At this point in the water birth series, we know there is no evidence in the comparative literature that infants born in the water have worse outcomes than those born through conventional hospital delivery. Our next question was about the decisions being made about who can and cannot use hydrotherapy.
An interesting phenomenon has happened with hydrotherapy in maternity care – and that is, we treat hydrotherapy differently than other pain management options. For example, one researcher in Australia looked at the informed consent procedures and paperwork for water birth in comparison to receiving an epidural. Though no one would deny that epidurals have risks, we accept that these risks can be safely managed with good care – and this comes across in the informed consent procedures that did not require the women to read the epidural policy to have an epidural and the pamphlets used to educate patients focused on the safety of the method. In contrast, women were required to read the water birth policy and agree to it to have a water birth. The pamphlets used to educate patients focused on the risks of being in the water.
In the United States, we have diverse interpretations of risk data that allow us to ignore the rare but severe adverse neonatal effects of analgesia and anesthesia (which offer no neonatal benefits); but interpret the same types of outcomes to argue for a blanket refusal for water birth. So I wondered, was there any evidence in the studies that could help us identify which women we could all agree were “safe” to be in the water. This turned out to be more difficult than we expected.
Who is “safe” for water birth?
As part of the meta-analysis, we reviewed the information in each study about who was, and was not, eligible for a water birth. We saw some trends in the research literature that can provide some evidence for basic guidelines.
There were some standard items – most studies limited hydrotherapy to women with singleton, term pregnancies, with cephalic presentation. These standards were also seen in the four water birth policy documents I reviewed. There is physiologic evidence to support the gestational age requirement because the dive reflex needs to be functioning to prevent aspiration. I was not able to find evidence that supported restriction to singleton, cephalic deliveries. These restrictions are likely a product of limiting water birth to the women with the least known risk instead of the scientfic comparison of the effects of the restriction.
There were ambiguous definitions about who was not eligible to use the water. The most common exclusion criteria was that the woman not have any pregnancy complications. And so the reader of the study has to interpret what it means for a woman to be at “increased risk” or to be “low risk”. This means the reader has to decide for herself what counts as pregnancy complications.
This might not be as easy a task as you initially suspect. For example, though women with pregnancy complications were already excluded, several studies also specifically excluded women with HIV, or other contagious diseases. This implies that these conditions were not included in the “pregnancy complications” category.
To make it even more confusing, some studies excluded both pregnancy complications and pre-existing conditions. But again, no defining of what those pre-existing conditions may be. While we may like to think any midwife client should meet the standard for low risk, midwives are often a part of clinical teams providing care for moderate and high risk women.
Another interesting criterion to highlight is the restriction of women who have had a prior cesarean, which again was apparently not excluded by defining women as “low risk” and not considered a pregnancy complication. But on a policy level, this is interesting because if our goal is to promote trial of labor after cesarean and reduce cesarean rates, we should be promoting policies and interventions that improve the chances of a successful VBAC.
We know that hydrotherapy is associated with increased odds of spontaneous delivery. This should give us reason to believe hydrotherapy can help women be successful with VABC. Importantly, the eligibility criteria for a trial of labor after cesarean in most facilities is similar to the eligibility criteria for hydrotherapy – which means there is likely no “risk” reason these women are restricted from hydrotherapy – it goes back to the problem of ambiguous definitions of risk.
But it also points to a bigger problem in the way we think about our hospital policies. We tend to think of them as independent with each one directing a specific part of care; but the reality is the care overlaps. If your facility is trying to increase rates of VBAC, your trial of labor after cesarean policy should be linked to all the policies about helping women achieve a physiologic birth so these practices can be promoted and your facility can reach the goals you’ve sent for successful VBAC.
Most of the studies that provided eligibility criteria also listed conditions under which a woman would be asked to leave the water. Most common for these were maternal fever, meconium stained fluid (sometimes differentiated as thick but not always), and any unusual or non-reassuring fetal heart tones. These conditions make sense given the physiology of the dive reflex because we know that hypoxia causes a dysfunction in this reflex so the neonate is not protected from taking that first breath while under water.
Do the policies make water birth safer?
We wondered if having these criteria increased the safety outcomes but we couldn’t actually test studies that did and did not have the criteria because, with a few exceptions, the studies didn’t tell us if they didn’t have a requirement. Instead it just wasn’t listed. If we assumed all studies that didn’t list specific criteria didn’t have those criteria, we would have accidentally grouped together studies that did not have the criteria, and studies that had the criteria but did not give us details about the criteria.
Instead, we had to use a sensitivity analysis technique to try to answer the question. To do this, we reran the analysis including only the studies that specifically told us they use a particular criteria. In this way, we could compare the results of the sensitivity analysis to the full analysis to see if it changed what we found. Ideally, any criteria that improved the outcomes could move the result in favor of waterbirth.
We first tested criteria that must be met for the women to be able to get into the water. Remember, every study reported different outcomes, so we could only report on outcomes that had at least two studies that met the criteria and also reported that outcome.
Criteria 1 – the woman must show a progressive labor pattern. This was sometimes described by papers as having to be in active labor. We found no difference for any of the outcomes when this restriction was applied.
Criteria 2 – the woman must have intact membranes. We found this to be a strange requirement, especially since meconium stained fluid was so often an exclusion criterion. This was a requirement for four studies, and we believe they may have used it only for the research (to ensure similar samples) but not necessarily for practice, but there is no way to know. We can contrast it with those studies that obviously did not require intact membranes, we counted a study in this category if “clear fluid” was a requirement for entering the pool. But still, either way, we restricted the data didn’t alter the outcome of no difference in neonatal outcomes.
The third criteria we tested was that an infant must be judged to be a normal fetal weight for gestation. While this outcome was associated with a favorable outcome for neonatal hyperthermia, remember that is the finding we had in the main study that favored water birth but by such a small amount you would not be able to tell the difference when you took an infants temperature. So this is not evidence of an improvement with this requirement.
Finally, we tested for the exclusion of prior cesarean. The results for this analysis were identical to the main analysis with both neonatal hypothermia and NICU admission favoring water birth.
We were able to test one criterion for removing a woman from the tub, meconium stained fluid. Most studies were not specific about whether it was any meconium or thick meconium, so we grouped them together. There was no difference in any of the restricted analyses to the full analyses. There are two ways to interpret this. The first is that, although the other studies didn’t specifically mention these criteria, they also follow them so the overall findings are the same as the restricted findings; or the other studies may not specifically mention these criteria and probably don’t follow them, but it doesn’t matter whether you follow them because the result is the same if you do or do not.We cannot tell from the study which interpretation is correct.
What about how to conduct a water birth?
We didn’t have a lot of information about the practices used by the midwives and obstetricians when conducting a water birth, but we could check one variable that seemed to make sense – the requirement that the infant be removed from the water immediately, within 10 seconds. Of the studies that gave us a time frame, only 2 gave a time longer than 10 seconds, and those were within 45 seconds to a minute. There was no difference from the main study when we restricted to this protocol. So again, we wonder, were the other studies doing this anyway, or did the timing not make a difference?