Caesarean section rates continue to increase worldwide. The appropriate caesarean section rate remains a topic of debate among women and professionals. Evidence-based medicine has not provided an answer and depends on interpretation of the literature. Overall caesarean section rates are unhelpful, and caesarean section rates should not be judged in isolation from other outcomes and epidemiological characteristics. Better understanding of caesarean section rates, their consequences and their benefits will improve care, and enable learning between delivery units nationally and internationally. To achieve and maintain an appropriate caesarean section rate requires a Multidisciplinary Quality Assurance Programme in each delivery unit, recognising caesarean section rates as one of many factors that determine quality. Women will always choose the type of delivery that seems safest to them and their babies. Professionals need to monitor the quality of their practice continuously in a standardised way to ensure that women can make the right choice.
Introduction
Caesarean sections continue to increase worldwide. No agreement has been reached on an appropriate caesarean section rate, and views are mixed on whether too many are being carried out. Many women enquire about caesarean section as an option for delivery, and a significant number request a caesarean section. Most women do not want an operation, they request a caesarean section because they do not want to labour and deliver vaginally. Nulliparous women request a caesarean section because they are worried about something that may happen. Multiparous women request a caesarean section because of something that did happen. National guidelines have reinforced the right of women to decide the mode of their delivery provided that they have been counselled appropriately. If an obstetrician disagrees with the woman’s decision to deliver by caesarean section, then she should be referred to an obstetrician who would be prepared to carry out the caesarean section.
Against this background there also seems to be a lower threshold of carrying out caesarean sections for medical reasons. These reasons include (1) different views on the management of labour and delivery, organisational issues; and (2) societal intolerance of poor outcomes and experience, and a culture of blaming individuals or systems (a significant concern among professionals). Caesarean sections are thought to be a procedure that protects both babies and mothers from adverse events. Although in some cases this is undoubtedly true, it needs to be continually justified and safely implemented. Extreme views on low or high rates of caesarean section are not helpful, especially if the arguments are based on selected evidence. An indifferent view on caesarean sections is not helpful as all procedures have their implications. Until women and professionals alike appreciate this, a polarised debate will continue that will do little more than confuse. In order to rationalise decision-making, more useful information is needed on a continuous and timely basis about the quality of care that is being provided in each delivery unit. Obstetricians and midwives may know less about events and outcomes in their own unit compared with their knowledge of published research. Professionals have a responsibility to practice evidence-based medicine but they should not forget their responsibility to collect the evidence to ensure that they are providing good quality care to their patients. Caesarean section rates have been at the centre of the debate for too long. Discussions about reducing caesarean section rates without taking other factors into account are at best inappropriate and at worst dangerous. The aim should be a Multidisciplinary Quality Assurance Programme (MDQAP) with women, babies and their family at the centre. Caesarean section rates are only one of many factors that determine quality. An appropriate caesarean section rate may change over time and vary in different delivery units. Ultimately, it will depend on the MDQAP that takes into account all the criteria used for assessing maternity care.
The purpose of this chapter is to reinforce and develop the principles that were described 10 years ago.
The Multidisciplinary Quality Assurance Programme
In order to achieve an appropriate caesarean section rate, the concept of an MDQAP needs to be implemented. This concept is described in Fig. 1 in the context of labour and delivery of the pregnant woman, and similar programmes have been suggested elsewhere. Quality assurance should be applied to the subject as a whole. Audit, classification of information, assessing management and modifying management, when applicable, should be applied to the processes involved in achieving it. All the above components are crucial to achieving quality, but the quality of information collection is paramount. At present, setting standards and benchmarking of interventions and outcomes are used as assessment of quality in a healthcare organisation. Good information collection itself must be the first quality standard. Information has to be easily available, quality controlled and validated. Four criteria will be used for the assessment of maternity care: level of interventions and outcomes (including safety), choice (experience), cost and efficiency. This philosophy can be extrapolated to the debate on caesarean section rates: ‘it is not that a caesarean section rate is high or low but rather whether it is appropriate or not, after considering all the relevant information’.
Audit
Audit is defined as the formal examination and recording of the results, and is divided into structure (representing resources), process (the way that resources are applied) and outcome (the result of intervention). Recently, more emphasis has been placed on auditing processes rather than outcomes, whereas patients are more interested in outcome. Quality is related to outcome, and outcome will guide processes. A more practical definition of audit is continuously looking at your outcomes in a standardised way at the most senior level on a regular basis, resulting in a formal written annual report documenting the quantity and quality of care.
High-quality audit has long been undervalued in developing and supporting clinical practice. The reason is that audit requires time and resource, but most of all discipline and leadership. The challenge from a practical point of view is to combine routine documentation of notes with audit and the ability to use same for teaching, education and research without duplication of effort. The information needs to be relevant, carefully defined, accurately collected, timely and available. Information collection needs adequate resources and meticulous organisation. A detailed description of Labour ward audit has been given elsewhere.
Labour events and outcome
Two main types of data are available for labour and delivery. First, epidemiological data, such as age, height, body mass index, medical conditions, ethnicity and other case-mix variables. Second, ‘interventions’, which refer to events (or outcomes) taken by professionals involved in the mothers’ care. Although these are carried out with the intention of improving care, many interpret them as interference with a normal physiological process. The difficulty with a generic term like ‘interventions’ is that no distinction is made in how the mother, midwife or medical staff perceive the particular event or outcome in question. Even more confusing is the fact that what may be an intervention to one woman may not be an intervention to another; indeed, it may be a desired event or outcome. In order to clarify matters, the term ‘intervention’ should be avoided. Instead all events that take place should be recorded whether they are processes carried out by professionals or occur as a result of the care provided. Some labour events are also labour outcomes, in that the mother, midwife or medical staff consider them to affect the health or satisfaction of either the mother or baby. All events and outcomes need to be defined in a standard way.
Caesarean section is a case in point. A caesarean section is an event that may take place in the process of labour and delivery. It may also be an outcome either negative or positive or indeed neither, depending on the circumstances of the delivery. Induction of labour, artificial rupture of membranes, use of oxytocin, and length of labour are other examples of events that may also be outcomes or may affect the incidence of other outcomes.
The most useful maternal and fetal information that needs to be collected is presented in Table 1 . Additional information would be helpful, but quality information should not be compromised and prioritisation of information to be collected is essential. The structure of information collection is important and, in particular, indications for procedures such as inductions and caesarean sections. It would be useful from an epidemiological point of view to classify all indications into fetal, maternal or no medical indication. This will not be easy and we need to look at ways in which this could be done.
| Maternal information | Fetal information |
|---|---|
| Age of women. Ethnicity. Booking weight and height (body mass index). Total number of women (to construct the 10 groups). Spontaneous labour. Inductions (fetal, maternal no medical reasons). Pre-labour caesarean section indications (e.g. fetal, maternal, no medical reason). Number of caesarean sections (to analyse distribution of caesarean sections). Number of caesarean sections in first-stage of labour (e.g. fetal, dystocia). Number of caesarean sections in second stage of labour (e.g. fetal, dystocia). Artificial rupture of membranes. Oxytocin (first stage). Oxytocin (second stage). Epidural. Vaginal operative delivery (ventouse or forceps). Duration of labour. Episiotomy. Third- or fourth-degree tears. Postpartum haemorrhage. Blood transfusion. Peripartum infectious morbidity. Peripartum hysterectomy. Days at facility for the mother. Maternal deaths. | Birth weight. Gestation. Apgar score (less than 7 at 5 mins). Cord pH (ph less than 7.0). Erb’s palsy. Encephalopathy. Admissions to intensive care unit. Admissions to intensive care unit over 24 h. Days at facility for the newborn. Stillbirths (less than 37 and 37 weeks or over). Intrapartum deaths. Neonatal deaths (7 weeks or less and 28 weeks or less). Cerebral palsy. |
The third type of information that is collected in labour and delivery is information used to classify the epidemiological data and the events and outcomes.
Caesarean sections: indications
The number of caesarean sections carried out can easily be recorded, but their indications have been difficult to define and implement consistently. A further problem is the increase in numbers of indications used and also the number used for each operation. Although clinically not an issue, this does present a problem for classification and obtaining an overview of care, why the procedures are being carried out and whether they can be justified in terms of other outcome. If an appropriate caesarean section rate is to be described, then indications for caesarean section have to be standardised. Pre-labour caesarean sections should be classified into fetal, maternal or no medical indication. If more than one indication exists, then one main indication should be chosen with the other indications added in a hierarchical manner.
A definition for no medical indication or maternal request is required. Practically, it may be best defined as ‘at the time of the request by the woman, in the opinion of the obstetrician there is a greater relative risk of a significant adverse outcome to mother or baby by carrying out a caesarean section than awaiting spontaneous labour and delivery or inducing labour’. The relevance in defining it in this way is that it places an onus on the delivery unit to make sure the relevant information from their own results together with external evidence is available to justify the use of the indication.
A medical indication for a caesarean section must be one that is used consistently in similar circumstances. Otherwise, the indication must be recorded as maternal request especially when the woman has requested it. This does not mean to say it is inappropriate care to carry out a caesarean section after counselling the woman, but only that it should be classified as maternal request and also includes the reason for that request. The most common example for this in clinical practice is the decision to deliver the baby of a woman who has had one previous caesarean section at 39 weeks who has no other medical reason for caesarean section. This should be recorded as maternal request. If this is recorded as a medical indication, then the logical extrapolation of this is that babies of all women with one previous caesarean section and no other medical indication for caesarean section should be delivered by caesarean section at 39 weeks. The same woman with one previous caesarean section reaching 41 weeks with an unfavourable cervix and delivered by caesarean section should be classified as a medical indication. Variances in the application of indications can be studied by analysing them in different groups of women. Importantly though, it is not inconceivable that an indication for caesarean section recorded as maternal request today may well, with change in practice and outcomes from labour and delivery, become a medical indication in the future and also vice versa.
The terms elective and emergency caesarean section are difficult to define and are rarely applied in a standard way. An elective caesarean section might best be defined as a planned procedure (greater than 24 h), carried out during routine working hours, at greater than 39 weeks, in a woman who is neither in labour or has had labour induced. All other caesarean sections would be audited as emergency or possibly more appropriately non-elective caesarean sections. The reasons why they were recorded as non-elective could be recorded using the reasons described above. This adds an organisational element as well as clinical to the definition of elective and emergency, and would be helpful in assessing an appropriate caesarean section rate.
Indications for caesarean sections in labour need to be simple, replicable and allow for improvement of care. Management of labour depends on ensuring fetal well-being and achieving efficient uterine action, and they are also the reasons why caesarean sections are carried out in labour. It is, therefore, logical that indications for caesarean sections in labour might be classified into fetal or dystocia so that management can be assessed. A fetal indication would be defined by convention when a caesarean section is carried out for suspected fetal distress (for whatever reason), but without the use of oxytocin. All other caesarean sections carried out in labour are classified as a form of dystocia. No formal definition of dystocia is suggested, as each delivery unit will have their own interpretation, but this will not preclude them from using the following classification. Rather, the sub-classification of dystocia will depend upon whether the progress in labour had been less than 1 cm/h (inefficient uterine action) or more than 1 cm/h (efficient uterine action). Inefficient uterine action is then subdivided into poor response (despite maximum treatment with oxytocin), inability to treat adequately (for fetal reasons), inability to treat adequately (because of the uterus over-contracting), or, lastly, no treatment (oxytocin not given because it is thought to be inappropriate, for example, in labour with a malpresentation, in a woman with a previous caesarean section, when a woman declines oxytocin or indeed declines labour itself).
This classification ( Table 2 ) differentiates between suspected fetal distress without oxytocin compared with suspected fetal distress after oxytocin was started, but when the primary problem was dystocia.
| Fetal distress (no oxytocin) | ||
|---|---|---|
| Dystocia | Inefficient uterine action less than 1 cm/h | Poor response. Maximum dose a reached. |
| Inability to reach maximum dose a because of fetal intolerance | ||
| Inability to reach maximum dose a because of over-contracting or not following unit protocol. | ||
| No oxytocin given. | ||
| Efficient uterine action over 1 cm/h | Cephalopelvic disproportion. | |
| Malposition (occipito posterior or occipito transverse). |
The distribution of the results in its use reflects the way that dystocia is diagnosed and how oxytocin is used in labour in the delivery unit. In particular, the incidence, timing, dose and regimen of oxytocin. Applying this classification to different groups of women gives different results that can be used to analyse caesarean section rates and their implications more rationally.
Classification of information
For the MDQAP to be successful, quality information is clearly needed, but as important is the need to classify and organise information so that it can be easily used by clinicians on a daily basis to assess and improve care.
Classification systems are used in medicine to transform crude data and information into useful information so that clinical care can be improved. They are based on the identification of different concepts that may each have several parameters. Different permutations of these parameters, and of their systematic arrangement, result in specific groups or categories that share some defined property feature or quality. The purpose of a classification system usually determines its structure, but the ideal classification will satisfy different purposes. The main groups of the classification must be robust enough to be unlikely to need changes. The groups or categories of the classification need to be prospectively identifiable so that outcomes can be improved in those same patients in the future. The groups or categories must be mutually exclusive, totally inclusive and clinically relevant. The classification system must be simple to understand and easy to implement.
Ten-group classification
The 10-group classification system (TGCS) complies with the principles of a classification system described above. If implemented on a continuous basis, it would allow the critical assessment of perinatal care leading to change if thought necessary. The obstetric concepts, with their parameters, used to classify the women in the TGCS, are the category of the pregnancy, the previous obstetric record of the woman, the course of labour and delivery, and the gestational age of the pregnancy. The concepts and their parameters are all prospective, mutually exclusive, totally inclusive, simple and easy to understand and organise ( Table 3 ).
| Obstetric concept | Parameter |
|---|---|
| Category of pregnancy | Single cephalic pregnancy. |
| Single breech pregnancy. | |
| Single oblique or transverse lie. | |
| Multiple pregnancy. | |
| Previous obstetric record | Nulliparous. |
| Multiparous (without a uterine scar). | |
| Multiparous (with a uterine scar). | |
| Course of labour and delivery | Spontaneous labour. |
| Induced labour. | |
| Caesarean section before labour (elective or emergency) | |
| Gestation | Gestational age in completed weeks at time of delivery. |
Importantly, they are clinically relevant to midwives and obstetricians because the information they depend on is required whenever an assessment is made of a pregnant woman who is either in labour or about to deliver. It therefore makes sense that all maternal and fetal information, as described in Table 1 , is viewed within these concepts and parameters or combinations of them, and the TGCS was formed as shown in Table 4 . They were chosen on the basis that they provide the best clinical and organisational overview relative to the number of groups. They allow a comparison to be made between delivery units, allowing more specific analysis of the labour events and outcomes, including their indications and epidemiological variables.
| Groups | Overall caesarean section (CS) rate (%) 1977/9250 (21.4%) National Maternity Hospital 2011 | |||
|---|---|---|---|---|
| Number of CS over total number of women in each group | Relative size of groups % | CS rate in each group % | Contribution made by each group to the overall CS rate % | |
| 1. Nulliparous, single cephalic, ≥37 weeks, in spontaneous labour | 179/2389 | 25.8 2389/9250 | 7.5 179/2389 | 1.9 179/9250 |
| 2. Nulliparous, single cephalic, ≥ 37 weeks, induced or CS before labour a | 475/1368 | 14.8 1368/9250 | 34.7 475/1368 | 5.1 475/9250 |
| 3. Multiparous (excluding prev. CS), single cephalic, ≥37 weeks, in spontaneous labour | 30/2751 | 29.7 2751/9250 | 1.1 30/2751 | 0.3 30/9250 |
| 4. Multiparous (excluding prev. CS), single cephalic, ≥37 weeks, induced or CS before a labour | 109/871 | 9.4 871/9250 | 12.5 109/871 | 1.2 109/9250 |
| 5. Previous CS, single cephalic, ≥37 weeks | 571/936 | 10.1 936/9250 | 61.0 571/936 | 6.2 571/9250 |
| 6. All nulliparous breeches | 204/219 | 2.4 219/9250 | 93.2 204/219 | 2.2 204/9250 |
| 7. All multiparous breeches (including prev. CS) | 113/133 | 1.4 133/9250 | 85.0 113/133 | 1.2 113/9250 |
| 8. All multiple pregnancies (including prev. CS) | 134/212 | 2.3 212/9250 | 63.2 134/212 | 1.5 134/9250 |
| 9. All abnormal lies (including prev. CS) | 35/35 | 0.4 35/9250 | 100 35/35 | 0.4 35/9250 |
| 10. All single cephalic, ≤36 weeks (including prev. CS) | 127/336 | 3.6 336/9250 | 37.8 127/336 | 1.4 127/9250 |
a Groups 2 and 4 are commonly divided into a (inductions) and b (prelabour caesarean sections).
Each of the 10 groups can and should be further subdivided when required. Groups 1 and 2 should be analysed separately and also together, as should Groups 3 and 4.
The philosophy of the TGCS in assessing maternity care is based on the premise that all epidemiological information, maternal and fetal events and outcomes will be more clinically relevant if first analysed within the 10 groups, their obstetric concepts or parameters. This is particular important in assessing caesarean section rates but also other perinatal outcomes. The TGCS can also be used to classify any group of women defined by data derived from Table 1 . For example, all women over the age of 35 years or different ethnic groups can all be classified into the 10 groups and analysed and compared with a standard population.
Classification of caesarean section and induction of labour
At present, no accepted classification system exists for caesarean sections. This is quite extraordinary considering the continuing volume of literature on the subject and the concern voiced by governments about the rise in the numbers of caesarean sections and possible implications on woman’s health. Many descriptive studies have been published, but no standard classification system has been used that fits the principles described above, and that has been used to make changes in specific prospective groups of women. Caesarean section rates have been analysed by comparing overall rates, by indication for caesarean section, by sub-groups of women and by primary and repeat caesarean section rates. They all have their disadvantages. Two national guidelines have been published on caesarean sections in the UK, but no reference was made to a classification of caesarean sections. The World Health Organization carried out a systematic review of classifications of caesarean sections and concluded that:
‘Women-based classifications in general, and Robson’s classification, in particular, would be in the best position to fulfill current international and local needs and that efforts to develop an internationally applicable caesarean section classification would be most appropriately placed in building upon this classification. The use of a single caesarean section classification will facilitate auditing, analyzing and comparing caesarean section rates across different settings and help to create and implement effective strategies specifically targeted to optimize caesarean section rates where necessary’.
The biggest single step to try and achieve and maintain appropriate caesarean section rates would be to agree a classification for caesarean sections and to use that classification in reporting for all delivery units.
The indications for caesarean section should be analysed within each group of women because the definition and management will vary in each group and will have different risk–benefit ratios. The TGCS can be used to assess any caesarean section rate in absolute terms, but also to compare with other lower or higher caesarean section rates either within the same delivery unit from previous years, or with other delivery units elsewhere. It would be possible to see how the sizes of the different groups vary, and also in which groups of women there is a difference in caesarean section rates. It will not immediately explain the reasons, and further analysis would be required, but it will allow a useful overview from which to start. From this, it will be possible to identify different groups of women and change the management according to available evidence.
The benefit of using the TGCS within the MDQAP is that the classification, although commonly used for analysing caesarean section rates, was originally devised so that all perinatal events and outcome, not only caesarean section rates, could be analysed within standard groups of women. Furthermore, it can also be used to analyse other quality indicators of maternity care, in addition to allowing for differences in case mix and other epidemiological variables.
Induction of labour, and the contribution it makes to caesarean section rates, remains a controversial issue. The TGCS allows a unique analysis of that contribution. The two groups of women that are relevant in the study of induction are single cephalic nulliparous women Group (2a) and single cephalic multiparous (without a previous scar) women Group (4a) ( Table 4 ). The denominator that is used to study the incidence and indications for the inductions is the total number of women in Groups 1 and 2, and Groups 3 and 4, respectively. Classifying the indications for inductions, initially at least, as shown in Table 5 , has proven useful to obtain an overview. More detail about each induction can be included in a hierarchical manner within these six groups.
| Fetal reasons |
| Pre-eclamptic toxaemia/hypertension |
| Post dates (42 weeks or over) |
| Spontaneous rupture of membranes |
| Maternal reasons and pains |
| Non-medical reasons or dates less than 42 weeks |
Assessment of management: interpretation of data
The use of the 10-group classification system to analyse a caesarean section rate
A simple stepwise way to analyse caesarean section initially is using the TGCS, as shown in Tables 4 and 6 in conjunction. When comparing data with other hospitals or within the same hospital over time, column 5 will immediately tell you the contribution of each group to the overall caesarean section rate. Columns 3 and 4 will tell you whether the difference in contribution is either a result of a change in the size of the group or the caesarean section rate within the group, or a combination of both. It is essential to remember the importance of the size of the groups.
