Background
During the second stage of labor, the maternal pelvic floor muscles undergo repetitive stretch loading as uterine contractions and strenuous maternal pushes combined to expel the fetus, and it is not uncommon that these muscles sustain a partial or complete rupture. It has recently been demonstrated that soft tissues, including the anterior cruciate ligament and connective tissue in sheep pelvic floor muscle, can accumulate damage under repetitive physiological (submaximal) loads. It is well known to material scientists that this damage accumulation can not only decrease tissue resistance to stretch but also result in a partial or complete structural failure. Thus, we wondered whether certain maternal pushing patterns (in terms of frequency and duration of each push) could increase the risk of excessive damage accumulation in the pelvic floor tissue, thereby inadvertently contributing to the development of pelvic floor muscle injury.
Objective
This study aimed to determine which labor management practices (spontaneous vs directed pushing) are less prone to accumulate damage in the pelvic floor muscles during the second stage of labor and find the optimum approach in terms of minimizing the risk of pelvic floor muscle injury.
Study Design
We developed a biomechanical model for the expulsive phase of the second stage of labor that includes the ability to measure the damage accumulation because of repetitive physiological submaximal loads. We performed 4 simulations of the second stage of labor, reflecting a directed pushing technique and 3 alternatives for spontaneous pushing.
Results
The finite element model predicted that the origin of the pubovisceral muscle accumulates the most damage and so it is the most likely place for a tear to develop. This result was independent of the pushing pattern. Performing 3 maternal pushes per contraction, with each push lasting 5 seconds, caused less damage and seemed the best approach. The directed pushing technique (3 pushes per contraction, with each push lasting 10 seconds) did not reduce the duration of the second stage of labor and caused higher damage accumulation.
Conclusion
The frequency and duration of the maternal pushes influenced the damage accumulation in the passive tissues of the pelvic floor muscles, indicating that it can influence the prevalence of pelvic floor muscle injuries. Our results suggested that the maternal pushes should not last longer than 5 seconds and that the duration of active pushing is a better measurement than the total duration of the second stage of labor. Hopefully, this research will help to shed new light on the best practices needed to improve the experience of labor for women.
Introduction
Human vaginal delivery is a complex event, often involving a tight fit between the fetus and birth canal, and never without risk. Maternal pelvic floor damage in the form of partial or complete tears of the levator ani muscles is still occurring, especially in first-time mothers during a difficult delivery. This injury has been well documented whether to the left, right, or both pubovisceral muscles (PVMs) near their origin on either side of the pubic symphysis, with a prevalence ranging between 13% and 41%. Lying near vital structures in a relatively inaccessible location deep within the pelvis, these injuries are not amenable to surgical repair. Women with this injury are at an increased risk of developing pelvic floor disorders later in life, including pelvic organ prolapse and urinary incontinence, both of which can significantly affect the quality of life and require surgery. As such, there is room for improving the management of labor to provide more satisfactory outcomes for the mother.
Why was this study conducted?
Passive tissues in tendons, ligaments, and muscles can accumulate microdamage to the point of rupture under repeated submaximal forces that in isolation would never cause failure. Derived from ex vivo animal experiments, a material constitutive model was developed to predict the damage accumulation in the pelvic tissues. A finite element model of the second stage of labor was employed to compare directed and spontaneous pushing during labor to determine whether one carries a higher risk of pelvic muscle injury.
Key findings
The finite element model predicted less damage accumulation when 3 maternal pushes per contraction, each lasting 5 seconds, were performed. This pushing style did not increase the predicted duration of the active second stage of labor while potentially carrying the lowest risk of injury.
What does this add to what is known?
Levator muscle tears during vaginal delivery are likely because of the accumulation of localized microdamage during the repetitive stretches of the active stage of labor.
Recently, it has been shown that highly specialized collagenous tensile structures, such as the anterior cruciate ligament (ACL), can tear under repetitive submaximal loading. With each loading cycle, such as landing from a strenuous jump, the origin of the ACL showed accumulation of microdamage at the collagen fibril and fiber levels, suggesting that microdamage can accumulate there and, if allowed to continue, can eventually cause a complete structural failure. Although the ACL can, of course, rupture under a single supramaximal load, it can also fail under a series of submaximal loads that in isolation would never cause failure. This type of injury is called a material fatigue failure. Motivated by these results, we performed a systematic review of the musculoskeletal literature and found that repetitive submaximal loading was reported at the musculotendinous junction of rabbit leg muscles, with histologically detectable microdamage accumulation. From observational studies, indirect evidence suggested material fatigue failure as a possible damage mechanism in the tendon and at the musculotendinous junction. This is because several nonimpact and nontraumatic injuries were reported within a cohort represented by professional and recreational athletes or people strongly engaged in sports activities, where strenuous repetitive movements are often performed without adequate rest.
Based on both the ACL data and indirect evidence found on our systematic review, we developed a laboratory study to test the hypothesis that damage could accumulate at the origin of the PVM because of repetitive submaximal loadings. Using sheep as an animal model, we applied 60 loading cycles at a 60% ultimate displacement a
a Ultimate displacement is the applied displacement that will cause the material to tear.
and found microdamage in the muscle, near the musculotendinous junction injury, in the form of elongated and crimped muscle fibers and individual muscle fiber rupture ( Figure 1 ). No visible tear was seen macroscopically; however, there was a visible permanent deformation in all specimens, averaging approximately 50% strain. Moreover, this damage accumulation was sufficient to affect the mechanical properties, in the absence of a complete tear, as captured by the strain-stress response.
During vaginal delivery, the second stage of labor commences during complete cervical dilation after which active maternal pushing can be used to augment each uterine contraction to facilitate the delivery. It is a common practice for the midwife or clinician to direct these efforts by instructing the mother when to begin the “active” stage of labor and thereupon to inhale deeply and to push with a closed glottis for 10 seconds, 2 to 3 times during each uterine contraction. This extended use of the Valsalva maneuver was popularized in the 1950s to accelerate the second stage of labor to avoid the need for forceps, which were in common use at the time. A popular alternative method is spontaneous pushing, which is carried out naturally by the woman according to her desire to push down progressively. There is no deep breathing, and it starts only when the full uterine contraction is well established. , Generally, 3 to 5 pushes occur near the peak uterine contraction for 3 to 5 seconds with approximately 2-second breathing between efforts, although 1 bigger push near the peak of the contraction is also used. Despite the directed pushing being the most common, it is not yet clear what the best pushing technique is. A summary of several studies that compare maternal and fetal outcomes for directed pushing and spontaneous pushing groups is shown in Table 1 . Overall, these studies have suggested that spontaneous pushing is not detrimental, although it seemed to cause longer mean second stages of labor. However, it is not clear whether directed pushing is detrimental or not, with some studies showing no substantial differences and other studies finding unfavorable outcomes. All these studies involved small sample sizes, which limited the generalizability of the results. Finally, in a review including 21 studies comparing spontaneous vs directed pushing, Lemos et al concluded that “there is no conclusive evidence to support or refute any specific style” and that “further properly well-designed randomized controlled trials, addressing clinically important maternal and neonatal outcomes are required to add evidence-based information to the current knowledge.”
| Study | Study design | Number of participants | Main outcomes |
|---|---|---|---|
| Schaffer et al, 2005 | Randomized trial | 128 |
|
| Araujo et al, 2021 | Clinical trial | 62 |
|
| Yeates and Roberts, 1984 | Pilot study | 10 |
|
| Paine and Tinker, 1992 | Clinical trial | 30 |
|
| Sampselle and Hines, 1999 | Retrospective study | 39 |
|
| Thomson, 1993 | Randomized controlled trial | 32 |
|
| Barnett and Humenick, 1982 | Pilot study | 10 |
|
| Koyucu and Demirci, 2017 | Randomized prospective study | 80 |
|
We demonstrated earlier that the most commonly injured structure in the levator ani muscles has the potential to accumulate fatigue damage under repeated submaximal loading cycles consistent with the combination of the uterine contraction augmented by strenuous maternal pushing. , , Thus, we wondered whether certain maternal pushing patterns could influence the prevalence of PVM tears. Our goal was to analyze the results of 4 different pushing methods to determine whether there is optimum management for labor that reduces the chance of a PVM tear because of material fatigue. A biomechanical approach was chosen to test this idea. Several finite element methods of the second stage of labor have already been developed to analyze factors, such as the pelvic floor strain, , episiotomy, , fetal head flexion, duration of the second stage of labor, fetal presentation, and fetal head molding. These models were developed to answer specific research questions, simplifying factors that do not strongly impact each research question. As such, material properties, such as viscoelastic effects, were often neglected (except for Vila Pouca et al and Moura et al who included viscoelastic effects to be able to analyze the influence of the duration of the second stage of labor). Moreover, none of these models included the ability to cyclically accumulate damage, an effect we have recently demonstrated can occur in collagenous and muscular tissues. , This is the main difference between the previous work and the finite element model applied in this work as it, for the first time, includes this ability to cyclically accumulate damage. In addition, it also includes viscoelasticity. Although an engineering approach inherently simplifies the problem, it also has the advantage of eliminating unwanted factors, which helps to achieve a baseline understanding of the problem. Moreover, it is a tool that can be used to better define the methodology of future clinical trials by helping to establish a more insightful hypothesis. The main questions we addressed were as follows: “Does the duration and frequency of maternal pushing influence the accumulation of damage near the origin of the PVM?” and “What frequency and duration of the maternal pushes minimize the risk of PVM tear?”
In light of our recent findings, , , our working hypothesis was that the maternal pushes influence PVM damage accumulation and that shorter pushes may be beneficial for the mother.
Materials and Methods
To achieve our goal, we developed a simplified finite element model of the second stage of labor during vaginal delivery. It incorporated 3-dimensional (3D) geometric models of the pelvic floor muscles and fetus and a user-defined material model to better reproduce pelvic floor muscle behavior. Different maternal pushing profiles were simulated to find the optimal management of labor concerning maternal pelvic floor health. The simulations were performed using Abaqus finite element software (version 2018; Dassault Systèmes) with an implicit solver (Abaqus/Standard).
Geometric model and boundary conditions
The 3D finite element model of the levator ani muscles was created on the basis of a geometric data point set acquired from a female cadaver measurement by Janda et al. The creation of this geometry is described in Parente et al. Briefly, the geometric point set was used to define a 3D surface, which was extruded to define a volumetric 3D finite element mesh. For the thickness, a constant value of 2 mm was assumed. Figure 2 , A shows this geometric model, which includes the levator ani muscles modeled as 1 structure. A schematic representation of the PVM is shown in Figure 2 , A with highlighted edges (in green) that were fixed. The main dimensions are described in Figure 2 , B along with the fiber directions of the model, which were defined on the basis of the literature. We noted that the area of the model hiatus was set to 32 cm 2 at the initial stage of the simulation (corresponding to the beginning of the second stage of labor) to be representative of pregnancy-induced changes and complete dilation by the beginning of the second stage of labor. This value ranged between 15 and 57 cm 2 , which was reported in late pregnancy. ,
Figure 3 shows the finite element model of the fetal head, which was obtained by a simplification of the model described by Parente et al. This simplification removed intricate details in the fetal head, such as the nose, which often caused numeric problems without being relevant to the simulations. Despite these simplifications, the main dimensions were maintained and represent a 50th percentile full-term fetus, with a cross-sectional area of 9199 mm 2 .
The initial position of the head was assumed as a vertex presentation with occiput anterior. The fetus descent was defined by the uterine and maternal forces, applied to a fetal head reference node. The head flexion, applied to the same reference node, was added so that the smallest head diameter was presented in the birth canal at all instances. Regarding the fetal boundary conditions, all displacements were free throughout the simulation. The lateral and transversal constraints were imposed by the rigid surface representing the pubic bone and pelvic floor muscles. Except for head flexion concerning the X-axis ( Figure 3 ), all other rotations were otherwise fixed. Please refer to the Appendix for further details on the finite element meshes.
Material models
The material model for the levator ani muscles incorporated several realistic properties, such as transverse isotropy, viscohyperelasticity, and the ability to weaken (accumulate damage) because of repetitive submaximal loading b
b Transverse isotropy occurs when the material stiffness is directionally dependent, which is the case for muscular tissues because of the presence of the fibers. Viscohyperelasticity describes the materials that have a nonlinear force-displacement behavior and time-dependent (viscous) properties. These materials have a force response to strain that, importantly, depends on the strain rate: faster strain rates induce higher forces, whereas lower strain rates induce smaller forces. Moreover, they can “relax”: particularly, if subjected to the same strain over some time, the stresses (force per unit area) developed initially decrease over time. Besides these properties, the material model incorporates a cumulative damage model: when submaximal loading cycles are performed, there is damage accumulation in the material, which negatively impacts its mechanical properties by weakening the material.
. In contrast, the fetal head and “pubic bone” surface, not so critical to the biomechanical model in terms of material properties, were simplified to increase computational efficiency. The fetal head was modeled as a hyperelastic material as the outer layer with a rigid interior. Hyperelasticity is a feature of all soft tissues, including the levator ani, as they can deform significantly. The “pubic bone” surface was modeled as an almost rigid elastic material with very high stiffness. Details on these material models can be found below, and in the Appendix .Levator ani muscles
A transversely isotropic viscohyperelastic model, including cumulative damage with nonlocal averaging, was used to simulate the levator ani muscle behavior. The transversely isotropic hyperelastic behavior was captured by the Holzapfel-Gasser-Ogden (HGO) model and viscoelastic behavior by the generalized Maxwell model, as described by Holzapfel and Gasser. These phenomenological models were chosen as they are relatively simple to implement (eg, compared with complex multiscale models) and have a reasonable number of material parameters (as the relationship between stress and strain was empirically modeled from the experimental observation). The HGO and generalized Maxwell model, although initially developed to characterize the behavior of arteries, , are robust and have been successfully applied to several soft biological tissues. , Moreover, the fatigue damage was simulated using a phenomenological model, whose evolution equation was previously developed on the basis of experimental evidence. Briefly, the damage accumulation is a function of the energy required to deform the material, which accumulates as the material suffers several deformations (loading cycles). As such, if the material deforms significantly, the energy required to deform it will be higher as will the damage accumulation. However, if a smaller deformation is applied several times, the energy required to deform will be the sum of the energy required for each deformation, also causing a considerable damage accumulation. The damage accumulation is measured quantitatively in the muscle fibers by the damage variable <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='Df1ˆ(0≤Df1ˆ≤1),’>𝐷𝑓1ˆ(0≤𝐷𝑓1ˆ≤1),Df1ˆ(0≤Df1ˆ≤1),
D f 1 ˆ ( 0 ≤ D f 1 ˆ ≤ 1 ) ,
with <SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='Df1ˆ=0′>𝐷𝑓1ˆ=0Df1ˆ=0
D f 1 ˆ = 0
representing the undamaged state and <SPAN role=presentation tabIndex=0 id=MathJax-Element-3-Frame class=MathJax style="POSITION: relative" data-mathml='Df1ˆ=1′>𝐷𝑓1ˆ=1Df1ˆ=1
D f 1 ˆ = 1
representing a complete failure. This damage variable will affect the stress response, mimicking the effect of reduced mechanical integrity observed experimentally, as <SPAN role=presentation tabIndex=0 id=MathJax-Element-4-Frame class=MathJax style="POSITION: relative" data-mathml='Df1ˆ’>𝐷𝑓1ˆDf1ˆ
D f 1 ˆ
approaches 1. In a previous work, these material models proved capable to capture the behavior of sheep pelvic floor muscle tissue, under repetitive submaximal loading.
Further details and equations regarding the constitutive model can be found in the Appendix with a table showing the material parameters used.
Fetal head
The hyperelastic part of the fetal head was modeled following the Neo-Hookean law (equation A.3 in the Appendix ). The parameter was obtained from the literature, <SPAN role=presentation tabIndex=0 id=MathJax-Element-5-Frame class=MathJax style="POSITION: relative" data-mathml='C10′>C10C10
C 10
=0.13 MPa, to represent a skinlike behavior.
Pubic bone surface
The surface that constrains the fetal head, the “pubic bone,” was modeled with a Young modulus of 1000 MPa and a Poisson coefficient of 0.25.
Maternal pushing
Rempen and Kraus measured the uterine pressure near the fetal head in 44 spontaneous deliveries with a vertex presentation. The results showed that the basal pressure averages 2.6 kPa and that a uterine contraction increased it to an average of 8.5 kPa. , A maternal push added an average of 10.5 kPa, thereby increasing the total pressure to 19 kPa at the peak of the uterine contraction. , These intrauterine pressures apply an expulsive force on the fetal head that can be calculated by multiplying each pressure by the fetal head cross-sectional area. The fetal head cross-sectional area is calculated in calculated in the Geometric model and boundary conditions section, and the resulting peak forces are listed in Table 2 .
| Peak uterine contraction (N) | 74 |
| Peak maternal push (N) | 101 |
| Peak uterine contraction + maternal push (N) | 175 |
To simulate directed pushing, three 10-second maternal pushes were added to the uterine contraction profile, according to the literature. , , , , As a key feature of the directed pushing can be seen as extended use of the Valsalva maneuver, we defined this pushing profile throughout the paper as the Valsalva-push pattern.
We assumed the volitional maternal effort would take 1 second to ramp up and 1 second to decrease ( Figure 4 , A). The profile of the combined efforts is shown in Figure 4 , B.
When spontaneously pushing, women reportedly push powerfully 1 time as the contraction reaches its peak or perform 3 to 5 smaller pushes (each lasting for 3–5 seconds with a 2-second interval). As such, we decided to standardize 3 possible spontaneous pushes, analyzing their effects separately. First, we tested the possible scenario of women pushing powerfully 1 time as the contraction reaches its peak. This is the designated 1-push pattern ( Figure 5 , A), with 1 powerful maternal push lasting 10 seconds ( Figure 4 , A) superposed symmetrically over the peak of the uterine contraction; moreover, for the remaining analyses, a shorter volitional maternal effort lasting 5 seconds was defined, with the profile represented in Figure 5 , B. As it is reported that women spontaneously perform between 3 and 5 smaller pushes, we tested separately the extremes and defined the 3-push pattern ( Figure 5 , C), with three 5 seconds pushes superposed on the uterine contraction and the 5-push pattern ( Figure 5 , D), with five 5-second pushes superposed on the uterine contraction.
Biomechanical model evaluation
To evaluate the complete biomechanical model, we analyzed the following parameters:
- 1.
The evolution of the vertical displacement of the fetal head: measured at the same reference node in the 4 simulations and plotted against time. This vertical displacement is useful for comparing the fetal descent between the 4 simulated pushing patterns; however, it does not correlate with any obstetrical measurements. We ended the simulations after the complete passage of the fetal head, which coincided with a vertical displacement of 153 mm.
- 2.
The stretch ratio of the hiatus contour: the contour of the hiatus was chosen ( Figure 2 , B: black dotted line) as it was shown to sustain the highest stretches. The stretch ratio ( <SPAN role=presentation tabIndex=0 id=MathJax-Element-6-Frame class=MathJax style="POSITION: relative" data-mathml='λ’>𝜆λ
λ
) was calculated by dividing the current length by the initial length: <SPAN role=presentation tabIndex=0 id=MathJax-Element-7-Frame class=MathJax style="POSITION: relative" data-mathml='λ=L/L0′>𝜆=𝐿/𝐿0λ=L/L0
λ = L / L 0
. This parameter can be related to strain, <SPAN role=presentation tabIndex=0 id=MathJax-Element-8-Frame class=MathJax style="POSITION: relative" data-mathml='ε’>𝜀ε
ε
, as: <SPAN role=presentation tabIndex=0 id=MathJax-Element-9-Frame class=MathJax style="POSITION: relative" data-mathml='ε=λ−1′>𝜀=𝜆−1ε=λ−1
ε = λ − 1
.
- 3.
The distribution of the damage variable in the muscle fibers, <SPAN role=presentation tabIndex=0 id=MathJax-Element-10-Frame class=MathJax style="POSITION: relative" data-mathml='Df1ˆ’>𝐷𝑓1ˆDf1ˆ
D f 1 ˆ
: this distribution was color coded as follows red, <SPAN role=presentation tabIndex=0 id=MathJax-Element-11-Frame class=MathJax style="POSITION: relative" data-mathml='1≤Df1ˆ<0.8′>1≤𝐷𝑓1ˆ<0.81≤Df1ˆ<0.8
1 ≤ D f 1 ˆ < 0.8
; orange, <SPAN role=presentation tabIndex=0 id=MathJax-Element-12-Frame class=MathJax style="POSITION: relative" data-mathml='0.8<Df1ˆ≤0.45′>0.8<𝐷𝑓1ˆ≤0.450.8<Df1ˆ≤0.45
0.8 < D f 1 ˆ ≤ 0.45
; yellow, <SPAN role=presentation tabIndex=0 id=MathJax-Element-13-Frame class=MathJax style="POSITION: relative" data-mathml='0.45<Df1ˆ≤0.35′>0.45<𝐷𝑓1ˆ≤0.350.45<Df1ˆ≤0.35
0.45 < D f 1 ˆ ≤ 0.35
; green, <SPAN role=presentation tabIndex=0 id=MathJax-Element-14-Frame class=MathJax style="POSITION: relative" data-mathml='0.35<Df1ˆ≤0.1′>0.35<𝐷𝑓1ˆ≤0.10.35<Df1ˆ≤0.1
0.35 < D f 1 ˆ ≤ 0.1
; and blue, <SPAN role=presentation tabIndex=0 id=MathJax-Element-15-Frame class=MathJax style="POSITION: relative" data-mathml='Df1ˆ≤0.1′>𝐷𝑓1ˆ≤0.1Df1ˆ≤0.1
D f 1 ˆ ≤ 0.1
. We defined a risk qualitative scale as follows: red areas represent “critical” risk, as the muscle fiber mechanical properties are affected up to 100% (complete tear); orange areas are “high” risk; yellow areas are “noteworthy” risk; green areas are “moderate” risk; and blue areas are “low” risk.
- 4.
The damage evolution at the critical location, which is the node of maximum predicted damage. We chose this approach instead of averaging around a set of nodes because a nonlocal averaging is already performed to reduce mesh dependency ( Appendix provides further details).
- 5.
The maximum damage at the critical location as a function of the total time of the simulation.
- 6.
The maximum damage at the critical location as a function of the time spent in active pushing. This time was measured by counting the number of maternal pushes performed and multiplying by their duration.
Results
The videos of the finite element simulations are available as Supplemental Material (Videos 1-4) .
The maximum stretch ratio was 1.87 for the 1-push and 5-push patterns, 1.83 for the Valsalva-push pattern, and 1.80 for the 3-push pattern. As the difference among the 4 simulations is merely the applied maternal pushes, it was expected to result in similar values of muscle stretches.
Figure 6 shows the vertical displacement of the fetal head with time for the 4 simulated pushing patterns. The spontaneous 3-push, 5-push, and the Valsalva-push patterns needed 4 complete contractions and reached the end in the first half of the 5th contraction. The spontaneous 1-push pattern took longer to reach the final stage, which occurred in the initial stage of the 6th contraction. Overall, the progress of the fetal head was similar. We can see that in the initial stage of each contraction, when the forces are slowly increasing, the fetal head did not descend. Descent only occurred when the maternal pushes were initiated. Moreover, one can see that the fetal head did not continuously advance but retreated briefly during the unloading and resting phases of the maternal push and during the uterine contraction resting phase. For the pushing patterns involving >1 maternal push, this effect was seen occurring more than once in a contraction, whereas the spontaneous 1-push pattern had a smoother pattern of fetal advance.
Figure 7 shows the damage distribution at the end of the simulation for the 4 simulated pushing patterns, with the location of the maximum value pinpointed. The critical risk location did not differ with the pushing patterns, consistently being near the right lateral origin of the PVM. The critical node is the point where the maximum damage was recorded, and this was the same for the 1-push, 3-push, and Valsalva-push patterns. The critical node of the 5-push pattern lay slightly (1.6 mm) further away from the other simulations’ critical nodes. Thus, the maximum location site was virtually the same for all tested pushing patterns.
Figure 8 shows a quantitative view of the distribution shown in Figure 7 . The boxplot includes the muscle fiber damage values predicted for every node of the pelvic floor muscles, grouping the values by the categories defined in the Materials and Methods section (critical, high, noteworthy, moderate, and low). The 5-push pattern has more critical, high, noteworthy, and moderate areas than the other pushing patterns, indicating that it is the pushing pattern that is predicted to cause the highest amount of damage accumulation. Moreover, it is followed by the Valsalva-push pattern. The 1-push pattern led to a similar damage distribution to the 3-push pattern, although more moderate-risk areas exist in the 1-push and more low-risk areas exist in the 3-push pattern.
1 ≤ D f 1 ˆ < 0.8
; high, <SPAN role=presentation tabIndex=0 id=MathJax-Element-17-Frame class=MathJax style="POSITION: relative" data-mathml='0.8<Df1ˆ≤0.45′>0.8<Df1ˆ≤0.450.8<Df1ˆ≤0.45
0.8 < D f 1 ˆ ≤ 0.45
; noteworthy, <SPAN role=presentation tabIndex=0 id=MathJax-Element-18-Frame class=MathJax style="POSITION: relative" data-mathml='0.45<Df1ˆ≤0.35′>0.45<Df1ˆ≤0.350.45<Df1ˆ≤0.35
0.45 < D f 1 ˆ ≤ 0.35
; moderate, <SPAN role=presentation tabIndex=0 id=MathJax-Element-19-Frame class=MathJax style="POSITION: relative" data-mathml='0.35<Df1ˆ≤0.1′>0.35<Df1ˆ≤0.10.35<Df1ˆ≤0.1
0.35 < D f 1 ˆ ≤ 0.1
; and low, <SPAN role=presentation tabIndex=0 id=MathJax-Element-20-Frame class=MathJax style="POSITION: relative" data-mathml='0≤Df1ˆ≤0.1′>0≤Df1ˆ≤0.10≤Df1ˆ≤0.1
0 ≤ D f 1 ˆ ≤ 0.1
.
Figure 9 , A shows the fatigue damage evolution at the critical node during each uterine contraction. For easier comparison, Figure 9 , B shows the critical node maximum value, reached at the end of the simulation. The Valsalva-push and 5-push patterns showed faster damage evolution and higher maximum values. The 3-push pattern induced the least maximum damage of all tested pushing patterns, although the damage accumulated faster than with the spontaneous 1-push pattern in almost all uterine contractions. Figure 9 , C and D shows an attempt to verify which time measurement is the better assessment of high-risk labor, in terms of maternal tissue damage. We can see that the simulation total time does not influence the maximum tissue damage ( Figure 9 , C), whereas the time spent in active pushing seems to exert some influence ( Figure 9 , D). However, we did not have enough data to confirm if there was a correlation between the time spent in active pushing and the maximum tissue damage.
Discussion
Principal findings
Our simulations demonstrated that maternal pushing efforts can influence the damage accumulation in the pelvic floor muscles. For all 4 maternal pushing patterns, the region with the most damage accumulation was near the origin of the PVM, which is in agreement with where the tears have been reported to occur using different imaging modalities. , , , The pushing pattern that minimized damage was, for each uterine contraction, 3 maternal pushes lasting 5 seconds. Performing the maternal effort for >5 seconds did not shorten the second stage of labor and caused more damage accumulation. The same happened when >3 pushes per contraction were performed.
Results in the context of what is known
This work allowed a quantitative view of the effect of vaginal delivery on regions of pelvic floor muscle that were under a remarkable degree of stretch, whichever maternal pushing pattern was used. The degree of tissue-stretch for all pushing patterns was >1.8, which represented strains >80%. This value mainly depended on the geometries of both the fetal head and pelvic floor, which explains why it was similar across the 4 simulations. Moreover, it was dependent on the degree of flexion of the fetal head, which was kept consistent across the 4 scenarios. However, as there was a different progression of the fetal head in each case, this caused slight differences in the degree of tissue strain calculated in the 4 scenarios. Our muscle stretch ratio results were consistent with published values that range from 1.64 to 3.26. ,
The predicted duration of the second stage of labor was similar for the 4 pushing patterns, ranging between 10 and 13 minutes. A systematic review that included 208,000 normal labors found the median durations for nulliparous women to range from 14 to 66 minutes (95th percentiles, 65–138 minutes) and multiparous women to range from 6 to 12 min (95th percentiles, 58–76 minutes). Although our results were shorter than the typical labor, they were still within the reported range, particularly for multiparous women. Moreover, our second stage of labor ended after the passage of the fetal head, and not the fetus, which would slightly increase our predicted second stage of labor duration.
All pushing patterns took the tissue to a critical point, with maximum damage values >0.8, at the same location of the maximum damage as to where tears have been reported to occur in the literature. , , , One can argue that the differences were not substantial enough to conclude; however, this was also in agreement with the literature. , , It may be one of the reasons that there is no consensus on the optimum pushing style from an injury standpoint, as there has not been a considerable difference in published studies (see Introduction). However, we can still answer our working hypothesis as the type of maternal pushing style did result in different damage accumulations and different maximum critical damage values. Regarding the second part of our hypothesis, which was that the smaller pushes are beneficial for the mother, the answer was not so straightforward. We found that the spontaneous 3-push pattern caused the least damage and the 5-push pattern was the riskiest. Hence, it seems that the frequency exerts an important influence and that >3 pushes per contraction is substantially detrimental without shortening the duration of the second stage of labor. In contrast, performing only 1 push per contraction might also not be adequate, because it lengthened the duration of the expulsion stage of the second stage of labor and the damage values were still higher than the 3-push pattern. Thus, we concluded that 3-pushes per contraction is the best approach, but each push should be <10 seconds. Maintaining each maternal push for a 10-second duration (Valsalva-push pattern) did not bring any benefits in terms of the duration of labor and caused more tissue damage than the 3-push pattern, for which each push lasted 5 seconds. From a broader perspective, as women do not breathe during the maternal push, reducing the maternal push to 5 seconds can help counteract the reduced fetal oxygenation and mean cerebral oxygen concentrations associated with directed pushing with approximately 10-second maternal pushes. , Moreover, shorter maternal pushes may be beneficial if we consider muscle fatigue c
c Note that muscle fatigue is a completely different process than the material fatigue discussed in this paper. Muscle fatigue is the decrease in the muscle’s ability to actively develop a sustained contractile force over time and when such contractions are repeated after rest intervals that prove insufficient for complete recovery of the muscle; it can result in a state of physical exhaustion. Material fatigue is a phenomenon in which structures fail when subjected to cyclic loading because of the progressive accumulation of damage at a microscopic level.
and maternal exhaustion. The peak force in a volitional push can only be maintained for 2 seconds, after which the force will decline over time. As such, maintaining the effort for 3 to 5 seconds will be more efficient from the point of view of maintaining a high expulsion force. Importantly, if the maternal pushes are performed for shorter periods, the intervening muscle recovery intervals are longer, helping to reduce the risk of maternal exhaustion. , Another interesting aspect of our results was that the total duration of the second stage of labor does not influence the maternal pelvic floor tissue damage. Although a prolonged second stage of labor is usually a factor to avoid, in terms of maternal injuries, it was not predicted to have an impact, and it might even be better than other management styles that cause a shorter stage of labor (ie, the 1-push pattern caused a longer simulated labor but showed less damage than the Valsalva-push and 5-push patterns). Although we do not have enough data to support this theory, our results did suggest that a better measurement for maternal risk is the time spent in active pushing, which is already well discussed in the literature.Clinical implications
There is currently no consensus as to the optimum pushing style for the management of the second stage of labor. Medical professionals usually favor the directed coaching (Valsalva-push pattern), although there is no evidence that it is the best approach. In contrast, the World Health Organization recommended that women in the expulsive phase of the second stage of labor should be supported to follow their natural desire to push as it helps them to feel more confident and in charge. Although focused on the maternal outcomes only, this study suggested that the goal should not be to shorten the second stage of labor but rather to provide conditions for the mother to sustain less damage to the pelvic floor with less risk of exhaustion. This has immediate implications for directed pushing, which, this work suggested, should be reevaluated to consider damage accumulation in the pelvic floor muscles. At least in terms of damage accumulation, the model suggests that women might benefit from a new, more flexible, coached pushing approach that reduces the active pushing time.
As this engineering approach did not allow us to consider geometric maternal and fetal variability, we chose the standard geometries of a gynecoid pelvis and a 50th percentile fetal head. The size of the fetal head is directly related to the imposed strain, so larger fetal heads have the potential to accumulate damage at a faster rate. A nongynecoid pelvis is known to be associated with a higher risk of labor complications, so if, as a consequence, the second stage of labor involves a higher number of maternal pushes, this would be riskier in terms of damage accumulation and tissue injury. As for the critical locations in those differently shaped pelvises, because of the large report of injuries at the origin of the PVM by different research groups, it is reasonable to assume it is most likely the critical location as well (although other levator ani muscles and pubic bone injuries have been reported to occur, although with lower incidence ).
Regarding the positioning of the fetus, it was shown by Parente et al that an occiput posterior presentation increases the degree of muscle strain. Hence, if in each maternal push the degree of muscle strain is higher, the damage accumulation in each contraction will also be higher. As such, an occiput posterior presentation can also be considered a riskier presentation in terms of material fatigue damage.
Research implications
We believe our study provided a visualization of the underlying mechanisms, present in the pelvic floor muscles during the second stage of labor. Unfortunately, it is beyond the scope of this article to verify the model damage predictions in vivo. However, magnetic resonance imaging already has been shown to be effective for visualizing the inflammation in the PVMs following a difficult labor, so higher resolution imaging could indeed be used to verify that the inflammation starts in the critical damage zones predicted by this study. These findings could undergird a clinical trial informed by these results to try to reduce the risk of these injuries in the future. In particular, we believe the outcomes of these 3 groups should be compared: directed pushing with 3 pushes per contraction lasting 10 seconds (Valsalva-push pattern), directed pushing pattern with 3 pushes per contraction lasting 5 seconds (3-push pattern), and a spontaneous pushing where the mother would be told to push when she feels the need to do so. Using G∗Power and methodology described in the literature, we estimated that the total sample size should be between 108 (effect size, 0.305; confidence level, 95%; power, 80%) and 420 (effect size, 0.152; confidence level, 95%; power, 80%). The effect sizes were obtained considering 3 of the results of maximum damage as means (0.82, 0.89, and 0.97) and equal individual standard deviations of 0.2 and 0.4. To better access maternal damage accumulation, the analyzed maternal outcomes must include follow-up appointments to assess hiatal area enlargement or other evidence of irrecoverable damage, besides levator ani tears. A clinical trial of this sort would differ from what is currently in the literature as it would compare different approaches for the directed pushing, with the potential to find a better approach for the maternal pushing style.
Strengths and limitations
The strengths of our approach included the use of a well-characterized material behavior for the pelvic floor muscles, which incorporated viscohyperelasticity, muscle fiber directions, and, for the first time, the ability to accumulate damage under repetitive submaximal loading. Although simplified, the geometric models were based on human data. The complete biomechanical model was able to identify the weakest region of the pelvic floor during the second stage of labor, and that region is in agreement with where the literature reports tears.
Limitations of our study included the simplified bowl-shaped pelvic floor geometry having a single hiatus. Moreover, we did not consider the resting pelvic floor geometry changing with age from that of a valley to more of a bowl shape or that the levator plate straightens in late pregnancy. In the future, it would be better to start the model with a pelvic floor geometry measured from a pregnant woman with both the genital and levator hiatuses represented. Nevertheless, we believe the simplifications we used would not affect the qualitative results from the four comparisons that we made. The second limitation concerned the material parameters used to describe pelvic floor tissue behavior, which were obtained from experimental data from nonpregnant ovine pelvic floor muscles. This decision was made because of the lack of any human experimental pelvic floor data regarding low-cycle damage accumulation, thereby pointing out a current knowledge gap. Further studies are needed to characterize the damage accumulation in human pelvic floor tissue in pregnant individuals (see below). Hence, quantitatively, our results must be viewed with care because of a possible overestimation of stress levels (as material parameters came from nonpregnant individuals, with less compliant tissues compared with pregnant). Nevertheless, as we compared 4 pushing patterns while keeping the remaining conditions constant (ie, geometry and material properties), we believe that qualitatively our results were valid and our conclusions would hold for different geometries and material properties. The third limitation was that we did not consider the effect of muscle fatigue c in the simulations, specifically the decrease in the peak contractile force that happens in a maximum isometric muscle contraction over time in addition to the decrease in peak force whenever isometric muscle contractions are repeated after intervening rest intervals. A fourth limitation was that although viscoelastic properties were included, we did not consider the effect of the large interindividual differences in levator viscoelastic properties that have been measured during labor ; these could affect the propensity for tissue injury. The fifth limitation was that we did not include the perineal body in our finite element model, which was shown to be protective, and so it could potentially lessen the damage accumulation in the levator ani. Minor limitations included the fact that we did not differentiate between muscle fibers and collagenous fibers in carrying the load through the ground substance of the pelvic floor muscle material, although the model-predicted locations of maximum damage were the origin of the PVM, where it connects to the pubic bone via an aponeurosis.
Conclusions
Fatigue damage accumulation was a possible PVM failure mode during the repetitive loading caused by the strenuous maternal pushing characteristic of the second stage of labor. Maternal pushing style influenced the risk of damage accumulation in the passive tissues of the PVM. Although the critical damage location did not differ across pushing styles, that location was confirmed to lie near the origin of the PVM. The optimal pushing style was 3-pushes per uterine contraction, maintaining each push for 5 seconds. That pattern caused the least amount of damage to the PVM without increasing the duration of the second stage of labor. We could not ascertain whether it is better to achieve that pushing pattern via directed or spontaneous pushing, but, if directed, the model suggests that women should sustain a push for <10 seconds.
Supplementary Data
Finite element simulation for the 1-push pattern.
Vila Pouca et al. Pathomechanics of levator injury during labor. Am J Obstet Gynecol 2022.
Finite element simulation for the 3-push pattern.
Vila Pouca et al. Pathomechanics of levator injury during labor. Am J Obstet Gynecol 2022.
Finite element simulation for the 5-push pattern.
Vila Pouca et al. Pathomechanics of levator injury during labor. Am J Obstet Gynecol 2022.
Finite element simulation for the valsalva-push pattern.
Vila Pouca et al. Pathomechanics of levator injury during labor. Am J Obstet Gynecol 2022.
Appendix
The finite element model of the pelvic floor muscles was composed of 20,804 hexahedral hybrid elements (C3D8H). The average volume of each finite element was 1.16 mm 3 , and the tissue volume of the pelvic floor muscles was 24,088 mm 3 , with a ratio between the average finite element volume and the total model volume of 0.005% ( Supplemental Figure 1 ) . Hexahedral elements were chosen because of their better precision than other 3-dimensional finite elements.
The fetal head was modeled using 19,520 tetrahedral elements (C3D4H). In this case, we considered tetrahedral elements the best choice because they perform faster, and no mechanical analysis was to be made of the fetal head. For both the fetal head and pelvic floor, hybrid elements were used as we believe it is the best choice when the material is incompressible or near incompressible. They have a mixed formulation that allows one to overcome numeric problems caused by incompressibility, such as volume strain “locking.” More details are available in the Abaqus Documentation.
The constitutive model applied to the levator ani muscles includes the Holzapfel-Gasser-Ogden model to capture the transversely isotropic hyperelastic behavior; the generalized Maxwell model, as described by Holzapfel and Gasser, to capture the viscoelastic behavior; and a fatigue damage model, previously developed based on experimental evidence, to capture the damage accumulation because of submaximal repetitive loadings. Here, we also included a nonlocal damage averaging of the integral type to attenuate the mesh influence on the damage contours.
As we wanted to simulate the behavior of a soft tissue, which is nearly incompressible because of the high water content, we can typically express the strain energy function in a decoupled form, defining a contribution for the volumetric and the isochoric components separately. In this way, applying a penalty function to the volumetric contribution will assure the material is incompressible. Moreover, the isochoric component can be further decoupled to add the specificities of the different constituents of the soft tissue. In particular, this model considers the presence of a single family of muscle fibers embedded in a ground substance (or ground matrix) material (a family of fibers is a group of fibers that are assumed to have a preferred direction). As such, the complete strain energy function can be defined in a decoupled form as:
Ψ=Ψvol+[(1−Dmˆ)⋅(Ψm¯+Ψmvsc)]+(1−Dfˆ)⋅(Ψf¯+Ψfvsc)
Ψ vol
is the volumetric contribution, <SPAN role=presentation tabIndex=0 id=MathJax-Element-23-Frame class=MathJax style="POSITION: relative" data-mathml='Ψm¯’>Ψm⎯⎯⎯⎯⎯Ψm¯
Ψ m ¯
is the isochoric contribution of the ground matrix, <SPAN role=presentation tabIndex=0 id=MathJax-Element-24-Frame class=MathJax style="POSITION: relative" data-mathml='Ψf¯’>Ψf⎯⎯⎯⎯Ψf¯
Ψ f ¯
is the isochoric contribution of the family of fibers, <SPAN role=presentation tabIndex=0 id=MathJax-Element-25-Frame class=MathJax style="POSITION: relative" data-mathml='Ψ∗vsc∗=m,f’>Ψvsc∗∗=m,fΨ∗vsc∗=m,f
Ψ ∗ vsc ∗ = m , f
are the viscous effects, and <SPAN role=presentation tabIndex=0 id=MathJax-Element-26-Frame class=MathJax style="POSITION: relative" data-mathml='D∗ˆ(0<D∗ˆ<1)∗=m,f’>D∗ˆ(0<D∗ˆ<1)∗=m,fD∗ˆ(0<D∗ˆ<1)∗=m,f
D ∗ ˆ ( 0 < D ∗ ˆ < 1 ) ∗ = m , f
are the nonlocal continuous damage variables. Both the viscous effects and damage variables affect each isochoric component of the strain energy. The indices m and f refer to the ground matrix and fibers contribution, respectively.
Each contribution can be defined as:
Ψvol=K(J2−12−ln(J))
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
