Despite the large number of stillbirths (SB) occurring every year (estimated at 2.6 million per year), global attention regarding this issue is still insufficient [1]. In Emilia-Romagna, a district in the North of Italy, a regional SB audit program has been implemented since 2014. SB rate in this area is one of the lowest in Europe (3.1 per 1000 births at 22 weeks, 2.3 per 1000 births at 28 weeks) [2]. There is a consensus among the reviewed guidelines that a thorough investigation is warranted, in case of SB, in order to identify the cause of death, ensure the appropriate management of the couple and prevent the recurrence in the subsequent pregnancies. As recommended by a recent review of the literature [3], this investigation should include a structured personal, obstetric, and family medical history of the mother, physical examination, laboratory tests (serology, microbiological investigations), genetic analysis of the fetal and the placental tissues, histological examination of the placenta and postmortem autopsy.

The choice of the mode of delivery, in case of SB (fetus non-viable $>$22 weeks’ gestation), should consider maternal preference, gestational age, bishop score, the clinical condition of the woman and her previous obstetric history [4, 5, 6]. As a rule, there is no need for an immediate birth and the timing of delivery should be planned according to women preferences, except for cases of signs of sepsis, preeclampsia, placenta abruptio or rupture of membranes at diagnosis of SB [7]. If delivery is postponed, it is recommended biweekly blood test execution to exclude DIC (disseminated intravascular coagulopathy) associated with prolonged retention of a dead fetus; uncommon event with an estimated risk of 10% after 4 weeks [8].

Vaginal birth, in the absence of absolute contraindications, is always recommended. The maternal request of cesarean section (CS), caused by the rejection of any “pain” related to the current pregnancy failure, should be discussed with the parents, giving them time to understand the situation and the risks for subsequent pregnancies such as placenta accrete or uterine rupture [9]. The CS should be performed only in particular circumstances (placental abruption, placenta previa, persistent transverse situation, etc.) or in presence of high risk of uterine rupture.

Several studies have evaluated methods for inducing termination of pregnancies in the second or third trimester [10, 11]. These are various and included some drugs such as Dinoprostone (analogue PgE2), Misoprostol, Mifepristone and Oxytocin [11, 12]; mechanical methods namely in women at high risk for uterine rupture, such as Cook Balloon or Foley [13] and alternative methods such as acupuncture, namely for women who refused pharmacological induction [14]. Misoprostol represents the most effective method under 28 weeks, as shown in the meta-analysis of Berghella V et al. [15] because was related to a shorter induction-delivery interval and a lower rate of complications respect with other methods. While Mifepristone should be used before Misoprostol for reduce the induction-delivery interval namely in cases of second trimester SB [16, 17, 18].

Finally, different studies [19, 20] showed that the pharmacological analgesia must be offered and performed in every birth unit in case of SB at every gestational age.

However, despite these clear indications on the most appropriate practices to be adopted in the management of the delivery of stillborn babies, data on the effective implementation of the latter (and on their effectiveness) (in Italy) are lacking.

Thus, the aim of our study is to evaluate the different mode of delivery in a SB Italian population through a prospective cohort study, according to gestational age, parity, causes of death, obstetric history, and maternal characteristics.

This is an area-based, prospective cohort study with information collected within the Surveillance System, active since 2014 in Emilia-Romagna, Italy. Data included cases of SB between January 2014 to December 2020. The diagnosis of SB was based on the World Health Organization (WHO) recommendation [4] and was defined as fetal death at 22 weeks (154 days) of gestation or greater, or a birthweight of 500 g if the gestational age was unknown. According to WHO’s recommendation, late SB was defined as a fetus of 1000 g and/or 28 weeks of gestation or greater, and early SB as a fetus with a gestational age between 22 and 27. Maternal demographics (including maternal age, country of origin, and education level), obstetric history, presence of risk factors, antenatal investigations were collected. Gestational age at delivery, birthweight, placenta weight, induction of labor, mode of delivery, analgesia, and circumstances of the SB were recorded together with the list of tests of the diagnostic work-up included placental histology, stillborn autopsy, microbiological evaluation, maternal blood tests, maternal serologic status for infections, cytogenetic analysis, flow cytometry for the research of fetal-maternal hemorrhage and neonate inspection by a neonatologist.

Causes of death were attributed through a multidisciplinary local audit and primary and associated relevant conditions at death were categorized using ReCODE classification [21]. The multidisciplinary team included at least an obstetrician, a neonatologist, and a pathologist.

The present analysis of data was performed in agreement with the Regional Council’s resolution [22] and requested by the Birth Regional Commission in order to evaluate mode of delivery in women with diagnosis of SB. Information was stored anonymously in a secure database. Informed consent for diagnostic work-up was not required because in Italy diagnostic investigation is mandatory by law in case of SB (D.M. 7/2014 and D.P.C. 170/99). Mother and fetus privacy was ensured during the phase of data collection and analysis.

From 2014 to 2020, 783 SB occurred out of a total of 232,506 births, with a SB rate of 3.3 per 1000. Of these SB, 199 (25.4%) occurred before 28 weeks and 584 thereafter (74.6%), yielding a late SB rate of 2.5 per 1000. Sixty-one cases (7.8%) occurred after the onset of labor and were considered intrapartum. Fifteen seven cases (7.3%) originated from multiple pregnancies. Mean maternal age was 32.6 years, in line with the last Birth Report of the same area (32 years). The rate of women $>$35 years was 38.2% in SB population comparing with 34.7% of the overall population [23].

Our population was heterogeneous in term of ethnicity: there were 416 (54.1%) Italian women, the others were from North Africa (11.6%), East Europe (13.9%), Sub-Saharan Africa (10.4%), Indian Subcontinent (6.8%) and other countries (3.2%). Other maternal characteristics and pregnancy details were reported in Table 1.

Data on the mode of delivery were recorded in 770/783 cases. Thus, labor was spontaneous in 85 (11%). Of remnant, 567 (73.6) were induced and 118 (15.3%) had no labor. Mode of delivery was vaginal in most of cases (649/770, 84.3%) and by cesarean section in 121/770 (15.7%) of cases. Emergency CS was most frequent and performed in 89/121 (73.5%) of total CS, representing 11.5% of SB deliveries.

The classification of causes of death by ReCODE found in our study is presented in Table 2.

Pregnancy details, risk factors according to mode of induction are showed in Table 3. Among induced women, the majority received prostaglandins (394/770), then 97/770 underwent to multiple methods induction and finally 72/770 received only Oxytocin. Mode of induction did not differ in relation with gestational age at stillbirth, while vaginal delivery was significantly higher in women induced with prostaglandins (p = 0.000) respect than other methods. Moreover, nulliparous women had a significantly higher need for multiple methods of induction (p = 0.000) respect multiparous and obese women used more frequently prostaglandins (p = 0.03) than other methods.

Mode of delivery in relation with obstetric characteristics are resumed in Table 4. Women with history of previous CS presented a significantly higher rate of repeated elective CS (p = 0.000) both in case of one or two previous CS. Moreover, emergency CS was performed more frequent in obese (p = 0.02), diabetic (p = 0.04) and hypertensive (p = 0.04) women. However, at the multivariate analysis including obesity, diabetes, hypertension and placental disorder, the risk of emergency CS resulted statistically significant only for placental disorders (OR = 2.03, 95% CI 1.24–3.28, p = 0.004).

Based on the primary causes of death (Table 3), women with SB caused by fetal or maternal causes used multiple methods for induction of labor (p = 0.008) and women with unexplained SB used more frequently prostaglandin as a method of induction (p = 0.008). Women with SB caused by placenta disorders did not use any method (p = 0.008) because they frequently underwent a CS.

When analyzing the different placenta disorders, we found that in the case of abruptio placentae women significantly did not use any method of induction (p = 0.000) and delivered by emergent CS (p = 0.000); while in the case of chorioamnionitis and funisistis women significantly were induced with prostaglandin (p = 0.000) and delivered vaginally (p = 0.000).

Our study on a large population of SB registered a rate of 3.3 per 1000, in line with other developed and European countries [4]. Most of those women delivered, as expected, with induction and only 15.7% with cesarean section. Appropriate methods for labor induction vary based on gestational age at the time of fetal death, previous obstetric histories such as cesarean section and maternal preference. In our population, SBs were induced equally with prostaglandins including misoprostol, multiple methods, and oxytocin regardless of gestational age, contrarily to what is reported by previous RCT data [18, 24] and metanalysis [15]. They indeed evaluated misoprostol (both oral and vaginal) as the most efficient method of induction before 28 weeks. In our population prostaglandins and oxytocin were used equally in second and third trimester as demonstrated by several studies [16, 25, 26, 27]. In our population Mifrepistone was not usually used, nevertheless Perritt et al. [17] demonstrated that Mifepristone (either 200 or 600 mg orally) can be used as an adjunct to misoprostol for induction of labor in the setting of SB because it reduces the time to delivery when compared with misoprostol alone.

Women with previous CS (either one or two) delivered in our study more frequently with elective CS. Moreover, our data showed that the mode of delivery is not influenced by the gestational age, indeed the rate of caesarean sections does not vary in relation to the gestational age classes. This datum does not confirm the ACOG guidelines [28], where appears that mode of induction and delivery are related with gestational age (28 weeks) and previous uterine scar. Indeed, in these patients the ACOG guidelines recommend the use of mechanical method with Balloon or Foley and the individualization of the delivery plan in case of previous multiple CS, based on the specifical circumstances and patient preference [6].

According to our results, we could speculate that Italian obstetric providers fear the uterine rupture, also in the management of stillborn delivery, in women with previous uterine scar. Nevertheless, in our population the rate of CS was about 16% of cases, lower respect with other Spanish study [29] where 22.9% of stillbirths $\ge$26 weeks gestation was delivered by CS, with a significant (p 0.001) difference between public hospitals (16.8%) and private hospitals (41.5%). Interestingly, in our stillborn population, the rate of emergency CS was 11%, and these were performed predominantly in obese, diabetic, hypertensive women, and in those cases of placental primary cause of death, confirming the higher incidence of acute events (i.e., abruptio placentae) in these pathologic women [30]. Thus, the need of emergent CS in these situations explains the physicians’ attempt to save the baby and the uterus through a fast delivery.

Epidural analgesia was performed in most cases of SBs (80%), considering also that 16% of women underwent a CS. This data showed that, in the same area, in case of SB the request of epidural analgesia was significantly higher respect with the women who deliver a live fetus, who used this method in 25% of cases, while benefit from non-pharmacological methos in 73% of cases [23]. This datum confirms that our assistance respects recent guidelines [5] which recommend the use of labor analgesia services in cases of stillbirth, showing a better and sensible care than that offered by Japanese health, where epidural analgesia is poorly used [19].

Our assistance offered the possibility to delay the admission to the hospital for these patients, nevertheless only few cases experimented this plausible option as demonstrated by Muin DA et al. [31] who did not find significant difference regarding delivery mode, labor duration, use of intrapartum analgesia, need for episiotomy and risk of perineal injury and antepartum hemorrhage between the groups of immediate and delayed admission in a population of stillbirths.

The limitation of our study is the lack of data about the time between induction to delivery and the absence of cases that experimented the delayed admission to the hospital.

To the best of our knowledge, this is among the largest prospective area-based study on a stillbirth population where each case was carefully investigated and classified, evaluating then the mode of delivery as well as the mode of induction in relation with the different causes of death.

In conclusion our data showed that the method of induction of labor and the mode of delivery in case of stillbirth did not depend on gestational age at the diagnosis of death, while they are related to placenta disorders, representing relevant conditions leading to emergency CS also after diagnosis of fetal death. These data could help obstetric providers in managing the deliveries of stillborn infants.

Authors obtained written informed consent for the study and publication.

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

FM and FF conceived the study. CS, FM, VD, BM and FM managed the data collection. DM and IN managed the analysis of the data. Drafting of the manuscript was led by DM, FM with input from FF, who give the final approval of the version to be published. All authors have read and approved the final manuscript.

The Ethical approval for this study was obtained from the local institutional review board (35265-24/11/21).

We would like to thank all the components of the Stillbirth Emilia-Romagna Audit Group* for the great work accomplished.

*The Stillbirth Emilia-Romagna Audit Group:

Group of Modena: Licia Lugli, Isotta Guidotti (Neonatal Unit), Tiziana Salviato, Luciano Mancini (Pathology Unit) from Policlinico of Modena, Paola Sparano, Francesco Torcetta (Neonatal Unit), Chiara Lanzoni, Giulia Pedrielli (Obstetrics and Gynecology Unit) from Ospedale Ramazzini di Carpi), Federica Ricchieri (Obstetrics and Gynecology Unit), Claudio Chiossi, Rossella Pagano (Neonatal Unit) from Nuovo Ospedale Civile di Sassuolo, Laura Sgarbi, Alessandro Ferrari, Cristina Pizzi (Obstetrics and Gynecology Unit, Ospedale S. Maria Bianca di Mirandola).

Group of Ferrara: Erica Santi, Cristina Banzi, Roberta Cappucci, Maria Grazia Cristofori (Obstetrics and Gynecology Unit), Raffaella Contiero, Gianpaolo Garani, Elisa Ballardini (Neonatal Unit), Sergio Fini (Genetic Unit), Massimo Pedriali, Sonia Rossi (Pathology Unit) from Azienda-Ospedaliero Universitaria di Ferrara, Massimo Di Bartolo, (Obstetrics and Gynecology Unit) from Ospedale del Delta, Daniele Radi (Obstetrics and Gynecology Unit) Ospedale SS. Annunziata di Cento.

Group of Bologna: Alessandra Vancini, Fabrizio Sandri, Luca Le Pera (Neonatal Unit), Anna Donati, Marinella Lenzi, Eleonora Guadalupi, Ilaria Cataneo (Obstetrics and Gynecology Unit), Angela Salerno (Pathology Unit), Francesca Righetti (Clinical Pathology Unit) from Ospedale Maggiore di Bologna, Francesca Fantuz, Raffaella Morandi (Obstetrics and Gynecology Unit), Giacomo Caprara, Nunzio Cosimo Mario Salfi (Pathology Unit), Guido Cocchi, Chiara Locatelli, (Neonatal Unit), Liliana Gabrielli (Microbiology and Virology Unit), Claudio Graziano, Marco Seri (Genetic Unit), from Ospedale S. Orsola-Malpighi di Bologna, Federica Ferlini (Obstetrics and Gynecology Unit), Deborah Silvestrini (Neonatal Unit), Elisa Righi (Pathology Unit) from Ospedale di Imola, Franco Foschi (Neonatal Unit) from Ospedale di Bentivoglio.

Group of Parma e Piacenza: Stefania Fieni, Tiziana Frusca, Alice Ferretti, Laura Angeli, Alissa Valenti, Letizia Galli, Arianna Commare (Obstetrics and Gynecolgy Unit), Cinzia Magnani (Neonatal Unit), Enrico Silini (Pathology Unit) from Azienda Ospedaliero-Universitaria di Parma, Letizia Balduzzi (Obstetrics and Gynecology Unit), Melissa Bellini (Neonatal Unit), Anna Maria Rodolfi (Pathology Unit) from Ospedale di Piacenza, Maria Paola Sgarabotto (Obstetrics and Gynecology Unit), Giorgia Fragni (Neonatal Unit) from Ospedale di Fidenza.

Group of Reggio Emilia: Giuseppina Comitini (Obstetrics and Gynecology Unit), Giancarlo Gargano, Melli Nives (Neonatal Unit), Maria Paola Bonasoni (Pathology Unit) from Arcispedale S. Maria Nuova di Reggio Emilia, Loredana Fioroni (Obstetrics and Gynecology Unit) from Ospedale Castelnuovo Monti, Marco Panteghini, Cristina Rozzi (Obstetrics and Gynecology Unit) from Ospedale di Montecchio Emilia, Antonella Tuzio (Obstetrics and Gynecology Unit) from Ospedale di Scandiano; Ida Vito (Obstetrics and Gynecology Unit) from Ospedale di Guastalla.

Group of Area Vasta Romagna: Palma Mammoliti, Marilù Capelli (Neonatal Unit), Elena De Ambrosi, Elisa Tidu (Obstetrics and Gynecology Unit), Monica Ricci (Pathology Unit) from Ospedale Infermi di Rimini; Angela Bandini (Obstetrics and Gynecology Unit) from Ospedale di Forlì; Chiara Belosi (Obstetrics and Gynecology Unit) from Ospedale di Faenza; Claudia Muratori, Giuliana Vania (Neonatal Unit), Tiaziana Arcangeli (Obstetrics and Gynecology Unit), Silvia Zago (Pathology Unit) from Ospedale di Ravenna; Susanna Giorgetti, Marisa Vitarelli (Obstetrics and Gynecology Unit), Arianna Leone (Neonatal Unit) from Ospedale di Cesena.

Regione Emilia-Romagna DG Cura della persona, salute welfare: Elena Castelli, Elisabetta Mazzanti, (Servizio assistenza territoriale), Camilla Lupi, Sergio Battaglia (Servizio ICT, tecnologie e strutture sanitarie).

This research received no external funding.

The authors declare no conflict of interest.