Waterbirth is rare in Taiwan.
Most people assume the reason is equipment: pools are expensive, rooms are small, hospitals don't want to invest. Equipment plays a part. After several years of attending births here, what I notice is that the harder thing to find is someone who has done it before.
That observation has research behind it, and I have not seen it written about for Taiwan.
Availability. In our guide to giving birth in Taiwan we could verify three places with documented waterbirth. There is no official register, so that is not a complete list. A hospital that mentions waterbirth in a brochure or blog may not offer it today. Call and ask.
Language. The 16 questions further down only work if you and the provider understand each other precisely: temperature limits, when you would be asked to leave the pool, how an emergency is handled. Ask for the written protocol. Bring an interpreter or a doula who speaks your language. I work in Mandarin, English and Japanese.
Two terms to separate first
Most confusion starts here, and this step is the one people skip.
| Water immersion in Mandarin: 水中陣痛 |
You labour in the pool and leave the water as the baby is about to be born, then give birth on land. The evidence on pain relief is clearer. |
|---|---|
| Waterbirth in Mandarin: 水中生產 |
The baby is born under the water. The evidence is newer and thinner, and the debate is concentrated here. |
Many conversations go wrong from the first sentence because one person means immersion and the other means waterbirth. When you talk to your care team, start by asking: which one are we talking about?
The pool is inflatable and the mother arranged the room herself. | Photo: Milliya
Benefits: most numbers point the same way
Peer-reviewed A 2024 systematic review and meta-analysis in AJOG pooled 52 studies. Compared with land birth, waterbirth was associated with the following:
| Outcome | Direction | Data pooled |
|---|---|---|
| Postpartum haemorrhage | ↓ about 20% (OR 0.80, CI 0.68–0.94) | 21 studies, 149,732 pregnancies |
| Newborn aspiration needing resuscitation | ↓ about 40% (OR 0.60, CI 0.43–0.84) | 19 studies, 181,001 pregnancies |
| Newborn infection | ↓ about 36% (OR 0.64, CI 0.42–0.97) | 15 studies, 53,635 pregnancies |
| Low Apgar score | ↓ about 31% (OR 0.69, CI 0.58–0.82) | 21 studies, 165,917 pregnancies |
| NICU admission | ↓ about 44% (OR 0.56, CI 0.45–0.70) | 30 studies, 287,698 pregnancies |
| Maternal infection | No difference (OR 0.93, CI 0.76–1.14) | 10 studies, 113,395 pregnancies |
Peer-reviewed The largest single study is Bovbjerg and colleagues (2022, BJOG). It used propensity-score matching to compare 17,530 waterbirths with 17,530 land births, balanced on more than 80 variables. These were low-risk births in US community settings, home and birth centre. Hospitalisation in the first six weeks was lower for both mothers (aOR 0.87) and babies (aOR 0.95). Neonatal death was not higher (aOR 0.56, CI 0.31–1.0). The authors' own summary is that waterbirth is neither as harmful as some guidelines suggest nor as harmless as some advocates claim.
Evidence gap Nearly all of this evidence comes from observational studies, not randomized trials. Evidence Based Birth points out one built-in bias: when complications appear during labour, the mother is asked to leave the pool, so the babies who remain in the "waterbirth" group tend to be the healthiest. That can make waterbirth look better than it is.
On pain, Evidence Based Birth summarises one small randomized trial in which 57% of the waterbirth group said the pain was less than they had expected, against 28% on land. It also notes that trials of this kind are small and that many participants crossed over between groups. Treat it as a hint, not a finding.
Five common myths
Healthy full-term newborns have a diving reflex and do not start breathing under water. Three conditions matter: water close to body temperature (overheating can trigger underwater breathing), the baby is born fully under the water and brought up gently with the face upward, and once the face has touched air, the baby must never go back under.
In the pooled data, newborn aspiration needing resuscitation was lower in the waterbirth group.
An inflatable pool works, as long as it is brand new or thoroughly disinfected before use. The guidelines I read pair it with a purpose-made disposable liner that is thrown away after each use. An ordinary leisure pool that has been used before and not properly disinfected is not suitable.
Guideline More important: never use a spa-style tub with circulation pipes or a heater pump. Those pipes can harbour Legionella and Pseudomonas.
Guideline In the first stage, go by comfort, but never above 37.5°C. In the second and third stages keep it at 37.0–37.5°C.
Water temperature is only half of it. The mother's own temperature should be taken every hour, and use stops if it rises 1°C above her baseline. An overheated mother raises the baby's temperature too, which can cause a fast fetal heart rate and may trigger underwater breathing.
Peer-reviewed In Bovbjerg's study, uterine infection was slightly higher in the waterbirth group (aOR 1.25, CI 1.05–1.48), but this did not lead to more hospital admissions. In the 2024 meta-analysis, newborn infection was lower and maternal infection was no different.
This assumes strict cleaning, disinfection and staff protection. How well infection control is done matters more than whether you are in water.
Water can ease pain, help you relax and make it easier to find positions. It is not anaesthesia.
If you treat it as a painless alternative, the gap between expectation and reality will be large. It changes your relationship with the pain. The pain itself is still there.
The one risk that goes up
Put all the numbers side by side and almost every one moves in a favourable direction. One does not.
Peer-reviewed Umbilical cord avulsion: the cord snaps during the birth. Bovbjerg's study found an adjusted odds ratio of 1.57 (CI 1.37–1.82), and the 2024 meta-analysis pooled an odds ratio of 1.75 (CI 1.38–2.24). In absolute terms that is 0.57% of waterbirths against 0.37% of land births, about 20 more per 10,000.
The mechanism is simple: the baby is lifted out of the water too fast, often together with a short cord.
Signs: the water turns deep red at once, you hear a snap, and the tension on the cord disappears.
Response: clamp the baby's end of the cord immediately. If it is recognised and clamped at once, it is usually managed without lasting effects. The review Sidebottom cites found no reports of long-term adverse effects. In Bovbjerg's data none of the avulsion cases led to death, and in the cases Sidebottom's team reported there were no adverse outcomes. It is still not trivial: one review cited in that paper estimated that roughly a quarter of avulsions ended in NICU admission and about one in eight in a transfusion.
"Recognised at once" means someone in the room knows what to look for.
One study followed what happened after a change in procedure
Peer-reviewed Sidebottom and colleagues (2020, Obstetrics & Gynecology) studied eight US hospitals. Among 384 second-stage water immersion births, cord avulsion occurred in 0.8% (3/384), against none in matched controls. That difference was not statistically significant.
The detail is in the timing: all three cases were in 2014, the first year of the study. Two happened in the water and one when a mother carried her newborn to the bed before the cord had been cut. The hospitals changed their teaching: the cord is cut before the mother leaves the pool, and the baby, if stable, is handed to the partner.
No subsequent cord avulsions were seen in the subsequent 4 years of the study.— Sidebottom et al., 2020, Obstetrics & Gynecology
Source: Sidebottom AC et al. (2020) Obstetrics & Gynecology 136(4):707-715, full text.
Evidence gap No randomized trial has tested a specific delivery technique against the avulsion rate. Advice such as bringing the baby up slowly, not pulling on the cord and cutting it before leaving the pool is clinical consensus and case-review level. It is not a proven result.
Sidebottom's study is also a before-and-after comparison within one set of hospitals. It is persuasive, and it cannot establish cause. Only one of the three cases is described as prompting the change, so the later drop to zero is encouraging rather than proof.
I mark this boundary because the next part of the article makes a stronger claim, and a strong claim needs to stand on firm ground.
Why your provider's familiarity is a safety question
One thing keeps showing up in my work: midwives and doctors in Taiwan and Japan often seem tense about waterbirth. I would not call it opposition. Most of them have simply never done it.
Not studied locally This is my own practice observation. I could find no research on this question from Taiwan or Japan. The international literature points in the same direction.
Peer-reviewed A 2024 mixed-studies review of 22 studies found that midwives were the most supportive of waterbirth, nurses next, and few physicians supported it. Reasons given for opposing it included insufficient training, lack of support from colleagues, concerns about work efficiency, and worries about safety and litigation (Chua et al., 2024, Journal of Advanced Nursing).
Insufficient training is the first reason listed.
Peer-reviewed A survey in northern Ontario (Orrantia & Petrick, 2021) points the same way. Of 362 women, 81.8% thought waterbirth was safe, as did 100% of the midwives surveyed, while none of the obstetricians did. The obstetrician group was very small, so the confidence interval is wide.
Source: Ulfsdottir H et al. (2020) Women and Birth 33(2):186-192.
Why clinicians hesitate
One number changed how I see this.
Peer-reviewed A study of midwives in the birth centre of a tertiary hospital in Western Australia (Lewis et al., 2018) found that after training, 93% (27 of 29 who answered) felt equipped to support waterbirth. The average number of waterbirths needed to feel confident was seven. It is a small survey in one unit, so read it as an illustration.
Seven.
Where waterbirth is uncommon, a midwife may take years to reach seven. That part is my inference. So a clinician who feels tense before the seventh is responding honestly to limited experience.
It is also why I think the useful question is "how many have you attended?", rather than "do you do waterbirth?":
How many waterbirths have you attended?
What to prepare
Guideline One thing many people do not know: there is no high-level international guideline written specifically for waterbirth. ACOG, RCOG and the RCM give principles. The operating detail, down to the centimetre and the minute, sits in the clinical guidelines of individual NHS trusts in the UK. Most of what follows comes from those documents. Where they disagree, I show both.
The pool
- There is no fixed depth. Guidelines say water up to the breastbone when you sit, or up to the nipple line with the belly fully covered. Depth is set by your body.
- An inflatable pool is acceptable if it is new or thoroughly disinfected. Guidelines usually add a purpose-made disposable liner, discarded after each use.
- No spa-style tubs with circulation pipes or heater pumps.
- For a fixed pool, run the tap before filling to flush the pipes (2 to 4 minutes depending on the unit), or for 1 minute if it has not been used for three days. For a home birth use a new or single-use hose.
- Fill the pool close to the birth. Do not let water stand.
Temperature and monitoring
| Water, first stage | Comfort-led, not above 37.5°C |
|---|---|
| Water, second and third stage | 37.0–37.5°C |
| Checking water temperature | Hourly in the first stage; every 15 minutes in the second stage (some units: every 30). Stir and measure in the middle of the pool |
| Mother's temperature | Hourly. Stop at a 1°C rise; leave the pool if >37.5°C twice or ≥38°C once |
| Baby's heart rate | Waterproof Doppler or wireless telemetry. Every 5 minutes in the second stage, listening for at least 1 minute straight after a contraction; feel the mother's pulse every 15 minutes to tell the two heartbeats apart |
| Room temperature | 23–24°C |
Who is a good fit, and who is not
Green light: 37–42 weeks, single baby, head down and engaged, spontaneous labour, clear waters, active labour, fetal heart rate 110–160 bpm, normal vital signs, BMI <35, intact skin, able to get in and out of the pool unaided.
Red light: serious medical conditions, blood-borne viral infection, placenta praevia, breech, multiples, preterm or post-dates, meconium-stained fluid, fever, active herpes, opioid pain relief within the last 2–3 hours or drowsiness, severely limited mobility. An epidural rules it out.
Amber (assess individually): BMI ≥35, GBS positive is allowed, with IV antibiotics given before or during immersion (the hand with the cannula is gloved and kept out of the water), history of third-degree tear, clotting risk, induction, high risk of postpartum haemorrhage, previous shoulder dystocia.
Guideline Guidelines disagree on previous caesarean (VBAC). Some units allow it (labour ward only, wireless fetal monitoring, normal CTG). Most do not list it as suitable. Discuss this point line by line.
When you will be asked to leave the water
This is the core of safety. Guidelines list more than ten reasons:
- Temperature >37.5°C twice (or ≥38°C once)
- Blood pressure 140/90 twice, 30 minutes apart, or 150/110 once
- Pulse persistently >100 bpm
- Abnormal or decelerating fetal heart rate, or it cannot be heard clearly
- Meconium-stained fluid
- Bleeding or clots
- Slow progress
- Need for pain relief beyond Entonox, an injection, or an episiotomy
- Once the baby's head has been exposed to air (no going back in)
- The baby is not born after one push
- Heavy bleeding in the third stage, or active management needed (most guidelines have oxytocin given out of the water or after leaving the pool; one Welsh guideline allows it in the water, with the leg lifted clear for the injection)
Technical points at the birth
- Hands off: verbal guidance, and no touching the baby under water, because touch can trigger a breath.
- No pulling on the cord. The baby is born fully under water, then brought up face upward, gently, straight away.
- Never return the baby to the water. Once the face meets air, breathing has started. Tell family beforehand. Well-meaning relatives sometimes want to "put the baby back where it was comfortable".
- Delayed cord clamping: never clamp or cut under water, which stimulates breathing. Hold the baby against your chest with the cord still attached, and deal with it out of the water once it stops pulsing.
- A practical note: blood loss cannot be estimated in water. It is recorded only as under or over 500 mL, and the mother's condition has to guide the decision.
In an emergency, how do they get me out of the pool?
People ask about temperature and cost. Hardly anyone asks about evacuation, yet guidelines differ on it. One Welsh protocol describes a team of at least five: the midwife, a helper to turn the mother, someone to ready the trolley and two to position the net. The principle common to the guidelines I read is to slide the mother out on a net and board and not to drain the pool, because buoyancy is what makes lifting her possible.
The same Welsh guideline wants two health care workers ideally available for a birth in water and asks that annual emergency evacuation training is evidenced. Milton Keynes includes evacuation drills in its routine skills sessions.
Also: every manoeuvre for shoulder dystocia has to be done out of the water.
Taiwan has no official waterbirth safety standard
Checked, none found I could find no dedicated waterbirth guideline or safety standard from the Taiwan Association of Obstetrics and Gynecology, the Taiwan Midwives Association, the Ministry of Health and Welfare or the Health Promotion Administration (checked August 2026).
In practice, each hospital or clinic that offers waterbirth follows its own internal rules.
As far as I could find, there is also no national competency certification or standardised training for waterbirth in Taiwan or in Japan. International courses exist, such as Waterbirth International's online certificate, but that is a course-completion certificate, not a licence to practise.
So asking "is the provider certified?" does not get you far here. Your main leverage is asking questions.
16 questions to ask when choosing a provider
You can print this page and bring it. Ask them in a language you both understand fully, or bring an interpreter.
About experience
(Whether the document exists is already an answer.)
About temperature and monitoring
About leaving the pool
(Ask them to list them. Someone who can name ten or more is usually doing it for real.)
About the pool and infection
(At a home birth, cleaning often falls to the family. Many people assume the midwife handles it.)
About the birth itself
(A "yes" is a red flag.)
About me
Some people worry that detailed questions will sound distrustful.
My experience is the opposite. People who really know waterbirth are glad when you ask. It shows you are serious, and that they have finally met a family who follows the details.
If someone is irritated by your questions or cannot answer them, that tells you something too.
To close
The evidence on waterbirth mostly points one way: less bleeding, less newborn infection, fewer babies needing resuscitation. The one risk that clearly goes up may well be one that training can address.
If that is right, how safe a waterbirth is depends in large part on how familiar the person with you is with it.
So "how many have you attended?" puts safety back into a place where it can be talked about.
And if the midwife or doctor around you seems hesitant, please do not read that as opposition straight away. The research suggests it takes about seven to feel confident. Their hesitation may be honest, and perhaps nobody has yet given them those seven.
Frequently asked questions
What is the difference between water immersion and waterbirth?
Water immersion means labouring in the pool and leaving the water to give birth on land. Waterbirth means the baby is born under water. The evidence is stronger for pain relief with immersion, and newer and thinner for waterbirth itself. Many conversations go wrong because each side means a different thing.
What are the benefits of waterbirth?
In a 2024 meta-analysis of 52 studies, waterbirth was associated with about 20% less postpartum haemorrhage, about 40% fewer newborn aspirations needing resuscitation, about 36% less newborn infection, about 31% fewer low Apgar scores and about 44% fewer NICU admissions. The largest single study (17,530 matched pairs) found lower hospitalisation for mothers and babies in the first six weeks. The evidence is observational.
Can the baby inhale water?
Healthy full-term newborns have a diving reflex and do not start breathing under water. What matters is water close to body temperature, bringing the baby up gently with the face upward right after a full underwater birth, and never returning the baby to the water once the face has met air. In pooled data, aspiration needing resuscitation was lower with waterbirth.
Is there a higher risk with waterbirth?
One: umbilical cord avulsion, with an odds ratio of about 1.6 to 1.75 in the main studies (0.57% against 0.37% in absolute terms, about 20 more per 10,000 births). It happens when the baby is lifted out too fast, sometimes with a short cord. In one US study of eight hospitals, all three cases were in the first year, and after the hospitals changed their teaching (cut the cord before leaving the pool) there were none in the next four years. That was not a randomized trial.
What should the water temperature be?
In the first stage, comfort-led but not above 37.5°C; in the second and third stages 37.0–37.5°C. The mother's temperature should be checked hourly and use stops at a 1°C rise. An overheated mother raises the baby's temperature, which can cause a fast fetal heart rate and may trigger underwater breathing.
What is the most important thing to ask a provider?
The question people ask least: "In an emergency, how do you get me out of the pool?" Guidelines differ on how many staff are needed (one Welsh protocol describes at least five), and the ones I read say not to drain the pool, since buoyancy is what makes lifting possible. The next is "how many have you attended?" A survey in Western Australia suggests about seven are needed before midwives feel confident.
I would like help asking these questions.
This is what I do. Whichever way you choose to give birth, I can help you look at the options, the risks and the preparation. See my services or send me a message. For places in Taiwan that have documented waterbirth, see our guide to giving birth in Taiwan.
Sources
Evidence and outcomes
- Bovbjerg ML et al. (2022) Maternal and neonatal outcomes following waterbirth: a cohort study of 17,530 waterbirths and 17,530 propensity score-matched land births, BJOG (link)
- Water birth: a systematic review and meta-analysis of maternal and neonatal outcomes, AJOG, 2024 (link)
- Cluett ER, Burns E, Cuthbert A (2018) Immersion in water during labour and birth, Cochrane (link)
- Sidebottom AC et al. (2020) Maternal and Neonatal Outcomes in Hospital-Based Deliveries With Water Immersion, Obstetrics & Gynecology 136(4):707-715 (PMC full text)
- Schafer R (2014) Umbilical cord avulsion in waterbirth, J Midwifery Womens Health (PubMed)
Clinicians' attitudes and training
- Chua JYX et al. (2024) Insights of healthcare professionals regarding waterbirths and water immersion during labour: A mixed studies review, J Adv Nurs (PubMed)
- Ulfsdottir H, Saltvedt S, Georgsson S (2020) Testing the waters – A cross-sectional survey of views about waterbirth among Swedish health professionals, Women and Birth 33(2):186-192 (PubMed)
- Lewis L et al. (2018) Midwives' experience of their education, knowledge and practice around immersion in water for labour or birth, BMC Pregnancy Childbirth 18:249 (PubMed)
- Cooper M et al. (2023) Understanding the barriers and facilitators related to birthing pool use, Reproductive Health (PubMed)
Practice standards and guidelines
- Evidence Based Birth: Waterbirth
- ACOG Committee Opinion No. 679 (2016)
- RCM, Evidence Based Guidelines for Midwifery-Led Care in Labour: Immersion in Water (2012)
- Milton Keynes University Hospital NHS, Water Birth Guideline (2022)
- Royal Devon University Healthcare NHS, Water Birth v1.6
- Wrightington, Wigan & Leigh NHS (2024); Royal Cornwall Hospitals Trust; NHS Shetland; Canterbury DHB (New Zealand): waterbirth clinical guidelines
Evidence labels. Peer-reviewed marks studies that went through academic review. Guideline marks formal clinical guidance from professional bodies or hospitals. Evidence gap, Not studied locally and Checked, none found mark gaps stated openly. Where I add my own inference, I say so in the text.
This article is a summary of information. It is not medical advice and does not replace a discussion with your doctor or midwife. Practice differs between hospitals, and birth carries risks that cannot be fully predicted. If anything unusual happens in pregnancy, seek medical care at once.