- Top of page
Objective To examine the relationship between the cervical dilatation at which women present in labour and the subsequent likelihood of caesarean section.
Design Retrospective cohort study.
Setting University teaching hospital.
Population 3220 women met the entry criteria from 14,050 deliveries between January 1995 and December 1999.
Methods Women meeting the following criteria were identified: those in spontaneous labour with a singleton pregnancy and a cephalic presentation at 37–42 weeks of gestation; all women delivering within 36 hours of first presentation were included. Women who had spontaneous rupture of the membranes before first attendance were excluded.
Main outcome measures The primary outcome was the rate of caesarean section. Secondary outcomes were operative vaginal delivery, fetal weight, cord pH, five minute Apgar score, length of labour, labour augmentation with oxytocin and epidural analgesia.
Results The risk of caesarean section decreased with increasing cervical dilatation at presentation. This was true for nulliparous (n=1168) and parous women (n=2052). The caesarean section rate of nulliparous women presenting at 0–3cm (n=812) was 10.3%, compared with 4.2% for those presenting at 4cm–10cm (n=356), and the mean duration of labour before presentation was 2.0 hours versus 4.5 hours, respectively (P=0.0001). For parous women the caesarean section rates were 5.7% and 1.3%, respectively (P=0.0001). There were significantly greater frequencies of use of oxytocin and epidural analgesia by women presenting earlier in labour. The caesarean section rate of 185 nulliparae (15.8%) who were initially allowed home was no different from those admitted immediately (9.2%vs 8.2%, P=0.67). Similarly, 196 (9.5%) of multiparae went home and had a caesarean section rate of 3.6%, compared with 3.1% if admitted immediately (P=0.76).
Conclusions Women who present to hospital at 0–3cm spend less time in labour before presentation and are more likely to have obstetric intervention than those presenting in more advanced labour. Outcomes were similar whether or not the woman was initially allowed home.
- Top of page
Rates of caesarean section have been a major public health concern in North America in recent years. Most caesarean sections for nulliparous women are performed for dystocia, and this is exceeded only by previous caesarean section as an indication for caesarean delivery in Canada1 and the United States2. However, the study of antenatal and intrapartum predictors of caesarean section in labour has not yet produced a model with high sensitivity and specificity3,4.
Knowledge of the patterns of normal and abnormal labour, and of women's behaviour, is fundamental to the formulation of strategies to reduce caesarean section rates. One of the difficult decisions that women have to make during a pregnancy is the decision when to go into hospital if they think that labour may be beginning. Nulliparous women in particular have no experience of labour and so may find the timing of presentation to hospital particularly difficult to judge. In addition, in many Western countries there is increasing centralisation of health resources, which may lead to increased journey times to the hospital. A previous study has shown that women admitted to hospital early (contractions of four hours or less) have a higher frequency of obstetric interventions in labour than those admitted later5.
The objective of this study was to determine how the caesarean section rate changes with the cervical dilatation at which women present in labour. Other outcome measures were operative and spontaneous vaginal delivery, fetal weight, cord pH, five-minute Apgar score, length of labour, labour augmentation with oxytocin and epidural analgesia. In addition the effect of deferring admission, in women presenting in early labour, was examined.
- Top of page
Data were extracted from the Ottawa Hospital labour and delivery database, for the five calendar years 1995 to 1999. The database includes delivery information, demographic data and electronically archived vaginal examination data, entered at the bedside. All cases in the database for this period satisfying the following criteria were identified; nulliparous or parous women in spontaneous labour with a singleton pregnancy, and cephalic presentation at 37–42 weeks of gestation. The cervical dilatation at first vaginal examination was identified from the database for each woman. Women were categorised as having presented early if they attended at <4cm cervical dilatation, or late if they presented with a cervical dilatation of ≥4cm. All women delivering within 36 hours of the first vaginal examination were included, regardless of whether they were initially admitted, or allowed home. This time span represents three standard deviations from the mean for duration of labour of nulliparous women in our population. Women who had ruptured membranes before attendance were excluded to create a uniform cohort, as this subgroup is managed by immediate induction of labour in our unit, in accordance with the findings of the TERMPROM study6. The outcome of labour was noted with regard to caesarean section, operative and spontaneous vaginal delivery, fetal weight, cord pH, five-minute Apgar score and neonatal death. Frequencies of labour augmentation with oxytocin and of epidural analgesia were also noted. The onset of labour is entered in the database by the admitting obstetric nurse, as the time when the woman reported the onset of strong regular contractions. This was used to calculate the length of labour and the duration of labour at home before the first vaginal examination.
We also looked at the effect of deferring hospital admission on the caesarean section rate and on the length of labour, by comparing women who were allowed home following their first assessment, with women who were admitted immediately. Travel time to the hospital was examined, classifying women as living locally if they were within 25 minutes driving time.
Statistical analysis was performed using StatView (Windows Version 4.57). Data were compared using χ2 tests or Fisher's exact test for nominal variables, and t tests for continuous variables. The P-value for correlations was calculated using Fisher's method. A P value of <0.05 was considered statistically significant.
- Top of page
Of 14,050 deliveries during the study period, 11,602 were singleton cephalic deliveries at 37–42 weeks. Of these, 4180 women were induced, 1247 had a caesarean section before labour onset and 2553 had prelabour rupture of the membranes. A further 243 women had a vaginal examination-to-delivery interval of >36 hours, and 159 women had no vaginal examination data. The remaining 3220 women formed the study population, of whom 1168 were nulliparous and 2052 were parous. Demographic data are shown in Table 1, and the number of women presenting at each dilatation in Fig. 1. The greater birthweight in infants of parous women presenting late was due to an excess of babies weighing 4.0kg or more when compared with parous women presenting early (19.1%versus 12.7%; P=0.001). The proportion of infants with a birthweight of <2.5kg was similar for the early and late presenters. There was no neonatal death in any group.
Table 1. Comparison of demographic data for women presenting at 0–3 cm dilated versus 4–10 cm dilated. Values are given as mean [SD] or n (%).GA= gestational age.
| ||Nulliparous women||Parous women|
| ||Presented at 0-3cm (n=812)||Presented 4-10cm (n=356)||Presented at 0-3cm (n=871)||Presented 4-10cm (n=1181)|
|Maternal age (years)||27.6 [5.2]||27.8 [6.5]||31.1 [4.6]||30.9 [4.6]|
|Live locally||620 (76.4)||275 (77.3)||663 (76.1)||935 (79.1)|
|GA (weeks)||39.7 [2.3]||39.4 [1.1]||39.4 [1.0]||39.3 [1.0]|
|Birthweight (g)||3440 ||3459 ||3513 ||3599 |
|Birthweight <2.5kg||11 (1.3)||2 (0.6)||4 (0.5)||7 (0.6)|
The risk of caesarean section was seen to fall with increasing cervical dilatation on presentation (correlation coefficient -0.9; P=0.0001) for both nulliparous and parous women (Fig. 2). Women presenting at 0–3cm spent less time at home in labour before their first vaginal examination, had longer labours and had a significantly smaller proportion of their labour at home (Table 2). They also had a higher caesarean section rate than the late presenters, as well as a higher rate of oxytocin and epidural usage. The odds ratio for caesarean section for women who presented at 0–3cm was 2.62 for nulliparous women (95% CI 1.49–4.61), and 4.73 (95% CI 2.64–8.49) for parous women. There was a reduction in operative vaginal delivery rates, particularly in parous women, but this did not reach statistical significance.
Table 2. Comparison of labour characteristics and interventions for women presenting at 0–3 cm dilated versus 4–10 cm dilated. Values are given as mean [SD] or n (%). CS=caesarean section.
| ||Nulliparous women|| ||Parous women|| |
| ||Presented at 0-3cm (n=812)||Presented 4-10cm (n=356)||P =(0-3 v.'s 4-10cm)||Presented at 0-3cm (n=871)||Presented 4-10cm (n=1181)||P =(0-3 v.'s 4-10cm)|
|Length labour (h)||14.26 [7.4]||11.16 [6.1]||0.0001||9.11 [5.9]||6.75 [4.7]||0.0001|
|Labour onset to 1st VE (h)||1.96 [5.8]||4.48 [4.4]||0.0001||1.57 [5.1]||3.13 [3.8]||0.0001|
|Proportion of labour at home||0.25 [0.18]||0.36 [0.23]||0.0001||0.33 [0.23]||0.44 [0.26]||0.0001|
|Oxytocin use||349 (42.9)||97 (27.2)||0.0005||174 (20.0)||100 (8.5)||0.0001|
|Epidural use||666 (82.0)||217 (60.9)||0.0003||506 (58.1)||464 (39.6)||0.0001|
|Operative vaginal delivery||216 (26.6)||89 (25.0)||0.56||70 (8.0)||75 (6.4)||0.14|
|CS||84 (10.3)||15 (4.2)||0.001||50 (5.7)||15 (1.3)||0.0001|
|Mean dilatation CS (cm)||6.5 [2.6]||7.5 [2.6]||0.001||6.2 [2.7]||7.3 [2.6]||0.001|
|Apgar score ≤ 7||31 (3.8)||8 (2.2)||0.22||25 (2.9)||27 (2.3)||0.48|
|Cord pH≤7.1||34 (4.2)||11 (3.1)||0.41||25 (2.9)||23 (1.9)||0.18|
Of the 2052 parous women, 202 (9.8%) had a history of previous caesarean section. When these women are analysed as a subgroup, 130 presented at 0–3cm of whom 32 (24.6%) had a caesarean section. Of the 72 women who presented at ≥4cm, seven (9.7%) had a caesarean section (P=0.01). If the remaining 1850 parous women, who had delivered vaginally in previous pregnancies, are analysed separately, 741 presented at 0–3 cm and 1109 presented at ≥4cm. The caesarean section rates for these groups are 2.4% and 0.7%, respectively (P=0.001).
A total of 371 women had admission deferred following their initial presentation. These women returned to hospital an average of 12.5 hours (9.7) (range 1–31 hours) after their first vaginal examination with a median increase in cervical dilatation of 1cm. Caesarean section rates were compared with women admitted immediately, for each cervical dilatation at presentation, and no difference was found at any single dilatation, or overall, for either nulliparous or parous women (Tables 3 and 4). Demographic factors and other outcomes for women with immediate versus deferred admission are shown in Table 5. Women admitted immediately had a slightly shorter mean length of labour (P=0.04 for nulliparous women and P=0.001 for parous women). There were no other statistically significant differences between the groups.
Table 3. The effect of deferred versus immediate admission on caesarean section (CS) rates in nulliparous women, according to cervical dilatation at first examination. Values are given as n (%).
|Dilatation||Deferred admission||Immediate admission|| |
|0 cm||30||5 (16.7)||36||3 (8.3)||0.45|
|1 cm||74||9 (12.2)||153||28 (18.3)||0.32|
|2 cm||53||3 (5.7)||190||18 (9.5)||0.36|
|3 cm||17||0 (0)||259||17 (6.6)||0.61|
|4-10 cm||11||0 (0)||345||15 (4.3)||0.48|
|Ttotal||185||17 (9.2)||983||81 (8.2)||0.67|
Table 4. The effect of deferred versus immediate admission on caesarean section rates in parous women, according to cervical dilatation at first examination. Values are given as n (%).
|Dilatation||Deferred admission||Immediate admission|| |
|0 cm||22||1 (4.5)||34||5 (14.7)||0.22|
|1 cm||38||4 (10.5)||100||12 (12)||0.81|
|2 cm||60||2 (3.3)||191||12 (6.3)||0.39|
|3 cm||42||0 (0)||384||14 (3.6)||0.21|
|4-10 cm||34||0 (0)||1147||15 (1.3)||0.50|
|Total||196||7 (3.6)||1856||58 (3.1)||0.76|
Table 5. Comparison of data for women with immediate versus deferred admission. Values are given as mean [SD] or n (%). GA= gestational age.
| ||Nulliparous women||Parous women|
| ||Deferred admission (n=185)||Immediate admission (n=983)||Deferred admission (n=196)||Immediate admission (n=1856)|
|Maternal age (years)||27.6 [5.3]||27.6 [5.7]||30.8 [4.7]||31.0 [4.5]|
|Live locally||149 (80.5)||746 (75.9)||149 (76.0)||1449 (78.1)|
|GA (weeks)||39.4 [2.1]||39.6 [1.0]||39.4 [0.94]||39.4 [1.0]|
|Birthweight (g)||3458 ||3444 ||3513 ||3570 |
|Birthweight <2.5kg||2 (1.1)||11 (1.1)||1 (0.5)||10 (0.5)|
|Labour length (h)||14.4 [8.6]||13.1 [6.9]||9.3 [7.2]||7.6 [5.1]|
|Oxytocin||77 (41.6)||369 (37.5)||30 (15.3)||244 (13.1)|
|Epidural||143 (77.3)||740 (75.3)||105 (53.6)||865 (46.6)|
|Operative vaginal delivery||41 (22.2)||264 (26.9)||18 (9.2)||127 (6.8)|
|PH< 7.1||3 (1.6)||42 (4.3)||4 (2.0)||44 (2.4)|
|Apgar Score < 7||5 (2.7)||36 (4.0)||4 (2.0)||48 (2.6)|
- Top of page
This study demonstrates that women who present to hospital early in labour have a higher risk of caesarean section, oxytocin augmentation and epidural analgesia than those who present later. There are a number of possible explanations for this observation. Firstly, women presenting early in labour may represent a higher risk group than those presenting late, perhaps having been instructed to attend early. In a retrospective study of this nature we cannot exclude this possibility; however, given the large sample size, the exclusion of inductions and the absence of a difference in the proportion of low birthweight infants between the early and late presenters, we feel that such an effect would be minimal in this population.
Secondly, those presenting early may have had a dysfunctional latent phase of labour; they might have laboured for as long, or longer, before attending hospital than women who presented late. A prolonged latent phase has previously been shown to be independently associated with an increased incidence of caesarean section and other labour abnormalities7. We examined this possibility by comparing the total length of labour, and the length of time from labour onset to first vaginal examination, between the groups of women. The women presenting early did have significantly longer labours, both in nulliparous and parous women, but they spent less time at home before presentation, with a smaller proportion of the labour at home, than women who presented late. This in turn may be related to higher levels of anxiety, or lack of support at home. It is well recognised that high levels of pain8 and anxiety9 are associated with increased intervention in labour, and that the provision of support, in hospital, for women in childbirth reduces both anxiety and obstetric intervention10,11.
Thirdly, early admission to hospital may itself have had an effect on labour through differences in maternal position and ambulation12, although this issue was not specifically addressed in our study. In addition there is evidence that the longer the labour is perceived to be by the physician, the higher the chances of intervention. One major distinction between the early and late presenters is the absence of data relating to the duration of the latent phase of labour in the late presenters. Intervention rates may be higher when physicians are provided with this information13.
Other potential weaknesses of this study include the absence of prospective data to determine the obstetric and psychosocial differences between women presenting early, allowed home, or presenting late. In addition, the timing of the onset of labour may be subject to bias as it is based on the nurse's questioning of the woman's recollection.
The women who attended early and had admission deferred had caesarean section rates similar to those admitted immediately, in keeping with other studies of early labour assessment in hospital14. This lack of a beneficial effect of deferred admission suggests that the increased intervention associated with early attendance is a result of intrinsic maternal or obstetric characteristics, and may not be due to unnecessary medical intervention. Indeed, it could be argued that this group of women are at higher risk of caesarean section and other interventions, and may benefit from early admission and close monitoring. However, there is also some evidence that home assessment in early labour can delay hospital attendance with a concomitant reduction in interventions15. We feel the factors which cause women to seek early admission are an important area for further study.