Table 2 Views synthesis driving the construction of model 1–3 and its truth tables

Model 1 – Implementing training to improve providers knowledge and clinical skills

The existing qualitative evidence synthesis and intervention component analysis indicated four different themes relating to training [18, 28, 34, 44, 86]. Firstly, healthcare providers are often reluctant to implement new CS programs or to implement overall change due to perceived insufficiency of skills and knowledge on labour and vaginal birth management, especially the younger generation [28]. Secondly, healthcare providers and other stakeholders (i.e. policy makers, hospital managers) emphasised the importance of implementing various training or education for healthcare providers [18, 28, 86]. This training includes clinical skills training in labour and vaginal birth, recommendations in practice, clinical audit and program content itself [18, 28, 86]. Thirdly, both providers [28, 86] and trialists [34, 44] mentioned that the underlying factor of success lies on providers’ beliefs about CS and vaginal birth, as well as whether providers are willing to step out of their comfort zone to. Lastly, providers mentioned that they preferred the intervention to be reflective in nature, instead of dictatorial and enforcing [28]

Model 2 – The audit and feedback process

In relation to audit and feedback process, the existing qualitative evidence synthesis and intervention component analysis revealed three different themes [28, 86]. Firstly, the process of conducting audit and feedback was considered critical by healthcare providers, as the content, methods of delivery, and timing of audit and feedback influenced how they feel about the intervention overall [86]. Secondly, some providers were concerned that audit and feedback may pose a threat to their identities and careers [28]. Therefore, the more acceptable the structure of feedback is to the providers (i.e. feedback delivered individually instead in group), the better they respond to it, thus increasing its effectiveness [86]. Thirdly, findings by Kingdon et al. also revealed organisations which were able to reduce CS are often characterised by having healthcare providers who valued continuous quality improvements, such as clinical audits, second opinion, continuing education [28]

Model 3 – Working relationship and environment

In terms of working relationship and environment, existing qualitative evidence syntheses by Kingdon et al. revealed three themes [18, 28]. Firstly, multi-disciplinary collaboration between doctors, midwives, nurses and other maternity care providers was pointed as a key element in optimising CS [18, 28]. Multi-disciplinary collaboration has been observed as very poor in health facilities with high CS rates, and actively present in health facilities with low CS rates [18, 28]. Secondly, healthcare providers reported about the unequal and hierarchical power relations when caring for women. Working relationships, collaboration and communication may also be diminished through hierarchy-driven fear, which may be present when for example midwives are considered to have fewer skills than doctors [18, 28]. Thirdly, Kingdon et al. also emphasized the effectiveness of interventions to reduce unnecessary CS is strongly mediated by stakeholder commitment and organizational buy-in or, systems and policy changes that facilitate vaginal birth [18]