Normal childbirth and evidence based practice
Article Outline
- Summary
- Introduction
- Normal childbirth
- Evidence-based practice
- Caesarean section
- Epidural analgesia during labour
- Ultrasound examination during pregnancy
- Patient request
- Anxiety
- Evidenced-based practice would increase the percentage of normal births!
- References
- Copyright
Summary
This paper was presented at a Health Conference in March 2007, celebrating the 150th birthday of the Royal Women's Hospital in Melbourne. It discusses the definition of “normal childbirth”, and the pros and cons of three medical technologiesa: caesarean section, epidural analgesia during labour and routine ultrasound screening during pregnancy, and whether clinical practices, in Australia and Sweden (author is Swedish), in relation to these methods are evidence based. It also discusses the impact of non-scientific reasons, such as anxiety, on clinical decision making.
Keywords: Evidence-based practice, Childbirth, Pregnancy, Caesarean section, Epidural pain relief, Ultrasound screening
Introduction
Throughout history, childbirth has taken place in a variety of contexts, related to economic development, social class and culture. The most dramatic changes in modern societies have occurred in recent times, when childbirth moved from home to hospital, and with the development of pharmacological methods to relieve labour pain, to induce labour and to speed up labour. The most remarkable technical development is the ultrasound machine, which allows us to look into the expectant mother's womb and see the unborn baby. The most remarkable change, from a biological perspective, is the increasing number of babies who come into this world via their mother's abdomen, and not her vagina. In the light of this development, one may ask: “What is normal childbirth today?”
Normal childbirth
Defining what is normal has been important to midwives, since our professional responsibility, in most countries, has been limited to the management of normal childbirth. The debate about normality has also been triggered by the development of medical technology. However, agreeing on a definition is not an easy task, since “normal childbirth” may be defined as something between “natural” and “common”. Some argue that a normal birth is a natural birth without any form of intervention, whereas others include as normal also a vaginal birth with epidural pain relief.
If normal is the absence of any intervention, only a small minority of women in Australia and Sweden has a normal birth. If, on the other hand, normal is defined as “common” practice, then, caesarean section is normal childbirth in some Latin American countries,1 and homebirth is normal from a global perspective.
“Normality” differs between cultures and changes over time. The reason why the use of a specific method is compatible with normal childbirth and another is not, may depend on
If midwifery is limited to normal childbirth, then the varying interpretation of “normality” is a problem. In clinical practice, however, the boundaries of midwifery practice have changed, and expanded with the development of medical technology. In Sweden, for example, where only about 10% of the births are natural, without any form of intervention, midwives care for about the same number of women as before, but now in closer collaboration with our medical colleagues. The question is then if we can still say that midwives are only responsible for the management of normal labour and birth.
One may ask whether it is important to define what is normal. Should not the focus be on how we can obtain best possible outcomes for mother and baby? Medical technology has made childbirth safer, and probably also a better experience than, for instance, hundred years ago. The course of nature alone would cause much higher rates of maternal and infant mortality and morbidity than modern obstetric care.
From a theoretical point of view, a simple solution would be to abandon the word “normal” in the context of modern childbirth, and only to talk about “evidence-based” practice, or “appropriate” level of medical technology. However, this would be problematic in cases when research evidence is lacking, or not possible to obtain. Also, the use of medical technology in modern obstetrics is not always driven by scientific evidence and rational arguments, but by other needs and interests. In this complex clinical context it is helpful to have a standard, against which different interventions are weighted.
This standard should be nature—a natural birth without medical interventions. Even if nature can be cruel in a few cases, it is mostly superior to anything that human nature has invented. When we introduce new methods, it is up to us to prove that they can do better than nature itself.
When it comes to the word “normal”, I believe that we need to be pragmatic and continue to use it—but then with a stronger emphasis on “natural” than on “common”. This interpretation also guides the recommendation by the World Health Organisation:
“The aim of the care is to achieve a healthy mother and child with the least possible level of intervention that is compatible with safety. This approach implies that: In normal birth there should be a valid reason to interfere with the natural process.”
However, if we define normal as “natural” in the clinical context, we still have the problem of how to define a vaginal birth with epidural pain relief in front of the woman. Has she had a normal birth or not? I am sorry to say that I do not have a good answer. Maybe we have to accept a double standard—to the woman: yes, you had a normal birth; to colleagues: no, she did not because she had an epidural and a syntocinon drip.
In Australia, only 59% of the women had a spontaneous vaginal delivery in 2004,2 and a large proportion of these women received epidural pain relief. This means that only a minority of women today have a normal birth, according to my “clinical” definition. The question is whether this is a problem or not. If the reduced number of normal deliveries means healthier mothers and babies, and better birth experiences, then the current situation should not be a problem. But is that the case? What does the evidence say?
Evidence-based practice
Regardless of our views about normality, there is today a growing awareness among health care professionals that clinical practices should be evidence-based. The emphasis on evidence may be looked upon as a new paradigm, replacing the traditional medical paradigm which is based on authority. A number of definitions are given on the Internet (Definitions of Evidence Based Practice, Web page from ScHARR). What they all have in common is that decision making should be based on the best evidence available. Best evidence is the result of a process of systematically finding, appraising, and using contemporaneous research findings. This process is dependent on the use of randomised controlled trials and systematic reviews, although not restricted to these.
In the following I will give three examples which illustrate the complexity of evidence-based practice: the increasing rates of caesarean section and epidural analgesia, and ultrasound screening during pregnancy. In all cases, there is evidence showing both pros and cons. And in all cases, clinical practice seems to be guided more by the pros than by the cons. These examples illustrate that medical technology is driven, not only by research findings, but also by a wide range of other factors. In order to understand clinical decision making, and make evidence-based practice a reality, it is also important to explore these other factors.
Caesarean section
Starting with my first example, there is a tendency for the caesarean section rate to increase more or less continuously in many countries. Figure 1 shows the most recent national figures from Australia2 and Sweden (Swedish Medical Birth Register3) which are from 2004, and the development over a 10-year period.
Fig. 1.
Caesarean section rates in Australia and Sweden: 1995–2004.
The figure illustrates national differences in relatively homogenous populations.
The appropriate caesarean section rate is widely debated. The WHO recommends 15%. On what evidence is such a recommendation based? For understandable reasons it is not easy to conduct randomised controlled trials where women are allotted a caesarean versus a vaginal birth on a random basis. This has only been possible when dealing with a specific clinical problem, such as breech delivery. We have to rely on observational studies.
When weighing up the pros and cons of caesarean section it is easy to identify the pros. Caesarean section may save infant and maternal lives and reduce morbidity. The operation is performed on several indications, such as fetal distress, multiple pregnancy, prematurity, malposition of the fetus, dystocia, placenta praevia, previous CS or on maternal request due to fear of childbirth. CS prevents perineal trauma and, to some extent, also stress incontinence after childbirth.4 The cons have not been paid the same attention, at least not in the public debate. Even if CS has become a much safer procedure than before, the procedure as such is still less safe than a vaginal delivery.
The Center of Epidemiology at the Swedish Board of Health and Welfare has published a study of all births between 1990 and 2001, altogether over 1 million.5 During this period the total CS rate increased by 60%. An increase was noted among multiple pregnancies, premature births (<37 weeks, singleton), babies in breech position (at term, singleton), and in the largest group of pregnancies, which were at term with only one baby, presenting the head first. The risk of different complications over the time period was examined in this large group of relatively normal pregnancies. Changes in the population during the 11 years, such as: increased maternal age, body mass index and infant weight; reduced smoking; a larger proportion of migrants; and variation in parity; were adjusted for in the statistical analyses. The following complications were more common in the CS group compared with the group of women with a vaginal delivery: extensive haemorrhage, endometritis, tromboembolic complications and uterine rupture in next pregnancy. The risks associated with the caesarean section as such, compared with a vaginal delivery were:
>1000ml: 1 in 12 CS.
No association was found between the CS rate and the use of vacuum extraction or forceps. No association was observed between the general CS rate and the risk of perinatal mortality. The author concluded that “it seems as if a large proportion of the dramatic increase has not been motivated by medical reasons”.
When summing up the pros and cons, my interpretation of the evidence is that the CS section rate, which in Sweden is now approaching 20%, is not evidence-based.
Epidural analgesia during labour
My second example which illustrates the complexity of evidence-based clinical practice is epidural analgesia during labour. Figure 2 shows the Swedish figures, with a continuous increase from the beginning of the 1970s, when the method was introduced—with the exception of a reduction during the 1980s, when “natural childbirth” was widely discussed in the media and alternative models, such as in-hospital birth centres, opened.3
Fig. 2.
Epidural analgesia in nulliparous and multiparous women in Sweden (vaginal births).
Australian epidural rates differ between the private and public sector.2 Primiparas in private care have epidural pain relief to the same extent as primiparas in Sweden, whereas the epidural rate is lower in public patients in Australia.
As with many other new methods, epidural analgesia was introduced before the method was thoroughly evaluated by means of randomised controlled trials. However, today the Cochrane Library includes 21 trials, which compare epidural analgesia with other methods of pain relief, or none at all. The systematic review shows that epidural analgesia is much better than non-epidural to relieve pain during labour.6 In spite of this finding, no difference was found between the randomised groups in women's satisfaction with the method of pain relief. Considering that the major aim of pharmacological pain relief is to reduce pain, and in this way improve women's overall experience of childbirth, it is surprising that only half of the 21 studies measured women's subjective perception of pain, and only one (!) measured women's experience of childbirth. No differences were found, either in these outcomes, or in women's perception of control during labour.
As with most medical procedures there are also some disadvantages, such as a longer second stage and an increased risk of instrumental vaginal delivery. More serious complications are rare, but may occur. In a study of 200,000 cases of epidural analgesia in Swedish obstetric care, serious neurological complications occurred in 1 out of 25,000 deliveries, such as spinal haematoma, epidural abscess, permanent abducense paresis.7 However, the major concern among anaesthetists that I have met is the risk of post-spinal headache, which may be very painful and last for days and even weeks after the delivery. The prevalence in Sweden is 0.5%.
When weighing up the pros and cons of epidural analgesia, it is necessary to consider that expected pain during pregnancy, perceived pain during labour, and memory of pain after the birth are slightly different things. During pregnancy, it may be reassuring that effective pain relief, such as epidural analgesia, is available. During labour, an epidural block may alter a woman's experience, from unbearable suffering to pain-free harmony, at least soon after the administration of the epidural and some time ahead. It is more complicated to assess the overall experience of labour and birth after the delivery. I have found no study showing that epidural analgesia protects against a negative birth experience—which can only be measured after the birth. In fact, some studies show a negative relationship.8 These findings have been explained in different ways. They may depend on the negative side effects of epidural analgesia, such as a longer second stage or an instrumental vaginal delivery. They may be caused by disappointment of not having met one's own expectations of a natural birth. They may have physiological causes related to decreased levels of oxytocin after the birth caused by the epidural.9 Oxytocin is a hormone associated with feelings of wellbeing and also with memory.10, 11
A review of the literature of women's experience of childbirth shows that support from the caregiver during labour is more important than pharmacological pain relief.12
When considering these different aspects of epidural analgesia one may ask if a rate of 50% in vaginal births of first-time mothers in Sweden,3 and 33% in Victoria,2 is evidence-based practice. We can also ask how many women are given continuous support in labour, as an alternative to pain relief. In contrast to epidural analgesia, continuous support has a documented positive effect on women's overall experience of the birth.13
Ultrasound examination during pregnancy
My third example illustrating the complexity of evidence-based practice is ultrasound screening during pregnancy. In Sweden and Australia, almost all pregnant women have at least one ultrasound scan during pregnancy. In this context I will only refer to examinations offered as screening to all pregnant women, not to examinations on medical indication. In Sweden, the initial aim of the routine ultrasound scan, offered in gestation week 17–20, was to measure fetal growth in order to estimate the date of delivery, detect multiple pregnancy, and localise the placenta. Screening for fetal malformations was not part of the aim, and this was also a controversial topic at the time. However, with the development of the ultrasound machine, the picture of the baby became more detailed, and in 1998 the majority of centres examined the anatomy of the fetus. Today, fetal screening is an established part of the examination in Sweden, as in many other countries.
What is the evidence regarding the effects of ultrasound screening in early pregnancy? The Cochrane Library includes a review of nine randomised controlled trials.14 The findings suggest that there are some advantages, such as earlier detection of twins, but most of all, fewer inductions for post-term pregnancy, due to more accurate estimation of the date of delivery. However, it has not been possible to show that these advantages affect infant or maternal health. The routine scan does not reduce the perinatal mortality rate, except in a study where terminations increased because of detected fetal abnormality.
Regarding the issue of safety, no negative short-term effects on infant growth or Apgar score have been found. Long-term follow-ups have been conducted of two trials, one in Sweden and one in Norway, and no adverse effects have been reported.15 The only concern was an increase in non-right-handedness in boys, and especially if the fetus was actually exposed to ultrasound. Two independent cohort studies have subsequently confirmed the same findings,16, 17, 18 which means that, of the four studies that have investigated the effect of ultrasound exposure during pregnancy on handedness, all have produced the same results. These findings have been controversial because no other explanation has been presented other than the possibility that ultrasound may affect the fetal brain.
The potential adverse effects of ultrasound exposure are poorly researched. A search in PUBMED showed that 2.4% of the publication about “ultrasound+pregnancy” were about fetal side effects, whereas 52% of the publications about “oral contraceptives” were about side effects.
Another problem is that the only long-term follow-up studies of ultrasound screening which are published today, evaluate the effect of ultrasound during a period when the machines were less powerful. High-energy ultrasound can induce biophysical effects when passing through tissue, for example, thermal effects and mechanical stress, causing cavitation. Modern equipment can produce energy intensities substantially greater than in the older machines, sometimes nearly eight times higher.19 Another change over time is that the responsibility for potential bio-effects, by regulating the energy output of the machine, has been transferred from the manufacturer to the health care professional who uses the equipment. The demands on the person who performs the scan have therefore increased. In this context, it was worrying that a survey of 199 doctors, sonographers and midwives trained in ultrasound, all of them using diagnostic ultrasound on a daily or weekly basis, had such problems giving correct answers to questions about safety. Only 22% gave a correct answer to the question: “How do you adjust the energy output on your own machine?”.20
Altogether, there seems to be common agreement that the issue of handedness needs further investigation. A Swedish research group concluded that “present knowledge is insufficient to state that the advantages of routine scanning outweigh the disadvantages”.17
Patient request
My interpretation is that there is insufficient scientific evidence to support current practices in our two countries regarding the use of caesarean section, epidural analgesia during labour and routine ultrasound screening during pregnancy. I could continue with additional examples from Sweden, such as the use of electronic fetal monitoring and syntocinon stimulation during labour, both of which are procedures that are used more frequently than recommended by our national guidelines for normal births.
Besides the risk of causing unnecessary harm and increase costs, non-evidence-based practices may reduce the number of normal births. Therefore, it is important to study the process of clinical decision making.
My three examples illustrate that the weighing of advantages against disadvantages is far from straightforward in clinical practice. Research findings need to be interpreted, and integrated assessments need to be made by professional groups that may have different interests. Even when scientific evidence is available, and the results are straightforward, they may not necessarily guide practice.
One popular explanation, for example of the high caesarean section rates, is patient request. Since clinical decision making should be based on the best evidence available, in consultation with the patient, current practices could possibly be justified by greater consideration of patient opinion. However, when this issue was investigated in Sweden, it was shown that even if there was an association between women's wishes in early pregnancy and subsequent mode of delivery,21 maternal request explained only 9% of the total increase in the caesarean section rate.5 A comparison between hospitals indicates that the views of doctors and midwives are more important. The caesarean section rate in Swedish university hospitals varies from 12.5 to 25%, and this difference is difficult to explain by the selection, or wishes, of the women.3 In Australia there is a difference in caesarean section rates between states, varying from 27% in Tasmanisa to 32% in Western Australia.2
Similarly, the epidural rate in women giving birth for the first time varies from 11 to 46%, when all hospitals in Sweden are included.3 Such a dramatic difference cannot possibly be explained by geographic variation in women's ability to cope with pain, or in their attitude to pain relief. I do not have national statistics of epidural rates in Australia, but a comparison between eight non-tertiary public hospitals in Victoria, with 1000 or more births per year, all of which have access to anaesthetists, showed a difference in primiparas ranging from 18 to 39%.22
Regarding the use of ultrasound during pregnancy, there is no national statistics in Sweden. However, also in this case, I guess that current practices in our two countries are driven more by the health care professionals than by the patients—even if ultrasound examination is a very popular procedure among expectant parents.
Regardless of how much patients and caregivers respectively influence the use of medical technology during childbirth, it is the responsibility of the caregiver to ascertain optimal use of scientific evidence in clinical practice. This also implies the responsibility to inform patients about advantages and potential risk associated with different interventions—in order for the patient to be able to make an informed choice. In the case of ultrasound examinations, my impression is that too little information is given about the uncertainties associated with the method.
Anxiety
Together with some colleagues I was interested in knowing the views of Swedish obstetricians about the current caesarean section rate, and took the opportunity to hand out a brief questionnaire at their annual conference in 2005. This conference was very well attended, and some midwives were also invited. One third of the participants gave their responses. Seventy-one percent of the obstetricians, and 92% of the midwives, said the CS rate in Sweden was too high.23 An open-ended question asked about possible reasons that could explain the increasing rate. The 418 different quotes given were analysed and grouped into categories as shown in Table 1.
Table 1. Obstetricians’ (166) and midwives’ (69) views about the reasons behind the increasing caesarean section rate in Sweden
| Reasons | Quotes (n=418), % of quotes |
|---|---|
| Related to the woman | 41 |
| Anxiety | |
| The woman's opinion has become more important | |
| Less tolerance of risk—belief that CS is safe | |
| Related to hospital staff | 21 |
| Anxiety | |
| Ignorance of evidence | |
| Lack of clinical competence | |
| Changes in attitudes to CS | 13 |
| In society, among staff and pregnant women | |
| Factors related to intrapartum care | 13 |
| Shortage of staff, lack of time, stress | |
| Monitoring, medical technology | |
| Large units | |
| Medical reasons | 12 |
Surprisingly, only 12% of the quotes were concerned with medical reasons. The most prominent single explanation was anxiety—both in women and doctors. I believe this is an important finding; an analysis should be made of the extent to which the use of medical technology during childbirth in general is anxiety driven, and whether something could be done about it. A British survey found that pregnant women had become more anxious.24 Twenty-six percent who were expecting their first baby were “very worried” by the thought of pain in labour in year 2000, compared with only 9% in 1987.
These findings make me think about cultural changes in our modern societies. Recently, a book was published in Sweden where a medical doctor argued that we have become addicted to security.25 He pointed at the paradox that so much focus is on protecting us from danger, during a period in history when we have never been safer. If pregnant women have become more afraid of labour pain, during a period in history when effective pain relief finally is available, then this description is relevant also to childbirth.
As a consequence of globalisation and the immense flow of information, and a more individualistic lifestyle, where family or other authorities are less important, we may have lost a certain degree of control, which makes us more vulnerable. Pregnant women today are exposed to much more information than any previous generation, and they have to make many more decisions. They are expected to be actively involved consumers, and at the same time deal with a life event that cannot be completely controlled. In this context, medical technology may be regarded as one way of getting control. Besides, pregnant women today have a more positive attitude to medical technology than for instance 20–30 years ago, when the natural childbirth debate was peaking.
Doctors and midwives are exposed to the same context, and the use of medical technology is one way of dealing with anxiety caused by the threat of litigation. When something goes wrong, it is always better to have done too much than too little. Besides, having a positive attitude to new technology is usually regarded as more “progressive” than to express scepticism because possible risks have not been ruled out.
Evidenced-based practice would increase the percentage of normal births!
Finally, back to normal childbirth! There are many good reasons for aiming at an increased rate of normal births. I have not discussed changes in the population that are threats to normal childbirth, such as increasing maternal age, and maternal weight. However, these are important public health issues that need to be addressed. My focus has been on the use of medical technology, and I have tried to illustrate that we still have a long way to go before maternity care in our countries is evidence-based. In some areas we still need more research, but in others we already have the necessary evidence and need to discuss the wide range of reasons that makes it difficult to apply them in practice. It is my conviction that evidenced-based practice would increase the percentage of normal births!
References
- . Rates and implications of caesarean section in Latin America: ecological study. BMJ. 1999;319:1397–1402
- . Australia's mothers and babies 2004. Perintal statistics series. No. 18. Sydney: AIHW National Perinatal Statistics Unit; 2007;
- Swedish Medical Birth Register, 1973–2005. Pregnancies, deliveries and newborn babies. National Board of Health and Welfare, Center of Epidemiology, Statistics; 2007:9.
- . Symptoms of stress incontinence 1 year after childbirth: prevalence and predictors in a national Swedish sample. Acta Obstet Gynecol Scand. 2004;83:928–936
- . Caesarean sections in Sweden 1990–2001. Stockholm, Sweden: Swedish Board of Health and Welfare; 2005;
- Anim-Somuah M, Smyth R, Howell C. Epidural versus non-epidural or no analgesia in labour. The Cochrane database of systematic reviews; 2005, Issue 4. Art. No.: CD000331.pub2. doi:10.1002/14651858.CD000331.pub2.
- . Severe neurological complications after central neuraxial blockades in Sweden 1990–1999. Pain Regional Anesth. 2004;101:950–959
- . A negative birth experience—prevalence and risk factors in a national sample. Birth. 2004;31:17–27
- . Plasma oxytocin levels in women during labor with or without epidural analgesia: a prospective study. Acta Obstet Gynecol Scand. 2002;81:1033–1039
- . Central functions of oxytocin. Neurosci Behav Rev. 1991;15:217–231
- . Intracerebral oxytocin is important for the onset of maternal behavior in inexperienced ewes delivered under peridural anesthesia. Behav Neurosci. 1992;106:427–432
- . Pain and women's satisfaction with the experience of childbirth. A systematic review. Am J Obstet Gynecol. 2002;186:S160–S172
- Hodnett ED, Gates S, Hofmeyr GJ, Sakala C. Continuous support for women during childbirth. Cochrane database of systematic reviews; 2003, Issue 3. Art. No.: CD003766. doi:10.1002/14651858.CD003766.
- Neilson JP. Ultrasound for fetal assessment in early pregnancy. The Cochrane database of systematic reviews; 1998, Issue 4. Art. No.: CD000182. doi:10.1002/14651858.CD000182.
- . Ultrasound during pregnancy and subsequent childhood non-righthandedness: a meta-analysis. Ultrasound Obstet Gynecol. 1999;13:241–246
- . Sinistrality—a side-effect of prenatal sonography: a comparative study of young men. Epidemiology. 2001;12:618–623
- . First trimester ultrasound scans and left-handedness (letter). Epidemiology. 2002;13:370–372
- . Ultrasound and adverse effects. Ultrasound Obstet Gynecol. 2002;20:102–103
- . Effects of repeated prenatal ultrasound examinations on childhood outcome up to 8 years of age: follow-up of a randomised controlled trial. Lancet. 2004;364:2038–2044
- . The output display standard: has it missed its target?. Ultrasound Obstet Gynecol. 2005;25:211–214
- Hildingsson I. How much influence do women in Sweden have on caesarean section? A follow-up study of women's preferences in early pregnancy. Midwifery, 2006; Dec 28 [Epub ahead of print].
- Davey MA. Consultative Council on Obstetric and Paediatric Mortality and Morbidity, and Perinatal Data Collection Unit, personal communication.
- . Obstetricians and midwives: the frequency of Caesarean section is too high. Swedish Med J. 2007;12:946–949
- Green J, Baston H, Easton S, McCormick F. Greater expectations? Summary report. Mother & Infant Research Unit, University of Leeds; October 2003.
- . I trygghetsnarkomanernas land. Prisma: Sverige och det nationella paniksyndromet; 2006;
- a Definition: The word “medical technology” is used as a general concept including medical and diagnostic devices, drugs and other therapies.
PII: S1871-5192(07)00093-5
doi:10.1016/j.wombi.2007.08.004
© 2007 Australian College of Midwives. Published by Elsevier Inc. All rights reserved.
