Crash caesarean section audit: an important tool for an obstetric unit to ensure safe quality care

Abstract

Background

Crash caesarean section (CCS) aims to achieve delivery within 30 min of decision in potentially life-threatening situations for mother, baby or both but is not without risk. Auditing CCS in any unit providing surgical delivery is important to ensure compliance with the recommended guideline while ensuring validity of indication for surgery and maintaining safe care. Here we present the results of a retrospective audit of CCS in a busy tertiary maternity unit in Singapore looking at the DDI and validity for surgery and reporting on outcomes.

Method

Maternal demographics and antenatal details of mothers delivered by CCS in the preceding 4 weeks, collected from the patient’s electronic medical records, along with perinatal events leading to the CCS were presented at a monthly meeting attended by obstetricians of all grades. The decision to delivery interval (DDI) was noted and management scored with respect to appropriateness of operative indication from 1 (appropriate) – 3 (not appropriate). The outcome of cases with a score > 1.5 was scrutinised for areas of care which might have had a negative impact on final outcome.

Results

One hundred and forty-eight CCS were identified. Twenty-one cases (14.2%) scored > 1.5. All CCS were performed within the recommended 30 min with a mean of 10.2 min. There were 15 maternal complications, 1 birth injury and 4 perinatal deaths. The top 3 indications for surgery were non-reassuring fetal status (83/148), cord prolapse (15/148) and placental abruption (10/148). Sixty-eight babies were admitted to the neonatal intensive care unit (NICU), 36 to the special care nursery (SCN) and 49 went to the general ward.

Conclusion

Regular audit of CCS for DDI and validity of surgical indication with the inclusion of perinatal outcomes is an important tool in improving quality of care ensuring compliance with guidelines while maintaining safety of care by identification and correction of deficiencies.

Introduction

The main aim of the audit was to examine whether all CCS performed over a 17- month period in a tertiary obstetric unit in Singapore with around 12,000 deliveries annually were performed within the 30 min (min) DDI recommended for a Category 1 Caesarean section (CS) and had a valid indication. Inclusion of perinatal maternal and neonatal outcomes allowed safety aspects to be studied and deficiencies in care to be recognised and addressed.

‘Crash’ caesarean sections are done in the hospital according to a protocol designed for use in critical situations [1, 2]. ‘Code green’, initiated through the switchboard by the attending obstetrician activates a multidisciplinary coordinated response to facilitate rapid delivery of the baby (Fig. 1) usually from the labour ward. This includes preparation and transport of the patient to the operating theatre (OT) where anaesthetic, surgical, nursing and neonatal teams take over management. At times it is also used to summon immediate assistance in OT if delivery is difficult. It is viewed as a priority safety initiative and an OT is always available. The system does allow fetal delivery in a very short time but is resource intensive and not without inherent safety risks. To ensure continuing effectiveness of the CCS protocol, a monthly audit is conducted to ensure compliance with guidelines and identify shortcomings which potentially compromise safety with an aim to reduce morbidity and mortality [3].

Fig. 1.

The flow of events after “code green” is activated

Recognised indications for Category 1 CS include suspected uterine rupture, cord prolapse, antepartum haemorrhage due to an abruption or vasa praevia and pathological fetal heart patterns, unresponsive to intrauterine resuscitation measures [4] all of which can result in fetal hypoxia and death and significant maternal morbidity. A group of Anglo-Americans [4, 5] recommended a DDI of 30 min for a Category 1 Caesarean section (CS) which should be achievable in any Level 3 maternity unit although this does not guarantee against long term fetal or maternal sequelae in every circumstance [6]. In some European countries, including Germany [7] and France [8, 9], an even shorter DDI has been recommended for life-threatening indications but achieving rapid operative delivery is not without risk and the German Society of Gynaecologists and Obstetricians’ recommendation of 20 min has been associated with both adverse neonatal and maternal outcomes [10]. Although ‘Category 1’ is conventionally used to indicate a critical indication the term ‘crash’ has been used to highlight urgency. Invariably a ‘crash’ situation is always stressful for the medical team and patient involved.

“Coding” is an effective way to activate a dedicated team and reduces the period of fetal compromise ensuring compliance with the DDI guidelines but depending on the size of the responding team it can lead to overcrowding and lack of leadership in OT and does not safeguard against maternal or fetal injury. The facility to ‘crash’ can also be subject to abuse by obstetricians who want to jump the existing surgery queue and may incur additional charges for the patient in places where healthcare provision is not free [11].

In our unit there is a monthly audit of all CCSs focussed on surgical indication and DDI at which cases deemed inappropriate for ‘crashing’, for whatever reason, are discussed further with the aim of reducing the associated risk and improving the safety and quality of care.

Methods and materials

The study looked at all CCSs performed between 1 st April 2021 and 31 st August 2022.

As it was a clinical audit, the Singhealth Centralised Institutional Review Board did not require informed consent.

The main outcome measures were validity of indication and DDI but neonatal and maternal outcome measures were also included.

The audit meeting took place on a monthly basis attended by a group of obstetricians at grades ranging from junior specialist trainee to senior consultant. The numbers and grades of staff present varied between meetings. All CCS performed in the preceding four weeks were discussed. A summary of each mother’s demographics, any past medical and obstetric history of note and antenatal course in current pregnancy including investigations, obtained from the contemporaneous electronic medical record, systemically reviewed and presented by an obstetric resident. Perinatal events and investigations leading to the CCS including the cardiotocogram, any anaesthetic involvement, time of day of surgery and neonatal outcomes were also recorded. The management of each case was reviewed by all the attending clinicians who assigned a score to the appropriateness of operative indication between 1 appropriate, 2 appropriate with reservations and 3 inappropriate which was submitted via mobile phone and processed by a central system (by summating and dividing by number of submitting clinicians) to produce an overall score. A score > 1.5 triggered further discussion into factors influencing decisions taken, adverse events and shortcomings of care. At the start of the audit, the baby’s condition was revealed before scoring took place but this was changed in April 2022 because of perceived bias from known outcome. The learning points arising from the meeting were subsequently presented at the Labour Ward Steering Committee meeting where discussion among senior obstetricians led to changes in protocol for safer practices. These were highlighted in a monthly departmental-wide ‘bulletin’.

The scoring system used in the audit was a surrogate for decision making around the need to ‘crash’ with an implication for final outcome. Scores of </=1.5 indicated consensus on need for ‘crashing’ with final outcome reflecting on validity while scores > 1.5 raised questions around the need to ‘crash’ with safety implications for final outcome. A score of </= to or > 1.5 was chosen as it was considered to discriminate between cases with valid indication for CCS which would contribute to final outcomes (</=1.5) and those where the indication was questionable with final outcome having implications for safety of care.

All analyses were performed using means and standard deviations and compared with a student’s t test using Stata 13.0.

Results

Over the 17-month study period there were 16 623 deliveries and 5364 caesarean sections. A total of 148 cases of CCS were identified and analysed for indication. All decisions for CCS were made at the level of obstetric registrar or above except 1, made by a midwife. The cohort of women undergoing CCS had a mean age of 31.9 ± 5.2 years. The majority were multiparous women (56%), and 29 (19.6%) had had a previous caesarean` section. The mean gestation at delivery was 36.3 ± 1.6 Weeks. Labour had been induced in 49 (33.1%). Table 1. shows the demographics of the women.

Table 1.

Demographics and baseline characteristics (Mean values)

	N = 148	SD
Maternal age, years, SD	31.9	±5.2
Primigravidae	65 (43.9%)
Gestational age, weeks, SD	36.3	±4.6
Previous Caesarean section	29 (19.6%)
Multiple pregnancy	10 (6.8%)
After hours delivery n (%)	107 (72.3%)
Induced labour	49 (33.1%)
Vaginal delivery after code	3 (2%)
Gestation < 27 weeks	12 (8.1%)
General anaesthesia	71 (47.8%)
Regional anaesthesia	74 (50%)

71(49%) cases had general anaesthesia and 74(51%) had regional anaesthesia with no statistically significant difference in average DDI for the CCS which were 10.3 ± 3.1 min and 10.4 ± 3.9 min respectively(p = 0.8716).

The recommended DDI of less than 30 min was met by all the CCS in our cohort with an overall mean of 10.5 min (range: 6 to 24 min). Six of the 148 cases were excluded from the calculation of the mean DDI. Three which were coded for difficult extraction of baby in the operating theatre and 3 who had vaginal deliveries shortly after code activation, none of the latter 3 having any anaesthesia and all in advanced labour, 2 with breech presentation and the third with a history of two previous caesarean sections.

This represented 2.8% of all the caesarean sections and 0.9% of the 16 623 deliveries over the study period. 72.3% of cases were performed outside of office hours with an average DDI of 10.4 ± 3.4 min, compared to 9.9 ± 4.3 min done during office hours. This difference in DDI did not reach statistical significance. (p = 0.4566).

Fetal distress (83/148) was the top indication for CCS followed by cord prolapse (15/148), mostly late preterm, and placental abruption (10/148). Table 2 shows the primary indications for all CCS with their respective DDI.

Table 2.

The primary indications for crash CS and the respective DDI

Primary indications for crash CS	No. of cases	DDI* (min) mean	Range (min)
Abruption	10	8.9	(7–12)
APH	2	5	(3–7)
Cord Prolapse	15	12.1	(6–22)
Cephalopelvic disproportion (os full)	2	9.5	(8–11)
Difficult extraction in Operating Theatre	3	3.3	(1–4)
Eclampsia	2	9	(7–11)
Failed instrumental	9	11.1	(7–14)
Maternal AFE/Pulmonary embolism/Maternal anaphylaxis	2	11.5	(8–15)
Non-cephalic presentation	8	14.5	(9–24)
Non-reassuring fetal status	83	10	(6–24)
Pre-eclampsia	2	10.5	(10–11)
Previous CS Scar pain/scar dehiscence	3	13	(10–19)
Multiple pregnancy	3	11.5	(9–13)
Vasa praevia	1	10	-

 DDI Decision to Delivery time, AFE Amniotic fluid embolism, CS caesarean section

All cases were scored for the appropriateness for indication. 21 (14.2%) cases had an overall score of > 1.5 and were subject to further discussion on perinatal actions taken and decision making (Table 3). For 18 cases, the decision for CCS was made by an obstetric consultant, two by an obstetric registrar and one by a midwife. The cited indications for these cases were NRFS including a fetal bradycardia of less than nine minutes without attempts at intrauterine resuscitation measures, failed instrumental delivery in the absence of fetal compromise, breech presentation, cephalopelvic disproportion (CPD), pre-eclampsia and caesarean scar pain.

Table 3.

Cases with score > 1.5

Case No.	Indication for CCS	Type of anaesthesia	DDI	Score	Learning points from discussion
67	Breech	GA	10	1.99	Consider Cat 2 CS, use of terbutaline prior to CS
42	Breech	ETU	14	1.66	Consider Cat 2 CS, use of terbutaline prior to CS
62	Breech delivery	NA	1	1.83	Successful breech delivery. CCS not indicated.
140	Breech delivery	NA	6	2	Successful breech delivery. CCS not indicated.
41	CPD (os full)	ETU	8	1.96	Normal CTG noted. CCS not necessary.
39	Failed instrumental delivery	ETU	12	1.91	Normal CTG noted. CCS not necessary.
83	Failed instrumental delivery	GA	10	2.23	Normal CTG noted. CCS not necessary.
89	Failed instrumental delivery	GA	14	1.66	Normal CTG noted. CCS not necessary.
92	Failed instrumental delivery	ETU	14	1.84	Normal CTG noted. CCS not necessary.
101	Failed instrumental delivery	GA	10	2	Normal CTG noted. CCS not necessary.
23	NRFS	ETU	10	3	CTG reviewed. CCS not necessary.
38	NRFS	GA	9	1.77	CTG reviewed. CCS not necessary.
48	NRFS	GA	8	1.62	CTG reviewed. CCS not necessary.
52	NRFS	ETU	10	1.6	CTG reviewed. CCS not necessary.
66	NRFS	GA	10	1.66	CTG reviewed. CCS not necessary.
80	NRFS	ETU	8	2.52	CTG reviewed. CCS not necessary.
84	NRFS	GA	8	1.62	CTG reviewed. CCS not necessary.
96	NRFS	ETU	9	2.14	CTG reviewed. CCS not necessary.
107	NRFS	ETU	7	2	CTG reviewed. CCS not necessary.
49	Pre-eclampsia	GA	11	2	No clinical indication for CCS
64	Scar pain	GA	10	1.78	No clinical indication for CCS.

A total of 154 babies were delivered. There were 9 twin deliveries. 68 babies were admitted to NICU, 36 to SCN and 49 went to the general ward. There was one birth injury and three perinatal deaths and one stillbirth. Table 4. summarizes the neonatal outcomes by ward admission status and average cord PH. There was 1 case of neonatal trauma, a displaced femoral fracture. All three perinatal deaths were in women who had the CCS for NRFS and the stillbirth was in a woman who was delivered at 23 weeks for malpresentation.

Table 4.

Summarized neonatal outcomes in terms of baby ward admission status and average cord PH

Primary indications for crash CS	No of babies	Admission ward status (no of cases)			Birth Trauma	Stillbirth	NND	Average cord pH
		NICU	SCN	P ward				Arterial	Venous
Abruption	10	6	2	2	0	0	0	7.10	7.19
APH	2	1	1	0	0	0	0	7.14	7.2
Cord prolapse	19	13	2	4	0	0	0	7.14	7.22
CPD (os full)	2	2	0	0	0	0	0	7.10	7.17
Difficult extraction in OT	3	2	1	0	0	0	0	7.05	7.19
Eclampsia	2	1	1	0	0	0	0	6.94	6.96
Failed instrumental	9	3	3	3	1	0	0	7.16	7.21
Maternal AFE/PE/Maternal anaphylaxis	2	2	0	0	0	0	0	6.86	6.92
Non-cephalic presentation	8	4	1	2	0	1	1	7.19	7.29
NRFS	85	26	23	36	0	0	3	7.11	7.16
Pre-eclampsia	2	1	1	0	0	0	0	7.10	7.02
Previous CS Scar pain/scar dehiscence	3	0	1	2	0	0	0	7.2	7.26
Multiple pregnancy	6	6	0	0	0	0	0	7.20	7.28
Vasa praevia	1	1	0	0	0	0	0	6.87	-

CS caesarean section, NICU Neonatal intensive care unit, SCN Special care neonate, NND Neonatal death, P Periphery APH Antepartum haemorrhage, OT operating theatre, AFE Amniotic fluid embolism, PE Pulmonary embolism, NRFS Non reassuring fetal status

There were 15 (10.1%) maternal peri-operative complications, including one case of bladder injury and five uterine extension tears. Eight patients had an intra-operative blood loss of at least one litre, and one patient underwent hysterectomy after an estimated blood loss (EBL) of four litres (Table 5).

Table 5.

The maternal complications with indication for CCS & DDI

Case No.	Indication for CCS	Type of anaesthesia	DDI (min)	Maternal complication
113	Cord prolapse	ETU	12	Inverted J incision
115	Cord prolapse	GA	9	EBL 1000 ml
129	Cord prolapse	GA	20	EBL 4000 ml
39	Failed instrumental delivery	ETU	12	2 cm vertical extension right
110	Maternal AFE/PE	GA	8	EBL4000ml hysterectomy, relaparotomy
2	NRFS	GA	7	extension left uterine angle
19	NRFS	GA	12	EBL1000ml
23	NRFS	ETU	10	EBL 5000 ml
28	NRFS	GA	10	EBL 1000 ml with cervical laceration
32	NRFS	ETU	12	lateral extension of uterine incision with laceration of the uterine vein EBL 1000 ml
105	NRFS	GA	10	EBL 1000 ml
3	Placenta abruption	GA	7	EBL 1000 ml
6	Placenta abruption	GA	9	EBL 1000 ml
49	Pre-eclampsia	GA	11	cx bladder serosal tear
40	Transverse lie	ETU	24	J Uterine incision with vertical extension

AFE Amniotic fluid embolism, PE Pulmonary embolism, NRFS Non reassuring fetal status, GA General anaesthesia, ETU Epidural top up, EBL Estimated blood loss

Discussion

There have been three previous studies on the DDI of CCS in Singapore. Kwek et al. and Lim et al. reported average DDI’s of 14.6 min and 7.7 min respectively [1, 5]. In the third study, which showed a mean DDI of 10.8 min Nair et al. suggested that an available operating theatre was essential to achieve this target [12] which led to the current set-up. This audit showed that the code green protocol has enabled our unit to continue to comply with the recommended DDI despite the increasing CS rate over time and compared to the previous audit [1] there has been a fall in the maternal morbidity rate from 17.5% to10.1%. This is almost certainly due to an increased awareness of safety in healthcare practices mandating the introduction of skills training in this area since those initial studies. In fact, the audit showed that a mean DDI of just over 10 min was achieved for our entire cohort which is well below the recommended international standard of 30 min for a Category 1 CS and meets the shorter recommendations for European countries for CS for life-threatening conditions.

Closer analysis of cases of CCS in our audit which scored > 1.5 allowed recommendations to be made on the need for crashing and when it should be avoided which were disseminated within the unit. We found that checking on anaesthetic and theatre availability before coding helped reduce the need for a CCS in certain circumstances.

The 21(14.2%) cases with questionable indications for coding suggests inappropriate use of resources with diversion of care from other areas towards an expensive, and possibly unnecessary, intervention and highlights the fact that an efficient CCS system is vulnerable to abuse and ignorant overuse. Bloc et al. [13] showed that regular audit with multidisciplinary input helps safeguard against the use of ‘crashing’ for inappropriate maternal and fetal complications. This patient safety measure will likely also reduce unnecessary increase cost of the CCS to the patient and the institution [14].

During review of the three cases delivered vaginally after the code was activated the patients were found to have been assessed by a junior member of the team and stage of labour at code activation was wrong This highlighted the need for additional training of junior staff in breech deliveries and second stage instrumental deliveries.

After fetal distress, we found cord prolapse was the second indication for CCS, followed by abruption. Hillemanns et al. [10] also found that cord prolapse along with abruption and CTG abnormalities were the three main indications for CCS. However, in that study more than 50% fell into the very premature group, which was at variance with our findings. Tan et al. [15] looked at the CCS DDI for cord prolapse and found that among 34 women the median time from diagnosis to delivery was 20 min (range 10 to 40 min) and that 76% were delivered within 30 min. In Tan’s study there were 3 NICU admissions but no perinatal deaths. In 2021 Gallagher et al. [16] reported multidisciplinary simulation led to a significantly improved DDI in patients with cord prolapse.

There have been concerns that the time of day of CCS may affect DDI, maternal and neonatal outcomes due to staffing issues and seniority of the medical personnel onsite. Our audit showed the majority were delivered outside hours and there was no statistical difference between DDI of CCS during and outside working hours. Moreover, the level of seniority of the doctor conducting the CCS was the same regardless of the time of day. This was similar to the findings of Amankawah et al. [17] who did a prospective study over 11 months on their “code 333” efficiency. This group reported a median DDI of 16 min with 98% of deliveries being completed within 30 min and found that it was not affected either by time of day or day of the Week. Even in the 2% of codes delivered beyond 30 min there were no adverse outcomes. In a study from Germany, where the recommended DDI is 20 min for CS for life-threatening situations, Brandt et al. [7] looked at the DDI for 437 women outside of core working hours and showed that it was significantly longer, although this did not seem to have a negative impact on neonatal outcome.

Studies have been done on whether the type of anaesthesia affects DDI and maternal and neonatal outcomes. Theoretically, general anaesthesia (GA) might help achieve the shortest DDI for a CCS but increases the risk of maternal complications. One of the local studies of 98 women (2.7% of all CS) showed the majority of patients (88%) had general anaesthesia and the mean DDI was 7.7±3 min with all deliveries occurring within 17 min [5]. In our audit, in which 49% of women had CCS under GA, there was no statistical difference in DDI between women having CCS under GA or regional anaesthesia. Bidon et al. looked at the DDI of 194 women who underwent CCS, 67 of whom (35%) were under GA. The median DDI was 10 min (8–12.5 min) and 174 (90%) of women had CS in under 15 min. Their study found that general anaesthesia was independently associated with a depressed 5-minute Apgar score in the neonate and suggested early insertion of an epidural catheter for women deemed likely to need an emergency caesarean section. In a retrospective review of 135 women with emergency caesarean section, Hein et al. [18] showed, that GA shortened the DDI significantly by 10 and 13 min over spinal anaesthetic and top-up epidural respectively. Despite this, the authors suggested that regional anaesthesia with its potential advantages should still be considered if time allows.

Conclusion

Crash caesarean section is a highly valuable but expensive resource. The results of this audit show that, with the activation protocol for ‘crash’ and close multidisciplinary effort, our CCS complied with the recommended 30 min DDI and the majority were performed in under 20 min. This compliance is essential in a busy tertiary teaching hospitals with a significant number of high- risk patients as it represents a way of decreasing maternal and perinatal morbidity and mortality in acute situations and by analysing outcome measures has potential to improve safety and quality of care. The approximate 14% “misuse” rate shows the importance of multidisciplinary review to highlight systemic issues which might need improvement with senior input. Our monthly audit forms an integral part of clinical governance for monitoring adverse events around delivery especially for cases potentially open to complaint and litigation and lends itself to guiding improvements in clinical practice.

This study was a large retrospective study conducted at a tertiary centre in Singapore. Although the results cannot be extended to centres in which reaching the target DDI is impossible, for example in third world countries lacking infrastructure and the staff and training required for a coordinated response, they may be generalisable to similar centres in the developed world and provide pointers towards setting up an efficient CCS system.

The discriminatory scoring system used in this audit to differentiate outcomes due to surgical indication from those reflecting safety of care might be considered contentious as level could be adjusted according to priority. In addition, the mix of grades of obstetric staff, which varied between meetings, scoring appropriateness of action could be considered a limitation of the study. Prospective studies looking at differing groups of women for incidence might be an alternative way of studying this aspect of care and remove the need for scoring.

Authors’ contributions

AW: Project development, manuscript writing and editing. NYHG: data analysis and manuscript editing. TLK: manuscript editing.

Funding

We would like to thank Benjamin Henry Sheares Professorship in Obstetrics & Gynaecology for the funding support.

Data availability

The datasets used and/or analyzed during the current study are not publicly available due to limitations involving the patient data and anonymity but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The need for Informed Consent was waived by the Singapore Singhealth Review Board due to the retrospective and audit nature of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.