective approach establishes, within observational constraints, the intrapartum and early neonatal death of planned home compared with planned hospital births.

They further analyzed the data by removing deaths due to the "Big 4": congenital anomalies, premature births, intrauterine growth retardation and low Apgar scores. The decision to analyze the data with these deaths removed is baffling. It's baffling because it removes patients who received potentially substandard midwifery care during pregnancy (failure to diagnose anomalies and intrauterine growth retardation as well as failure to appropriately refer patients delivering before term) and it is baffling because it removes babies in need of expert resuscitation. This group is of prime concern when investigating the safety of homebirth because it is this group that faces the greatest risk when born outside the hospital.

Of note, both groups (real-world and hypothetically perfect) differed substantially by maternal characteristics.

Compared with women who planned birth in the hospital or with an unknown location, the women with a planned home birth were more likely to be multiparous, 25 years of age or older, of Dutch origin, and to live in a privileged neighborhood (all of which are favorable conditions). In home birth women, neonatal case mix compared also favorably. Premature delivery was less common as was the prevalence of a Big 4 condition (natural prospective approach home birth 8.7% compared with hospital 10.8% compared with unknown 10.5%; perfect guideline approach home birth 6.5% compared with hospital 8.2% compared with unknown 7.5%, P <001 in both cases).

In other words, the homebirth group was much lower risk than the hospital birth group.

What did the investigators find?

In the natural prospective approach population, crude mortality risk was significantly lower for women who planned to give birth at home (relative risk 0.80, 95% confidence interval [CI] 0.71– 0.91) ... compared with those who intended to give birth in hospital (P <.05). All maternal and neonatal risk factors, except living in a deprived neighborhood, showed significant effect sizes in agreement with the expected direction. Mortality was significantly increased in neonates with a Big 4 outcome, especially in those with multiple Big 4 conditions (relative risk 276.6, 95% CI 240.3–318.3).

When looking at what actually happened, the death rate at home was lower than in the hospital, but that reflects both the difference in risk factors between the two groups and the difference in "Big 4" bad outcomes between the two groups. After adjusting for these risk factors and differences in Big 4 outcomes, home and hospital had similar perinatal mortality rates:

The nested multivariable logistic regression analysis showed that in the presence of adjusting maternal factors only (model 2), the intended place of birth had no significant effect on outcome. The maternal factors showed risks similar to the univariable (crude) analysis. The addition of Big 4 case mix adjustment (model
3) showed the intended place of birth to be a significant covariable, yet the contrast of planned home birth (odds ratio 1.05, 95% CI 0.91–1.21) compared with a hospital birth (reference <1) turned out to be nonsignificant. The effect of maternal risk factors was affected to a limited degree by the introduction of the Big 4 case mix.

The perfect guideline approach yielded similar results.

What does this mean? It means that when nothing goes wrong at homebirth, it is just as safe as midwife attended hospital birth. Since complications are uncommon, the overall rates of homebirth and hospital birth perinatal mortality are very similar. However, in the event of an unanticipated bad outcome, homebirth has a much higher perinatal mortality rate than midwife attended hospital birth.

In a nod to the BMJ study published by their colleagues last year, which showed that low risk birth with a Dutch midwife (home or hospital) has a higher mortality rate than high risk hospital birth with a Dutch obstetrician, the authors acknowledge that the mortality rate for midwife attended births, both home and hospital, are higher than expected:

... The data from an otherwise very similar country such as Flanders suggest that more favorable results may be expected in low-risk women in general from a hospital-based system. In Flanders, perinatal mortality is approximately 33% less than in The Netherlands, whereas the cesarean delivery rates show little difference.

The authors compare their results with other homebirth studies, noting that home and hospital populations differ markedly in risk profile and that any study of homebirth outcomes must correct for these differences.

Our conclusions apparently contradict those of De Jonge et al who concluded equal intrapartum and early neonatal outcome of planned home birth compared with hospital birth in apparently the same population... Our principal approach (natural prospective approach) compares neonatal mortality in the actual populations delivering at home compared with the hospital, whereas the approach of De Jonge et al compares neonatal mortality in a hypothetical group resembling our perfect guideline approach population.

What's the bottom line?

Homebirth is as safe as hospital birth when nothing goes wrong. But when complications occur unexpectedly at birth, hospital is much safer than home.

This piece first appeared on The Skeptical OB in October 2011.

Missouri: homebirth has a 20 fold increase in intrapartum death

The homebirth statistics keep on coming and they keep demonstrating the same thing: homebirth increases the risk of death.

I've written extensively about the appalling rate of perinatal death at the hands of licensed midwives in Colorado, and, of course, the overall US statistics show that homebirth with a direct entry midwife triples the neonatal death rate. The latest data comes from Missouri and the trend continues. Homebirth has a risk of intrapartum death that is more up to 20 times higher than hospital birth.

Birth outcomes of planned home births in Missouri: a population-based study by Chang and Macones published in the American Journal of Perinatology in August 2011 is notable for careful methodology.

... We obtained data from the Missouri live birth and fetal death files that have been linked together by the Missouri Department of Health and Senior Services... The Missouri vital record system is considered very reliable and has been adopted as a "gold standard" to validate other vital statistic datasets in the United States...

Our study sample consisted of women who delivered singleton pregnancies between 36-44 weeks of gestation ... Pregnancies complicated by major fetal anomalies and breech presentation were excluded ...

The authors divided the more than 800,000 births by place of birth and attendant creating three groups: hospital/birth center births attended by physicians and CNMs, homebirths attended by physicians and CNMs and homebirths attended by non-CNM midwives. (Of note, unplanned homebirths and births attended by non-midwives were excluded.) The groups differed significantly by maternal characteristics.

Women who had planned home deliveries attended by either non-CNMs or physicians/CNMs were more likely to be older, to be white, to have more children, to be overweight and to deliver at greater than 41 weeks gestational age, but less likely to be a Medicaid recipient or unmarried, to smoke during pregnancy, or to have a maternal medical risk factor.

Despite this, the outcomes in the homebirth groups (both those managed by non-CNMs and those managed by physicians or CNMs) had much poorer outcomes.

... [W]e observed that rates of newborn seizures were 4 per 1000 births among planned home births attended by non-CNMs, 0.6/1000 among planned home births delivered by physicians/CNMs and 1.1 per 1000 births among deliveries made by physicians/CNMs in hospitals and birthing centers. The rates of intrapartum fetal death were 0.9 per 1000 births among planned homebirths attended by non-CNMs, 1.7/1000 among planned home births delivered by physicians/CNMs, and 0.1 per thousand among deliveries made by physicians/CNMs in hospitals or birthing centers.

Rates of neonatal death were 1.4/1000 among planned homebirths attended by non-CNMs, 0 among planned homebirths attended by physicians/CNMs and 0.6/1000 among hospital/birth center births attended by physicians CNMs. This difference does not reach statistical significance, however.

After the authors employed multivariable logistic regression models:

... We observed that planned home births attended by non-CNMs remained positively associated with odds of newborn seizures after controlling for confounders. Specifically, the adjusted OR of newborn seizure among births delivered by non-CNMs was more than 5 times as much as the odds in hospital/birthing center births delivered by physicians/CNMs ... For intrapartum fetal death, planned home births attended by non-CNMs and physicians/CNM yielded adjusted ORs of 11.24 and 20.33 respectively relative to hospital/birthing center births attended by physicians/CNMs ...

The authors conclude:

Our analysis demonstrates cause of concern about safety of planned home births attended by non-CNMs and physician/CNMs. The results of our study suggest that planned home births are associated with increased likelihood of intrapartum fetal death and newborn seizures, despite the fact that the lowest risk women choose this birthing option.

As is the case with most homebirth studies, this study could not separate out home birth transfers from the hospital birth group. Therefore, the study likely underestimates the magnitude of the increase risk posed by homebirth.

This piece first appeared on The Skeptical OB in October 2011.

Cochrane childbirth reviews riddled with statistical errors

Lay people love Cochrane pregnancy and childbirth reviews. They always include plain language summaries, are systematic reviews or meta-analyses that are easy to understand, and are generally written by volunteers, many with an natural childbirth ax to grind.

Doctors are not nearly so enamored of Cochrane pregnancy and childbirth reviews. Although they agree in principle with the aims of the Cochrane project (synthesizing scientific evidence), the reviews are limited by the fact that papers included in a review are often poorly done, underpowered and differ markedly from each other in what results are measured and how they are measured. Moreover, Cochrane Childbirth reviews are often written by self-selected volunteers with an ax to grind, and therefore suffer appear to start with the conclusion and work back to include only papers that support it.

As a general matter, systematic reviews and met-analyses suffer serious limitations, some of which can be overcome with appropriate statistical analysis. However, as a new paper on the Cochrane Childbirth Reviews reveals, most are riddled with serious errors of statistical analysis that render their conclusions suspect or even useless.

Statistical methods can be improved within Cochrane pregnancy and childbirth reviews by Riley, Gates, Neilson, and Alfirevic was published in this month's issue of the Journal of Clinical Epidemiology. Coincidentally, I recently referenced Alfirevic as the author of the Cochrane Review on electronic fetal monitoring (EFM), the review that he acknowledged was underpowered to determine if EFM saves lives.

The intrinsic problems of systematic reviews have been summarized elsewhere as follows:

• There are numerous ways in which bias can be introduced in reviews
and meta-analyses of controlled clinical trials.

• If the methodological quality of trials is inadequate then the findings
of reviews of this material may also be compromised.

• Publication bias can distort findings because trials with statistically
significant results are more likely to get published, and more likely to
be published without delay, than trials without significant results...

• Criteria for inclusion of studies into a review may be influenced by
knowledge of the results of the set of potential studies...

These limitations can be summarized by the pithy phrase "garbage in, garbage out." A meta-analysis or systematic review is only as good as the quality of the papers reviewed.

The Cochrane Childbirth Reviews suffers from these problems and more:

There are deficiencies in the use of statistical methods within the Cochrane Pregnancy and Childbirth Group (CPCG) reviews. The issue of publication bias is
rarely addressed; the process of measuring, investigating, and accounting for heterogeneity is often limited or inadequate; and random-effects analyses are
not correctly interpreted. The large number of metaanalyses per review also raises the concern of multiple testing. These problems need to be urgently
addressed...

Improved use of statistical methods is urgently needed within Cochrane reviews. Although we have only assessed CPCG reviews in the article, our findings have general implications for all Cochrane reviews... The Cochrane Collaboration must seek to engage more statisticians and methodologists within individual reviews ...

The problems identified in the Cochrane Reviews were not limited to a small subset of the reviews. For example, in assessing publication bias, the authors note:

Just 6 (7%) of the 75 reviews stated in their Methods section how they would assess publication bias; only 7 (9%) described a publication bias assessment in their Results or Discussion section or justified why publication bias assessments were not possible; and only 3 reviews described a publication bias assessment plan in their Methods section and subsequently reported an assessment in their Results or Discussion section...

The authors acknowledge that errors such as these seriously limited the validity of Cochrane pregnancy and childbirth reviews:

... It is clear that CPCG reviews must now consider the issue of publication bias in more detail, both when planning their review and when interpreting their results. This is particularly important for their primary analyses, as else misleading or overly strong conclusions may be made...

Unless and until these issues are addressed, Cochrane pregnancy and childbirth reviews will continue to dazzle lay people with incorrect conclusions, and be dismissed by doctors as poorly done and riddled with statistical errors.

This piece first appeared on The Skeptical OB in June 2011.

Electronic fetal monitoring halves early neonatal death rate

The American Journal of Obstetrics and Gynecology has just published a "Report of Major Impact" that demonstrates that electronic fetal monitoring saves lives.

Electronic fetal heart rate monitoring and its relationship to neonatal and infant mortality in the United States by Chen et al. is the largest study ever done on electronic fetal monitoring (EFM). The authors reviewed 1,732,211 singleton live births (elective C-sections and congenital anomalies were excluded. Of these births); 89% involved EFM while 11% did not. They found:

The corrected early neonatal, late neonatal, postneonatal, and infant mortality rates for all subjects were 0.9, 0.5, 1.7, and 3.1 per 1000 births, respectively... The risk of corrected mortality rate was different between those with vs without EFM during the early neonatal period (0.8 vs 1.7 per 1000 births, respectively; P ‹ .001), but not in late (0.5 vs 0.6; P ‹ .402) or postneonatal periods (1.7 vs 1.8; P ‹ .296).

In other words, EFM cut the rate of early neonatal death in half (death from birth to 7 days), but had no impact on late neonatal death (from 7-28 days of life) or postneonatal death (from 1 month to 1 year of age). This is exactly what you would expect to find if EFM prevents peripartum death from hypoxia (lack of oxygen). The results are represented graphically below.

There were secondary findings as well:

... Use of EFM was associated with an increased likelihood of operative vaginal delivery for all indications, as well as for fetal distress. In addition, use of continuous monitoring was associated with an increased risk of primary cesarean delivery for fetal distress ...

...[U]se of EFM was associated with a lower likelihood of 5-minute Apgar score ‹4...

The secondary analysis also indicates that the rate of neonatal seizure was significantly lower only among high-risk women who had EFM...

What are the differences between this study and the Cochrane review that purported to show that EFM increases operative delivery rates without improving neonatal survival? The Cochrane review, encompassing 37,000 women, was simply too small.

The combined sample size of 12 RCTs is insufficient to determine whether EFM can significantly lower neonatal mortality. Alfirevic [the author of the Cochrane review] noted that to test the hypothesis that continuous monitoring can prevent 1 death in 1000 births, more than 50,000 women need randomization...

Moreover, as the Cochrane review authors themselves noted, of the 12 RCTs included in the analysis, only 2 were high quality studies.

Chen et al. conclude:

According to the Cochrane review and the most recent ACOG recommendation, the use of fetal heart rate monitoring increases operative delivery rate without a concomitant decrease in longterm neonatal outcomes. Thus, understandably there has been continued angst about using fetal heart rate monitoring during labor. The main implication of our study is that now there is reassuring evidence for the use of EFM; its use is linked with ... a significant decrease in early neonatal and infant mortality ... [F]etal heart rate monitoring can be used in daily practice with some assurance.

A conclusion of the study is the large sample size necessary to demonstrate improvement in neonatal outcomes. One reason the ... Cochrane review did not demonstrate benefit of EFM is small sample size of published reports. Alfirevic et al acknowledged that over 50,000 women need to be randomized to demonstrate improvement in mortality. The issue of sufficient sample size ... remains unachievable in modern day obstetrics... Thus, when the outcomes are uncommon and randomized trials are not plausible, we should consider evidence from "reality-based medicine," for it, along with this study, demonstrates improvement in mortality
with EFM.

Of note, this study, which contains only live births, almost certainly underestimates the benefit of EFM. EFM appears to save lives by decreasing the risks of hypoxic brain injuries associated with low Apgar scores (‹4). The study did not include intrapartum deaths (Apgar 0), where the benefits of EFM are similar or even larger.

The bottom line is that the largest study of electronic fetal monitoring to date shows that EFM cuts the rate of early neonatal death in half. That is a dramatic benefit.

This piece first appeared on The Skeptical OB in June 2011.

A stunning indictment of midwives in The Netherlands

Homebirth and midwifery advocates point with pride to a recent study that showed that homebirth with a midwife in the Netherlands is as safe as hospital birth with a midwife (Perinatal mortality and morbidity in a nationwide cohort of 529 688 low-risk planned home and hospital births). They tout this study as evidence that homebirth is as safe as hospital. A new study suggests an entirely different explanation: Dutch midwives have unacceptably high rates of perinatal mortality both at home and in the hospital. Indeed, the perinatal mortality rate for LOW risk women cared for by Dutch midwives is HIGHER than the perinatal mortality rate for HIGH risk women cared for by Dutch obstetricians!

The new study, Perinatal mortality and severe morbidity in low and high risk term pregnancies in the Netherlands: prospective cohort study, appears in this week's issue of the British Medical Journal. The authors explain that the study was undertaken to investigate why the Netherlands has highest perinatal mortality rate in Europe.

Several factors are mentioned as possible explanations for this high mortality, such as differences in registration and maternal characteristics of the Dutch childbearing population, restricted management of premature babies, and the absence of standard screening for congenital anomalies. The numbers of older mothers, multiple pregnancies, and mothers belonging to an ethnic minority are relatively high in the Netherlands. However, this can only partly explain the high perinatal mortality. Whether the Dutch obstetric care system contributes to this relatively high mortality remains unclear.

This is an important question because the Dutch system of maternity care relies primarily on midwives and those midwives perform a relatively high number of homebirths. This study, a cohort study of severe morbidity and mortality of term fetuses or neonates, called ATNICID (Admission of Term Neonates to Intensive Care or Intrauterine Death), was begun in 2007 with the express intent of examining the relationship between the organization of the Dutch maternity care system and the high rate of perinatal mortality.

The study ultimately enrolled 37,735 term infants without congenital anomalies:

16,672 (44.2%) infants of nulliparous women (including 143 (0.9%) twin pregnancies) and 21,063 (55.8%) infants of multiparous women (including 226 (1.1%) twin pregnancies). Data on 91 (0.2%) infants were missing; we excluded these from further analysis... 18,686 (49.5%) infants were born to women who started labour in primary care as low risk, of whom 5492 (29.4%) were referred to secondary care during labour; 13,194 (35.0%) infants were born under the supervision of a midwife in primary care, and 24,450 (64.8%) infants were born under the supervision of a gynaecologist.

The results were astounding:

Of the 60 antepartum stillbirths, 37 occurred in primary care and 23 in secondary care...

Twenty-two intrapartum stillbirths and 14 delivery related neonatal deaths occurred. Infants of pregnant women at low risk had a significantly higher risk of delivery related perinatal death (relative risk 2.33, 1.12 to 4.83), compared with infants of women at high risk whose labour started in secondary care under the supervision of an obstetrician. Infants of women who were referred to secondary care during labour had a 3.66 times higher risk of delivery related perinatal death than did infants of women who started labour in secondary care (relative risk 3.66, 1.58 to 8.46)...

A total of 210 infants were admitted to the NICU:

... resulting in an overall incidence of admission to NICU of 5.58 (4.83 to 6.33) per 1000 live births.... Half of the women (51%, n=107) started labour in primary care. Of these, 70% (n=75) were referred to secondary care during labour... The incidence of admissions to the NICU was 2.43 per 1000 term births in primary care, 13.7 per 1000 term births if referral to secondary care during labour occurred, and 5.45 per 1000 term births managed exclusively in secondary care.

Nearly half the NICU admissions were the result of one cause: asphyxia. Among the 17 infant deaths:

71% (n=12) [were] because of asphyxia and 29% (n=5) because of an infection. Fourteen cases were classified as directly related to circumstances during labour.

Of the 26 deaths related to labor presided over by midwives, 65% were attempted homebirths.

These results are deeply shocking.

We found that delivery related perinatal death was significantly higher among low risk pregnancies in midwife supervised primary care than among high risk pregnancies in obstetrician supervised secondary care. This difference was even greater among the cases that were referred from primary to secondary care during labour. Unfortunately, we were unable to adjust for confounding variables because we used aggregated data from a large birth registry database. However, the results are unlikely to have been overestimated, because risk factors such as low socioeconomic status, higher age, or non-Western ethnicity were more prevalent among the women at high risk. (my emphasis)

The authors express their concern:

In summary, the Dutch obstetric care system is based on the assumptions that pregnant women and women in labour can be divided into a low risk group and a high risk group, that the first group of women can be supervised by a midwife (primary care) and the second group by an obstetrician (secondary care), and that women in the primary care group can deliver at home or in hospital with their own midwife. When complications occur or risk factors arise antenatally, during labour, or in the puerperium in primary care, the women is referred to secondary care. We found that the perinatal death rate of normal term infants was higher in the low risk group than in the high risk group, so the Dutch system of risk selection in relation to perinatal death at term is not as effective as was once thought. This also implies that the high perinatal death rate in the Netherlands compared with other European countries may be caused by the obstetric care system itself, among other factors. A critical evaluation of the obstetric care system in the Netherlands is thus urgently needed.

In contrast to the claims of homebirth and midwifery advocates, the Netherlands is far from being the ideal model of obstetric care. The Netherlands has the highest perinatal mortality in Europe, and midwifery care may very well be the cause of this calamity.

This piece first appeared on The Skeptical OB in November 2010.

Induction is rising and perinatal mortality is falling

A new paper in the journal Obstetrics and Gynecology calls attention to the rising rate of induction and the falling birth weight which appears to be a consequence. Decreased term and postterm birthweight in the United States: impact of labor induction implies that this trend is worrisome:

From 1992 through 2003, mean BW fell by 37 g, mean GA by 3 days, and macrosomia rates by 25%. Rates of induction nearly doubled from 14% to 27%. Our ecological state-level analysis showed that the increased rate of induction was significantly associated with reduced mean BW (r = –0.54; 95% confidence interval [CI], –0.71 to –0.29), mean GA (r = –0.44; 95% CI, –0.65 to –0.17), and rate of macrosomia (r = –0.55; 95% CI, –0.74 to –0.32)...

Increasing use of induction is a likely cause of the observed recent declines in BW and GA. The impact of these trends on infant and long-term health warrants attention and investigation.

The following graph illustrates the trend of decreased birth weight.

The decrease in birth weight appears dramatic because of the scale of the y-axis, which displays weights from 3400 gm to 3500 gm. Indeed the fall in birth weight over the entire study was less than 50 gm from 3492.3 gm to 3455.3 gm.

As the authors note:

A reduction of 40 g in BW or 3 days in GA may not matter for an individual infant, but represents a substantial change for a population.

But that change is not necessarily bad. for example, as the authors point out, the rate of macrosomia has dropped by 25% in the same time span. Nonetheless, the authors are worried:

Recent systematic reviews and metaanalyses ... concluded that labor induction may reduce perinatal mortality but without increasing the risk of cesarean delivery. As observed in this study, increasing and earlier use of labor induction appears to have shortened the duration of gestation and thus reduced both mean BW and rates of macrosomia. Although several studies have reported increased risks of some causes of neonatal morbidity and maternal complications with increasing GA at term, more and more infants are being delivered at early term gestation (37-38 weeks), up from 19% in 1992 to 29% in 2003. Earlier term birth is associated with increased risk of sudden infant death syndrome, and we have recently documented increases in several adverse birth outcomes among early term births, including increased risks of infant mortality and some types of neonatal morbidity. Thus the impact of these recent trends requires further investigation, including large randomized trials, to ensure that the rise in induction is doing more good than harm.

It's rather surprising then that the authors did not investigate the trend in perinatal mortality during the same period. The main purpose of labor induction is to reduce stillbirth, which will be reflected in the perinatal death rate (death from 28 weeks of pregnancy to 7 days of life). All their data comes from CDC databases and perinatal mortality is available from the same source.

As the following graph shows, perinatal mortality did indeed drop by 21%.

Correlation, of course, is not causation, and it is possible that perinatal mortality has been steadily dropping for other reasons. And as the authors of the paper point out, there are risks associated with delivery at earlier gestation. Nonetheless, the existing evidence suggests that the increasing induction rate has not led to an increase in perinatal mortality. The primary reason for induction is to reduce perinatal mortality and that is exactly what seems to have happened.

This piece first appeared on The Skeptical OB in September 2010.

The high perinatal mortality rate in The Netherlands.

Homebirth and natural childbirth advocates often approvingly cite the maternity care in the Netherlands. Homebirth rates are the highest in the world (30%, but down considerably and falling every year) and midwives are the mainstay of the system, caring for any woman who does not require the care of a doctor.

What homebirth and natural childbirth advocates fail to realize is that The Netherlands has one the highest perinatal death rate in Europe and a high and rising rate of maternal mortality. Indeed, the Dutch have become so alarmed at the perinatal and neonatal death rates that the government has convened a variety of investigations to determine the cause.

The paper Higher perinatal mortality in The Netherlands than in other European countries: the Peristat-II study, published in a Dutch journal, brought the issue of perinatal mortality into focus:

... In Peristat-II from 22 weeks gestation, after France, The Netherlands had the highest fetal mortality rate (7.0 per 1,000 total number of births). Of all western European countries, The Netherlands had the highest early neonatal mortality rate (3.0 per 1,000 live births). Over the past 5 years the perinatal mortality rate in The Netherlands has dropped from 10.9 to 10.0 per 1,000 total births but this drop has been faster in other countries. CONCLUSION: The Netherlands has a relatively high number of older mothers and multiple pregnancies, but this only partly explains the high Dutch perinatal mortality rate which still ranks unfavourably in the European tables. More research is necessary to gain insight into the prevalence of risk factors for perinatal mortality compared with other European countries. In addition, perinatal health and the quality ofperinatal healthcare deserve a more prominent position in Dutch research programmes.

The government has commissioned researchers at the Erasmus Medical Centre in Rotterdam to oversee the investigation. From the Erasmus MC website:

The Netherlands has a relatively poor position in Europe when it comes to health at the time of birth, in other words, perinatal health. Approximately 10 out of every 1000 children die around the time of birth. In similar other countries this mortality rate can be as much as 30% lower. Of the perinatal deaths in the Netherlands, 70% are stillbirths when counted from the 22nd week of pregnancy. Thirty percent of the perinatal deaths take place in the first week after birth. In Flanders, that is socio-democratically and economically comparable to the Netherlands, the perinatal death rate has been two-thirds of that in the Netherlands for at least 10 years. This means that instead of 1700 cases of perinatal death that occur per year among the 175,000 newborns in the Netherlands, only 1150 cases should occur; an unprecedented large difference. Moreover, within the Netherlands, and particularly in the larger cities such as Rotterdam and The Hague, there are distinct differences between groups of pregnant women.

The ZonMw has commissioned Erasmus MC to carry out the Descriptive study Pregnancy and Childbirth. The aim of the study is to determine knowledge questions and research opportunities to improve the perinatal care in the Netherlands. Aspects studied include patient-related risk factors such as diseases already present, lifestyle and social factors on the one hand and the role of the midwife practices including use of care, risk selection, and quality of care in the Netherlands on the other. The preliminary conclusion is that the unfavorable European position is probably mainly caused by factors in the care system while the differences within the Netherlands and the larger cities are linked to large risk differences between groups on the basis of ethnicity, social deprivation and the neighborhood in which people live. A research agenda has been formulated based on this.

In other words, the government investigation found that one of the main reasons for the high perinatal death rate is the midwife care system including use of care, risk selection, and quality of care.

Not only is perinatal mortality unacceptably high, maternal mortality is high and rising. According to the paper Rise in maternal mortality in the Netherlands, published in the British Journal of Obstetrics and Gynaecology earlier this year:

The overall maternal mortality ratio was 12.1 per 100 000 live births, which was a statistically significant rise compared with the maternal mortality ratio of 9.7 in the period 1983–1992 (OR 1.2, 95% CI 1.0–1.5). The most frequent direct causes were (pre-)eclampsia, thromboembolism, sudden death in pregnancy, sepsis, obstetric haemorrhage and amniotic fluid embolism. The number of indirect deaths also increased, mainly caused by an increase in cardiovascular disorders (OR 2.5, 95% CI 1.4–4.6). Women younger than 20 years and older than 45 years, those with high parity or from nonwestern immigrant populations were at higher risk. Most substandard care was found in women with pre-eclampsia (91%) and in immigrant populations (62%).

Conclusions
Maternal mortality in the Netherlands has increased since 1983–1992. Pre-eclampsia remains the number one cause. Groups at higher risk for complications during pregnancy should be better identified early in pregnancy or before conception, in order to receive preconception advice and more frequent antenatal visits. There is an urgent need for the better education of women and professionals concerning the danger signs, and for the training of professionals in order to improve maternal health care.

In an accompanying commentary, British obstetrician JJ Walker notes that there has been pressure in the UK to adopt the Dutch system of maternity care:

... The fact that there are areas of concern in the Netherlands over rising maternal death ratios, despite their generally high socio-economic profile, as well as the previously documented high level of perinatal mortality, suggests that we should be cautious about moving our pattern of care towards theirs without careful consideration of a potentially adverse effect on maternal and perinatal mortality and morbidity. The UK has improved its safety for both mothers and babies by careful audit and guideline development. Care should be taken not to undo these changes by striving for political correctness.

American homebirth advocates and natural childbirth advocates who point to the Dutch system as a model would do well to heed Dr. Walker's advice.

This piece first appeared on The Skeptical OB in July 2010.

The Wax study has serious flaws

I've been crunching the numbers myself for years. I've never been in any doubt about them, but it's nice to see confirmation in the literature. The largest homebirth study ever done shows that homebirth triples the rate of neonatal death.

Maternal and newborn outcomes in planned homebirth: a meta-analysis will be published in the September issue of the American Journal of Obstetrics and Gynecology. I was fortunate to obtain an advance copy. More than 342,000 homebirths were compared to more than 207,000 hospital births. The data was obtained by pooling 12 major studies from a variety of countries. These studies include the recent DeJonge study from The Netherlands and the Janssen study from Canada. There are also studies from the US, the UK, Australia and Sweden.

The 12 studies were culled from a search of the scientific literature comparing planned homebirth to hospital birth. Only those studies that used the intended place of birth, as opposed to the actual place of birth, are included. That's important because many homebirth studies look at actual place of birth and thereby include homebirth transfers in the hospital group, skewing the results.

Not surprisingly, the rate of interventions was lower in the homebirth group:

Planned home births experienced significantly fewer medical interventions including epidural analgesia, electronic fetal heart rate monitoring, episiotomy, and operative vaginal and cesarean deliveries. Likewise, women intending home deliveries hadfewer infections, 3-degree lacerations [or greater], perineal and vaginal lacerations, hemorrhages, and retained placentas.

Unfortunately, that lower rate of interventions resulted in a higher rate of neonatal death:

... the overall neonatal death rate was almost twice as high in planned home vs planned hospital births, and almost tripled among nonanomalous neonates. Importantly, these latter observations were consistent across all studies examining
neonatal mortality, regardless of the covered time period...

These findings are troubling:

Of concern, this investigation identified a doubling and tripling of the neonatal mortality rate overall and among nonanomalous offspring, respectively, in planned home compared to planned hospital births. This finding is particularly robust considering the homogeneity of the observation across studies. It is especially striking as women planning home births were of similar and often lower obstetric risk than those planning hospital births. The planned home delivery group commonly exhibited fewer obstetric risk factors such as excessive body mass index, nulliparity, prior cesarean, and previous pregnancy complications.

Why was the neonatal death rate higher in the planned homebirth group?

... 2 cohort studies implicated intrapartum asphyxia in 31% and 52% of planned home delivery perinatal deaths. The past 2 decades have seen a significant decrease in such deaths, with evidence suggesting fewer fetuses experiencing intrapartum anoxia. Speculative explanations for the trend include more liberal use of ultrasound, electronic fetal heart rate monitoring, fetal acid-base assessment, labor induction, and cesarean delivery. Our findings, considered in light of these observations, raise the question of a link between the increased neonatal mortality among planned home births and the decreased obstetric intervention in this group.

I'd like to tell you that this study is rock solid, since it confirms what I have been writing for years. Unfortunately, the study has some serious flaws. First, it includes some data collected more than two decades ago. Second, it includes some papers that looked at very small numbers of births. Third, while it found a dramatically increased risk of neonatal mortality, it found no difference in perinatal mortality. This is not what we would expect if the excess deaths were due to intrapartum stillbirths or failed resuscitations.

The meta-analysis include two large studies from The Netherlands and Canada, both of which showed no difference in perinatal and neonatal mortality rates. The other 10 studies (from the US, the UK, Sweden and Australia) did show increased rates of perinatal and neonatal death. It seems to me that the take home message of the study is not that homebirth is unsafe, but that homebirth can only be safe when practiced by highly trained midwives, fully integrated into the hospital system in countries with strict criteria for homebirth and dedicated transport systems for emergencies. In other words, homebirth is safe in The Netherlands and Canada, but no where else.

This piece first appeared on The Skeptical OB in July 2010.

Waterbirth fatalities

Waterbirth has become a central component of "natural" childbirth dogma, despite the fact that for primates giving birth underwater is entirely unnatural. You don't need a medical degree to appreciate the idiocy of birth in water. The most critical task for the newborn is to take its first breath. Inhaling a mouthful of fecally contaminated water instead of air is profoundly dangerous. Not surprisingly, as the popularity of waterbirth has grown, the number of neonatal deaths directly attributable to it has grown as well.

A new paper in the American Journal of Forensic Medical Pathology discusses the tragic case of a term newborn who died of Pseudomonas pneumonia and sepsis as a result of waterbirth. The authors review the existing literature on fatalities associated with waterbirth and the underlying processes leading to neonatal death.

The case report:

A normally formed 42-week gestation male infant was born underwater in a birthing tank to a 29-year-old primigravida mother. The Apgar scores were 9 and 10 at 1 and 5 minutes, respectively. The infant was covered with thick meconium and demonstrated intercostal recession with peripheral cyanosis. He was transferred to hospital where his respiratory status worsened and a chest x-ray demonstrated generalized opacity. Presumed sepsis was treated with broad-spectrum antibiotics. There was no evidence of hyponatremia. Despite maximal therapy he developed respiratory failure with disseminated intravascular coagulation and died at 4 days of age.

... Death was due to extensive P. aeruginosa pneumonia and sepsis associated with meconium aspiration and water birth.

The authors reviewed the literature:

Underwater birth has been promoted as a means of improving the quality of delivery... While the benefits of immersion are said to include increased comfort and relaxation for mother and infant, with greater maternal autonomy, fewer injuries to the birth canal, reduced need for analgesia, with decreased instrumentation and operative intervention, this has been disputed with no clear advantages or disadvantages over conventional births being demonstrated. In addition, other reports of underwater births have documented significant morbidity and even death. Problems have included infections, near drowning/drowning, hyponatremia/water intoxication, seizures, infections, respiratory distress, fevers, hypoxic brain damage, and cord rupture with hemorrhage.

Natural childbirth advocates have a terrible habit of inventing scientific "facts" and waterbirth is a classic example. According to NCB advocates, newborns will not attempt to breathe while immersed in warm water and will wait to take a first breath until they are in direct contact with air. That theory has no basis in neonatal physiology.

It has been postulated that newborns will not breath or swallow while immersed in warm water, and that respiration will only be initiated on exposure to cold air. This has been used to support assertions that drowning and aspiration of water cannot occur with underwater delivery. However, animal studies have demonstrated that this reflex can be over-ridden, and given that respiratory movements occur in utero, it is difficult to see why this process would not continue in a neonate delivered into water. The documentation of cases of near drowning and respiratory distress with apparent aspiration of fluid would also be supportive of the occurrence of breathing under water. In addition, the finding of hyponatremia in certain of these infants would be in keeping with inhalation of fresh water, as lowered sodium levels have resulted from fresh water drowning.

It is ironic that NCB advocates, the self appointed guardians of "physiologic birth" would embrace a practice that is profoundly non-physiologic. Not surprisingly, the consequences can be devastating. Neonates can and do inhale copious amounts of fecally contaminated water during waterbirth. Indeed, they have been found to inhale such large quantities of water that the water dilutes the concentration of sodium in the bloodstream to fatally low levels (hyponatremia). Even small amounts of inhaled water can introduce significant amounts of bacteria into the neonatal lungs leading to pneumonia and other infections as the authors explain:

Sepsis has also arisen from underwater deliveries, ranging from umbilical and ear infections to septicemia and pneumonia. The source of such infections has been contamination of birthing tubs, hoses, and taps with virulent organisms such as P. aeruginosa and Klebsiella pneumoniae. These bacteria have been found despite careful cleaning of systems between deliveries. Lethal Legionella infection has occurred in an underwater birth reported from Japan and other organisms such as amoeba and Mycobacterium avium have been found in spa baths...

The bottom line is that waterbirth kills babies.

As the death of a newborn from entirely preventable factors is of great concern, parents who elect to have an underwater delivery must be appraised of the risks that characterize an aquatic birth, and should have access to resuscitation equipment to enable rapid suctioning of the airway.

The avoidable tragedies of waterbirth cast a harsh light on the fundamental weakness of "natural" childbirth philosophy. "Natural" childbirth advocates pick and choose desired elements of "natural" birth without regard to whether those elements are truly natural. Despite the claims of NCB advocates that their philosophy is "evidence based," they routinely ignore scientific evidence and make recommendations without ever performing safety testing on those recommendations. Moreover, they are not above fabricating scientific "facts" to bolster claims that have no scientific support. Finally, and most egregiously, babies die as a result of their "advice" and they don't know and apparently don't care.

This piece first appeared on The Skeptical OB in May 2010.