The data of the present study shows that supraglottic airway devices are common in clinical practice and are used in the majority of cases, not only in low-risk cases, but to a considerable extent even in vital interventions. A comparison of patients who received supraglottic airway management versus those who were intubated shows that intubation was preferred in non-fasting patients, in patients with ASA 4 classification, in patients undergoing urgent procedures, as well as in patients undergoing highly complex surgical interventions. This leaves room for the assumption that tracheal intubation still seems to be the method of choice and indeed the standard in high-risk cases. LMA and bag mask ventilation was predominantly used in obese patients and smokers, as well as those with history of gastroesophageal reflux disease, diabetes, asthma and hypothyroidism. Although the use of laryngeals masks, including safety and efficacy in pregnant women has been addressed in more recent studies, those reports were limited to caesarean sections [7, 16].

The postoperative status of patients receiving intubation was judged as unstable in 1.7 %. In those managed with a supraglottic airway device or with bag mask ventilation, 0.2 % were judged as unstable. Subsequently, the reason for this assessment cannot be determined any more. However, it seems likely that unstable patients are preferentially managed by tracheal intubation.

We intended to select patients who present risk factors for pregnancy associated pulmonary aspiration, such as an enlarged uterus, a decrease in the lower oesophageal sphincter with consecutive gastroesophageal reflux or increased BMI. However, several studies could not confirm the BMI as an independent risk factor for pulmonary aspiration [17, 18]. Interestingly, obese patients were managed predominantly with LMA and bag mask ventilation, although higher grades of obesity are said to be associated with a higher risk of aspiration. Not only elective but also emergency procedures that are believed to have a higher risk of aspiration were considered [9, 17]. In our study, 32.5 % of all interventions were emergency procedures. A selection of other factors generally thought to increase the risk did not lead to aspiration either: At night, 868 (36.3 %) operations were performed. Out of the 2390 patients, 4.8 % were classified as risk class ASA 3 or higher. Overall, it was found that neither pregnancy alone, nor in combination with one or more risk factors, promoted aspiration in our retrospective analysis.

In the literature on pulmonary aspiration in pregnant women, incidences range from 0.01 – 0.23 %, [19, 20] more recent data describes an incidence of 0.022 % for pulmonary aspiration during pregnancy or immediately postpartum in association with general anaesthesia [21]. Our sample size of 2390 patients was not high enough to safely detect a single case of pulmonary aspiration. When referring to data that has been recorded more than 25 years ago, one must keep in mind that technical achievements, growth in knowledge and training in the field of anaesthesia have been further developed. Second-generation supraglottic airway devices that have better, although not complete, protection against aspiration are now recommended as rescue airway devices after failed tracheal intubation, particularly in patients at increased risk of aspiration [22].

It is difficult to compare studies investigating the incidence of pulmonary aspiration in an obstetric surgical population, since the definition of pulmonary aspiration is inconsistent, as are other factors, such as patient populations and interventions considered, and the level of training of the anaesthesiologist performing the procedure. Several authors have already referred to the ambiguous definition of pulmonary aspiration [17, 23].

Difficult intubation is another independent risk factor for aspiration [4, 24]. In our study population, we found 3 difficult intubations and one failed intubation (finally managed with LMA) within the Cormack and Lehane score 1 & 2 group and 5 patients with Cormack and Lehane score 3 & 4.

We found an incidence of difficult/failed intubation of 1.4% (9 out of 639) in our study population. Difficult intubation using Wilson and Arné scoring was predicted in 10 out of 1582 cases (0.6%). However, not all anaesthesiologists seem to agree that a Cormack and Lehane score of 3 should be considered as being difficult to intubate [25].

Existing data regarding difficult intubation is poor because most studies examine the incidence of failed intubation [25, 26]. Values in the range from 3.3 % - 7.9 % are documented for the incidence of difficult intubation in obstetric anaesthesia, but these do not coincide with those of the present work, where an incidence of 1.4% has been found [25, 27].

In general, it is said that obstetric patients are more difficult to intubate than general surgery patients [28, 29]. The definitions used for airway problems are inconsistent and potentially misleading. Definitions include failure to intubate, three or more laryngoscopies, four or more attempts to pass the endotracheal tube or longer duration of endotracheal placement (>10 minutes), or a poor view of the vocal cords via direct laryngoscopy following anaesthesia induction [30]. Due to inconsistencies in defining difficult intubation, comparability of our data with other studies seems to be hardly possible [25, 26]. The incidence of difficult intubation in our study is lower than those published by the mentioned authors.

The incidence of failed intubations in the field of obstetric anaesthesia is described as being between 0.08 - 0.45 % and 0.1 % - 0.47 % in the general surgical population [26, 30, 31]. In this retrospective study, the rate of failed intubations is 0.16 % (1 out of 638) and thus in the range of the discussed values in the obstetric population.

When comparing the number of potential pulmonary aspiration events in relation to specific surgical characteristics, the highest risk is found in ASA 2 patients that are fasting for more than 6 hours and undergo a removal of retained placenta procedure in the setting of an emergency surgery. However, the absolute figures are put into perspective when the frequency distribution is considered, so that no intervention-specific risk profile can be identified.

Despite the fact that our data was collected in a single centre, it might have implications. The risk of aspiration during general anaesthesia for obstetric procedures managed without tracheal intubation in our study population was not higher than that reported in the literature. In the majority of our patients, airway management was safely performed with supraglottic airway devices. Especially when it comes to a "cannot intubate" situation, one should be aware that in our work many cases could be treated with a supraglottic airway without complications. Therefore, our data might support the concept that intubation should not be forced, especially in pregnant women, even if the physiological changes of pregnancy are already present. In our opinion, the gestational age should be rated to a smaller amount in deciding whether RSI induction is indicated. Rather, clinical assessments and anaesthetist’s experience level should be given greater consideration. The necessity of endotracheal intubation to secure the airway in pregnant women [14, 32, 33] and the procedure of RSI [32, 34] has been questioned by different authors.

Our study has several limitations. First, as it is a retrospective analysis our results are prone to misclassification bias. Data was manually extracted from the anaesthesia database with the greatest possible diligence, but definitions and procedures were not standardised a priori.

Second, since this was a single-centre study, caution is advised when extrapolating our findings to other centres. Third, the sample size was relatively small compared to reports, [16, 19,20,21] which may increase the risk of selection bias. Since the incidence of pulmonary aspiration is lower than 1:2000, [17, 21, 35] we cannot completely rule out that absence of this adverse event is attributable to chance. However, this limitation could be overcome through the use of national databases or by collaborating in a multicentre setting to retrospectively examine aspiration complications.