Pregnant women’s clinical characteristics, intrapartum interventions, and duration of labour in urban China: a multi-center cross-sectional study

Childbirth is a normal psychological process for the majority of women. The World Health Organization (WHO) defines normal labour as low risk throughout, spontaneous in onset, with the fetus in vertex position and ending with the mother and fetus in good condition following a spontaneous delivery [1]. Over the decades, however, there is an increasing global trend towards the widespread over-medicalisation of labour and childbirth in many parts of the world [2‐4] . A range of labour practices have been applied to initiate, accelerate, terminate, regulate or monitor the physiological process of labour [5], thus leading to a technocratic model of care for normal childbirth with frequent use of interventions during labour and childbirth [6].

Miller et al. termed the excessive medicalisation as too much, too soon (TMTS), which has become a global threat to maternal, fetal, and newborn wellbeing [7]. For instance, rising rates of clinical interventions, such as caesarean section, routine cardiotocography, labour induction, augmentation, and routine episiotomy, have been shown to cause avoidable harm if overused [7‐9]. Unnecessary childbirth interventions could be linked to negative consequences in maternal and child health and may cause substantial health-care costs [10, 11].

The causes of the rise in intrapartum interventions emerge as a complex multifactorial labyrinth involving health systems, care providers, women, societies, and even fashion and media [12‐14]. Many common obstetric practices, however, are of limited or uncertain benefit for low-risk women in spontaneous labour [15]. Unnecessary routine interventions in labour are actually associated with further interventions and result in decreased rates of spontaneous vaginal birth [16, 17]. Bohren et al. reported that more than a third of women experienced mistreatment and were particularly vulnerable around the time of birth [18]. The increasing medicalisation of childbirth processes tends to undermine the woman’s own capability to give birth and negatively impacts her childbirth experience [5]. Satisfaction with women’s childbirth experience is related to personal expectations, support from caregivers, bonding with professionals, and women’s involvement in decision making [15, 19]. Hence, there is a growing call from the international community for considering reducing interventional approaches for intrapartum management of childbearing women in spontaneous labour [15, 17].

In China, the overall annual rate of caesarean deliveries reached 34.9% between 2008 and 2014 [20]. In some urban areas like Shanghai, the caesarean section rate declined from 67% in 2009 to 52% in 2014 [20], but still far beyond the recommended maximum level of 15% from the WHO [21]. The third phase of the WHO global survey reported that intrapartum caesarean section (ICS) rate in China accounted for 24.2% of the women who attempted trial of labour, much higher than that in other countries [22]. For example, it was reported that the ICS rate was 14.6% in India and 6.9% in Japan [22]. In Bernitz et al.’s study conducted in Norway, the ICS rate was 5.9% in the WHO partograph (control) group and 6.8% in Zhang’s guideline (intervention) group [23]. A national explorative study in Dutch reported the ICS variation was 13–15% for nulliparas and 5–6% for multiparas, respectively [11]. The above figures show that China is confronted with a more serious situation than other countries in terms of childbirth medicalisation. In addition, since the implementation of China’s two-child policy, numbers of women with advanced maternal age and other high risks have increased and posed great challenges on promoting normal labour and childbirth [24, 25].

In view of the substantial changes in women’s characteristics and medicalisation of labour management over the decades, Zhang et al. conducted a large contemporary cohort study in US and has established a new labour guideline since then [26]. Within this context, however, data on the characteristics of labour among Chinese women, their use of intrapartum interventions, and the association between women’s clinical characteristics and duration of labour are less well investigated. Therefore, the aim of this study was to investigate pregnant women’s characteristics, intrapartum interventions, duration of labour and its associated factors; and to compare the differences of these variables between nulliparas and multiparas in urban China.

A total of 1523 women admitted to the labour and delivery unit were approached and assessed for their eligibility and consent for the study, and 1209 participants provided informed consent. The sociodemographic characteristics of the study population are presented in Table 2. Of the 1209 consented participants, 841 (69.6%) women were nulliparous and 368 (30.4%) were multiparous women. The participants’ ages ranged from 20 to 48 years, with the mean maternal age of 29.86 (SD3.94) years. The majority of the participants were well educated, with 65.9% (n = 797) having completed a college or higher level degree. The mean gestational week at admission of this study was 39.56 (SD1.02) weeks, ranging from 37.0 to 41.5 weeks. The overall prevalence of underweight women was 18.9% (n = 228), whereas overweight and obesity accounted for 12.2% (n = 147).

Different intrapartum interventions were performed for the labouring women, including 27.4% (n = 331) of the amniotomy, 37.9% (n = 458) of use of oxytocin, 53.0% (n = 641) of the continuous electronic fetal monitoring (EFM), and 52.9% (n = 639) of use of epidural analgesia (Table 3). Of the 1209 participants, 983 women eventually achieved successful vaginal birth after trial of labour while 226 (18.7%) women ended in an ICS. These women undergoing ICS were censored during labour observation. Kaplan-Meier survival analysis unfolded that the duration of total stage of labour was significantly longer in nullliparous group than that in multiparous group (χ²=81.805, P < .001) (Fig. 2).

Of the 983 participating women who delivered vaginally, 706 women were nulliparous and 277 women were multiparous. The median duration of first stage of labour was 8.50 (5.50, 13.00) hours in the nulliparous group, and 4.50 (2.58, 7.00) hours in the multiparous group, making significant differences between the two groups (P < 0.001). The median duration of second stage of labour in the nulliparous group [0.78 (0.45, 1.18) hours] was significantly longer than that in the multiparous group [0.28 (0.16, 0.53) hour](P < 0.001). The median duration of total stage of labour was 9.38 (6.33, 14.10) hours in the nulliparous group, and 5.08 (3.00, 7.83) hours in the multiparous group, giving significant differences between the two groups (P < 0.001). We found no significant differences in the neonatal birth weight between the two groups (P > 0.05) (Table 3).

Compared with nulliparous women, multiparous women in this study were less likely to have an amniotomy (75 [20.4%] versus 256 [30.4%]; P < .001), less likely to be administered oxytocin (67 [18.2%] versus 391 [46.5%]; P < .001), less likely to receive continuous EFM (119 [32.3%] versus 522 [62.1%]; P < .001), and less likely to use epidural analgesia (101 [27.4%] versus 538 [64.0%]; P < .001) (Table 3).

Of the 983 women who eventually achieved successful vaginal birth after trial of labour, we analysed the associations between potential predictors and the outcome variable of labour duration. Due to skewness of the distribution of labour duration, data was z-transformed into normal distribution before analysed using multivariable linear regression. The potential associated factors were then entered into the multivariable linear regression model. In the final model, the following factors were independently associated with longer duration of total stage of labour: epidural analgesia (P < .001), primiparity (P < .001), continuous EFM (P = .035), and increased birth weight (P = .005) (Table 4).

This was a multi-center cross-sectional study conducted in three tertiary university hospitals in Shanghai, China. We found that intrapartum medical interventions become common obstetric practices in urban China. Also multifactorial variables associated with duration of labour were identified in our study.

It was found in this study that the intrapartum interventions accounted for 27.4% in amniotomy, 37.9% in oxytocin administration, and 53.0% in the use of continuous EFM (Table 3), unfolding that these procedures are common interventional practices in China’s obstetric settings. Meanwhile, we found in our study that the overall epidural rate was 52.9%, whereas an Australian study reported that the overall rate of epidural analgesia was 35.9% [31]. Beyond that, our study showed that the rate of ICS was 18.7%. This rate is lower than Lumbiganon et al.’s report of 24.2% in China [22], but still much higher than Bernitz et al.’s report of 5.9–6.8% in Norway [23]. Our findings illustrated a technocratic model of care for normal childbirth in China, which could be explained by the fact that China’s maternity care is an facility based obstetrician-led model [32‐34]. The regulations and organisational framework of maternity services constitute contextual factors fostering medical interventions during labour [35].

We found in our study that compared with nulliparas, multiparas were less likely to receive intrapartum interventions including amniotomy, oxytocin, continuous EFM, and epidural analgesia (Table 3). Similar findings have been previously reported by Grylka-Baeschlin et al.’s study, where there were lower intervention rates in multiparas than in primiparas [36]. For neonatal birth weight, we found no statistically significant differences between the nulliparas group and the multiparas group, which indicated that parity did not affect this variable significantly. Admittedly, further studies still need to be conducted on account of the causality challenge of this study. Meanwhile, our study also identified the independent multifactorial variables associated with labour duration, ranging from epidural analgesia, parity, continuous EFM, and birth weight (Table 4) .

Firstly, by adjusting for confounding, we found an association between labour duration and epidural analgesia. Our study indicated that use of epidural was associated with longer duration of labour. This finding was in concordance with Turner et al.’s findings [31]. Although epidurals may reduce pain during labour more effectively than any other form of pain relief, it may also be associated with unwanted effects like prolonged labour, hypotension, drowsiness and fever [37]. Therefore, to ensure a positive childbirth experience, women should be allowed to make decisions about their pain management. Epidural analgesia can be one of the options to choose when required by labouring women [5]. Beyond that, nonpharmacologic methods such as some relaxation therapies and continuous labour support should also be offered to women, in order to help them feel more in control and satisfied with their labours [38].

Secondly, the results of our study demonstrated that labour duration was significantly longer in nulliparous women than that in multiparous women, which was consistent with Chen et al.’s finding [39]. Concurrently, durations of the first, second, and total stage of labour in our study was similar to those reported by Chen et al. [39]. Then we found that parity was an independent variable associated with labour duration.

In addition, we found an association between continuous EFM and duration of labour. On the one hand, continuous EFM might restrict women from moving freely during labour and could be associated with prolonged labour. Previous studies demonstrated that freedom to move and adopt upright positions in labour results in a range of physical and psychological benefits for women, including reduced risk of caesarean section, increased sense of control during labour and increased satisfaction with the birth experience [40]. Therefore, women should be encouraged to be mobile and to adopt comfortable positions of their choice so as to gain a positive childbirth experience. On the other hand, continuous EFM might be performed for labouring women because of prolonged labour. Therefore, temporal direction of the association between continuous EFM and duration of labour could not be clearly ascertained.

Furthermore, we found in our study that increased birth weight was an independent factor associated with longer duration of labour, which is consistent with Leftwich et al.’s study conducted in the US [41]. Based on this research evidence, it is reasonable to allow longer time for labour progression when a larger fetus is suspected [41]. Beyond that, this finding also highlighted the importance to design and implement weight management interventions that may prompt women to conceive an appropriate size of fetus and to experience a normal labour progression.

However, it is notable that we did not find any association between duration of labour and amniotomy (Table 4). It can be assumed that amniotomy might be performed for women who had presented a prolonged labour but labour was not prolonged due to amniotomy. Interestingly, however, Vadivelu et al. have shown that amniotomy was associated with a shorter labour duration compared with conservative management in women with singleton pregnancies [42]. Conversely, Chen et al. found that in nulliparas, the average time of first stage of labour and total labour duration increased due to amniotomy [39]. The association between amniotomy and labour duration varied due to different research designs and inter-study heterogeneity. Yet Chen et al.’s study illustrated that women with medical interventions were more likely to have prolonged labour processes [39]. And there is no evidence that these interventions can improve childbirth experience for women who have had a prolonged labour [43]. Therefore, amniotomy, as one of the most common interventions in modern obstetric practice, should be performed with caution instead of being used routinely during labour management.

Meanwhile, we found in our study that maternal age was not associated with duration of labour (Table 4). However, this does not mean that the variable of age is unrelated to labour duration and management of labour. Greenberg et al. stated that older women had a persistently higher likelihood of experiencing longer labour and prolonged labour than younger women [44]. Given these contradictory findings, further research is needed to clarify the association between maternal age and duration of labour.

From all discussed above, we can see in this study that various medical interventions prevail in Chinese urban obstetric settings. In concordance with our findings, previous studies have also noted that there is a continually increasing trend in routine use of medico-technical interventions [17, 45, 46]. Lu et al. also pointed out the current situation of frequently used intrapartum interventions which may substantially distort the labour pattern [47]. According to International Confederation of Midwives (ICM), midwives are the most appropriate care providers to attend childbearing women [48]. One Cochrane Review by Sandall et al. suggests that women who received midwife-led continuity models of care were less likely to experience intervention and more likely to be satisfied with their care than women who received other models of care [49]. As such, models of care with lower intervention rates such as midwife-led models of care should be developed and implemented in order to rectify the medicalisation of childbirth in China.

Currently in urban China, intrapartum interventions for childbearing women mainly consist of various medico-technical measures during labour. Among them, continuous EFM, amniotomy, and oxytocin treatment have become common practices. Also, the magnitude of epidural analgesia in our study represented a higher proportion among labouring women, indicating that women in urban China are on the one hand having more choices of and access to labour analgesia services, and on the other hand are receiving more pharmacologic methods for labour pain management. Meanwhile, epidural analgesia, primiparity, continuous EFM, and increased birth weight are associated with longer duration of labour. Further research is required to validate these variables and to determine the modifiable factors for labour management. In order to limit unnecessary intrapartum medical interventions and to improve womens’ childbirth experience, obstetric care providers should take full account of these factors during prenatal counseling and in the process of labour management for childbearing women. Furthermore, there is an urgent need for developing models of care with lower intervention rates such as midwife-led models of care to rectify the medicalisation of childbirth in China.