Abstract
Background/Aim: At the beginning of the 21st century, obstetric medicine took a turn from interventional to restrictive in low-risk birth. The present study examined the changes in peripartum management over the past 20 years at the Women’s University Hospital Cologne. The attitudes of the becoming mother and physicians towards anesthesia, episiotomy, and vaginal-operative deliveries were compared and the factors influencing the duration of birth over the past 20 years were examined. Patients and Methods: In this retrospective study, the low-risk singleton birth of 955 in 2000/2001 and 944 births in 2018 at the Women’s University Hospital Cologne were analyzed. Results: The age of women who tended to give birth has significantly increased at present compared to 20 years ago. In 2018, labor was induced significantly more often than in 2000/2001. The rate of vaginal operative deliveries has fluctuated between 15% and 20% in the last 20 years. Forceps are no longer used. The use of episiotomy has taken a fundamental turn in the last 20 years. Prophylactic episiotomy is not performed anymore, most vaginal operative deliveries take place without the episiotomy. The birth duration has been significantly shortened at present compared to 20 years ago. Conclusion: Pregnancy and childbirth over the last years are not considered as a disease, but as a natural course, and the trend of minimizing interventions in low-risk delivery has a positive effect on childbirth.
- Evidence-based medicine in obstetrics
- low-risk birth
- change in trend in obstetrics
- epidural anesthesia during birth
- episiotomy
- vaginal-operative deliveries
- birth injuries
- birth induction
- duration of birth
At the beginning of the 21st century, obstetric medicine took a turn from interventional to restrictive. With the introduction of “evidence-based medicine” at the end of the 1980s, there has been an effort to introduce this in obstetrics. In 1989, the book “A Guide to Effective Care in Pregnancy and Childbirth” by Enkin, Keirse and Chalmers, written on the basis of several studies, was published and became the standard for evidence-based medicine in obstetrics (1). In their work, the authors define two principles: There must be no intervention in a physiologic process unless it is certain that the intervention is more effective than nature. The disadvantages of the intervention must not be higher than its benefits. They showed that interventions during birth, that were performed routinely, have no benefit in low-risk births and these are also supported by the WHO (2). Based on these insights, there has been a gradual decrease in the use of vaginal operative delivery, involving the utilization of mechanical instruments in the expulsion phase, in Germany over the years (3). Nowadays, the tools used for vaginal - operative delivery are the vacuum extractor and the forceps. In most European countries, the frequency of vacuum extraction among vaginal - operative deliveries is increasing, while that of forceps is becoming less frequent (4). In 2017, the overall rate of vaginal operative deliveries as a proportion of all deliveries was approximately 6%, with a forceps delivery rate of 0.8% (5, 6).
While the WHO guideline of 1996 considered vaginal - operative delivery as an indication for episiotomy, results of diverse studies show that episiotomy does not provide protection for anal sphincter injuries and even increases the risk for higher grade of perineal tears (5). Approximately 40% of all women suffer birth injuries during vaginal delivery (7). Episiotomy was performed since early 19th century and has been one of the most common surgical procedures in gynecology. The aim of episiotomy is to shorten the birth path by incising the perineum and to prevent severe perineal tears. Despite a substantial evidence from multiple meta-analyses demonstrating the unfavorable outcomes associated with episiotomy and its inability to prevent severe perineal tears, as well as the Cochrane recommendation in 2000 advocating for its selective rather than routine use (8-22), prophylactic episiotomy has been performed in approximately half of women in Germany until the early 21st century (23). According to DIMDI statistics, in 2006, 30% of women in Germany who delivered vaginally had episiotomy, whereas in 2016 the episiotomy rate was 20% (23). The administration of oxytocin for induction of labor has become much seldom at present (24). Pharmacological options for pain relief in childbirth have been provided since the 19th century. Since the 1970s, the most modern, common, and effective method for pain relief in childbirth has been epidural administration. Its main disadvantage is the prolonged expulsion period. Both labor and the woman’s sense of pushing contractions are suppressed. A Cochrane Review of meta-analysis from 21 studies including 6664 women showed a higher rate of vaginal - operative delivery with epidural administration compared to natural delivery (23). As opposed to that, a randomized placebo-controlled prospective study did not show significant difference in the duration of the expulsion period with and without epidural anesthesia (25).
Although obstetrics is one of the oldest fields in medicine and vaginal - operative deliveries as well as episiotomy have been standardized and widespread worldwide since the 19th century, both the methods and the indications have been changed over time. Prophylactic episiotomy has been particularly controversial in recent years. Vaginal - operative methods and attitudes toward anesthesia are also under ongoing change.
The present study examines the changes in peripartum management that have developed over the past 20 years, using data from the Women’s Hospital at the University of Cologne, a maximum care hospital. The sociodemographic data of the mothers and the children were collected, the choice of delivery position as well as the attitudes of the becoming mother and the physicians towards anesthesia, episiotomy and vaginal-operative deliveries were compared. Finally, the factors influencing the duration of birth over the past 20 years were examined.
Patients and Methods
Patient population. The present study is a retrospective analysis of patients, who had vaginal delivery in 2000, 2001 and 2018 at the Women’s Hospital at University of Cologne. Only live singleton deliveries of gestational age between 37 and 43 weeks with low risk are included in the study. Here, low risk pregnancies are defined as pregnancies for which no increased risks to mother and/or unborn child have been identified and for which there is no need or benefit for intervention (15). In 2000, there were 496 deliveries that met our inclusion criteria, in 2001 there were 459 deliveries, and in 2018 there were 944 deliveries, so that two groups with comparable number of patients from 2000/2001 and 2018 were formed. The differences in characteristics of mothers and newborns, delivery methods and birth injuries from each group were analyzed. Patient data were collected using GE’s PIA fetal database software (ViewPoint ™ 5, GE HealthCare Technologies, Inc., Chicago, IL, USA) and transferred to SPSS (statistical software package SPSS Statistics 26.0, SPSS Inc., Chicago, IL, USA) for the statistical analysis.
Statistical analysis. All statistical analyses were performed using IBM SPSS version 24 software. The patients were divided into two independent groups: delivery year 2000/2001 and 2018. For group comparison of nominal scaled variables, crosstabs were used and the chi-square test or, in the case of four-field tables, the exact Fisher test was applied. Ordinal variables such as severity of birth injury were compared using a Mann-Whitney U-test. For the comparison of two groups of interval-scaled metric variables, the t-test was used. A multivariate analysis using linear regression was performed. The influence of the independent metric and dichotomous variables on the dependent interval-scaled variable “birth duration” was tested. Consequently, a univariate multifactorial analysis of variance was applied to examine the association between the factors “group” and “epidural administration (EA)” with birth duration. Because of the sample size of 1899 patients, the significance level here was set at 1%.
Results
The number of patients with low-risk singleton deliveries in 2000 and 2001 was 955 (group 1) and in 2018 944 (group 2). Mothers of group 2 were significantly older at the time of delivery than those of group 1 (p<0.001). Thus, the proportion of patients from 18 to 35 years decreased from 70.6% to 62.6% and the proportion of patients older than or equal to 35 years increased from 28.5% in 2000/2001 to 37.1% in 2018. The average age of mothers with first born was 30.6±5.6 in 2000/2001 and 31.8±5.3 in 2018 (p=0.002). The age difference was even greater for multiparous women: 32.5±4.9 vs. 34.1±5.0 (p<0.001).
The gravidity and parity did not differ significantly between the groups: approximately 1/3 of the women had their first pregnancy, 1/3 their second, under 1/5 their third, and the number of women with more than three pregnancies was less than 14% in both groups (Figure 1). Regarding parity, approximately 45% of each group were primipara and 35% in group 1 and 36% in group 2 bipara, respectively, and less than 20% were multipara in both groups (Figure 2).
Number of gravidity in 2000/2001 and 2018.
Number of parity in 2000/2001 and 2018.
The mean gestational age at the time of delivery was 40 weeks in both groups. Figure 3 shows the comparison of the gestational age of the newborns between the two groups by means of boxplots. Although the interquartile range of group 2 was wider (p<0.001), the gestational age of both groups was within normal range, as we only included low risk births (Figure 3). The weight, length, and head circumference of the newborn at birth remained unchanged over the last 18 years (Table I).
The mean gestational age in 2000/2001 and 2018.
Characteristics of newborns in 2000/2001 and 2018.
In 2000/2001, the most preferred delivery position was lying on the delivery bed with 86.8%, whereas in 2018 only 62.9% of women choose to deliver in the same position. 37% of the women in group 2 benefited from the alternative positions, such as sitting position, quadruped stand, stool, water birth, and others (p<0.001) (Table II).
Various delivery positions in 2000/2001 and 2018.
The mean duration of labor was almost twice as long in 2000/2001 compared to 2018 (8.9 vs. 4.6, p<0.001). This difference was highly significant. Birth was induced in 243 (25.4%) patients from group 1 and in 324 (34.3%) patients from group 2 (p<0.001). The mean duration of labor was not significantly different between induced birth (induced birth 6.5 h vs. not induced birth 6.8 h, p=0.384).
There was a significant difference in delivery mode between 2000/2001 and 2018. Spontaneous deliveries were similarly distributed, but there were significantly more vacuum extractions in 2018 (12.1% in group 1 vs. 19.8% in group 2). At present, forceps were no longer used at the University Hospital of Cologne; in 2000/2001, there were still a total of 31 (3.2%) forceps deliveries. Among patients delivered using vacuum extraction, 86.1% (n=99) in group 1 and 33.5% (n=63) in group 2 received episiotomy (OR=12.3, 95%CI=6.7; 22.6, p<0.001). In case of forceps deliveries in group 1, the episiotomy rate was 87.1% (27 of 31 patients) (Table III).
Delivery modes in 2000/2001 and 2018.
A total of 271 (28.4%) patients from group 1 and 430 (46.2%) patients from group 2 did not use anesthesia during delivery. This difference was highly significant (p<0.001) and indicates a strong trend toward natural childbirth. Previously, the most preferred anesthesia was epidural administration (47.5% in group 1 vs. 28.7% in group 2, p<0.001). 1.3% of women in group 1 and 0.3% in group 2 used pudendal block. Local anesthesia, which is particularly used to treat the birth injuries, was similar in both groups (Table IV).
Use of anesthesia of mothers in 2000/2001 and 2018.
53% of patients in group 1 had no birth injuries, whereas 31% of patients in group 2 did not have any birth injuries (p<0.001). However, considering median episiotomy as a 1st degree perineal tear and mediolateral episiotomy as a 2nd degree perineal tear (15), patients in group 1 did not have a significantly lower rate of perineal injuries than those in group 2 and even had a significantly higher rate of overall birth injuries (83% vs. 76%, p<0.001). Severe perineal tears were rare and similarly distributed in both groups (3rd degree perineal tear: 19 vs. 19, 4th degree perineal tear: 3 vs. 5, respectively group 1 vs. 2 in number of patients). 75 patients from group 1 and 171 patients from group 2 had more than one birth injury. Analysis of the severity of injuries per patient with regard to episiotomy showed significantly more mild to moderate injuries in group 1, whereas a significantly higher rate had no birth injuries in group 2 (p<0.001) (Table V).
Birth injuries between primipara and multipara women in 2000/2001 and 2018.
Considering birth injuries with regard to parity, it was noticeable that primiparous women have more 3rd and 4th degree perineal (3.2% vs. 1.1%), vaginal (21.3% vs. 11.5%) and labial tears (12.6% vs. 7.3%) than multiparous women (p<0.001). Episiotomies were also performed more frequently in primiparous women (43.4% vs. 20.6%, p<0.001). The majority of patients had mild injuries during delivery (70% primiparas and 61% multiparas, respectively). Moreover, the large difference between primiparas and multiparas without birth injuries was observed (9% in primipara vs. 32% in multipara, p<0.001). Moderate (17% vs. 6%) and severe injuries (5% vs. 1%) occurred significantly more often in primiparous patients (p<0.001). It is clear that in group 2, where episiotomy was rarely indicated, the proportion of multiparous women without birth injuries was significantly higher than that in group 1 (27% vs. 36%, p<0.001).
In total, 464 (48.6%) of the patients in group 1 and 125 (13.2%) of the patients in group 2 received episiotomy (p<0.001). In contrast to 2000/2001, episiotomy was performed only in restrictive indication in 2018. With few exceptions, mediolateral episiotomy was performed in 2018, whereas in group 1 the most common variant was median episiotomy (Table VI).
Types and frequency of episiotomy in 2000/2001 and in 2018.
Patients, who underwent episiotomy, were more likely to get severe perineal (4.2% vs. 1.6%, p=0.001) and vaginal tears (14.4% vs. 8.6%, p<0.001) than patients who did not receive episiotomy. Significantly more moderate birth injuries with episiotomy were observed in the 2000/2001 (17.8% vs. 4.7%; p<0.001) and 2018 (27.6% vs. 8.2%; p<0.001) cohorts, and significantly more severe injuries with episiotomy occurred in 2000/2001 (4.8% vs. 1.7%; p=0.009), and a similar trend was observed in 2018 as well (6.5% vs. 2.1%; p=0.012).
In 2000/2001, episiotomy was performed in nearly 90% of vaginal - operative deliveries (87.1% for forceps and 86.1% for vacuum extraction delivery). In 2018, 66.5% of vacuum extractions (VE) are performed without episiotomy (p<0.001). The most common injury from VE as well as forceps deliveries was 2nd degree perineal tear with 50.8% followed by vaginal tear with 48.5% in 2000/2001. In 2018, these injuries were observed in 42.1% and 24.3% of patients, respectively. All other injuries among vaginal - operative deliveries remained below 10% and did not differ significantly between the groups (Table VII).
Birth injuries of vaginal- operative deliveries in 2000/2001 and in 2018.
Next, we examined whether the year of delivery, maternal age, infant size (weight, length, and head circumference), parity, induction of labor, mode of delivery (spontaneous or vaginal-operative), epidural administration, and episiotomy influenced the duration of birth. As it is well known that epidural administration (EA) suppresses labor and delays birth, an additional univariate multifactorial ANOVA was performed with the variable group as a fixed factor and with EA as a random factor. This analysis of variance aimed to examine whether the group difference in birth duration is due to the use of EA. Our results showed a significantly longer duration of birth in the 2000/2001 cohort (p<0.001). Women who gave birth spontaneously had a highly significant shorter duration of birth than those who had a vaginal - operative delivery (p<0.001). In this case, however, the cause - effect relationship was unclear. Maternal age, infant length, weight, and head circumference, as well as induction of labor and episiotomy, did not play a significant role in birth duration. Multiparous women had a shorter duration of labor than primiparous women (β=−1.758; p<0.001) and EA prolonged the duration of labor by an average of 3.4 hours (p<0.001) (Table VIII). Analysis of variance showed both a highly significant difference in duration of labor between group 1 and 2 (p<0.001) and between patients with and without epidural administration (p<0.001). The interaction of these two factors, however, was not significant, indicating that birth duration differs in 2000/2001 vs. 2018 regardless of epidural administration.
Linear regression with birth duration in hour as dependent variable.
The analysis of the impact of maternal age, divided as younger than or equal to 35 and older than 35, on birth duration between two groups showed a significantly longer duration of birth in the 2000/2001 (p<0.001) and younger (p=0.005) cohorts. However, the interaction of these two factors was not significant (p=0.254), indicating that maternal age does not explain the difference between the 2000/2001 and 2018 groups (Table IX).
Birth duration in hours according to the maternal age in 2000/2001 and in 2018.
Discussion
The age of women who tend to give birth has significantly increased at present than 20 years ago. The average maternal age in our study was 31.6 years in 2000/2001 and 33.0 years in 2018. The proportion of women giving birth older than 35 years increased from 29% in 2000/2001 to 37% in 2018.
A prolonged delivery period is associated with more complications and increased morbidity. Especially, it is associated with a higher rate of acidosis of the newborn or increased birth injury of the mother (26, 27). Therefore, it is the interest of the physician and the patient to determine the causes of prolonged labor and shorten it. In 2000/2001, the duration of birth was almost twice as long as it is at present. Multivariate analysis shows a significant delay in birth duration in patients who used EA. The number of vaginal births with EA dropped highly significantly from 48% in 2000/2001 to 29% in 2018, so this could explain the shorter birth duration. However, our analysis of the subgroups showed a significantly shorter birth in 2018 than in 2000/2001 regardless of the use of EA. The fact that epidural anesthesia prolongs the expulsion period is confirmed again in our study but does not provide an explanation for the difference in birth duration between the groups.
Maternal age showed a negative effect with respect to birth duration. The duration of birth tended to be longer in younger mothers. Similarly, child head circumference increased birth duration. Multiparous women showed a shorter duration of birth than primiparous women. Women who deliver spontaneously have a shorter birth duration, although this is more likely because prolonged labor is an indication for vaginal operative delivery. It should be noted that episiotomy, which was indicated in case of prolonged birth until 20 years ago, showed no significant effect on birth duration (23).
In 2018, labor was induced significantly more often than in 2000/2001. Which labor induction agents were used then and now, and whether the difference in birth duration of labor is due to the different method of labor induction between 2000/2001 and 2018, requires further investigation (24, 28, 29).
The rate of vaginal operative deliveries has fluctuated between 15% and 20% in the last 20 years. This number has increased by about 5% in the last 20 years (p<0.001). Forceps are no longer used. After numerous studies have shown forceps delivery to be riskier and more traumatic than vacuum extraction (5, 6, 12, 30), vacuum assisted deliveries have gained more acceptance in the western European countries.
The use of episiotomy has taken a fundamental turn in the last 20 years. Although as early as the 1970s, numerous studies demonstrated routine use of episiotomy to be ineffective against severe perineal tears, this was performed continuously in European countries and in the United States (13, 15, 19, 22). The average episiotomy rate in Germany at the end of the 20th century in all vaginal births was 60%. Nowadays, most vaginal operative deliveries take place without the episiotomy. In 2018, only 1/3 of women had an episiotomy during a vacuum assisted delivery, whereas in 2000/2001, this rate was 87%. The ineffectiveness of episiotomy in avoiding severe perineal tears in vaginal operative deliveries has been refuted often since the 1990s (30, 31). Our analysis shows significantly more severe perineal tears (p=0.001) and significantly more vaginal tears (p<0.001) in patients who underwent episiotomy. These findings are consistent with the results of several large studies and meta-analyses since the end of the 20th century (10-13, 16, 22, 32-34).
As limitation of retrospective study, the phase of labor epidural administration was applied could not be evaluated. Furthermore, the question remains whether it was administered because of a prolonged course of labor or whether a prolonged duration of labor was due to the EA.
The restriction of our patient population to vaginal low-risk live singleton births should be noted. That is, rates of birth injury, vaginal operative births, or episiotomies are not related to all patients but to women with low-risk live singleton births who deliver vaginally.
It remains unexplained why the birth duration in the younger group (younger than or equal to 35 years) is about half as long as in the older cohort (older than 35) (Ø 4.57 h vs. 8.90 h). Although EA and maternal age play a significant role in birth duration and are distributed differently in the 2000/2001 and 2018 groups, their interaction with the time of the survey is not significant, indicating that these variables do not explain the difference in birth duration between the groups. We assume that one of the causes for that is the difference in labor induction in 2000/2001 and present. In the past, oxytocin was often administered as a labor induction agent even though the cervix was not mature yet, which could result in prolonged labor. Today, prostaglandins are used to mature the cervix, which in many cases is sufficient to induce labor. In this procedure, the birth proceeds in a more natural way. Oxytocin can be administered to support contraction after cervix has opened. Unfortunately, the administration of oxytocin is not sufficiently documented in our database, thus this hypothesis cannot be quantitatively tested in the present study.
As a conclusion of our study, we can state that during the last years pregnancy and childbirth are not considered a disease, but a natural course, and that the trend of minimizing interventions in low-risk delivery has a positive effect on the course of childbirth.
Acknowledgements
The Authors would like to thank all the patients that participated in the study.
Footnotes
Authors’ Contributions
DR designed the study and wrote the final manuscript. AH performed the data collection and wrote the draft of the manuscript. EG performed the statistical analysis. SL and JR contributed to the data analysis and interpretation. NM contributed to the data collection. BG was involved in conceptualizing the study. PM approved the final version. SB contributed to writing the final version of the manuscript and supervised the data analysis.
Conflicts of Interest
All Authors declare that they have no conflicts of interest in relation to this study.
- Received August 4, 2023.
- Revision received August 28, 2023.
- Accepted August 29, 2023.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.