Appendix D: Benefits and risks of birth out of water and water birth
Where has this data come from?
The information used to generate the tables below comes from 4 published studies: Aughey et al. (2021), Bailey et al. (2020), Bovbjerg et al. (2022) and Sanders et al. (2024). Two studies were conducted in the UK (Aughey et al. 2021, Sanders et al. 2024) and two studies were conducted in the USA (Bailey et al. 2020, Bovbjerg et al. 2022). Evidence review Q includes full information on each primary study, but overall the committee agreed that the evidence was broadly applicable to the current UK context. The populations reported in the studies were recorded as women and no further breakdown given, so that terminology is used in this appendix.
How was risk difference in the table calculated?
In order to estimate risk differences in these tables, adjusted or unadjusted relative effects of water birth compared to birth out of water were used to calculate these values for each outcome. Because these differences are based on relative effects, they are not always the same as the mathematical difference between the risk with birth out of water and the risk with water birth.
Does this mean that water birth causes these outcomes?
As the evidence in these tables is all derived from non-randomised studies, it is not possible to conclude that water birth definitively causes any outcomes. Each study has adjusted for some potential confounding factors, which makes it more likely that the difference in risk is related to water birth, but it is impossible to say this for certain without large randomised studies.
| Outcomes | Risk with birth out of water | Risk with water birth | Risk difference |
|---|---|---|---|
|
Perineal trauma (third-degree or fourth-degree tear at birth): multiparous women (Sanders et al. 2024) |
About 17 women out of 1,000 |
About 11 women out of 1,000 |
About 6 fewer women out of 1,000 who had a water birth would be expected to have perineal trauma (third-degree or fourth-degree tear at birth). So for about 994 women out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Postpartum haemorrhage (500 ml or more): nulliparous and multiparous women (Sanders et al. 2024; follow-up period not reported) |
About 160 women out of 1,000 |
About 131 women out of 1,000 |
About 29 fewer women out of 1,000 who had a water birth would be expected to have postpartum haemorrhage (500 ml or more). So for about 971 women out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Postpartum haemorrhage (major: 1,000 ml or more): nulliparous and multiparous women (Sanders et al. 2024; follow-up period not reported) |
About 38 women out of 1,000 |
About 29 women out of 1,000 |
About 4 fewer women out of 1,000 who had a water birth would be expected to have postpartum haemorrhage (1,000 ml or more). So for about 996 women out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Postpartum haemorrhage (major: 1,500 ml or more): nulliparous and multiparous women (Sanders et al. 2024; follow-up period not reported) |
About 13 women out of 1,000 |
About 11 women out of 1,000 |
About 2 fewer women out of 1,000 who had a water birth would be expected to have postpartum haemorrhage (1,500 ml or more). So for about 998 women out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Postpartum haemorrhage (major: 1,500 ml or more): nulliparous and multiparous women (Aughey et al. 2021; follow-up period not reported) |
About 12 women out of 1,000 |
About 9 women out of 1,000 |
About 4 fewer women out of 1,000 who had a water birth would be expected to have postpartum haemorrhage (1,500 ml or more). So for about 996 women out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Postpartum haemorrhage (defined as "diagnosed with haemorrhage in third or fourth stage of labour" by the study, regardless of estimated blood loss): nulliparous and multiparous women (Bovbjerg et al. 2022) |
About 54 women out of 1,000 |
About 47 women out of 1,000 |
About 7 fewer women out of 1,000 who had a water birth would be expected to have postpartum haemorrhage (defined as "diagnosed with haemorrhage in third or fourth stage of labour" by the study, regardless of estimated blood loss). So for about 993 women out of 1,000 the outcome was the same irrespective of the method of birth. |
RR: risk ratio.
Notes: risk with birth out of water and water birth were based on raw data from the studies, and hence there is a discrepancy with risk difference calculated in GRADE, which considers the unadjusted relative effect or the relative effect adjusted for confounding factors in the calculation.
Risk difference has been calculated in GRADE, which takes into account the unadjusted relative effect or the relative effect adjusted for confounding factors in the calculation (for example, risk difference out of 1,000 = 1,000 × control group risk × (1 − RR)). Confounding factors are things such as maternal age, parity, ethnicity, birth weight, and socioeconomic status (see evidence review Q for full details) which may affect the results for different outcomes after birth out of water or water birth.
Bovbjerg et al. (2022) did not provide definitions for third or fourth stages of labour.
| Outcomes | Risk with birth out of water | Risk with water birth | Risk difference |
|---|---|---|---|
|
Neonatal unit admission (respiratory distress syndrome): nulliparous and multiparous women (Bovbjerg et al. 2022; follow-up period not reported) |
About 16 babies out of 1,000 |
About 15 babies out of 1,000 |
About 1 fewer baby out of 1,000 whose mother had a water birth would be expected to be admitted to a neonatal unit because of respiratory distress syndrome. So for about 999 babies out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Neonatal unit admission (respiratory support): nulliparous and multiparous women (Sanders et al. 2024; follow-up period not reported) |
About 15 babies out of 1,000 |
About 8 babies out of 1,000 |
About 6 fewer babies out of 1,000 whose mothers had a water birth would be expected to be admitted to a neonatal unit for respiratory support. So for about 994 babies out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Neonatal unit admission (administration of intravenous antibiotics commenced within 48 hours of birth; prospective study population): nulliparous and multiparous women (Sanders et al. 2024) |
About 29 babies out of 1,000 |
About 18 babies out of 1,000 |
About 7 fewer babies out of 1,000 whose mothers had a water birth would be expected to be admitted to a neonatal unit for intravenous antibiotics commenced within 48 hours of birth. So for about 993 babies out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Neonatal unit admission (administration of intravenous antibiotics commenced within 48 hours of birth; National Neonatal Research Database): nulliparous and multiparous women (Sanders et al. 2024) |
About 29 babies out of 1,000 |
About 18 babies out of 1,000 |
About 9 fewer babies out of 1,000 whose mothers had a water birth would be expected to be admitted to a neonatal unit for intravenous antibiotics commenced within 48 hours of birth. So for about 991 babies out of 1,000 the outcome was the same irrespective of the method of birth. |
RR: risk ratio.
Notes: risk with birth out of water and water birth were based on raw data from the studies, and hence there is a discrepancy with risk difference calculated in GRADE, which considers the relative effect adjusted for confounding factors in the calculation.
Risk difference has been calculated in GRADE, which takes into account relative effect adjusted for confounding factors in the calculation (for example, risk difference out of 1,000 = 1,000 × control group risk × (1−RR)). Confounding factors are things such as maternal age, parity, ethnicity, birth weight, and socioeconomic status (see evidence review Q for full details) which may affect the results for different outcomes after birth out of water or water birth.
| Outcomes | Risk with birth out of water | Risk with water birth | Risk difference |
|---|---|---|---|
|
Snapped umbilical cord (cord avulsion at birth): nulliparous and multiparous women (Sanders et al. 2024) |
About 3 babies out of 1,000 |
About 10 babies out of 1,000 |
About 8 more babies out of 1,000 whose mothers had a water birth would be expected to have snapped umbilical cord (cord avulsion at birth). So for about 992 babies out of 1,000 the outcome was the same irrespective of the method of birth. |
|
Snapped umbilical cord (cord avulsion at birth): nulliparous and multiparous women (Bovbjerg et al. 2022) |
About 4 babies out of 1,000 |
About 6 babies out of 1,000 |
About 2 more babies out of 1,000 whose mothers had a water birth would be expected to have snapped umbilical cord (cord avulsion at birth). So for about 998 babies out of 1,000 the outcome was the same irrespective of the method of birth. |
RR: risk ratio.
Notes: risk with birth out of water and water birth were based on raw data from the studies, and hence there is a discrepancy with risk difference calculated in GRADE, which considers the relative effect adjusted for confounding factors in the calculation.
Risk difference has been calculated in GRADE, which takes into account relative effect adjusted for confounding factors in the calculation (for example, risk difference out of 1,000 = 1,000 × control group risk × (1−RR)). Confounding factors are things such as maternal age, parity, ethnicity, birth weight, and socioeconomic status (see evidence review Q for full details) which may affect the results for different outcomes after birth out of water or water birth.
Further outcomes
In addition to the outcomes from Tables 1, 2 and 3, there were a number of outcomes where the evidence identified showed there was no difference between water birth and birth out of water and where there was conflicting evidence about the risk with water birth or birth out of water (see evidence review Q for full details).