The objective of this systematic review is to examine the global prevalence, structure and effectiveness of existing IP SBT programmes in managing combined obstetric and neonatal emergencies. Specifically, this review seeks to answer the following research question: what is the global landscape of IP SBT for combined maternal-neonatal emergencies, and how do these programmes impact practitioner-level performance and organizational-level outcomes?

This review followed PRISMA guidelines [10] and was prospectively registered on PROSPERO (CRD42024591980) [13].

The initial database search was performed in PubMed, EMBASE, CINAHL and MEDLINE from database inception to March 2023, repeated in October 2024, and most recently updated in March 2026 to ensure the inclusion of the most current literature (Table 2). No additional studies meeting the inclusion criteria were identified in this final update.

To ensure strict blinding during the review and extraction stages, all records were managed using Covidence Systematic Review Software (Veritas Health Innovation Ltd, Melbourne, Australia). Covidence facilitated the automatic removal of duplicate studies, supplemented by manual deduplication by the authors. Study selection was conducted in iterative phases to maintain methodological rigour: •

Initial Searches (through October 2024): two authors (JS and KXZ) independently screened titles and abstracts, followed by full-text screening, with conflicts resolved by a third independent reviewer (MB). Following the identification of the 40 included studies during this phase, JS and KXZ independently extracted data into separate Microsoft Excel spreadsheets, which were then combined into a single results table following the process recommended by The Cochrane Handbook [22].

This rigorous, multi-stage screening process ensured that all identified records were assessed by at least two independent reviewers while maintaining procedural transparency and strict blinding.

The ROBINS-I tool was used to assess risk of bias in quantitative studies and the JBI critical appraisal tool was used for qualitative studies [23,24]. Two authors (JS, KXZ) independently evaluated risk of bias with conflicts resolved by a third reviewer (MB). The certainty of evidence was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach [25].

A PRISMA flow diagram (Figure 1) outlines the selection process. The combined database searches identified a total of 7473 records. Following the removal of 2087 duplicates, 5386 studies were screened at the title and abstract stage. Full-text review was conducted for 257 studies, of which 217 were excluded. The updated search in March 2026 confirmed that no new studies meeting the inclusion criteria had been published since the previous search iteration. Consequently, 40 studies remained for final data extraction and synthesis [10,15,18–20,26–60]. We summarized the key characteristics and findings of these studies (Appendix 1).

Various SBT programmes were identified, with 14 studies conducted without a specific programme name [26,28,37,40,41,44,46–48,51–54,58]. The most commonly reported programmes included variations of the Obstetrics and Neonatal Emergency Simulation programme (ONE-Sim, n = 7) [18,29,35,42,49,50,60], Simulation and Team Training for Obstetric and Neonatal Emergencies programme (PRONTO, n = 6) [10,30,45,55–57] and Enhancing Training and Technology for Mothers and Babies in Africa programme (ETATMBA, n = 2) [15,31]. Other named programmes included PROMPT and EONC [20,39]. Kogutt simulated delivery in a unique biocontainment delivery [37].

Few studies detailed the exact methods and depth of facilitator training. Where reported, training ranged from specialist-led workshops to simple peer-led facilitation [32].

Programme duration ranged from short 65-minute workshops to programmes that spanned over several months and included repeated modules, lectures, simulations and mentorship. Fourteen studies implemented programmes that were able to be completed within one day, with durations ranging from 1 hour to half a day [26,29,33,35,37,39,42,43,48–50,52,53,59].

Scenarios simulated included obstetric and neonatal emergencies as well as ‘borderline’ conditions – pre-eclampsia was simulated in 13 studies [10,19,26–28,30–32,39,45,46,54,56] whilst PPROM and chorioamnionitis were specifically addressed in the study by Miller et al. [45]. Most frequently simulated conditions included PPH (n = 27) [10,18,19,26–36,39–41,45,46,48,49,54,56–60], shoulder dystocia (n = 25) [10,18–20,26–31,35,36,43,46,48–52,56–61], neonatal resuscitation (n = 22) [10,18,20,26–30,32–37,39,43,46,48,49,54,56–58,60] and pre-eclampsia (n = 13) [10,19,26–28,30–32,39,45,46,54,56]. Programmes typically combined team-based simulation followed by structured debriefs. Most studies did not report the exact protocol that simulations followed, and few (n = 5) [26,37,47,51,58] specifically reported the delivery of concurrent maternal-neonatal emergencies requiring simultaneous resuscitation. In many of the programmes where obstetric and neonatal care were integrated into a single curriculum, training was frequently delivered through sequential, same-day skills stations or linked but separate scenarios rather than a unified simultaneous drill. This modular approach ensures that neonatal resuscitation education is present within the curriculum without the operational complexity of a full-scale dual resuscitation.

Programmes spanned 18 countries (see Figure 2), with the highest prevalence in the USA (n = 8), India (n = 6) and Australia (n = 5). Other regions included Asia (n = 9), Europe (n = 5), Oceania (n = 5), North America (n = 13) and Africa (n = 8) [10,15,18–20,26–60].

The earliest implementation of IP SBT for neonatal and obstetric emergencies was identified in 2005, in the United Kingdom [28].

Setting and modality varied across the 40 studies. In-situ simulations conducted in real patient care environments were the most frequent explicit setting (n = 10) [10,28,30,40,41,44,47,55–57], followed by online delivery (n = 4) [35,42,49,50], while specialized simulation centres (n = 1) [43] and community-based home settings (n = 1) [48] were also utilized. The remaining studies (n = 24) [15,18–20,26,27,29,31–34,36–39,41,46,51–53,55,58–60] were described more broadly as clinical workshops or training courses conducted within hospital or educational facilities. Four studies utilized online delivery of workshops [35,42,49,50]. While simulation modalities were often not explicitly stated, six studies [10,18,45,55–57] utilized low-technology materials whilst others such as Meeker and Pak employed high-technology simulation [43,47].

Participants spanned a broad range of health professionals at various training levels involved directly in childbirth, including obstetricians, neonatologists, paediatricians, anaesthesiologists, nurses and midwives. Specialities that participated in IP SBT but are not commonly involved in childbirth included general practitioners, internal medicine specialists and allied health staff (including paramedics and respiratory therapists).

Ten programmes involved non-clinical staff, such as blood bank personnel and operating room staff, with one study including army corpsmen [20,28,33,37,38,40,41,43,47,48].

Across the studies, doctors participated most frequently (n = 26), followed by nurses (n = 20) and midwives (n = 20). Participant experience ranged from students to senior consultants, and facilitator experience varied from highly experienced consultants (n = 18) to peer trainers (n = 22) [10,15,18–20,26–60].

In African studies where the ETATMBA programme was implemented, NPCs or associate clinicians (ACs) were included. These are trained practitioners who provide medical and nursing care in regional and rural centres across Africa [15,31]. As defined in the Methods, these participants were included to reflect the integrated interprofessional response required in these clinical contexts.

The outcomes of the 40 included studies were categorized using the Kirkpatrick hierarchy of educational evaluation [21]. Evaluation methods varied significantly across the four levels of the framework, using both qualitative and quantitative approaches. Qualitative data were collected through interviews (n = 5) [27,31,38,54,60], focus groups (n = 3) [28,41,53], self-reflections (n = 3) [42,44,51] and surveys (n = 6) [26,29,35,36,48,49]. Participant reactions (Level 1) were explicitly captured in 17 studies through these methods [18,26–31,34–36,39,42,46,48,49,53,60]. Learning and clinical knowledge (Level 2) were assessed in 23 studies using quantitative pre- and post-intervention tests [10,18–20,31,36,39,40,44–46,49,52,54,56,59], while at least ten studies utilized objective structured clinical examination (OSCE) rubrics or validated skill performance checklists to evaluate technical performance [5,10,32,34,43,45,47,54,57,58]. Behavioural changes (Level 3), specifically involving teamwork and communication, were reported in 15 studies [10,27–30,33,35,38,41,43,44,54,57,58]. Finally, Organizational results (Level 4) were evaluated in 13 studies, including organizational outcomes (n = 7) [20,28,40,42,47,54] and patient-level outcomes (n = 6) [15,32,40,45,55,57]. For the purposes of this review, organizational outcomes were defined as system-level changes or insights identified through simulation, such as the detection of latent safety threats (LSTs), the development of new clinical protocols or improvements in equipment and resource logistics.

Across 17 studies assessing participant reactions [18,26–31,34–36,39,42,46,48,49,53,60], over 90% of participants rated the interprofessional simulation-based training (IP SBT) as relevant and useful to their clinical practice. Specifically, Adams reported that 92.9% of residents and nursing students found the training helpful, while Van der Scheer found that over 95% of participants agreed the training was relevant for improving clinical outcomes [26,53]. Regarding the simulation design, Afulani reported unanimous agreement (100%) among providers that the scenarios utilized were realistic [27].

The simulation environment was also found to foster psychological safety. Kumar and Wu noted that participants described the environment as ‘safe’, ‘supportive’ and non-judgemental, which allowed them to practise clinical skills without the pressures typically associated with real-time patient care [39,58]. Furthermore, Kumar found that mimicking clinical scenarios in a controlled setting allowed participants to strengthen their skills without feeling rushed [38].

Interprofessional learning was identified as a primary strength. Midwives in the Pauley study highlighted the value of training alongside paramedics to better understand community-based emergency management, with one participant stating it was ‘good to prompt me to think about managing emergencies in community’ [48]. Similarly, medical and nursing students in the Khot study reported that the interprofessional aspect was ‘the best part of the workshop’ [35].

Regarding online simulation delivery, while Malhotra reported a generally positive reception during the COVID-19 pandemic, participants noted that this format was less conducive to hands-on skill development [42]. Additionally, Prasad found that some students perceived participating via platforms like Zoom as ‘awkward’, which occasionally hampered effective communication [49].

This level explored the objective acquisition of clinical competencies. Regarding quantitative knowledge gains, significant improvements were noted across maternal and neonatal domains in 23 studies [10,10,18–20,31,36,39,40,44–46,49,52,54,56,59]. For example, knowledge scores rose from 51% to 72.6% in the Miller study, with the largest gains seen in pre-eclampsia and neonatal resuscitation topics [45]. Rosenberg reported median knowledge score improvements from 60% to 92% and 52% to 96% across two course iterations [20].

Technical skill performance was assessed in ten studies using objective scoring [5,10,32,33,43,45,47,54,57,58]. Gomez demonstrated OSCE score increases of 40–44% immediately post-training, with significant retention maintained at one year [32]. Varghese showed significantly higher skill levels in intervention groups compared to controls (55% vs. 46% for obstetric skills; 58% vs. 48% for newborn skills) [54].

A widespread increase in confidence was also reported. Li reported significant overall increases in confidence (p < 0.01) [19], while remote health workers in the Kildea study described a decreased fear of childbirth emergencies [36]. Meeker found that 81% of participants perceived improved clinical performance and 70% reported practice changes after high-technology simulations [43].

Teamwork and leadership gains were explicitly reported in 15 studies [10,27–30,33,35,38,41,43,44,54,57,58], including a clearer understanding of interdisciplinary roles and improved situational awareness. Madden established that a single leader is often insufficient for dual-patient emergencies and that serial drills improved role allocation [41]. In Varghese, a PPH drill prompted collaborative tray preparation, with a participant noting that teamwork ensured others would remind them if they forgot a step [54].

Communication improvements were reported in 14 studies [27,33,37,39–41,43–45,47–49,53,58], focusing on the value of ‘thinking out loud’ [27] and the adoption of standardized tools like SBAR and closed-loop communication [44,47]. Miller found that simulations allowed teams to identify and address ‘teamwork failures’, including a lack of protocol adherence [44]. Further, Meeker found significant improvement in participant communication (p = 0.031) following high-technology maternity simulations [43].

However, interprofessional barriers persisted. Walker and Prasad reported that pre-existing professional hierarchies and interdisciplinary tensions occasionally limited open communication or caused ‘frustrating’ interactions between medical and midwifery students [10,49].

Seven studies reported on organizational outcomes [20,28,40,42,47,54]. Lutgendorf and Pak identified critical LSTs, such as gaps in medication access (e.g. tranexamic acid) and blood product procedures, leading to immediate system corrections before adverse events occurred [40,47]. The PRONTO programmes in Mexico and Guatemala also demonstrated the achievement of facility-specific quality improvement goals [10,56].

Patient outcomes were explicitly reported in six studies [15,32,40,45,55,57]. Walker reported a 21% reduction in Caesarean section rates over a 12-month period (p = 0.005) [55]. Gomez and Miller found significant reductions in newborn mortality and intrapartum stillbirth (up to 34% reduction in some groups) following simulation-based interventions [32,45]. Finally, Lutgendorf observed a decreasing trend in PPH cases following the implementation of multidisciplinary drills [40].

The risk of bias assessment was performed using ROBINS-1 (for 16 quantitative studies) and the JBI checklist (for 24 qualitative studies) (see Table 3). Overall, nine studies were rated moderate risk and six rated as serious risk, mainly in the measurement of outcomes (Figure 3, Risk of Bias Results). Using the GRADE framework, the overall certainty of evidence was assessed to be moderate.

While established models like the PRONTO trial [10] have improved neonatal outcomes, they largely utilize modular or sequential scenarios, much like the Helping Babies Breathe curriculum, which focuses on neonatal resuscitation in isolation or like the Merién et al., which shows established multidisciplinary elements across the perinatal continuum, but lacking of a standardized approach to simultaneous resuscitation [11,12]. In many of the programmes where obstetric and neonatal care were integrated into a single curriculum, training was frequently delivered through sequential, same-day skills stations or linked but separate scenarios rather than a simultaneous drill. This modular approach ensures that neonatal resuscitation education is present within the curriculum without the operational complexity of a full-scale dual resuscitation.

The lack of simultaneous simulations likely reflects significant practical issues and resource implications. Running a dual-patient crisis requires a larger facilitator team, specialized equipment for both maternal and neonatal simulators, and simultaneous availability of staff from diverse backgrounds who must be released from clinical duties. Logistical issues in acquiring the size of the facilitator team and physical space required for a double resuscitation likely contribute to the preference for sequential models.

Furthermore, the feasibility of large team simulation requirements may limit transferability to smaller centres. Resource-intensive, high-technology-demanding simulations may work in large hospitals but do not reflect the reality of resource-limited centres. Future training models should consider how to scale down these integrated principles appropriately while ensuring the interprofessional training is relevant for practitioners who must manage these emergencies with a limited number of staff, such as in small rural maternity units or home-birth settings.

Our findings regarding teamwork and communication align with broader literature showing consistent improvements in team behaviours. Kiessling reported sustained improvements in confidence and team function after SBT in high-acuity environments (p < 0.0001), while Sezgin and Bektas found that IP SBT enhanced communication skills among healthcare students (p < 0.001) [61,62]. This is further supported by Weaver et al., which demonstrates that team training reduces errors in obstetric emergencies [63]. However, as Salas et al. note, there is significant uncertainty around how these behaviours translate to real clinical outcomes due to disparate effect sizes [64].

This review highlights that technical training alone may be insufficient due to persistent professional hierarchies. For instance, Prasad identified ‘interprofessional tensions’ and ‘slandering’ between medical and midwifery students, which hampered effective collaboration [49]. This is a well-documented phenomenon, as identified in Hall – hierarchy as a persistent barrier that disrupts communication in acute care. O’Daniel and Rosenstein directly link these poor dynamics to adverse outcomes [65,66]. To overcome these barriers, simulation must evolve to explicitly address psychological safety, a concept introduced by Edmondson as essential for team performance [67]. In this review, participants in the Kumar and Wu studies specifically valued the ‘safe’, ‘supportive’ and non-judgemental environment of simulation as a key factor in their learning [39,58].

While patient-level outcomes remain infrequently reported, the significant reductions in newborn mortality (up to 34%) and Caesarean section rates (21%) provide evidence of clinical translation [45,55]. These results are consistent with the PRONTO outcomes and low-dose, high-frequency training, which both show improved practices and reduced mortality [32,55]. However, a cautionary approach is required. Fransen et al. found limited high-quality evidence linking simulation directly to mortality outcomes, and Masoomi et al. emphasize that translation into practice is affected by multiple factors and is not an automated process [68,69].

Finally, a unique organizational benefit identified in this review is the role of simulation as a safety audit. As shown by Lutgendorf and Pak, simulation allows teams to identify LSTs – such as gaps in medication access or blood product procedures – before they result in adverse events [40,47]. This ‘dual role’ is strongly supported by Patterson et al., who demonstrated LST identification via in-situ simulation to enhance readiness for future emergencies [70]. Furthermore, Miller et al. report that simulation is a robust tool for uncovering system-level hazards [71]. Integrating these system-level improvements with practitioner training will represent an effective path forward for IP SBT.

Several limitations regarding both the review process and the included primary studies must be acknowledged. First, our search was restricted to full-text articles available in English and grey literature (including conference abstracts) was excluded, which may have introduced language and publication biases. While Medical Subject Headings (MeSH) were employed to increase sensitivity, there is a risk regarding term sensitivity; simulations that utilized multidisciplinary teams but did not explicitly use labels such as ‘teamwork,’ ‘interprofessional’ or ‘multidisciplinary’ may have been inadvertently omitted, reflecting the challenge of capturing inconsistently labelled interdisciplinary literature. Notably, an updated search conducted in March 2026 identified no additional studies meeting the inclusion criteria, confirming the stability of the current dataset.

Regarding synthesis, a meta-analysis was not feasible due to the high degree of heterogeneity in study designs, simulation programmes and outcome measures. The use of a narrative synthesis approach was further challenged by the variability of qualitative assessment tools, such as Likert scales, which were not universally applied.

The primary studies themselves exhibited significant variability in fidelity, structure, modality and facilitator quality. Facilitator expertise ranged from highly experienced clinicians to peer practice coordinators. A critical gap in the existing literature is the lack of focus on simultaneous resuscitation; only five studies explicitly reported simulating maternal and neonatal emergencies concurrently. No studies directly compared simultaneous resuscitation with separate, focused scenarios, limiting conclusions about the optimal sequencing of delivery room simulations. Additionally, data regarding the cost-effectiveness and feasibility of these programmes remain sparse.

Methodological biases within the included studies also influenced the certainty of evidence. Many studies used small sample sizes, featured unequal representation of disciplines or suffered from attrition at follow-up. Reporting bias was a concern, as participants may have provided ‘socially acceptable’ feedback rather than objective assessments of performance. Furthermore, the short duration of most programmes and a lack of long-term follow-up (frequently less than 6 months) raise concerns regarding skill retention. While some evidence suggested successful translation into clinical practice, many studies reported uncertainty regarding real-world sustainability and patient-level outcomes, highlighting the need for more longitudinal, robust evaluation.