Births at home or in hospital: risks explained

A major study published today has examined the risks of planned home births, comparing them against planned deliveries in hospitals and midwife units. The research was covered by several newspapers, some of which highlighted a low risk while others said the practice carries a high risk.

The extensive study looked the risks of serious complications and how these differed according to the setting women initially chose for their delivery. It found that, generally, births planned to take place at home, in hospital and in midwife units all carried a low level of risk. However, when the researchers focused solely on women planning to have their first baby at home, they found they were almost three times more likely to suffer complications than if they went to hospital. It's important to note, however, that even this raised risk equated to a low chance of serious complications.

While some newspapers have suggested home births are dangerous, this study supported the safety of a range of birth practices, with serious complications seen in only 4.3 births out of 1,000 overall. This report will no doubt be of interest to parents who are planning where to have their baby and wish to discuss their options with their midwife or GP.

What did the study look at?

This large English study was designed to take a detailed look at the risks associated with different settings where women with low-risk pregnancies planned to give birth. A low-risk pregnancy is one where the mother and baby are not affected by conditions or circumstances that can complicate the birth (see What is a low-risk pregnancy? for further details).

The study compared home births, midwifery units run outside of a hospital setting, obstetric unit births in hospitals and births in ‘alongside midwifery units’, which are midwife-led units on a hospital site that also have an obstetric unit. Its analysis featured data on almost 65,000 women using maternity services across England.

The researchers were primarily focused on a composite of the rates of mortality at or just after birth, and on injuries that may occur during birth such as broken bones, traumatic nerve injury, brain injury and a type of respiratory infection called meconium aspiration syndrome.

These outcomes were used to derive the composite measure, as they may be related to the quality of care while giving birth. In particular they reflect complications associated with oxygen starvation and birth trauma. The researchers also looked at the mode of delivery and whether women were transferred from their planned place of birth.

What is a ‘low-risk pregnancy’?

In this study the researchers defined a ‘low-risk pregnancy’ as one where women were not identified as having particular medical factors before the onset of labour. These medical or obstetric risk factors were defined as those listed in the NICE birthing care guideline, which may indicate that a hospital setting would be the most appropriate setting for delivery. They included:

• long-term medical conditions such as heart disease, high blood pressure, severe asthma, cystic fibrosis, diabetes and blood disorders such as sickle-cell disease
• infections such as HIV or hepatitis B or C, or current infection with chickenpox, german measles or genital herpes
• psychiatric disorders requiring current inpatient care
• complications with previous pregnancies
• complications during their current pregnancy, such as multiple births, placental praevia (where the placenta is positioned over the cervix), pre-term labour, pre-eclampsia, onset of gestational diabetes, damage to the placenta, induction of labour, and breech position of the baby. Risks may also include a ‘small for gestational age’ baby, or if the baby had an abnormal foetal heart rate.

NICE guidelines on this matter are extensive, and so the list above is not exhaustive.

How was the study performed?

The study aimed to collect data from every NHS trust in England providing home birth services, every freestanding midwifery unit, every alongside midwifery unit (attached or close to a hospital), and a random sample of obstetric units (using a system designed to ensure that large and small units from different parts of the country were included).

A total of 64,538 women with low-risk pregnancies were recruited between April 1 2008 and April 30 2010. They were assigned to different groups dependent on where they originally planned to give birth, regardless of whether they were transferred during labour or immediately after birth. The extensive study then proceeded to record key information on their pregnancy, birth and complications.

Is it safe for women to have a baby at home?

The overall rate of negative outcomes (a composite of outcomes of death or serious complications) was 4.3 per 1000 births (95% confidence interval [CI] 3.3 to 5.5) and there was no difference between non-obstetric unit settings compared with obstetric units. This indicates that as a whole, home births are as safe as ones in medical settings.

The researchers then looked only at women who were going through their first pregnancy. They found that women having their first birth at home had a greater chance of complications leading to injury in the child than women who had planned to go to an obstetric unit in a hospital. This risk was almost doubled (odds ratio [OR] 1.75, 95% CI 1.07 to 2.86).

Furthermore, when the sample was restricted to women who had no complicating conditions at the start of labour, there was almost a three-times greater risk for women with planned home births than for women having planned hospital births (OR 2.80, 95% CI 1.59 to 4.92). There was no difference in the rates of such complications in either type of midwife-led unit compared with hospital units.

An important point to note is that even though the risk associated with home births seems greatly elevated in women going through their first pregnancy, the absolute risks were still relatively low. To put this into context, they occurred in 39 of the 4,488 women who delivered their first child at home, and 36 of the 4,063 women who delivered their first child at home without complicating conditions at the start of labour.

It is important to highlight that the Daily Mail ’s headline that first-time mothers who opt for a home birth ‘triple the risk of death or brain damage’ may be misleading: the study had used a composite score of a variety of birth-related injuries. Overall, of the 250 events they saw in this study, early neonatal deaths accounted for 13% of events, brain damage 46%, meconium aspiration syndrome 30%, traumatic nerve damage 4% and fractured bones 4%. Some of events these would be treatable.

For women who had a previous pregnancy, the rates of such events did not differ between women who had planned a home birth, hospital birth or birth at a midwife-led centre.

How many women planning home deliveries or midwife unit delivery ended up in hospital?

Among women having their first pregnancy who opted for a home birth, 45% were transferred to hospital before or after delivery. For women attending a freestanding midwifery unit, 36% were transferred, and 40% of women attending an alongside midwifery unit were transferred.

For women who had previous pregnancies, 12% with a planned home birth were transferred, 9% at a freestanding midwifery unit and 12.5% at an alongside midwifery unit were transferred.

The chance of receiving caesarean section for these low-risk women was lower in all three non-obstetric unit settings, with the lowest rates seen among women who had planned to deliver at home or at freestanding midwifery unit births.

• obstetric unit: 11.1% (95% CI 9.5 to 13.0)
• home: 2.8% (95% CI 2.3 to 3.4)
• freestanding midwifery unit: 3.5% (95% CI 2.8 to 4.2)
• alongside midwifery unit: 4.4% (95% CI 3.5 to 5.5)

How do I choose where to give birth?

The researchers of this study said that their results support the policy of offering a choice of where they give birth to healthy women, both to those having their first baby and those who have had previous pregnancies. It is important to note that the type of adverse events described here are uncommon in all settings. Women choosing where to give birth can discuss the matter and these particular risks with their midwife or GP when deciding where they would feel most comfortable giving birth.

One key element that may influence the decision of where to give birth is pain management. The study had found that the proportions of women who received epidural or spinal analgesia were lower in non-obstetric units than in a hospital. For example, 30% of women attending a hospital, 8% of women with a home birth, 11% in women attending a freestanding midwife unit and 15% of women attending an alongside midwifery unit received epidural or spinal analgesia. There are many options for pain management besides an epidural, and this is something that can be planned with a doctor and midwife, and taken into account when planning where to give birth.