Birth trauma: short and long term effects of forceps delivery compared with spontaneous delivery on various pelvic floor parameters

Authors


Correspondence Dr S. Meyer, Urogynecology Unit, Department of Obstetrics and Gynecology, CHUV, 1011 Lausanne, Switzerland.

Abstract

Objective To compare the effects of forceps delivery and spontaneous delivery on pelvic floor functions in nulliparous women.

Design A longitudinal prospective study with investigations during the first pregnancy, 10 weeks and 10 months after delivery.

Setting Antenatal clinic in a teaching hospital.

Population One hundred and seven patients aged 28.4 years, divided into those with forceps (n= 25) or spontaneous (n= 82) delivery.

Methods Investigations with a questionnaire, clinical examination, assessment of bladder neck behaviour, urethral sphincter function, intra-vaginal/intra-anal pressures during pelvic floor contractions.

Results The incidence of stress urinary incontinence was similar in both groups at 9 weeks (32%vs 21%, P= 0.3) and 10 months (20%vs 15%, P= 0.6) after delivery, as was the incidence of faecal incontinence (9 weeks: 8%vs 4%, P= 0.9; 10 months: 4%vs 5%, P= 1) and the decreased sexual response at 10 months (12%vs 18%, P= 0.6). Bladder neck behaviour, urethral sphincter function and intra-vaginal and intra-anal pressures were also similar in the two groups. However, 10 months after delivery, the incidence of a weak pelvic floor (20%vs 6%, P= 0.05) and the decrease in intra-anal pressure between the pre- and post-delivery values (−17±28 cm H2O vs 3±31 cm H2O, P= 0.04) were significantly greater in the forceps-delivered women.

Conclusions Forceps delivery is not responsible for a higher incidence of pelvic floor complaints or greater changes in bladder neck behaviour or urethral sphincter functions. However, patients with forceps delivery have a significantly greater decrease in intra-anal pressure and a greater incidence of a weak pelvic floor.

INTRODUCTION

Urinary incontinence is a common worldwide problem, which costs more than 10 billion dollars annually in the United States alone1.

Of the various aetiologies of incontinence, stress urinary incontinence is the predominant type, with prevalence rates of 15–52%2. Incontinence is common during pregnancy, its prevalence being estimated as 30–60%3. Although urinary incontinence (stress or mixed) can already be present during pregnancy, it is generally agreed that stress urinary incontinence results from pelvic floor trauma during vaginal delivery. The importance of increasing parity and pelvic floor birth trauma is emphasised by some authors4,5, but other studies did not confirm these findings6,7. In addition, although the link between vaginal delivery and the development of urinary incontinence and other pelvic floor complaints (i.e. diminished intensity of sexual vaginal response and faecal incontinence) is well established, the precise aetiology of such symptoms depends on many parameters and frequently remains unclear. Forceps delivery has been suggested as sometimes being responsible for major pelvic floor trauma8,9, with the subsequent appearance of stress urinary incontinence and faecal incontinence.

The objective of our study was to assess the effects of forceps-assisted delivery and spontaneous vaginal delivery on urethral sphincter and pelvic floor function in nulliparous women investigated during pregnancy and at 9 weeks and 10 months after delivery.

METHODS

One hundred fifty-one white nulliparous women (29 ±4 years of age) gave informed consent before being enrolled in this study, which had been previously approved by the Ethics Committee of the University of Lausanne. The enrollment period was two years and each eligible subject was asked to participate. Women with pregnancy complications (twin pregnancy, haemorrhage due to low-inserted placenta, diabetes or premature labour), those in early labour and those with a history of urinary infections were excluded from the study. The acceptance rate was about 80%. Seventy-four percent of women enrolled in this study were seen in the private practice of the obstetrician-gynaecologist who was the principal investigator, 26% in the outpatient clinic of the obstetrics department. The subject population was composed as follows: 15% teachers—university researchers, 14% nurses and health-related professions, 42% office workers and saleswomen, 3% craftswomen, and 26% housewives.

In this cohort of 151 women, 33 (22%) had a caesarean section, 82 (54%) had a spontaneous delivery, 25 (17%) had a forceps-assisted delivery. Eleven women did not leave complete the study (drop-out rate: 7%), mostly because they moved out of the area.

The principal investigator (S.M.) explained the details of the study and, with the help of three nurses trained in urodynamics, investigated the women.

Firstly, a careful history was taken, with particular attention being paid to urinary infections, urinary incontinence (particularly stress urinary incontinence) and fecal incontinence. The classification of anorectal symptoms was as follows: grade 1: flatus incontinence, grade 2: liquid stools incontinence, grade 3: solid stools incontinence. The same questionnaire was repeated monthly to detect the appearance of new urinary or anorectal symptoms. Pre- and post delivery specific sexual anamnestic data (presence of vaginal climax before delivery, time to reach vaginal climax and intensity of vaginal climax after delivery) were compared using this questionnaire.

A conventional clinical examination was then performed with the mobility of the uterus-vaginal walls being determined by the halfway system of Baden and Walker10 (grade O: normal position for each respective site, grade 1: descent halfway to the hymen, grade 2: descent to the hymen with the patient straining hard) and the strength of pelvic floor muscular contraction being assessed subjectively on the normal scale of 0–511. A weak pelvic floor was defined as a pelvic floor muscular contraction tested with a grading inferior or equal to 3 (Pelvic floor contraction assessment: Grade 0: nil, Grade 1: flicker, Grade 2: weak and Grade 3: moderate).

Assessment of the bladder neck position at rest and stress and its mobility during the Valsalva manoeuvre and pelvic floor contraction (perineosonography), was performed by the same observer with the patient in the supine and standing positions with the bladder containing 200–300 mL (Bladder volume was estimated using the formula: transverse diameter × horizontal diameter × sagittal diameter × 0.7)12.

A curved linear array ultrasound scanner (Tosbee, Toshiba Corporation Medical Systems SA, Tokyo, Japan) with a 3.75 MHz transducer was placed on the vulva in a sagittal orientation to provide a complete view of the bladder base and neck, the urethra and the whole pubic symphysis. The method of bladder neck localisation in space was that described by Schaer et al.13 using a rectangular co-ordinate system, taking the inferior border of the pubic symphysis as the reference point.

Finally, each patient underwent conventional urodynamic testing of urethral sphincter function by means of supine and standing stress urethral pressure profiles using a Microtip transducer (Gaeltec Microtip, 7 F diameter, Gaeltec Manufactory, Isle of Sky, Scotland) placed at the 9 o'clock position in the urethra and using a withdrawal rate of 1 mm/second. The following parameters were determined: functional urethral length, maximum urethral closure pressure at stress and area of continence at stress (defined as the area in mm2 between the baseline and a line connecting each cough spike on the urethral closure pressure cough profile). Intra-vaginal and intra-anal pressures during pelvic floor contractions were also measured using an air-inflated balloon connected to a microtip transducer (Sedia Se 250, SEDIA AG, Fribourg, Switzerland) inflated to touch the vaginal and anal canal walls.

This examination was carried out at 12–20 weeks in 11 women, 21–30 weeks in 41 women and 31–39 weeks in 55 women. On average, this first examination was done at 30.7 weeks of pregnancy in the patients delivered-spontaneously and at 28.6 weeks of pregnancy in forceps-delivered patients (P= 0.3). As it has been demonstrated in a previous study, the timing of this first examination has no bearing on the different pelvic floor parameters because values show no significant differences in the three trimesters of pregnancy14.

Follow-up examinations were at 9 ±2 weeks and 10 ±2 months after delivery for each patient. The assessors were blinded to mode of delivery. The patients were divided into two groups depending on their mode of delivery, with group 1 (n= 25) consisting of women with forceps deliveries, and group 2 (n= 82) of women with spontaneous vaginal deliveries. The percentage of patients who underwent urogynaecologic re-education between 3 and 6 months after delivery was not significantly different in the two groups (group 1:16 patients, group 2:35 patients, P= 0.1)

The incidence of median and medio-lateral episiotomy and epidural anaesthesia was significantly higher in the forceps delivered group compared with the group of women with spontaneous deliveries (100%vs. 80%, P= 0.02 and 90%vs. 50%, P= 0.0001, respectively). Baby's mean birthweights and head circumferences (SD) were similar in the forceps and spontaneous delivery group, 3290 gr (390) vs. 3620 gr (425) and 34 cm (1.6) vs. 35 cm (1.3), respectively.

The duration of the second stage of labour was greater in the forceps-delivery group compared with the spontaneous-delivery group, but this difference was not significant (Forceps-delivered group: 35±25 min / Spontaneous deliveries group: 25±14, P= 0.1). The indications for forceps deliveries were a prolonged second stage of labour (90%) and fetal distress (10%).

Changes in the various indices according to the time of examination and mode of delivery were tested with an ANOVA model. Parameters measured in the two groups of patients were compared with Student's unpaired two-tailed t test and, in the case of non-normal distribution, with a Mann-Whitney nonparametric test. For comparison of frequencies, a two-sided Fisher's exact test was used. A P value < 0.05 was considered significant.

RESULTS

The incidence of stress urinary incontinence, decreased intensity of sexual vaginal response and faecal incontinence is listed in Table 1. There were no significant differences between the two groups. As 48% of the women in the forceps-delivered group and 40% in the spontaneously-delivered group had not resumed a regular sexual activity two months after delivery, this parameter was not considered at the examination two months after delivery.

Table 1.  Incidence of stress urinary incontinence, decreased vaginal sexual response and faecal incontinence in forceps-delivered and spontaneously-delivered women during pregnancy (1), two months after delivery (2) and 10 months after delivery (3). Values are given as n (%).
 Forceps delivery (n= 25)Spontaneous delivery (n= 82)
ExamStress incontinenceDecreased sexual responseFaecal incontinenceStress incontinencePDecreased sexual responsePFaecal incontinenceP
14 (16)0028 (34)0.10 0 
28 (32)2 (8)17 (21)0.3 3 (4)1
35 (20)3 (12)1 (4)12 (15)0.515 (18)0.64 (5)1

The incidence of cystocoeles Grade 1 (group 1:40%, group 2:43%, P= 1), grade 2 (group 1:0%, group 2:10%, P= 0.2), uterine prolapse Grade 1 (group 1:4%, group 2:8%, P= 0.7) and rectocoeles grade 1 (group 1:4%, group 2:2%, P= 0.5) was the same in both groups. However, the incidence of a weak pelvic floor was significantly increased at both 2 and 10 months after delivery in the forceps-delivered women (Table 2).

Table 2.  Incidence of patients with a weak pelvic floor (i.e. grade 0–3) 2 months (2) and 10 months (3) after delivery in forceps-delivered and spontaneously-delivered women. Values are given as n (%).
ExamForceps delivery n= 25)Spontaneous delivery (n= 82)P
100 
221 (84)29 (35)0.0001
35 (20)5 (6)0.05

The values for bladder neck position in the standing position at rest and its mobility during bearing down efforts, measured in the supine position after delivery are shown in Table 3. There were no significant differences between the groups. However, the ANOVA model calculation for bladder neck position in the standing position shows a significant difference between the two groups considering the time factor and the group factor (P < 0.05). This difference can be attributed to the higher values of bladder neck position at rest in the forceps group, which were present from the beginning of the study. Furthermore, the decrease between pre and post delivery bladder neck position is not significant when two group were compared (group 1: 3.7±7 mm, group 2: 1±8 mm, P= 0.2)

Table 3.  Comparison of bladder neck position at rest and at stress in the standing position and of bladder neck mobility in the supine position in forceps-delivered and spontaneously-delivered women during pregnancy (1) and 2 months (2) or 10 months (3) after delivery. Numbers are mean (SD).
 Forceps delivery (n= 25)Spontaneous delivery (n= 82)
 1231P2P3P
Bladder neck         
 Position at rest (standing) (mm)30 (5)25 (4)28 (5)26 (5)0.00925 (6)0.427 (4)0.4
 Mobility (supine) (mm)10 (4)13 (6)13 (5)12 (5)0.0615 (5)0.115 (6)0.08

The values for urethral functional length, maximal urethral closure pressure and area of continence at rest and at stress in the standing position are shown in Table 4. Again, there were no significant differences between the groups. The significantly higher urethral function length value in spontaneous delivery compared with the forceps group measured during pregnancy was also incidental.

Table 4.  Comparison of urethral functional length (FL), maximal urethral closure pressure (MUCP) and area of continence at rest and at stress in the standing position in forceps-delivered and spontaneously-delivered women during pregnancy (1) and 2 months (2) or 10 months (3) after delivery. Numbers are mean (SD).
 Forceps delivery (n= 25)Spontaneous delivery (n= 82)
 1231P2P3P
Functional length (mm)31 (8)29 (7)31 (7)35 (5)0.0530 (7)0.731 (7)0.7
MUCP at rest standing (cm H2O)74 (27)83 (29)81 (26)82 (32)0.284 (27)0.887 (28)0.4
Area of continence at rest at standing (mm2)1441 (600)1313 (600)1386 (641)1385 (598)0.71348 (514)0.81474 (561)0.5
MUCP at stress standing (cm H2O)81 (29)84 (28)86 (24)75 (31)0.581 (30)0.786 (28)0.9
Area of continence at stress standing (mm2)813 (767)676 (362)605 (260)597 (462)0.2610 (337)0.4601 (330)0.9

Intra-vaginal and intra-anal pressures during pelvic floor squeezing are listed in Table 5. There were no significant difference between the two groups two months and 10 months after delivery. However, when the changes between the pre- and late post-delivery values were calculated, forceps group showed a significantly greater decrease in intra-anal pressure measured 10 months after delivery compared with the spontaneous delivery group (Table 6).

Table 5.  Comparison of intra-vaginal and intra-anal pressures during pelvic floor squeezing in forceps-delivered and spontaneously-delivered women during pregnancy (1), 2 months (2) or 10 months (3) after delivery. Numbers are mean (SD).
 Forceps delivery (n= 25)Spontaneous delivery (n= 82)
 1231P2P3P
Intra- vaginal pressure (cm H2O)44 (29)31 (20)35 (22)37 (20)0.131 (22)0.938 (26)0.6
Intra-anal pressure (cm H2O)53 (28)32 (18)37 (20)44 (21)0.134 (21)0.741 (23)0.5
Table 6.  Comparison of intra-vaginal and intra-anal pressure changes (decrease or increase) during pelvic floor squeezing before delivery and 10 months after delivery (1–3) in forceps-delivered and spontaneously-delivered women.
 Forceps delivery (n= 25)delivery Spontaneous (n= 82)P
Intra-vaginal pressure   
changes (cm H2O)−8 (36)1 (31)0.2
Intra-anal pressure   
changes (cm H2O)−17 (28)−3 (31)0.04

DISCUSSION

Vaginal delivery is known to be associated with pelvic floor trauma, caused by pressure and distension/stretching from the fetal head and shoulders during the maternal expulsive efforts in the second phase of labour. These functional and anatomic changes in the connective tissues, muscles and nerves of the pelvic floor are believed to be responsible for the subsequent development of stress urinary incontinence, anal incontinence and a diminished sexual vaginal response. In a previous study15, we found that in women investigated two months after spontaneous vaginal delivery, stress urinary incontinence and faecal incontinence were present in 21% and 5%, respectively. In these women, bladder neck mobility was significantly increased, whereas intra-vaginal and intra-anal pressures during pelvic floor squeezing were significantly decreased when compared with their pre-delivery levels.

The obstetrician's aim is to seek to minimise the risk of muscular and neurological injury to the pelvic floor. Many risk factors have been suggested as being responsible for such birth trauma; including median episiotomy, nulliparity, birthweight, fetal occiput posterior position. We found that only the weight of the baby was weakly, but significantly, correlated with a decreased intra-anal pressure (r: 0.24, P= 0.01)14.

Forceps delivery is thought to be a major factor in the development of such trauma. Forceps delivery represents about 15% of vaginal deliveries and is used to quickly deliver an engaged baby or to shorten the second phase of labour16. Although the usefulness of such a technique is evident, effects on different pelvic floor parameters, especially urinary continence function and intra-vaginal pressures during pelvic floor squeezing, have not previously been studied and compared with the normal nonassisted delivery.

Sultan et al.* have described the traumatic effects of vaginal delivery on ano-rectal function and found that 3% of primiparous women had clinically apparent injuries to the anal sphincter after delivery, however, endosonography revealed occult anal sphincter damage in 35% of women six weeks after delivery8. Eighty percent of women who undergo forceps deliveries have subclinical sphincter defects not seen in women delivered by vacuum. In a further study, the same authors underlined the specific traumatic effects of forceps delivery causing third or fourth degree tears, the primary repair of which has been shown to result in a bad outcome with subsequent anal incontinence in 29–48% of women 3 months to 3 years after such repair17–19. Furthermore, in another study of 43 women with instrumental delivery, Sultan et al. found20 that 81% of forceps deliveries compared with 24% of vacuum deliveries were associated with sonographic anal sphincter damage. Johannson et al.21, in a randomised study of 600 women, also found a significantly higher incidence of maternal injury after forceps delivery compared with vacuum delivery. This occult sphincter damage can persist for many years, so the same authors performed a study to determine the prevalence of occult anal sphincter trauma 5 years after randomisation to forceps and vacuum delivery. They found occult sphincter defects in 82% of forceps, and 48% of vacuum deliveries, with a significant fall in maximum squeeze anal pressure in the forceps group compared with the vacuum group (56 vs 36 mmHg;P= 0.0007)22.

Our findings in general agree with those in other studies. Although the prevalence of faecal incontinence and the values for intra-anal pressure during pelvic floor squeezing were almost the same in the two groups when compared 10 months after delivery, the intra-anal pressure decrease was significantly higher after forceps delivery than after spontaneous delivery when pre-delivery and 10 month post-delivery values were compared (−17±28 cmH2 O vs−3±31 cmH2 O, P= 0.04). In addition, clinical digital testing of pelvic floor muscular strength also showed a significantly higher incidence of a weak pelvic floor at 10 months in the forceps delivery group (20%vs 6%, P= 0.05)

Although the problems of patent and occult ano-rectal function changes have been extensively studied for several years, the specific alterations in urethral sphincter function induced by forceps delivery and spontaneous delivery have not previously been compared. As forceps delivery is considered to be more traumatic to the pelvic floor than spontaneous delivery, we were surprised to find no significant differences in stress urinary incontinence incidence between the two groups 10 months after delivery. A similar result was obtained by Foldspang et al.23 who were unable to demonstrate a significant relationship between stress urinary incontinence and forceps delivery in a large cohort of 4345 women who were mailed a self-administered questionnaire focusing on urinary incontinence and other health variables. Similarly, Mellier et al.24, in a study of 265 patients questioned in immediate postpartum, found that postpartum incontinence did not increase significantly when forceps were used, while Dimpfl et al.25 concluded that patients with increased trauma to their pelvic floor, as in forceps delivery, show a higher, but not statistically significant incidence of persistent postpartum stress urinary incontinence.

Similarly, forceps delivery seemed not to have more deleterious effects than spontaneous delivery on the bladder neck anchoring system, judging by the fact that the bladder neck position in the standing position or bladder neck mobility in the supine position were not significantly different in the two groups. Surprisingly, forceps deliveries had no greater deleterious effects on urethral sphincter function measured by urethral pressure profiles at rest and at stress in the standing position. Urethral functional length, maximal urethral closure pressure and especially the area of continence at rest and at stress were not significantly decreased by forceps delivery. Sacco et al.26 also found in a retrospective study of 553 patients investigated with urodynamic evaluation that forceps delivery was not associated with decreased urethral competence and a significant risk of subsequent incontinence.

The quality of sexual vaginal response was assessed 10 months after delivery when all the patients had regained normal sexual activity. The incidence of changes in sexual vaginal response, i.e. the time to reach climax and/or intensity of the climax, was the same in the two groups (forceps group: 12%, spontaneous delivery group: 18%, P= 0.6). Three patients in the spontaneous delivery group described a complete absence of sexual vaginal response when investigated 10 months after delivery. These three patients recovered a sexual vaginal response, although decreased as compared with the pre-delivery level, after 12 sessions of electromyostimulation and biofeedback.

The pudendal nerves are of central importance in normal sexual activity, carrying excitatory information from the perineum, parasympathetic stimuli to the clitoris and somatic impulses to the bulbo-cavernosus, ischiocavernosus and pubo-coccygeous muscles. During vaginal orgasm, the striated muscles around the vagina contract, thereby producing an increased pressure in the vagina27. Because of this, the sexual response can be modified by pelvic floor trauma due to vaginal delivery. However, intra-vaginal pressure during pelvic floor contractions did not correlate with changes in sexual vaginal response in our patients with a diminished sexual vaginal response, who had a higher intra-vaginal pressure than patients with an unmodified sexual response (44±8 cm H2O vs 37±3 cm H2O, P= 0.4) This lack of correlation between intra-vaginal pressure during pelvic floor squeezing and the quality of the sexual response is due to the complexity of the physiological pathways necessary for establishing a sexual vaginal climax, which involve vaginal mechanical receptors (mostly located in the anterior vaginal wall), normal sensory pathways to the S2–S3 dorsal roots of the medullar cord, a synaptic connection to the pudendal nerves innervating the striated muscles of the pelvic floor, and, finally, connections to the CNS27. These structures can be involved either separately or in combination in the birth trauma mechanisms of a vaginal delivery.

Considering the patients with forceps deliveries involved in this study, we can say that this mode of delivery seems not to be responsible for greater birth trauma than spontaneous delivery in terms of ‘pelvic floor complaints’. In addition, forceps delivery seems not to induce greater deleterious effects on urethral sphincter function or on the bladder neck anchoring system than spontaneous delivery. The greatest effect of such assisted delivery is seen on the posterior compartment of the pelvic floor, with more frequent weak pelvic floor muscle contraction and subsequently a greater intra-anal pressure decrease without an increase in the incidence of faecal incontinence.

Ancillary