Evaluation of Suzor forceps training by studying obstetric anal sphincter injuries: a retrospective study

Instrumental deliveries are an unavoidable part of obstetric practice, with an operative delivery rate that depends on the medical center (varying from 5.3 to 34.1% of all deliveries) [1]. The choice of the instrument differs depending on local habits and personal selection. The key instrument at the Limoges University Hospital is the Suzor forceps (SF; forceps with parallel shafts). The second choice is the Kiwi OmniCup vacuum extractor (VE) and Thierry’s spatula is hardly used.

The main maternal complications associated with forceps delivery (FD) are perineal tears, which are more frequent than with the VE [2, 3]. Perineal tears can be classified into four categories of severity [4]. In the long run, anal incontinence appears to be more common after FD than after VE or spontaneous vaginal delivery [5‐7].

The main purpose of the study was to determine how many operative deliveries by Suzor forceps, and how many semesters (or months), it would take for a resident to be entrusted with instrumental delivery, by evaluating the rate of obstetric anal sphincter injuries (OASIS). Maternal, obstetrical and fetal factors that could affect the outcome of perineal complications were also studied.

This retrospective study was conducted in the Gynecology and Obstetrics Unit of the Limoges Mother and Child Hospital (LMCH), in France, a level 3 maternity unit, from November 1st 2008 to November 1st 2016. The study was approved by the Limoges Regional University Hospital institutional review board (306–2019-72).

The authors retrieved each resident’s teaching plan, including every semester completed until then. Residents were assumed to be the main delivering physicians every time their names were written on the patient’s file. The authors considered the first semester of attendance at LMCH to be the resident’s first level of experience. Each resident increased his/her obstetrical experience from one level to another every semester during which he/she performed FD at LMCH.

Inclusion criteria were instrumental deliveries using the SF, or sequential use of the Kiwi OmniCup° vacuum and then forceps by residents who started their training after November 2008, without mentoring by a senior physician, in singleton pregnancies with a cephalic presentation, regardless of the term or the level of engagement of the fetus in the genital tract. Sequential use of the vacuum extractor and then forceps was studied as well (VESF).

FD that involved a senior physician were excluded, as were failed operative deliveries, twin pregnancies, fetuses in the breech or transverse position, in utero fetal death, and medical termination of pregnancy. All files with incorrectly filled reports or that neglected to state clearly whether the senior physician participated in the instrumental delivery were also excluded.

The main outcome was the rate of severe perineal tears, 3rd or 4th degree, which are considered as obstetric anal sphincter injuries (OASIS). Sultan’s classification, used since 2007 by the Royal College of Obstetricians and Gynaecologists [4], was used to grade the perineal lesions. These OASIS were clinically suspected by residents and confirmed by a senior obstetrician.

Several risk factors for OASIS were studied (delivering physician’s experience, and maternal, fetal and obstetrical criteria). Obstetric practices did not change during the study period (episiotomy policy, for example).

Data were analyzed using JMP 12.0.1 software (SAS Institute, Cary, USA) and the results were presented as mean values, medians, standard deviations and percentages. Univariate and multivariate analysis was performed to assess the risk of OASIS given the obstetric experience of the delivering physician, controlling for confounding factors. Comparison of variables was done using the Chi-2 test with a significant cut-off determined as 5%. Odds ratios were implemented for some variables. A logistic regression analysis was performed.

During the 8-year study period, 21 residents performed 1253 instrumental deliveries, which represented 45% of all instrumental deliveries. Forceps were chosen as the delivery instrument in 1164 cases (92.9%), and in 89 cases (7.1%) the VE was used first and then forceps (Fig. 1).

At the time of the study, 10 senior residents completed their training and 11 junior residents were still undergoing training. Each of them had attended the LMCH for at least one semester during which they performed at least one FD.

One resident performed on average 11.6 (+/− 6.5) FD each semester. Older residents performed more instrumental deliveries per semester than younger ones (840 FD in total for the older residents and 12.9 FD per semester (± 6.4) vs 413 and 9.6 (± 6.0) per semester for the younger residents). The number of instrumental deliveries performed by residents each semester increased with their seniority and thus so did their obstetrical and instrumental experience (Fig. 2).

The mean time lapse before the first operative delivery by forceps was 84 days (+/− 81 days). The older residents took on average 49 days before their first use of forceps, and younger residents about 116 days (95% CI [53–180]; p <  0.018) (Table 1).

At Limoges university hospital, the operative delivery rate was approximately 13.2% (Fig. 1), which is consistent with the rate of 12% reported in national perinatal surveys in France, a rate which has been stable since the 1980s, and with the wide range of values among centers reported in the international literature: 5.3 to 34.1% [1‐8]. The key instrument at LMCH is the Suzor forceps, the second choice is the Kiwi Omnicup vacuum extractor, and Thierry’s spatula is hardly used. Although questionable, the choice of this instrument depends on local habits and personal selection. Contraindicated in some countries, the sequential use of VE and then forceps is controversial in France and occasionally practiced in our maternity unit.

Few studies have focused on the learning curve of residents concerning operative delivery. A team in Nice (south of France), reported a cut-off of 20 operative deliveries using Thierry’s spatula after which residents subjectively felt that they could perform operative deliveries without supervision [9]. The same team also showed in another study based on objective criteria, such as those used in the present study (maternal perineum), that there was an additional risk of OASIS if the operative delivery by spatula was performed by an inexperienced resident compared to a resident who had completed at least 5 semesters [10].

In the present study, we found that residents caused statistically fewer OASIS after 2.36 semesters (14 months) or after 24 FD. A clear decrease in OASIS was noted between the 3rd and 4th semesters (Fig. 3). These results are therefore comparable to those of the team in Nice [9]. In fact, the ranking of the different semesters of attendance by the residents started with the first semester of training at LMCH, which is mostly during the 2nd or 3rd semester of the whole residency program. Given the increasing number of residents in each year, the 4th semester of obstetrical experience in Limoges would match Nice’s 5th semester, which is when the authors of the Nice study consider residents to be sufficiently trained to perform operative deliveries.

The OASIS rate after at least 3 completed semesters was about 3.7%, which matched the rates reported in the literature after FD. A study of 284,783 births in Holland showed stage 3 and 4 perineal tear rates of 1.7% without operative delivery, 4.6% with FD, and 7.8% after VESF [11].

There are few recommendations about the number of operative deliveries to be performed by residents before the empowerment phase. Dupuis et al. consider that 40 operative deliveries including VE, FD, and breech delivery are needed [12]. In our hospital, this would be possible after 3 completed semesters, since a resident performs on average 10 FD per semester (for the new generation), but also operative delivery by VE or breech delivery.

Instrumental delivery is conventionally taught via bedside training, notably with close supervision of the resident by a senior physician. This is referred to as ‘mentoring’. However, because of the significant maternal and fetal morbidity that can be caused by FD, it is essential that it is “never the first time on the patient”. There are many questions concerning ways to improve training of obstetricians without increasing the risks for patients and their newborn. Furthermore, mindset changes concerning the medical field and legal aspects do not facilitate the learning curve [10].

The establishment of a ‘resident log’ of all instrumental deliveries (simulated and in vivo) would allow the senior physician to evaluate the experience of the mentored resident. Furthermore, it would let residents be entrusted with instrumental deliveries by forceps, as long as they had performed at least 24 operative deliveries with close mentoring [13]. There is also the additional optional training of simulated instrumental delivery. In the near future, the French national training program in obstetrics will include obstetrical simulation sessions.

The increasing number of medical students admitted to medical school raises new issues: it is becoming harder during residency to accumulate the minimum number of operative deliveries required for self-sufficiency. The mean number of operative deliveries per semester subsequently decreases with each new generation, as already observed in the present study.

To date, in France, there is no performance evaluation allowing an operator to be declared as fit to perform operative deliveries, or not. Nevertheless, each instrument (vacuum, forceps, spatula) requires proper theoretical and practical training in order to master its specific and proper use (handling, articulation, positioning, and traction).

The practical recommendations of the French national college of obstetricians and gynecologists in 2008 [5] stipulate that teaching and learning of operative delivery must include teaching of the use of forceps, vacuum and spatulas, as these instruments are complementary, and acknowledge that the dangers of operative deliveries are related to the experience of the operator performing them. Knowledge of 2 operative delivery methods is recommended [14, 15], and the choice of instrument should be guided by the clinical indications and not by the operator’s preferences.

The inclusion of simulation in teaching programs would help to raise the number of operative deliveries per resident and would enable experience to be acquired outside an emergency context, while evaluating the performance of the obstetricians during their initial and further education. Vieille et al. proved that it would allow residents to achieve a gain in both theoretical knowledge and practical skills [16]. Dupuis et al. observed the necessity of performing 31 FD in the occipito-posterior position and 62 in the inclined position, at least in simulation, in order to master operative delivery [12]. Improvement of residents’ training programs would eventually lower the rate of avoidable fetal and maternal complications [13]. Access to simulation could make it possible to become empowered more quickly. In the USA, simulation is well established [17]. It is a proper teaching tool and simulation centers work as a network to enhance their resources. It is used to certify health professionals and to accredit medical centers giving proper references. In France and in Europe, those centers lack resources even though they are increasing in prevalence [18]. The national health authority plans to promote the expansion of simulation.

An operator’s lack of experience is not the only cause of fetal and maternal morbidity and mortality. Other known OASIS risk factors were analyzed in our study [10‐16, 18‐24]. We found statistically more OASIS for FD in the posterior position, in line with the literature data [20, 21]. The systematic use of intrapartum ultrasound to detect fetal position could be helpful in order to decrease OASIS by optimization of the direction of the operative delivery. A protective effect of medio-lateral episiotomy was observed. Concerning this particular factor, studies are inconsistent. Several reviews have shown no positive effect of systematic episiotomy on OASIS [22‐25]. Others report that reducing the indication for episiotomy would raise the rate of first and second degree perineal tears, but not third and fourth degree perineal tears [26]. De Leeuw et al. reported a protective effect of episiotomy on the perineal area during FD [27]. Their article was widely discussed in France, because it went against the national clinical recommendations of 2006 concerning instrumental deliveries (against the liberal use of episiotomy during instrumental deliveries; allow the operator to assess clinically whether or not to use episiotomy) [28]. Nowadays in France, whether or not an episiotomy is performed during instrumental delivery still depends on the operator’s clinical assessment during the birth. In our study, there was no episiotomy in 10% of cases, because of the resident’s clinical assessment or a lack of time resulting in limited manual perineal control, which is essential to decrease the risk of OASIS [29].

In our study, more OASIS were reported in patients who gave birth to newborns with a small cranial perimeter. These results are inconsistent with the literature, where increase in cranial perimeter is associated with a statistically higher rate of OASIS [13]. We hypothesize that in such cases there may be dystocia (transverse or improperly flexed position), which complicates the manipulation of forceps, with a different axis of traction, which is riskier for the maternal perineal area.

Obesity (BMI ≥30) was protective against OASIS and was still significant after multivariate analysis (p = 0.0013). Nevertheless, obese patients have a higher risk of conceiving macrosomic newborns, a well-known risk factor for OASIS [30]. Two studies on large populations also showed that obesity was protective against OASIS [31, 32]. This could be explained by the fact that obese patients have a larger anus-to-vulva distance because of their fat tissue. Indeed, a short anus-to-vulva distance is a recognized risk factor for OASIS [33].

In our study, there was no significant change in the rate of OASIS according to other risk factors usually associated with OASIS: macrosomia, primiparity, instrumental delivery practiced in the pelvic inlet position. Shoulder dystocia was not studied because of its low incidence (0.5 to 1% of vaginal births) [34]. Maternal ethnic origin is sometimes reported as a factor implicated in maternal perineal complications (Indian and Asian women) [19]. Such factors were not taken into account here, as these ethnicities are not widely represented in our region. Maternal blood loss was also significantly higher in the case of OASIS, which would increase morbidity.

The present study in 1253 cases of FD is the largest to date. However, it has some limitations as it was single-center and retrospective. It was also difficult to study improvement in residents’ FD during semesters spent in other medical centers (4 of them). However, this bias had few effects on the main criteria since residents usually complete those semesters at the end of their training, which means long after the three semesters (18 months) of training for FD in Limoges and after conducting 24 or more FD.

Further analyses could be performed in order to compare perineal outcome after FD in Limoges to another university hospital where simulation is a proper part of the training program, and where residents are assisted by a senior physician until their last semester. Experience with SF should not be generalized to other instruments.

Operative deliveries by SF can occasionally cause OASIS, which involve a substantial risk of maternal morbidity.

Residents should receive proper practical and theoretical training, including good knowledge of instruments. Our results should encourage obstetricians to update their methods of teaching operative delivery, with proper mentoring and empowerment of residents after completion of at least 24 FD or 3 semesters (18 months) under a senior physician’s supervision.

Simulator training could be a way to raise residents’ experience and, therefore, to lower maternal morbidity in the immediate postpartum period.