- Email: [email protected]
Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level
G Model
JOGOH-175; No. of Pages 8 J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Original Article
Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level S. Desurmont *, A. Houze´ de l’Aulnoit, G. Brabant, D. Houze´ de l’Aulnoit Lille Catholic Hospitals, Obstetrics Department, Lille Catholic University, Lille, France
A R T I C L E I N F O
A B S T R A C T
Article history: Received 16 January 2018 Received in revised form 13 May 2018 Accepted 17 May 2018 Available online xxx
Purpose. – Engagement of the fetal head is a determinant element when deciding on operative vaginal delivery. In routine practice, engagement is a clinical diagnosis based on transvaginal digital examination. Transperineal ultrasound might provide complementary information useful for measuring the fetal head-perineum distance (HPD). The purpose of this work was to determine the cutoff HPD distinguishing engagement from non-engagement. Materials and methods. – This single-center prospective study approved by the institutional review board was conducted between December 25, 2012 and August 31, 2015 in 411 nulliparous women; 20 did not provide informed consent and were excluded; analysis concerned 391 patients. Clinical diagnosis – engagement or non-engagement depending on results of the transvaginal digital examination (Farabeuf’s and Demelin’s signs) – was compared with the ultrasound HPD measurement. Results. – The clinical diagnosis was non-engagement at complete dilatation in 96 patients (24.6%). The cutoff HPD distinguishing between engagement and non-engagement was 57 mm (AUC 83.5% [95%CI 79.3–87.8]), with 75.0% [65.5–82.6] sensitivity, 75.9% [70.7–80.5] specificity, 50.3% [42.2–58.4] positive predictive value, and 90.3% [86.0-93.4] negative predictive value. Conclusions. – In this series, the HPD cutoff distinguishing between engagement and non-engagement was 57mm. Below this cutoff level, the head should be considered engaged, beyond non-engaged. Nevertheless, the pertinence of this cutoff level is hampered by the imprecision of the gold standard used for the clinical diagnosis (transvaginal digital examination). In case of doubt, we recommend, in addition to considering the obstetrical setting, to combine transperineal ultrasound with transvaginal digital examination to avoid deleterious failure of operative vaginal delivery.
C 2018 Elsevier Masson SAS. All rights reserved.
Keywords: Transperineal ultrasound Head-perineal distance Vaginal digital examination Fetal presentation
1. Introduction The fetal presentation is said to be engaged when the apex of the cephalic pole passes through the plane of the ischial spines (ACOG level 0) [1]. In France, the diagnosis of fetal engagement is classically based on Farabeuf’s and Demelin’s clinical signs determined by transvaginal digital examination [2,3]. When there is an indication for immediate delivery, fetal engagement is a determinant element in the crucial choice between operative vaginal delivery and C-section. For the fetus, the risk of intracerebral hemorrhage is two-fold higher for Csection performed after unsuccessful vaginal delivery, and for the mother there is an increased risk of blood loss, bladder wounds, and damage to the uterine arteries [4–6].
* Corresponding author. E-mail address: [email protected] (S. Desurmont).
Several studies investigating ways of improving the diagnosis of fetal engagement have found that ultrasound (US) assessment of the fetal head-perineum distance (HPD) can be useful [7,8]. In these studies conducted by selected operators examining small cohorts of patients, fetal engagement was chosen as the main outcome. The purpose of the present study based on transperineal US and transvaginal digital examination data was to determine the HPD distinguishing between engagement and non-engagement. 2. Material and methods This prospective interventional single-center double-blind study was conducted in a level IIb maternity ward (SaintVincent-de-Paul Hospital, Lille, France) from December 25, 2012 through August 31, 2015. The study protocol was approved by the local ethics committee (No. CPP 2012/027) and registered
https://doi.org/10.1016/j.jogoh.2018.05.003 C 2018 Elsevier Masson SAS. All rights reserved. 2468-7847/
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 2
S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
Fig. 1.
with the national clinical trial authorities (NCT 02456753). Participating patients received oral and written information and provided signed consent forms. Nulliparous women aged over 18 years beyond 37 weeks gestation were eligible for inclusion at the time of clinical diagnosis of complete cervical dilatation with ruptured membranes and a cephalic presentation. Fetal head station – reported as engagement or non-engagement based on the results of the digital vaginal examination – was determined clinically by the attending midwife. According to Farabeuf [2], a distance between the most advanced point of the presentation and the sacral concavity insufficient to accommodate more than the width of two fingers is a sign of engagement. According to Demelin [9], an index finger introduced into the vagina perpendicularly to the pubic symphysis in the second plane of Hodge [10] (about the middle of the second sacral vertebra) that comes in contact with the presentation is a sign of engagement (Fig. 1). Other data recorded were: type of presentation; the presentation’s degree of flexion; progression of the descent after engagement (high, intermediate, low); presence of asynclitism; and presence of caput succedaneum. The patient had an empty bladder and was in the gynecological position for the vaginal examination performed during a pause between contractions. Thirty-nine midwives participated in the inclusion process. The US study was performed by the attending delivery room head physician during a pause between contractions 5 min after the digital vaginal examination. The patient had an empty bladder and was in the gynecological position. The physician placed the sterile abdominal probe on the perineum transversally without pressure (Fig. 2). The delivery room US machine (Voluson E, GE Medical Systems Ultrasound and Primary Care Diagnostics, LLC, Wisconsin, US) was used. The ultrasound probe was settled on 26 Hz frequency. As described by Eggebø et al. [10] and MaticotBaptista et al. [7], the HPD was measured perpendicularly to the anovulvar space between the beginning of the insonation and the outer limit of the osseous hyperechogenicity of the cephalic pole. Distance to caput succedaneum was measured between the beginning of the insonation (on the perineum) and the scalp in all fetuses. The difference of these two distances made it possible to calculate the caput succedaneum thickness (Fig. 3). Eight gynecologist–obstetricians participated in the study and were blinded to the results reported by the midwives. The midwife did
the digital vaginal examination first, then left the room. Thereafter, the obstetrician entered the room and performed the US exam. Results were sealed in envelopes, separately for the midwife and for the obstetrician. Delivery route was not noted. If the question of immediate birth had to be discussed, the patient was excluded from the study. If the decision was made late after sealing the envelopes, the obstetrician made the decision on the basis of a new digital vaginal examination only. 2.1. Statistical analysis Non-engagement at complete dilatation was chosen as the main outcome. The number of patients necessary for the study was
Fig. 2.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
Fig. 3.
3
based on the results of three earlier studies comparing digital vaginal examination and transperineal US findings [7,8,11]. These studies reported clinically diagnosed non-engagement in 20% of patients with 80% sensitivity and 95% specificity. The optimal cutoff was estimated considering a ROC curve to define the two groups ‘‘engagement’’ and ‘‘no-engagement’’. Given the proposed formula for the calculation of the confidence interval of the AUC (Area Under the Curve) from Machin et al. [12], the estimated prevalence (20%), the expected sensitivity and specificity values (Se 80%, Sp 95%), the width of the confidence interval (CI) of 95% and 5% risk of error (0.06) were taken into account. The computation gave 411 patients as the total necessary to include (82 ‘‘expected’’ cases of non-engagement and 328 in the engagement group). Data were processed with R software. The Shapiro–Wilk test was applied to test quantitative variables for normal distribution. Data sets with a normal distribution are expressed as mean sstandard deviation, otherwise as median and interquartile interval. For qualitative variables, data are presented as number and percentage.
Fig. 4.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
4
Table 1 Comparison of maternal and obstetrical characteristics between groups without and with clinical engagement.
Age (years) Weight at the beginning of pregnancy (kg) BMI at the end of pregnancy (kg/m2) Overweight BMI >25 kg/m2 Term pregnancy Presentation type Anterior Posterior Transverse Measure of caput succedaneum (mm) Birth weight (g) 1 min Apgar <7 5 min Apgar <7 Arterial pH <7.10 Umbilical cord’s lactates >10 mmol/l Length of labor (h) Extension of time CD/ultrasound (min) Extension of time ultrasound-birth (min)
Missing data
Non-engagement, n = 96
Engagement, n = 295
p
0 0 0 0 0 0
27.0 [25.0; 63.0 [58.0; 28.6 [25.7; 84 (87.5) 40.0 [39.1;
27.0 [24.0; 60.0 [55.0; 27.1 [24.7; 214 (72.5) 39.9 [39.1;
0.792 0.004*** 0.001*** 0.005*** 0.194 <0.001****
(0) (0) (0) (0) (0) (0)
30.0] 73.2] 32.4] 40.7]
65 (67.8) 25 (26.0) 6 (6.2) 20.0 [16.8; 23.0] 3480 360 8 (8.3) 1 (1.0) 3 (3.1) 0 7.71 [6.50; 10.0] 5 [3; 5] 139 [102; 168]
0 (0) 0 (0) 0 (0) 0 (0) 32 (8.1) 46 (11.7) 0 (0) 0 (0)
30.0] 68.0] 30.1] 40.6]
247 (83.7) 28 (9.5) 20 (6.8) 17.0 [12.0; 21.0] 3310 410 16 (5.4) 1 (0.3) 14 (4.7) 7 (2.3) 7.00 [5.67; 8.50] 5 [3; 5] 97.0 [57.0; 142]
<0.001**** <0.001**** 0.619 0.431 0.773 0.201 0.001*** 0.114 <0.001****
BMI: body mass index; CD: complete dilatation. Continuous variables with normal distribution are given as mean and standard deviation. Continuous variables with non-normal distribution are given as median and 25th and 75th percentile [in brackets]. Percentages are given in parentheses. *** < 0.01. **** < 0.001.
Table 2 Sensitivity, specificity, and positive and negative predictive values for HPD beyond or under the cutoff level. Digital vaginal examination
Ultrasound measure
Non-engagement, n = 96
Engagement, n = 295
HPD 57.4 mm
72
71
HPD < 57.4 mm
24
224
Se (%) 75.0 [65.5–82.6]
Sp (%) 75.9 [70.7–80.5]
PPV (%) 50.3 [42.2–58.4] NPV (%) 90.3 [86.0–93.4] Total = 391
HPD: head-perineum distance; Se: sensibility; Sp: specificity; PPV: positive predictive value; NPV: negative predictive value.
Fig. 5.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
5
exact test were applied to qualitative variables. Receiver operating characteristic (ROC) curves were plotted to determine the sensitivity and specificity of HPD for the diagnosis of nonengagement. Differences with p < 0.05 were considered statistically significant.
3. Results
Fig. 6.
Fig. 7.
For the bivariate analysis, Student’s t test or the Mann– Whitney–Wilcoxon test were applied as appropriate for quantitative variables to search for a link between the main outcome and explanatory variables. Similarly, the chi-square test or Fisher’s
Of the 411 patients included in the study, 391 were retained for analysis; a signed consent form was not available for 20 patients who were excluded from the analysis. The prevalence of nonengagement diagnosed clinically in 391 patients at complete dilatation was 24.6% (Fig. 4). The results of the bivariate analysis are presented in Table 1. Variables associated with non-engagement at complete dilatation were: higher body mass index (p = 0.001); fetus in a posterior position (p < 0.001); presence of caput succedaneum (p < 0.001); and higher birth weight (p < 0.001). No correlation was found for any of the other neonatal variables studied. Inter-observer comparisons failed to disclose any significant different in mean HPDs. However, comparisons between midwives revealed a significant difference (p < 0.001) in the frequency of non-engagement reports. This difference arose from six ‘‘pessimistic’’ midwives who reported higher percentages of nonengaged presentations. This subgroup of midwives was not significantly different from the others in terms of years of experience. The ROC curve comparing HPD with digital vaginal examination results showed that the cutoff level discriminating between engagement and non-engagement was 57.4 mm, with 75.0% [95% confidence interval (CI) 68.5–82.6] sensitivity, 75.9% [70.7– 80.5] specificity, 90.3% [86.0–93.4] negative predictive value (NPV), and 50.3% [42.2–58.4] positive predictive value (PPV) (Table 2). Beyond 57.4 mm, 50.3% of presentations were nonengaged (PPV); under 57.4 mm and 90% were clinically engaged (NPV) (Fig. 5). Nevertheless, the area under the curve (AUC) provided 83.6% [95%CI 79.3–87.8] discrimination with 75.7% of well-classed patients (Fig. 6). Withdrawing from the analysis patients whose engagement was measured by the subgroup of ‘‘pessimistic’’ midwives reporting higher rates of non-engagement did not improve sensitivity or specificity. The perineum-caput succedaneum distance had little discriminating power as seen by the inter-group comparison (Fig. 7). The two groups defined by engagement or non-engagement were significantly different for delivery route. C-section was performed more often among patients exhibiting non-engagement at complete dilatation (6% vs 0.7%, p < 0.001). Similarly, operative vaginal delivery was more common in the non-engagement group (38.1% vs 20.4%, p < 0.001), even up to 270 min after the diagnosis of complete dilatation (Table 3). In three cases, activity overload
Table 3 Obstetrical issues compared with clinical engagement. Lack of engagement since complete dilatation is achieved Duration before delivery (min) 30 60 C-section 0 1 nrfhr Operative vaginal delivery 1 1 Spontaneous vaginal delivery 1 4 Total 2 6 Engagement since complete dilatation is achieved Duration before delivery (min) 30 60 C-section 0 0 Operative vaginal delivery 4 5 Spontaneous vaginal delivery 35 36 Total 39 41
90 1 6 3 10 90 1odf 15 40 56
nrfhr
120 2 6 9 17
nrfhr, odf
120 0 10 46 56
150 0 7 16 23
180 1neng 9 13 23
210 0 2 4 6
240 0 5 2 7
270 1odf 0 1 2
150 0 11 34 45
180 0 10 25 35
210 0 3 13 16
240 0 2 4 6
270 1neng 0 0 1
6 (6.2%) 37 (38.1%) 53 (55.7%) 96 (100%)
2 (0.7%) 60 (20.4%) 233 (78.9%) 295 (100%)
nrfhr: non-reassuring fetal heart rate; odf: operative delivery failure; neng: non-engagement.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
6
Fig. 8.
led to a long latency (with reassuring fetal rates) before childbirth. Otherwise, we noticed three cases (1%) of C-section after failed operative vaginal delivery with forceps and in the non-engagement group two C-sections were performed 90 min and 270 min after diagnosis of complete dilatation. In the engagement group, one C-section was performed 90 min after diagnosis of complete dilatation in a patient classed in the engaged group but reclassed in the non-engagement group by the attending physician. The Kaplan–Meier curves, plotted for 57.4 mm engagement distance, revealed a significant difference between spontaneous and operative vaginal delivery (p = 0.008) (Fig. 8). 4. Discussion Our study demonstrated that the HPD discriminating between engagement and non-engagement is about 57 mm; below this cutoff, the fetal head should be considered engaged, and beyond non-engaged. We chose non-engagement as the main outcome for this study because non-engagement is a counter-indication for operative vaginal delivery. Two prior studies also investigating the HPD chose engagement as their main outcome [7,8]. In these studies, the prevalence of engagement at complete dilatation was 82% and 90% respectively, or in other words non-engagement was diagnosed in 18% and 10% of patients. In our study, the prevalence of non-engagement (24.6%) was significantly higher (p = 0.003).
The computation determining the number of subjects required demonstrating a non-random difference between engagement and non-engagement is dependent on these prevalence figures. If we had chosen engagement as the main outcome, the statistical computation would have given 232 as the number of patients to be included to establish the presence of a significant difference. By considering non-engagement as the main outcome, the number of subjects necessary to include rose to 411, giving our study more statistical power and reliability. In order to facilitate comparisons, the data from the literature are presented in Table 4 as originally reported and also recalculated considering non-engagement as the outcome (Table 4). The HPD discriminating between engagement and non-engagement was 60 mm for Maticot-Baptista et al. [7] and 55 mm for Rivaux et al. [8] and Dimassi et al. [11]. Our 57 mm thus lies between the levels reported earlier. Applying a 25% non-engagement rate to the data reported by Rivaux et al. [8], their predictive values are very close to those we report here. Like others, we used the digital vaginal examination as the gold standard for assessing fetal head progression. Demelin’s sign can lack precision, especially if the index finger is not introduced perfectly perpendicular to the symphysis pubis. Similarly, Farabeuf’s sign may be affected by the presence of caput succedaneum or extensive modeling of the fetal head. In our study, we combined the two signs in order to limit the risk of imprecision. In routine practice, the problem is to differentiate high cavity engagement – related to fetal skull deformation and modeling in the maternal pelvis – from real non-engagement. Nevertheless, the transvaginal digital examination, previously considered as a gold standard, seems to have a poor reproducibility [13,14]. Dupuis et al. [13] investigated transvaginal digital examination using a delivery simulator and found 30% interobserver discordance for the diagnosis of engagement, with no significant difference between senior physicians and residents. Elsewhere, a numerical error in the fetal station of the American College of Obstetrician and Gynecologist (ACOG) station occurred in 36–80% of cases, and Dupuis reported 12% of engagement errors equally distributed between false diagnosis of engagement and non-engagement. The same level of discordance was reported by Buchmann [14] in a sample of 500 patients who also found that operator experience had no impact. In our study, the midwives were equally qualified and the ‘‘pessimistic’’ midwives had the same number of years of experience as their ‘‘less pessimistic’’ counterparts. Since we wanted our potential conclusions to be applicable in routine clinical practice, we did not restrict participation to a few selected sonographists and midwives. For the US assessments, we did not observe any significant interobserver difference between the eight participating sonographists (p = 0.620). Kasbaoui et al. also reported excellent reproducibility
Table 4 Articles comparing clinical exam with transperineal US.
Maticot-Baptista 2009
Effective
obstetricians
midwifes
Main outcome
Prevalence (p) of lack of engagement, n (%)
Cutoff (mm)
Se (%)
62 measures/ 45 patients
1
n.s.
Engagement
18 (29%)
60
97.8
44 (71%) Engagement 90 (90%) Engagement 55 (78.5%) Non-engagement 96 (24.5%)
if p = 25% 10 (10%) if p = 25% 15 (21.4%) if p = 25%
55
73.3
55
93
87
57.4
75.0
75.9
Rivaux 2012
100
4
n.s.
Dimassi 2014
70
2
n.s.
391
8
39
Our series 2015
96 (p = 25%)
Sp (%)
PPV (%)
NPV (%)
89.0
95.6
94.1
93.0 100 55.5 67 70.5
96.3 29.4 100.0 98 97.3
50.3
90.3
100
AUC (%)
n.s.
n.s. 95
83
Se: sensitivity; Sp: specificity; PPV: predictive positive value; NPV: negative predictive value; AUC: area under the curve, n.s.: not specified, p: prevalence. The data from the literature are as originally reported and also recalculated (in bold) considering non-engagement as the outcome.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
for transperineal US with significant inter-observer correlation (r = 0.96) [15]. Conversely, for the 39 midwives who participated in our study, there was highly significant inter-observer difference (p < 0.001) mainly due to the six ‘‘pessimistic’’ midwives. This lack of clinical concordance is a drawback for this gold standard examination, constituting a confounding bias. To date, we still do not have any exam giving a precise discrimination of engagement that could be compared with transperineal US results. Tutschek et al. [16] analyzed different US parameters and compared them together and with digital palpation. They showed a large degree of correlation between HPD and the angle of progression, but only moderate correlation with vaginal digital examination to assess fetal head station during labor. Several studies described the angle of progression (AoP) for assessing fetal engagement [17–19]. The AoP is the angle between the long axis of the pubic symphysis axis and a line extending from its most inferior portion tangentially to the fetal skull. This involves a three-dimensional relationship dependent upon an imaginary line drawn between the often inaccessible ischial spines. Arthuis [20] tried to determine the AoP value assessing fetal engagement with a computed tomography scan model analyzing the anatomical relationship between the pubic symphysis and the ischial spines to determine ACOG station 0 and transpose it to translabial US. The linear method would be more practical than the angle method since it is easier to use a ruler than a scale protractor. In addition, lack of symmetry and unclear US contours for the symphyseal landmark [21] is a drawback for translabial US, a rather difficult exploration for routine practice. Moreover, fetal progression is linear high in the cavity, following the umbilical-coccygian axis before becoming curvilinear during mid-cavity rotation [22]. Angle measurements would thus be more adapted to curvilinear mid- and low-cavity progression [17]. For high cavity progression, transperineal US as described by Maticot-Baptista [7] and Eggebo [10] would appear to be sufficient. Torkildsen [23] and Youssef [24] tried three-dimensional transperineal ultrasound methods. Torkildsen [23] concluded that two-dimensional methods are simpler to learn and two-dimensional equipment more common in the labor room. Intraobserver repeatability was good. The purpose of our study was to determine the cutoff level at which US-measured HPD distinguishes engagement from nonengagement, irrespective of the type of delivery and with no attempt to predict potential problems. Despite the fact that we measured the HPD at complete dilatation and not at indication for immediate delivery, the difference in the rate of C-section was still significantly higher above the 57.4 mm cut-off, even after 270 min of complete dilatation. Kasbaoui et al. [15] studied a cohort of 659 patients to determine prospectively the risk of operative vaginal delivery. Their work concerned HPD measured for engaged presentations with an indication for operative delivery. These authors found a significant correlation between HPD 40 mm and risk of difficult operative vaginal delivery (OR = 2.38 [95%CI 1.51– 3.74] p < 0.0001). For a transvaginal examination, this level corresponds to a mid-cavity presentation. For a cutoff >60 mm, close to the 57.4 mm found here, these authors reported the OR was 3.02 [1.68–5.43], sensitivity 18.3%, specificity 93.3, PPV 37.9, NPV 83.1), with high risk of difficult extraction such that Csection would be indicated; between 40 and 60 mm, operative vaginal delivery would still be indicated for an experienced operator. By studying the deliveries issues based on indirect criteria (difficult extraction, C-section after failure of operative vaginal delivery), the authors clearly demonstrated the usefulness of transperineal US.
7
5. Conclusion US measurement of the fetal head-perineum distance is a useful easy-to-perform technique to assess fetal head progression. The cutoff between engagement and non-engagement appears to be 57 mm. The pertinence of this cutoff level is hampered by the imprecision of the clinical gold standard (transvaginal digital examination) used to establish it. In case of doubt, we recommend, in addition to considering the obstetrical setting (parity, macrosomia, fetal position, etc.), combining transperineal US with transvaginal digital examination in order to avoid deleterious failure of operative vaginal delivery. Other trials must be performed to ascertain whether this strategy indeed has a positive impact on the health of mothers and babies by enabling appropriate use of operative vaginal delivery and a lower rate of cesarean section. Generalized use of US in the delivery room would be useful to check the type of presentation and assess fetal head progression, especially during the early phase of expulsion efforts, to better determine the fetal head-perineum distance defining fetal and maternal risk. Disclosure of interest The authors declare that they have no competing interest. Acknowledgement We thank all patients, physicians and midwives who participated in this study.
References [1] Hagadorn-Freathy AS, Yeomans ER, Hankins GD. Validation of the 1988 ACOG forceps classification system. Obstet Gynecol 1991;77:356–60. [2] Farabeuf F, Varnier H. Introduction a` l’e´tude clinique eet a` la pratique des accouchements. Paris: Masson; 1923. [3] Schaal J-P, Riethmuller D, Maillet R. Me´canique et techniques obste´tricales. Sauramps me´dical; 1998. [4] Towner D, Castro MA, Eby-Wilkens E, Gilbert WM. Effect of mode of delivery in nulliparous women on neonatal intracranial injury. N Engl J Med 1999;341:1709–14. [5] Fasubaa OB, Ezechi OC, Orji EO, Ogunniyi SO, Akindele ST, Loto OM, et al. Delivery of the impacted head of the fetus at caesarean section after prolonged obstructed labour: a randomised comparative study of two methods. J Obstet Gynaecol 2002;22:375–8. [6] Murphy DJ, Liebling RE, Verity L, Swingler R, Patel R. Early maternal and neonatal morbidity associated with operative delivery in second stage of labour: a cohort study. Lancet 2001;358:1203–7. [7] Maticot-Baptista D, Ramanah R, Collin A, Martin A, Maillet R, Riethmuller D. Diagnostic e´chographique d’engagement de la pre´sentation fœtale. A` propos d’une se´rie prospective pre´liminaire franc¸aise. J Gyne´cologie Obste´trique Biol Reprod 2009;38:474–80. [8] Rivaux G, Dedet B, Delarue E, Depret S, Closset E, Deruelle P. Engagement de la teˆte fœtale: e´chographie transpe´rine´ale, un nouvel outil diagnostique? Gyne´col Obste´trique Fertil 2012;40:148–52. [9] Gernez L. The ‘‘engaged’’ or ‘‘non-engaged’’ head: reconsideration of Farabeuf’s sign. J Gynecol Obstet Biol Reprod (Paris) 1981;10:365–8. [10] Eggebø TM, Gjessing LK, Heien C, Smedvig E, Økland I, Romundstad P, et al. Prediction of labor and delivery by transperineal ultrasound in pregnancies with prelabor rupture of membranes at term. Ultrasound Obstet Gynecol 2006;27:387–91. [11] Dimassi K, Ben AA, Ben MK, Derbel M, Ben NA, Triki A, et al. [Diagnosis of fetal engagement by transperineal sonography: a preliminary Tunisian study]. Gynecol Obstet Fertil 2014;42:399–403. [12] Machin D, Campbell MJ, Tan S-B, Tan S-H. Sample size tables for clinical studies. John Wiley & Sons; 2011. [13] Dupuis O, Silveira R, Zentner A, Dittmar A, Gaucherand P, Cucherat M, et al. Birth simulator: reliability of transvaginal assessment of fetal head station as defined by the American College of Obstetricians and Gynecologists classification. Am J Obstet Gynecol 2005;192:868–74. [14] Buchmann E, Libhaber E. Interobserver agreement in intrapartum estimation of fetal head station. Int J Gynecol Obstet 2008;101:285–9. [15] Kasbaoui S, Se´verac F, Aı¨ssi G, Gaudineau A, Lecointre L, Akladios C, et al. Predicting the difficulty of operative vaginal delivery by ultrasound
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
8
[16]
[17]
[18]
[19]
measurement of fetal head station. Am J Obstet Gynecol 2017;216:507. e1-507.e9. Tutschek B, Torkildsen EA, Eggebø TM. Comparison between ultrasound parameters and clinical examination to assess fetal head station in labor. Ultrasound Obstet Gynecol 2013;41:425–9. Barbera AF, Pombar X, Perugino G, Lezotte DC, Hobbins JC. A new method to assess fetal head descent in labor with transperineal ultrasound. Ultrasound Obstet Gynecol 2009;33:313–9. Bultez T, Quibel T, Bouhanna P, Popowski T, Resche-Rigon M, Rozenberg P. Angle of fetal head progression measured using transperineal ultrasound as a predictive factor of vacuum extraction failure. Ultrasound Obstet Gynecol 2016;48:86–91. Dietz HP, Lanzarone V. Measuring engagement of the fetal head: validity and reproducibility of a new ultrasound technique. Ultrasound Obstet Gynecol 2005;25:165–8.
[20] Arthuis CJ, Perrotin F, Patat F, Brunereau L, Simon EG. Computed tomographic study of anatomical relationship between pubic symphysis and ischial spines to improve interpretation of intrapartum translabial ultrasound. Ultrasound Obstet Gynecol 2016;48:779–85. [21] Fouche´ CJ, Simon EG, Potin J, Perrotin F. Ultrasound in monitoring of the second stage of labour. Gynecol Obstet Fertil 2012;40:658–65. [22] Kamina. Anatomie gyne´cologique et obste´tricale, 3e`me, Paris: Maloine; 1979. [23] Torkildsen EA, Salvesen KA˚, Eggebø TM. Agreement between two- and three-dimensional transperineal ultrasound methods in assessing fetal head descent in the first stage of labor. Ultrasound Obstet Gynecol 2012;39:310–5. [24] Youssef A, Bellussi F, Montaguti E, Maroni E, Salsi G, Morselli-Labate AM, et al. Agreement between two- and three-dimensional transperineal ultrasound methods for assessment of fetal head–symphysis distance in active labor. Ultrasound Obstet Gynecol 2013;43:183–8.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
JOGOH-175; No. of Pages 8 J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Original Article
Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level S. Desurmont *, A. Houze´ de l’Aulnoit, G. Brabant, D. Houze´ de l’Aulnoit Lille Catholic Hospitals, Obstetrics Department, Lille Catholic University, Lille, France
A R T I C L E I N F O
A B S T R A C T
Article history: Received 16 January 2018 Received in revised form 13 May 2018 Accepted 17 May 2018 Available online xxx
Purpose. – Engagement of the fetal head is a determinant element when deciding on operative vaginal delivery. In routine practice, engagement is a clinical diagnosis based on transvaginal digital examination. Transperineal ultrasound might provide complementary information useful for measuring the fetal head-perineum distance (HPD). The purpose of this work was to determine the cutoff HPD distinguishing engagement from non-engagement. Materials and methods. – This single-center prospective study approved by the institutional review board was conducted between December 25, 2012 and August 31, 2015 in 411 nulliparous women; 20 did not provide informed consent and were excluded; analysis concerned 391 patients. Clinical diagnosis – engagement or non-engagement depending on results of the transvaginal digital examination (Farabeuf’s and Demelin’s signs) – was compared with the ultrasound HPD measurement. Results. – The clinical diagnosis was non-engagement at complete dilatation in 96 patients (24.6%). The cutoff HPD distinguishing between engagement and non-engagement was 57 mm (AUC 83.5% [95%CI 79.3–87.8]), with 75.0% [65.5–82.6] sensitivity, 75.9% [70.7–80.5] specificity, 50.3% [42.2–58.4] positive predictive value, and 90.3% [86.0-93.4] negative predictive value. Conclusions. – In this series, the HPD cutoff distinguishing between engagement and non-engagement was 57mm. Below this cutoff level, the head should be considered engaged, beyond non-engaged. Nevertheless, the pertinence of this cutoff level is hampered by the imprecision of the gold standard used for the clinical diagnosis (transvaginal digital examination). In case of doubt, we recommend, in addition to considering the obstetrical setting, to combine transperineal ultrasound with transvaginal digital examination to avoid deleterious failure of operative vaginal delivery.
C 2018 Elsevier Masson SAS. All rights reserved.
Keywords: Transperineal ultrasound Head-perineal distance Vaginal digital examination Fetal presentation
1. Introduction The fetal presentation is said to be engaged when the apex of the cephalic pole passes through the plane of the ischial spines (ACOG level 0) [1]. In France, the diagnosis of fetal engagement is classically based on Farabeuf’s and Demelin’s clinical signs determined by transvaginal digital examination [2,3]. When there is an indication for immediate delivery, fetal engagement is a determinant element in the crucial choice between operative vaginal delivery and C-section. For the fetus, the risk of intracerebral hemorrhage is two-fold higher for Csection performed after unsuccessful vaginal delivery, and for the mother there is an increased risk of blood loss, bladder wounds, and damage to the uterine arteries [4–6].
* Corresponding author. E-mail address: [email protected] (S. Desurmont).
Several studies investigating ways of improving the diagnosis of fetal engagement have found that ultrasound (US) assessment of the fetal head-perineum distance (HPD) can be useful [7,8]. In these studies conducted by selected operators examining small cohorts of patients, fetal engagement was chosen as the main outcome. The purpose of the present study based on transperineal US and transvaginal digital examination data was to determine the HPD distinguishing between engagement and non-engagement. 2. Material and methods This prospective interventional single-center double-blind study was conducted in a level IIb maternity ward (SaintVincent-de-Paul Hospital, Lille, France) from December 25, 2012 through August 31, 2015. The study protocol was approved by the local ethics committee (No. CPP 2012/027) and registered
https://doi.org/10.1016/j.jogoh.2018.05.003 C 2018 Elsevier Masson SAS. All rights reserved. 2468-7847/
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 2
S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
Fig. 1.
with the national clinical trial authorities (NCT 02456753). Participating patients received oral and written information and provided signed consent forms. Nulliparous women aged over 18 years beyond 37 weeks gestation were eligible for inclusion at the time of clinical diagnosis of complete cervical dilatation with ruptured membranes and a cephalic presentation. Fetal head station – reported as engagement or non-engagement based on the results of the digital vaginal examination – was determined clinically by the attending midwife. According to Farabeuf [2], a distance between the most advanced point of the presentation and the sacral concavity insufficient to accommodate more than the width of two fingers is a sign of engagement. According to Demelin [9], an index finger introduced into the vagina perpendicularly to the pubic symphysis in the second plane of Hodge [10] (about the middle of the second sacral vertebra) that comes in contact with the presentation is a sign of engagement (Fig. 1). Other data recorded were: type of presentation; the presentation’s degree of flexion; progression of the descent after engagement (high, intermediate, low); presence of asynclitism; and presence of caput succedaneum. The patient had an empty bladder and was in the gynecological position for the vaginal examination performed during a pause between contractions. Thirty-nine midwives participated in the inclusion process. The US study was performed by the attending delivery room head physician during a pause between contractions 5 min after the digital vaginal examination. The patient had an empty bladder and was in the gynecological position. The physician placed the sterile abdominal probe on the perineum transversally without pressure (Fig. 2). The delivery room US machine (Voluson E, GE Medical Systems Ultrasound and Primary Care Diagnostics, LLC, Wisconsin, US) was used. The ultrasound probe was settled on 26 Hz frequency. As described by Eggebø et al. [10] and MaticotBaptista et al. [7], the HPD was measured perpendicularly to the anovulvar space between the beginning of the insonation and the outer limit of the osseous hyperechogenicity of the cephalic pole. Distance to caput succedaneum was measured between the beginning of the insonation (on the perineum) and the scalp in all fetuses. The difference of these two distances made it possible to calculate the caput succedaneum thickness (Fig. 3). Eight gynecologist–obstetricians participated in the study and were blinded to the results reported by the midwives. The midwife did
the digital vaginal examination first, then left the room. Thereafter, the obstetrician entered the room and performed the US exam. Results were sealed in envelopes, separately for the midwife and for the obstetrician. Delivery route was not noted. If the question of immediate birth had to be discussed, the patient was excluded from the study. If the decision was made late after sealing the envelopes, the obstetrician made the decision on the basis of a new digital vaginal examination only. 2.1. Statistical analysis Non-engagement at complete dilatation was chosen as the main outcome. The number of patients necessary for the study was
Fig. 2.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
Fig. 3.
3
based on the results of three earlier studies comparing digital vaginal examination and transperineal US findings [7,8,11]. These studies reported clinically diagnosed non-engagement in 20% of patients with 80% sensitivity and 95% specificity. The optimal cutoff was estimated considering a ROC curve to define the two groups ‘‘engagement’’ and ‘‘no-engagement’’. Given the proposed formula for the calculation of the confidence interval of the AUC (Area Under the Curve) from Machin et al. [12], the estimated prevalence (20%), the expected sensitivity and specificity values (Se 80%, Sp 95%), the width of the confidence interval (CI) of 95% and 5% risk of error (0.06) were taken into account. The computation gave 411 patients as the total necessary to include (82 ‘‘expected’’ cases of non-engagement and 328 in the engagement group). Data were processed with R software. The Shapiro–Wilk test was applied to test quantitative variables for normal distribution. Data sets with a normal distribution are expressed as mean sstandard deviation, otherwise as median and interquartile interval. For qualitative variables, data are presented as number and percentage.
Fig. 4.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
4
Table 1 Comparison of maternal and obstetrical characteristics between groups without and with clinical engagement.
Age (years) Weight at the beginning of pregnancy (kg) BMI at the end of pregnancy (kg/m2) Overweight BMI >25 kg/m2 Term pregnancy Presentation type Anterior Posterior Transverse Measure of caput succedaneum (mm) Birth weight (g) 1 min Apgar <7 5 min Apgar <7 Arterial pH <7.10 Umbilical cord’s lactates >10 mmol/l Length of labor (h) Extension of time CD/ultrasound (min) Extension of time ultrasound-birth (min)
Missing data
Non-engagement, n = 96
Engagement, n = 295
p
0 0 0 0 0 0
27.0 [25.0; 63.0 [58.0; 28.6 [25.7; 84 (87.5) 40.0 [39.1;
27.0 [24.0; 60.0 [55.0; 27.1 [24.7; 214 (72.5) 39.9 [39.1;
0.792 0.004*** 0.001*** 0.005*** 0.194 <0.001****
(0) (0) (0) (0) (0) (0)
30.0] 73.2] 32.4] 40.7]
65 (67.8) 25 (26.0) 6 (6.2) 20.0 [16.8; 23.0] 3480 360 8 (8.3) 1 (1.0) 3 (3.1) 0 7.71 [6.50; 10.0] 5 [3; 5] 139 [102; 168]
0 (0) 0 (0) 0 (0) 0 (0) 32 (8.1) 46 (11.7) 0 (0) 0 (0)
30.0] 68.0] 30.1] 40.6]
247 (83.7) 28 (9.5) 20 (6.8) 17.0 [12.0; 21.0] 3310 410 16 (5.4) 1 (0.3) 14 (4.7) 7 (2.3) 7.00 [5.67; 8.50] 5 [3; 5] 97.0 [57.0; 142]
<0.001**** <0.001**** 0.619 0.431 0.773 0.201 0.001*** 0.114 <0.001****
BMI: body mass index; CD: complete dilatation. Continuous variables with normal distribution are given as mean and standard deviation. Continuous variables with non-normal distribution are given as median and 25th and 75th percentile [in brackets]. Percentages are given in parentheses. *** < 0.01. **** < 0.001.
Table 2 Sensitivity, specificity, and positive and negative predictive values for HPD beyond or under the cutoff level. Digital vaginal examination
Ultrasound measure
Non-engagement, n = 96
Engagement, n = 295
HPD 57.4 mm
72
71
HPD < 57.4 mm
24
224
Se (%) 75.0 [65.5–82.6]
Sp (%) 75.9 [70.7–80.5]
PPV (%) 50.3 [42.2–58.4] NPV (%) 90.3 [86.0–93.4] Total = 391
HPD: head-perineum distance; Se: sensibility; Sp: specificity; PPV: positive predictive value; NPV: negative predictive value.
Fig. 5.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
5
exact test were applied to qualitative variables. Receiver operating characteristic (ROC) curves were plotted to determine the sensitivity and specificity of HPD for the diagnosis of nonengagement. Differences with p < 0.05 were considered statistically significant.
3. Results
Fig. 6.
Fig. 7.
For the bivariate analysis, Student’s t test or the Mann– Whitney–Wilcoxon test were applied as appropriate for quantitative variables to search for a link between the main outcome and explanatory variables. Similarly, the chi-square test or Fisher’s
Of the 411 patients included in the study, 391 were retained for analysis; a signed consent form was not available for 20 patients who were excluded from the analysis. The prevalence of nonengagement diagnosed clinically in 391 patients at complete dilatation was 24.6% (Fig. 4). The results of the bivariate analysis are presented in Table 1. Variables associated with non-engagement at complete dilatation were: higher body mass index (p = 0.001); fetus in a posterior position (p < 0.001); presence of caput succedaneum (p < 0.001); and higher birth weight (p < 0.001). No correlation was found for any of the other neonatal variables studied. Inter-observer comparisons failed to disclose any significant different in mean HPDs. However, comparisons between midwives revealed a significant difference (p < 0.001) in the frequency of non-engagement reports. This difference arose from six ‘‘pessimistic’’ midwives who reported higher percentages of nonengaged presentations. This subgroup of midwives was not significantly different from the others in terms of years of experience. The ROC curve comparing HPD with digital vaginal examination results showed that the cutoff level discriminating between engagement and non-engagement was 57.4 mm, with 75.0% [95% confidence interval (CI) 68.5–82.6] sensitivity, 75.9% [70.7– 80.5] specificity, 90.3% [86.0–93.4] negative predictive value (NPV), and 50.3% [42.2–58.4] positive predictive value (PPV) (Table 2). Beyond 57.4 mm, 50.3% of presentations were nonengaged (PPV); under 57.4 mm and 90% were clinically engaged (NPV) (Fig. 5). Nevertheless, the area under the curve (AUC) provided 83.6% [95%CI 79.3–87.8] discrimination with 75.7% of well-classed patients (Fig. 6). Withdrawing from the analysis patients whose engagement was measured by the subgroup of ‘‘pessimistic’’ midwives reporting higher rates of non-engagement did not improve sensitivity or specificity. The perineum-caput succedaneum distance had little discriminating power as seen by the inter-group comparison (Fig. 7). The two groups defined by engagement or non-engagement were significantly different for delivery route. C-section was performed more often among patients exhibiting non-engagement at complete dilatation (6% vs 0.7%, p < 0.001). Similarly, operative vaginal delivery was more common in the non-engagement group (38.1% vs 20.4%, p < 0.001), even up to 270 min after the diagnosis of complete dilatation (Table 3). In three cases, activity overload
Table 3 Obstetrical issues compared with clinical engagement. Lack of engagement since complete dilatation is achieved Duration before delivery (min) 30 60 C-section 0 1 nrfhr Operative vaginal delivery 1 1 Spontaneous vaginal delivery 1 4 Total 2 6 Engagement since complete dilatation is achieved Duration before delivery (min) 30 60 C-section 0 0 Operative vaginal delivery 4 5 Spontaneous vaginal delivery 35 36 Total 39 41
90 1 6 3 10 90 1odf 15 40 56
nrfhr
120 2 6 9 17
nrfhr, odf
120 0 10 46 56
150 0 7 16 23
180 1neng 9 13 23
210 0 2 4 6
240 0 5 2 7
270 1odf 0 1 2
150 0 11 34 45
180 0 10 25 35
210 0 3 13 16
240 0 2 4 6
270 1neng 0 0 1
6 (6.2%) 37 (38.1%) 53 (55.7%) 96 (100%)
2 (0.7%) 60 (20.4%) 233 (78.9%) 295 (100%)
nrfhr: non-reassuring fetal heart rate; odf: operative delivery failure; neng: non-engagement.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
6
Fig. 8.
led to a long latency (with reassuring fetal rates) before childbirth. Otherwise, we noticed three cases (1%) of C-section after failed operative vaginal delivery with forceps and in the non-engagement group two C-sections were performed 90 min and 270 min after diagnosis of complete dilatation. In the engagement group, one C-section was performed 90 min after diagnosis of complete dilatation in a patient classed in the engaged group but reclassed in the non-engagement group by the attending physician. The Kaplan–Meier curves, plotted for 57.4 mm engagement distance, revealed a significant difference between spontaneous and operative vaginal delivery (p = 0.008) (Fig. 8). 4. Discussion Our study demonstrated that the HPD discriminating between engagement and non-engagement is about 57 mm; below this cutoff, the fetal head should be considered engaged, and beyond non-engaged. We chose non-engagement as the main outcome for this study because non-engagement is a counter-indication for operative vaginal delivery. Two prior studies also investigating the HPD chose engagement as their main outcome [7,8]. In these studies, the prevalence of engagement at complete dilatation was 82% and 90% respectively, or in other words non-engagement was diagnosed in 18% and 10% of patients. In our study, the prevalence of non-engagement (24.6%) was significantly higher (p = 0.003).
The computation determining the number of subjects required demonstrating a non-random difference between engagement and non-engagement is dependent on these prevalence figures. If we had chosen engagement as the main outcome, the statistical computation would have given 232 as the number of patients to be included to establish the presence of a significant difference. By considering non-engagement as the main outcome, the number of subjects necessary to include rose to 411, giving our study more statistical power and reliability. In order to facilitate comparisons, the data from the literature are presented in Table 4 as originally reported and also recalculated considering non-engagement as the outcome (Table 4). The HPD discriminating between engagement and non-engagement was 60 mm for Maticot-Baptista et al. [7] and 55 mm for Rivaux et al. [8] and Dimassi et al. [11]. Our 57 mm thus lies between the levels reported earlier. Applying a 25% non-engagement rate to the data reported by Rivaux et al. [8], their predictive values are very close to those we report here. Like others, we used the digital vaginal examination as the gold standard for assessing fetal head progression. Demelin’s sign can lack precision, especially if the index finger is not introduced perfectly perpendicular to the symphysis pubis. Similarly, Farabeuf’s sign may be affected by the presence of caput succedaneum or extensive modeling of the fetal head. In our study, we combined the two signs in order to limit the risk of imprecision. In routine practice, the problem is to differentiate high cavity engagement – related to fetal skull deformation and modeling in the maternal pelvis – from real non-engagement. Nevertheless, the transvaginal digital examination, previously considered as a gold standard, seems to have a poor reproducibility [13,14]. Dupuis et al. [13] investigated transvaginal digital examination using a delivery simulator and found 30% interobserver discordance for the diagnosis of engagement, with no significant difference between senior physicians and residents. Elsewhere, a numerical error in the fetal station of the American College of Obstetrician and Gynecologist (ACOG) station occurred in 36–80% of cases, and Dupuis reported 12% of engagement errors equally distributed between false diagnosis of engagement and non-engagement. The same level of discordance was reported by Buchmann [14] in a sample of 500 patients who also found that operator experience had no impact. In our study, the midwives were equally qualified and the ‘‘pessimistic’’ midwives had the same number of years of experience as their ‘‘less pessimistic’’ counterparts. Since we wanted our potential conclusions to be applicable in routine clinical practice, we did not restrict participation to a few selected sonographists and midwives. For the US assessments, we did not observe any significant interobserver difference between the eight participating sonographists (p = 0.620). Kasbaoui et al. also reported excellent reproducibility
Table 4 Articles comparing clinical exam with transperineal US.
Maticot-Baptista 2009
Effective
obstetricians
midwifes
Main outcome
Prevalence (p) of lack of engagement, n (%)
Cutoff (mm)
Se (%)
62 measures/ 45 patients
1
n.s.
Engagement
18 (29%)
60
97.8
44 (71%) Engagement 90 (90%) Engagement 55 (78.5%) Non-engagement 96 (24.5%)
if p = 25% 10 (10%) if p = 25% 15 (21.4%) if p = 25%
55
73.3
55
93
87
57.4
75.0
75.9
Rivaux 2012
100
4
n.s.
Dimassi 2014
70
2
n.s.
391
8
39
Our series 2015
96 (p = 25%)
Sp (%)
PPV (%)
NPV (%)
89.0
95.6
94.1
93.0 100 55.5 67 70.5
96.3 29.4 100.0 98 97.3
50.3
90.3
100
AUC (%)
n.s.
n.s. 95
83
Se: sensitivity; Sp: specificity; PPV: predictive positive value; NPV: negative predictive value; AUC: area under the curve, n.s.: not specified, p: prevalence. The data from the literature are as originally reported and also recalculated (in bold) considering non-engagement as the outcome.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
for transperineal US with significant inter-observer correlation (r = 0.96) [15]. Conversely, for the 39 midwives who participated in our study, there was highly significant inter-observer difference (p < 0.001) mainly due to the six ‘‘pessimistic’’ midwives. This lack of clinical concordance is a drawback for this gold standard examination, constituting a confounding bias. To date, we still do not have any exam giving a precise discrimination of engagement that could be compared with transperineal US results. Tutschek et al. [16] analyzed different US parameters and compared them together and with digital palpation. They showed a large degree of correlation between HPD and the angle of progression, but only moderate correlation with vaginal digital examination to assess fetal head station during labor. Several studies described the angle of progression (AoP) for assessing fetal engagement [17–19]. The AoP is the angle between the long axis of the pubic symphysis axis and a line extending from its most inferior portion tangentially to the fetal skull. This involves a three-dimensional relationship dependent upon an imaginary line drawn between the often inaccessible ischial spines. Arthuis [20] tried to determine the AoP value assessing fetal engagement with a computed tomography scan model analyzing the anatomical relationship between the pubic symphysis and the ischial spines to determine ACOG station 0 and transpose it to translabial US. The linear method would be more practical than the angle method since it is easier to use a ruler than a scale protractor. In addition, lack of symmetry and unclear US contours for the symphyseal landmark [21] is a drawback for translabial US, a rather difficult exploration for routine practice. Moreover, fetal progression is linear high in the cavity, following the umbilical-coccygian axis before becoming curvilinear during mid-cavity rotation [22]. Angle measurements would thus be more adapted to curvilinear mid- and low-cavity progression [17]. For high cavity progression, transperineal US as described by Maticot-Baptista [7] and Eggebo [10] would appear to be sufficient. Torkildsen [23] and Youssef [24] tried three-dimensional transperineal ultrasound methods. Torkildsen [23] concluded that two-dimensional methods are simpler to learn and two-dimensional equipment more common in the labor room. Intraobserver repeatability was good. The purpose of our study was to determine the cutoff level at which US-measured HPD distinguishes engagement from nonengagement, irrespective of the type of delivery and with no attempt to predict potential problems. Despite the fact that we measured the HPD at complete dilatation and not at indication for immediate delivery, the difference in the rate of C-section was still significantly higher above the 57.4 mm cut-off, even after 270 min of complete dilatation. Kasbaoui et al. [15] studied a cohort of 659 patients to determine prospectively the risk of operative vaginal delivery. Their work concerned HPD measured for engaged presentations with an indication for operative delivery. These authors found a significant correlation between HPD 40 mm and risk of difficult operative vaginal delivery (OR = 2.38 [95%CI 1.51– 3.74] p < 0.0001). For a transvaginal examination, this level corresponds to a mid-cavity presentation. For a cutoff >60 mm, close to the 57.4 mm found here, these authors reported the OR was 3.02 [1.68–5.43], sensitivity 18.3%, specificity 93.3, PPV 37.9, NPV 83.1), with high risk of difficult extraction such that Csection would be indicated; between 40 and 60 mm, operative vaginal delivery would still be indicated for an experienced operator. By studying the deliveries issues based on indirect criteria (difficult extraction, C-section after failure of operative vaginal delivery), the authors clearly demonstrated the usefulness of transperineal US.
7
5. Conclusion US measurement of the fetal head-perineum distance is a useful easy-to-perform technique to assess fetal head progression. The cutoff between engagement and non-engagement appears to be 57 mm. The pertinence of this cutoff level is hampered by the imprecision of the clinical gold standard (transvaginal digital examination) used to establish it. In case of doubt, we recommend, in addition to considering the obstetrical setting (parity, macrosomia, fetal position, etc.), combining transperineal US with transvaginal digital examination in order to avoid deleterious failure of operative vaginal delivery. Other trials must be performed to ascertain whether this strategy indeed has a positive impact on the health of mothers and babies by enabling appropriate use of operative vaginal delivery and a lower rate of cesarean section. Generalized use of US in the delivery room would be useful to check the type of presentation and assess fetal head progression, especially during the early phase of expulsion efforts, to better determine the fetal head-perineum distance defining fetal and maternal risk. Disclosure of interest The authors declare that they have no competing interest. Acknowledgement We thank all patients, physicians and midwives who participated in this study.
References [1] Hagadorn-Freathy AS, Yeomans ER, Hankins GD. Validation of the 1988 ACOG forceps classification system. Obstet Gynecol 1991;77:356–60. [2] Farabeuf F, Varnier H. Introduction a` l’e´tude clinique eet a` la pratique des accouchements. Paris: Masson; 1923. [3] Schaal J-P, Riethmuller D, Maillet R. Me´canique et techniques obste´tricales. Sauramps me´dical; 1998. [4] Towner D, Castro MA, Eby-Wilkens E, Gilbert WM. Effect of mode of delivery in nulliparous women on neonatal intracranial injury. N Engl J Med 1999;341:1709–14. [5] Fasubaa OB, Ezechi OC, Orji EO, Ogunniyi SO, Akindele ST, Loto OM, et al. Delivery of the impacted head of the fetus at caesarean section after prolonged obstructed labour: a randomised comparative study of two methods. J Obstet Gynaecol 2002;22:375–8. [6] Murphy DJ, Liebling RE, Verity L, Swingler R, Patel R. Early maternal and neonatal morbidity associated with operative delivery in second stage of labour: a cohort study. Lancet 2001;358:1203–7. [7] Maticot-Baptista D, Ramanah R, Collin A, Martin A, Maillet R, Riethmuller D. Diagnostic e´chographique d’engagement de la pre´sentation fœtale. A` propos d’une se´rie prospective pre´liminaire franc¸aise. J Gyne´cologie Obste´trique Biol Reprod 2009;38:474–80. [8] Rivaux G, Dedet B, Delarue E, Depret S, Closset E, Deruelle P. Engagement de la teˆte fœtale: e´chographie transpe´rine´ale, un nouvel outil diagnostique? Gyne´col Obste´trique Fertil 2012;40:148–52. [9] Gernez L. The ‘‘engaged’’ or ‘‘non-engaged’’ head: reconsideration of Farabeuf’s sign. J Gynecol Obstet Biol Reprod (Paris) 1981;10:365–8. [10] Eggebø TM, Gjessing LK, Heien C, Smedvig E, Økland I, Romundstad P, et al. Prediction of labor and delivery by transperineal ultrasound in pregnancies with prelabor rupture of membranes at term. Ultrasound Obstet Gynecol 2006;27:387–91. [11] Dimassi K, Ben AA, Ben MK, Derbel M, Ben NA, Triki A, et al. [Diagnosis of fetal engagement by transperineal sonography: a preliminary Tunisian study]. Gynecol Obstet Fertil 2014;42:399–403. [12] Machin D, Campbell MJ, Tan S-B, Tan S-H. Sample size tables for clinical studies. John Wiley & Sons; 2011. [13] Dupuis O, Silveira R, Zentner A, Dittmar A, Gaucherand P, Cucherat M, et al. Birth simulator: reliability of transvaginal assessment of fetal head station as defined by the American College of Obstetricians and Gynecologists classification. Am J Obstet Gynecol 2005;192:868–74. [14] Buchmann E, Libhaber E. Interobserver agreement in intrapartum estimation of fetal head station. Int J Gynecol Obstet 2008;101:285–9. [15] Kasbaoui S, Se´verac F, Aı¨ssi G, Gaudineau A, Lecointre L, Akladios C, et al. Predicting the difficulty of operative vaginal delivery by ultrasound
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003
G Model
JOGOH-175; No. of Pages 8 S. Desurmont et al. / J Gynecol Obstet Hum Reprod xxx (2018) xxx–xxx
8
[16]
[17]
[18]
[19]
measurement of fetal head station. Am J Obstet Gynecol 2017;216:507. e1-507.e9. Tutschek B, Torkildsen EA, Eggebø TM. Comparison between ultrasound parameters and clinical examination to assess fetal head station in labor. Ultrasound Obstet Gynecol 2013;41:425–9. Barbera AF, Pombar X, Perugino G, Lezotte DC, Hobbins JC. A new method to assess fetal head descent in labor with transperineal ultrasound. Ultrasound Obstet Gynecol 2009;33:313–9. Bultez T, Quibel T, Bouhanna P, Popowski T, Resche-Rigon M, Rozenberg P. Angle of fetal head progression measured using transperineal ultrasound as a predictive factor of vacuum extraction failure. Ultrasound Obstet Gynecol 2016;48:86–91. Dietz HP, Lanzarone V. Measuring engagement of the fetal head: validity and reproducibility of a new ultrasound technique. Ultrasound Obstet Gynecol 2005;25:165–8.
[20] Arthuis CJ, Perrotin F, Patat F, Brunereau L, Simon EG. Computed tomographic study of anatomical relationship between pubic symphysis and ischial spines to improve interpretation of intrapartum translabial ultrasound. Ultrasound Obstet Gynecol 2016;48:779–85. [21] Fouche´ CJ, Simon EG, Potin J, Perrotin F. Ultrasound in monitoring of the second stage of labour. Gynecol Obstet Fertil 2012;40:658–65. [22] Kamina. Anatomie gyne´cologique et obste´tricale, 3e`me, Paris: Maloine; 1979. [23] Torkildsen EA, Salvesen KA˚, Eggebø TM. Agreement between two- and three-dimensional transperineal ultrasound methods in assessing fetal head descent in the first stage of labor. Ultrasound Obstet Gynecol 2012;39:310–5. [24] Youssef A, Bellussi F, Montaguti E, Maroni E, Salsi G, Morselli-Labate AM, et al. Agreement between two- and three-dimensional transperineal ultrasound methods for assessment of fetal head–symphysis distance in active labor. Ultrasound Obstet Gynecol 2013;43:183–8.
Please cite this article in press as: Desurmont S, et al. Assessment of fetal head engagement with transperineal ultrasound: Searching for the cutoff level. J Gynecol Obstet Hum Reprod (2018), https://doi.org/10.1016/j.jogoh.2018.05.003