Factors influencing normalisation and sustainment of the Birmingham Symptom-specific Obstetric Triage System (BSOTS): a qualitative implementation evaluation study with UK maternity health professionals

Abstract

Background

Maternity triage is the emergency portal to access maternity services for pregnant or newly postnatal women experiencing unexpected complications or concerns. Traditionally women were seen in the order in which they attended maternity services without appropriate organisational and clinical systems in place to prioritise the clinical urgency of the women presenting. The Birmingham Symptom-specific Obstetric Triage System (BSOTS) is a standardised triage system co-developed by clinicians and researchers to address this need. Since its inception in 2013, BSOTS is now the recommended triage system and has been widely implemented in the UK with international adoption also occurring. This study aimed to explore the factors influencing the normalisation and sustainment of BSOTS in UK maternity units.

Methods

A qualitative evaluation study involving semi-structured interviews with 43 maternity professionals across 9 sites representing a range of implementation experiences (normalised n=3, partially normalised n=3, and probably not normalised n= 3). Participants were maternity triage staff of varying levels including midwives, obstetricians and senior leaders. Interviews were held virtually, and recorded and transcribed verbatim. Analysis was based on Normalisation Process Theory.

Results

Contextual factors were key to influencing implementation and sustainment of BSOTS. In sites where BSOTS was normalised and integrated into routine practice, organisational and leadership buy-in helped mobilise support for other contextual factors required for implementation fidelity and success, including establishing a clear identity of the triage department, sufficient midwifery and obstetric staffing, appropriate space, a dedicated and protected core team, adequate equipment and resources. Accommodating these factors often meant complex system-level changes were required to implement BSOTS for which strategic intentions and organisational support were integral. In some cases, such support was facilitated by Care Quality Commission regulatory recommendations.

Conclusions

Our study found successful implementation and normalisation of BSOTS was driven by contextual factors, particularly organisational buy-in and leadership support. While regulatory requirements enabled the prioritisation of triage within maternity services, our evaluation emphasised the importance of both leadership and frontline staff support for effective integration and sustainment. Evaluation findings were used to inform an implementation toolkit for clinical triage teams seeking to implement or reinvigorate BSOTS in their sites.

Contribution to the literature

• Implementation research involving maternity service interventions is often criticised for lacking a theoretical basis.

• This is the first in-depth qualitative evaluation study exploring the factors influencing normalisation and sustainment of the recommended and widely adopted Birmingham Symptom-specific Obstetric Triage System (BSOTS) using Normalisation Process Theory (NPT).

• Although buy-in was required at all staff levels, organisational and leadership support were essential to enabling the necessary capacity, resource and environmental conditions for successful implementation.

• Future research should draw upon implementation science frameworks to inform strategies to enhance organisational and leadership support in maternal and neonatal health interventions.

Introduction

Triage systems are designed to ensure patients receive the level and quality of care appropriate to their clinical needs. In the past, there was no such standardised triage system available in maternity care to identify, prioritise and treat women attending maternity services with unscheduled pregnancy related complications or concerns. The generic parameters of standard triage tools could not be extrapolated to pregnancy as the physiological changes of pregnancy are associated with a higher resting heart rate, lower blood pressure and increased respiratory rate [1, 2]. There was also no means of assessing the condition of the unborn baby using existing emergency medicine triage tools. This, together with the underlying good health of the maternity population, may mask the severity of maternal illness, unless a specific assessment is undertaken by an appropriately trained healthcare professional. The need for a nationally standardised maternity triage system has been highlighted in several UK confidential enquiries as well as in many local Care Quality Commission (CQC) and Health Services Safety Investigations Body (HSSIB) reviews ([3,4,5,6].

The Birmingham Symptom-specific Obstetric Triage System (BSOTS)

BSOTS is a triage system designed for maternity. It was co-produced by obstetricians and midwives at one English maternity unit in 2013 and consists of a standardised brief clinical assessment (triage) by a midwife within 15 min of a woman’s attendance [7]. Clinical priority is determined using standardised symptom-specific algorithms for the eight most common reasons for attendance. Based on the algorithms, the level of urgency is indicated as immediate/urgent care (categorised as red or orange), or less urgent care (categorised as yellow or green). An initial evaluation showed an increase in the numbers of women assessed within 15 min of attendance, and women identified as having greater urgency were seen more quickly by a doctor. The system had excellent inter-operator reliability and was reported to improve the management of the department and increased patient safety [7]. BSOTS has now been adopted in over 100 maternity units across the UK and 4 sites internationally. In 2023, BSOTS became the recommended triage system for maternity in England [8].

Aims and objectives of this evaluation

Given the wide scale adoption of BSOTS across maternity settings, an in-depth implementation evaluation is timely and integral to facilitate deeper understanding of how and why BSOTS works in practice. Through our evaluation, we aim to not only analyse implementation challenges and solutions, but also the contextual factors associated with sustaining implementation of a widely adopted evidence-based intervention. Context is defined as a set of characteristics and circumstances that consist of active and unique factors that surround the implementation effort [9]. As such it is not a backdrop for implementation but interacts, influences, modifies and facilitates or constrains the intervention and the implementation effort [9].

Given the importance of context for successful implementation, an intervention aimed to improve delivery of care is likely to be adapted as it is adopted thus challenging the fidelity of the original research-based intervention. We therefore need to pay particular attention to how an intervention aligns with, and is normalised within, frontline clinical practice. This is particularly pertinent when we seek to sustain the intervention once resource and attention dissipates following the pilot stage [10]. A relevant framework to address such a research challenge is Normalisation Process Theory (NPT) [11]. This represents a widely used theoretical framework to understand how new interventions are implemented and integrated into routine practice. It provides a structured approach to examine how different individuals, teams, and organisations interact with new practices and how these practices become routinely normalised or embedded in everyday work [11]. Often used interchangeably, the terms normalisation and embeddedness refer to the work actors do to engage with new or changed ways of working and how they become routinely integrated within existing structures and practices [12]. According to NPT, normalisation or embeddedness are dependent on how participants make sense of a new intervention (coherence), commit to it (cognitive participation), undertake work to establish the intervention (collective action), and evaluate the costs and benefits (reflexive monitoring) [11]. The model was expanded in 2022 to include further operationalisation of contextual factors (strategic intentions, adaptive execution, negotiating capacity, reframing organisational logics) alongside the core constructs [13]. The use of NPT facilitated the analysis of implementation challenges and solutions associated with sustaining the implementation of a widely adopted evidence-based intervention.

Methods

Design and setting

This was a qualitative study conducted as part of a wider mixed methods evaluation study. Semi-structured interviews were undertaken with frontline staff and senior leaders including midwives and obstetricians from selected NHS maternity sites implementing BSOTS in the UK.

Sampling and recruitment

Sampling was based on an earlier survey with triage leads from all NHS maternity sites implementing BSOTS in autumn 2022 (n = 75 sites). The survey aimed to assess the implementation and normalisation of BSOTS from the perspective of triage leads who were senior midwives responsible for the day-to-day management of the triage department. Sixty-six sites completed the survey yielding an 88% response rate, of which 60 sites consented to be contacted for this evaluation and 6 declined. Site leads self-reported how normalised BSOTS was in their units using a question from the validated NPT Normalisation MeAsure Development questionnaire [14] (NoMAD): “Do you feel BSOTS is currently a normal part of your practice?”, with the options Completely, Somewhat, or Not at all. Responses ranging from completely normalised (n = 37), partially normalised (n = 18), and probably not normalised (n = 5). Nine sites were purposively sampled for in-depth qualitative evaluation to include a range of sites of varying sizes, assessed via their birth rate, regions, time since initiation of BSOTS, and degree of self-reported normalisation of BSOTS. Further details comparing the participating sites and non-participating sites can be found in Supplementary Information 1.

Midwife Triage Leads from those sites that had consented to be contacted were invited to be interviewed in the first instance. They were then asked to circulate the study information to relevant colleagues including obstetricians and midwives working clinically in triage, or those responsible for managing the department including Obstetric Triage Leads, Matrons and Heads/Directors of Midwifery. In a few cases, additional participants were identified by interviewees. Sampling aimed to ensure a diverse range of staff roles and levels per site where possible.

Data collection

Virtual semi-structured video interviews were conducted between March—July 2023. Willing participants provided verbal consent to participate. The topic guide was based on NPT (Supplementary Information 2) and explored the implementation, adoption, and both practical and contextual challenges to the sustainment of BSOTS. Participants were offered coffee vouchers as a token of thanks. Interviews were conducted by two experienced female applied health service researchers [AL and ND], digitally recorded, transcribed verbatim and anonymised.

Data analysis

Interview transcripts were uploaded to NVivo (QSR International Pty Ltd) for management. A two- stage approach to analysis was applied following familiarisation with the data. In the first stage transcripts were independently and inductively coded using thematic analysis by three researchers (ND, AL, SD). Codes were reviewed and discussed by the researchers throughout the coding process, where codes were consolidated or clarified where necessary. As coding of the data began whilst data collection was still underway, there was flexibility to sample further sites and participants, however this was not required as data saturation was reached. Once the initial coding was complete, the codes were deductively mapped to NPT constructs in the second stage of analysis. We used the updated NPT coding manual featuring 22 constructs that had been mapped onto the context-mechanisms-outcome configuration for realist evaluation by May et al. [13]. Using the NPT framework enabled us to identify the key contextual factors influencing the implementation and normalisation of BSOTS, and their interplay with each other.

Preliminary validation was sought by sharing initial findings with members of the wider maternity research team, the creators of BSOTS (SK and NJ), and two midwives who were employed to support implementation and had conducted visits to all sites implementing BSOTS in 2022–2023. The final results were also shared at a national BSOTS dissemination event, which was attended by representatives from 64 sites who had implemented BSOTS, as well as Patient and Public Contributors.

Patient and public involvement and engagement

Patient and public contributors from the NIHR ARC West Midlands Maternity Theme were involved in the evaluation design and kept updated on progress throughout the project. The findings were shared with them for comments as they developed.

Results

Forty-three participants were interviewed across 9 NHS provider sites identified from an earlier implementation survey with triage leads to include sites of a variety of characteristics as shown in Table 1. Sample sizes ranged from 2–8 participants per site. Table 2 shows the interview participants’ professional roles.

Table 1 Key characteristics of participating sites (presented in random order to protect anonymity of sites)

Table 2 Professional roles of participants

Implementation, adoption and fidelity of BSOTS

All sites had been implementing BSOTS prior to the evaluation with implementation start dates ranging from 2013 to 2021. Most participants reported encountering some degree of challenge with implementing BSOTS, including those who had reported BSOTS was normalised in their sites in the earlier survey, and those who had been implementing BSOTS for several years. There were many examples of implementation being more cyclical through initial launches and relaunches of BSOTS as triage teams, leaders and resources changed and fidelity to the recommended implementation model wavered as a result. The following sections provide an overview of the range of factors influencing the implementation of BSOTS through the lens of NPT where theoretical constructs are italicised within the text.

“I think it’s gone through cycles…it embedded itself quite well, and then it seemed to lose its way a bit, because lots of other concerns that were separate from BSOTS as such began to take over for it. There were lots of issues to do with the staffing levels, about the staff dedicated… not being dedicated for triage, and then being called to cover delivery suite… So, it was almost like we had to have another relaunch for saying let’s get back to basics again” (Obstetric Triage Lead, site 2).

Coherence – how do teams make sense of BSOTS?

Understanding of the core elements of BSOTS

Key aspects of implementation fidelity of BSOTS such as women being triaged within 15 min of arrival and allocating a category of urgency (Red–Orange-Yellow-Green) to women were broadly well implemented. However, there were examples of women who were allocated a category of urgency not matching their actual clinical needs, resulting in them being over or under prioritised. This could suggest a lack of coherence in midwives’ understanding and interpretation of the algorithms as well as timings for subsequent review. There also a tension in midwives needing to balance using the BSOTS algorithms with their own clinical judgement, particularly when multiple women had been allocated the same category of urgency, but their presenting complaints varied as demonstrated in the quotation below.

“If I’m on triage I might have somebody that I’m more worried about, I might feel like from a triage point of view they’re all oranges, but I’d prefer you to see this one first. You’re still using your clinical judgement.” (Triage midwife, site 1).

BSOTS also requires the roles of the midwife undertaking the initial triage to be distinct from the midwife providing subsequent ongoing care. This was a recurrent area of challenge for several sites as many midwives were not accustomed to this new way of working, and most preferred working within traditional models of providing continuity for the duration of a woman’s care episode. This again suggested a lack of coherence and sense-making of a key aspect of implementing BSOTS by midwives which further impacted their cognitive participation and engagement with the intervention.

“The whole separation of duties into triage midwife and ongoing care midwife, is not always as separate as the system is designed to be, and I think some of it has to do with the individual midwives, and how they like to work, some of them don’t particularly like that style of work, some of it just has to do with the volume of work” (Obstetric Triage Lead, site 2).

Overall, BSOTS was reported to be a valuable addition to practice as it provided a structure to prioritise and manage women using shared language resulting in robust decision-making, improved safety and efficient assessment of pressures on the department. This was in stark contrast to previous ways of working prior to BSOTS which were described as chaotic for staff with no systematic way of prioritisation, unmanageable workloads and long waiting times for women. In NPT terms, adoption of BSOTS was aided by the absence of a competing similar intervention as staff were able to recognise the benefits of BSOTS and differentiate it from previous ways of working.

“Actually I think although it was an awful place to be at that point before we relaunched BSOTS, it actually gave us springboard to say this is going to make a massive difference, and this is why we need to do it. It was almost like it was welcomed.” (Triage lead midwife, Site 4).

Several sites had also developed bespoke information for women to ensure they understood how they would be prioritised based on clinical need which further helped strengthen implementation and mitigate potential complaints.

“I think that you need good communication to the women, with, about how triage works. Because I think that…triage works best also if the women understand what’s going on. Because I think otherwise they may feel that they’re being dismissed or fobbed off when actually no, they’re just being prioritised on clinical need.” (Obstetric Lead, site 1)

Cognitive participation and collective action – how do teams work together to engage with, and implement BSOTS?

Dedicated area leads

Many of the strategies driving successful implementation of BSOTS focused on enhancing cognitive participation and engagement of triage staff. These included having a dedicated midwifery and obstetric triage lead where the former provided the day-to-day leadership of the triage department, supported staff with addressing challenges, reminded them of the principles of BSOTS and role-modelled desired practices. According to NPT, this dedicated, multi-disciplinary on the ground leadership presence aided the initiation of BSOTS through legitimation and reinforcement of the importance of BSOTS to staff.

“You need leads, you need a multi-professional group of people to lead, so that the medical staff and the midwifery leadership understand that it’s a very, very important high risk area.” (Director of Midwifery, site 9).

Multi-disciplinary team (MDT) buy-in and engagement

MDT engagement and buy-in from both midwives and obstetricians was reported as essential for long term adoption and sustainment. In sites where BSOTS was normalised, it was included in mandatory training for the whole maternity team to ensure collective understanding (communal specification) of the requirements of BSOTS and shared language, particularly for rotational or temporary staff who may not frequently work in triage. There was increasing acknowledgement of the level of experience and confidence required to work in triage and so several sites had started to upskill and familiarise new and junior staff with triage. This ensured the skill-set workability of BSOTS as more MDT staff were trained and prepared to work in triage using BSOTS.

"If you’re working in maternity you should be able to understand how the whole unit works. If I have to go to labour ward I should know how the Syntocinon protocol works, and vice versa. I think as well it’s important because if you handover to a midwife on the labour ward and you’re like, “She is an orange,” they look at you and like, “What?” (Triage midwife, site 5).

However, the lack of communal specification and low awareness of BSOTS by obstetricians was a significant barrier in sites where BSOTS was either partially or not at all normalised.

“I think we need more consultant to medical team buy-in, and their buy-in on the system is not good, and their awareness, and I think that is a big challenge and barrier to it, because they’re trying to work in a different way, they’re not aware of it, they don’t know it, and I think if they’re on-board with it, it becomes even more robust.” (ACP, Site 2).

With regards to midwives, a recurrent challenge was changing habits around usual working practices and staff resistance to new ways of working that accompanied BSOTS such as the separation of roles of the triage and ongoing care midwives. This was particularly exacerbated when senior midwives directly influenced junior midwives to not use BSOTS, signalling differences in communal and individual (coherence) understanding impacted staff’s engagement and willingness to implement BSOTS.

“So there’s definitely been a shift with that, but yeah that was a challenge, especially when you’re… at the time I was a band five midwife, I’m not going to stand there and argue with a band seven midwife, that hierarchy was really not working in my favour.” (Triage midwife, site 4).

However, in sites where triage leads and managers continued to reinforce the importance and fidelity principles of implementing BSOTS, the initial resistance from staff appeared to dissipate after a period of implementing BSOTS as intended and realising its impact and benefits.

“We just would be like, “I’ll see that woman, you see the next woman,” and do it like that. I was very stuck in my ways and I was like, “No you surely can’t work any other way, how are we supposed to…” just not liking change, we’ve done it for three years or so. But they really, really pushed for us to try and do it properly, and in the last month I’ve got on-board, and I really like it.” (Triage midwife, Site 7)

Other activities promoting cognitive participation and collective action of staff included regular team meetings to feedback on key fidelity markers such as ‘the number of women triaged within 15 min of arrival’, listening and involving frontline staff in implementation – particularly during relaunches of BSOTS to understand previous implementation challenges, and nominated ‘champions’ to support the ongoing roll-out and embedding of BSOTS.

“As I say I think in BSOTS champions, so our core girls making them champions of BSOTS. So they can escalate that information, and also I think if you give them something like that as well that they care about and are passionate about you are going to see results, because they really want it to work.” (Triage midwife, site 9).

Reflexive monitoring – how is BSOTS appraised and assessed?

Continuous monitoring and feedback

Sites where audit data on key implementation markers were routinely collected, fed-back to staff and senior leaders appeared to gain the most momentum in terms of implementing and sustaining BSOTS. This was particularly evident in sites where BSOTS had been integrated into the electronic patient record system allowing systematisation and communal appraisal. There were many examples of triage leads and matrons using audit data to continuously adapt and improve implementation (reconfiguration) across areas such as midwifery and obstetric staffing, space and staff training.

“Also it [audit] enabled us to have a look at staff that maybe might be struggling with certain things, or if there’s a theme that was going on that we were missing things. It gave us that overview of where we needed to focus our attention on, because it is definitely something that just keeps developing. Even now we’re still looking and driving it forward. It’s not something oh that’s in we can just sit back.” (Triage lead midwife, site 4).

Examples of positive reconfigurations to strengthen the implementation of BSOTS as a result of audit and feedback included clearly distinguishing the roles of the triage and ongoing care midwife through badges, adding computers to the triage rooms and redesigning the whiteboard to give a better overview of the department.

Impact of contextual factors on implementation

Many of the implementation challenges highlighted by participants reflected contextual and environmental challenges such as organisational buy-in, leadership support, multiple services being delivered within the triage area, midwifery and obstetric staffing shortages, insufficient space, IT challenges, and the triage telephone being in the clinical area, all hindering the implementation and fidelity of BSOTS. The following section will further examine the impact of the key contextual factors on the implementation and sustainment of BSOTS. Table 3 provides an overview of how contextual factors, mapped onto NPT, facilitated the implementation of BSOTS.

Table 3 Contextual factors for successful implementation and sustainment

Organisational buy-in and leadership support

In most cases, strategic intentions to implement BSOTS or modify implementation to recommended standards (RCOG 2023) were underpinned by external regulatory inspection from the Care Quality Commission (CQC). Participants at all levels reported being aware of implementation challenges such as staff being redeployed from triage to other service areas, however felt unable to influence change in the top management level until raised by the CQC. In many cases CQC recommendations expedited organisational support resulting in instantaneous improvements to triage e.g. moving the triage telephone away from the clinical area and proving additional staff.

“We’d been crying out for a long time, a long time to say it’s just not safe, we can’t leave one person down there, it’s just… you would just leave a shift just demoralised, because it was just so exhausting. But yeah no definitely when the CQC came and just went. “This is not… this cannot be,” and everyone went oh right okay.” (Triage midwife, site 7).

Aside from external scrutiny provided by CQC, for some sites an increasing number of serious incidents relating to triage served as internal triggers to delve deeper into the root causes and inform strategic intentions for improvement. Whilst as an intervention, the components of BSOTS are discrete and well-defined, implementation is a complex process requiring system-level changes where organisational buy-in and support is integral to the normalisation and sustainment of BSOTS. This was echoed by senior leaders who highlighted the importance of organisational support for protecting and mobilising key implementation resources such as staffing and space (reframing organisational logics; negotiating capacity). This was partially achieved through increased visibility and direct reporting channels to the Trust board which strengthened buy-in for maternity triage from executive leaders and board members.

Furthermore, senior leaders’ understanding and presence in triage further enhanced the legitimation and contextual integration of BSOTS for frontline staff. Several senior leaders spoke about having direct conversations with triage staff to elicit solutions and gain their support and buy-in for BSOTS and its accompanying working practices. Moreover, the presence and support of middle-level managers such as matrons was pivotal for resolving day-to-day challenges and ensuring BSOTS continued to be implemented with fidelity, thus shaping implementation through reframing organisational logics.

In sites where a lack of leadership presence and organisational support were reported, implementation was challenging and led to BSOTS not being adopted, often requiring a subsequent re-launch. The quotation below reflects the importance of both senior and middle leadership level buy-in, particularly at implementation initiation, to endorse the legitimation of BSOTS as the right thing to do.

“It was pre-COVID… the manager that we had at that point in time was off on long term sick, our deputy manager was off on sick I think, and our matron was on annual leave, so it was a really bad time to roll BSOTS out. There was no management in the sense I really didn’t feel like it was a good time to roll it out, we needed to either delay it a little bit…” (Triage lead midwife, site 1).

Clear identity of the triage department

Successful implementation was evident in sites where triage had been reframed as the emergency portal for maternity, recognised by senior leaders as a separate area with staff protected from redeployment, a clear escalation policy, defined boundaries of service and moving non-triage activity away from the department (reframing organisational logics; adaptive execution; negotiating capacity). In contrast, in sites were BSOTS appeared to be challenging to normalise, multiple competing priorities were evident for staff as they were also required to provide non-triage services such as scan reviews, post-natal checks and women attending scheduled appointments. This added pressure and burden on the triage team was reported as a significant barrier to implementing BSOTS and often reflected how triage was perceived by the rest of the maternity service as an extension of other departments, also referred to as the “dumping ground.” Furthermore, recognition of triage as a separate area (reframing organisational logics) led to its inclusion in the obstetric ward round, which further legitimised the importance of triage and enabled contextual integration in sites where BSOTS was normalised.

Capacity and resources

Capacity and resource challenges in relation to midwifery and obstetric staff shortages was one of the most repeatedly reported challenges to implementing and sustaining BSOTS. In many settings, staffing shortages were prevalent across the whole maternity service leading to midwives being redeployed to support other areas thus depleting the triage team, consequently resulting in delays in triaging and providing ongoing care for women (negotiating capacity). The acute impact of staffing shortages was more pronounced in the context of national recruitment challenges to midwifery and obstetric training posts alongside periods of industrial action. Long delays for obstetric reviews were widely reported across all sites irrespective of how normalised BSOTS was.

Some units developed workarounds to help manage challenges, however not all workarounds supported the fidelity of BSOTS, which triage leads were aware of (adaptive execution). For example, in 3/9 sites maternity support workers undertook women’s observations as part of the initial triage due to staffing shortages. In other sites, women were redirected to other departments during busy periods where the volume of women exceeded capacity.

A key difference between sites where BSOTS was normalised and those where it was partially/not normalised was the location and handling of triage telephone calls. Sites where the triage telephone was located away from the clinical area and staffed separately had the most success in implementing and sustaining BSOTS. Conversely, in sites where the triage telephone was located in the clinical area, midwives were expected to attend to calls and see women which added further capacity challenges to an already stretched area.

Increasing capability and team stability

Increasing the capability of the triage team was integral to balancing challenges related to staffing shortages. In sites where BSOTS was normalised, a core team of trained, experienced and enthusiastic midwives was established to ensure the triage team adhered to the principles of BSOTS. This was particularly pertinent in areas where there was high staff turnover leading to temporary or rotational staff working in triage. In NPT terms, having a consistent team brought collective understanding and action to ensure BSOTS continued to be implemented with fidelity (communal specification; enrolment; activation; skill-set workability).

Triage environment

Having sufficient clinical space to match the volume of women attending triage, determined by the overall birth rate, was another integral factor to the successful implementation of BSOTS. In areas where the triage space was reconfigured to match capacity (negotiating capacity) organisational support and leadership buy-in to mobilise resources and funds to make the necessary changes was evident. Conversely, in sites who lacked dedicated triage space including separate triage and ongoing care rooms, and sufficient waiting space for women close to the clinical area, implementation was more challenging and correlated with a reported lack of organisational support.

Aside from physical space, another widely reported environmental challenge related to IT challenges in the triage department. For some this was a short-term issue whilst they migrated to a new electronic patient record, but for others it was a more long-standing challenge due to IT connectivity issues which impacted the time taken to triage women and document the tasks. Having adequate equipment and IT systems once again reflected how context impacted implementation and the importance of organisation support in addressing the practical challenges facing staff.

Discussion

This in-depth evaluation study explored the factors influencing how BSOTS, the first national maternity-specific triage system, could become normalised within UK maternity settings. Using NPT [13], we were able to identify key contextual factors influencing implementation and sustainment including organisational buy-in and leadership support, identity of triage, capacity and resources, capability and the triage environment. Implementation success and the fidelity of BSOTS varied both within and across sites with several examples of cyclical and repeated launches and re-launches, often due to contextual constraints such capacity, resources and the environment.

The role of contextual factors in the implementation of evidence-based interventions is well documented within the implementation science literature across various theoretical frameworks and research studies [9, 15, 16]. Organisation studies might direct us towards application of translation theory, which orientates towards a more sociological understanding of practice on the ground regarding implementation of evidence-based intervention, within which context is seen as an active part of translation [17]. In our study, we follow such lines, but apply NPT, which similarly focuses upon who is translating evidence-based intervention and the role of context in shaping this, but which has been derived from healthcare settings, and crucially which focuses as ‘it says on the tin’ upon normalisation (embedding) of the intervention so it sustains. Sustaining of interventions is crucial in healthcare settings, characterised by a plethora of innovation (“1000 flowers bloom”), but which often dissipates when pilot funding is removed [18].

Detailed examination of contextual factors can help explain the reasons behind implementation variation across settings [19]. We were able to understand and unpick the precise contextual factors allowing normalisation of BSOTS in some settings and failure to normalise in others. Our findings also go a step further by demonstrating that not all contextual factors have an equal impact: organisational buy-in and leadership support were overwhelmingly critical in determining the normalisation and sustainment of BSOTS, mediating the impact of all other contextual factors in Table 3. In a scoping review of contextual factors in implementation, Nilsen and Bernhardsson [15] found there were two broad dimensions of context within implementation theories and frameworks: those that functioned as necessary conditions for implementation and those that could be viewed as the active driving forces required to achieve successful implementation. From our findings, the first of these was key as organisational buy-in and leadership support functioned as the necessary conditions for the successful implementation of BSOTS and all other factors such as identity, capacity, space and resources as the active driving forces. In sites where BSOTS was normalised, organisational and leadership buy-in appeared to ‘unlock’ support for other contextual factors required for implementation fidelity and success including establishing a clear identity of the triage department (seeing only women attending for unscheduled pregnancy related concerns), sufficient midwifery and obstetric staffing, appropriate space, a dedicated and protected core team, adequate equipment and resources. In many cases, this meant complex system-level changes were required to embed BSOTS and balancing the implications for the whole maternity service for measures such as protecting triage staff from being redeployed to other areas and only providing care to women attending with unscheduled pregnancy related concerns. These findings were echoed in a Dutch study where multi-disciplinary buy-in at all levels facilitated the implementation and normalisation of a structured triage system within a protected obstetric triage department [20].

Research recognises the role of both organisational and leadership support as key enablers of implementation context [15, 21,22,23,24]. Leaders not only have a strategic role in mobilising funding and resources, but they also contribute to the implementation climate of an organisation [24] where, in NPT terms, there is collective understanding (coherence) and action to ensure an evidence-based intervention is implemented and normalised with fidelity. In our study, leadership presence and support for BSOTS also legitimised its initiation for staff and further enabled contextual integration. Furthermore, the role of the hybrid middle-level manager such as triage leads and matrons were also pivotal in driving implementation and acting as bridges between frontline staff and senior management [23].

Triage is an integral component of maternity services and often the first port of call for women experiencing unscheduled pregnancy related concerns. However, it was not recognised and prioritised as such in sites where BSOTS was not normalised, partly reflecting a broader context of national maternity staff shortages and budget constraints [25]. Lack of resources and staff shortages were also reported as barriers to implementation of BSOTS in Australia [26]. Researchers in Sweden similarly found that an organisational and individual mindset shift to one comparable with general Emergency Departments was required to reframe triage as the emergency portal for maternity services [27].

As well as understanding the factors contributing to the normalisation of BSOTS, our evaluation also demonstrated how the implementation fidelity of BSOTS was eroded when there was a lack of organisational buy-in and leadership support to provide the necessary resources (staffing, space) and absence of triage being recognised as the emergency portal for maternity. There were many examples in the data of sites implementing BSOTS with adaptive execution and enacting workarounds against fidelity to mitigate the impact of such challenges, resulting in a lack of normalisation and sustainment. Since its inception in 2013, the core elements of BSOTS are now well understood and defined with guidance available for best practice, although published after the data collection phase of this study [8].

A noteworthy finding of this study is the role of a national regulatory body, the Care Quality Commission, as part of their maternity inspection programme, facilitated strategic intentions for change resulting in immediate access the required resources to implement BSOTS with fidelity. It was not uncommon for triage staff and some senior leaders to report being aware when they were not implementing BSOTS with fidelity yet repeated calls for additional resources and capacity by frontline staff were not acted upon until flagged as an urgent safety priority by the CQC. Whilst frontline staff welcomed CQC recommendations, this top-down approach to requiring change may not apply to other interventions and settings. Lack of authority and power to change practice amongst frontline staff has been reported as an aspect of organisational culture hindering implementation of evidence-based practice [28]. Our study reveals the implementation (and sustaining) of evidence-based intervention as a journey, within which leadership influence is dynamic [10]. Future implementation research should explore strategies to enhance organisational support for maternity service interventions.

Strengths and limitations

Our systematic, comprehensive theory-based approach using the recently updated NPT coding framework [13] highlighted the role of key contextual factors and the dynamics between them, particularly how strategic intentions to implement BSOTS with fidelity were formed. Whilst we applied NPT retrospectively to evaluate BSOTS, using it prospectively during intervention development and piloting may have formalised both the necessary conditions (organisational buy-in and leadership support) and active driving forces (identity, capacity, space, resources) of successful implementation. Use of NPT as the underlying evaluation framework also addressed a common criticism of implementation research in maternity often lacking a theoretical basis to explaining and maximising the uptake of evidence-based interventions [29]. Our sample represented a diverse group of sites, regions, sizes and varying professional groups. Obtaining detailed accounts from multiple staff of varying roles and levels from the same unit allowed us to build a comprehensive understanding of the factors influencing the normalisation of BSOTS both within and across sites. Although recruitment of triage staff participants via site leads risked selection and social desirability biases, all participants spoke openly about the challenges of implementing BSOTS suggesting this did not influence the results. While ethnography and longitudinal analysis of implementation would have given the opportunity to see ‘work as imagined vs work as done’, findings were presented and validated with 64 triage teams from across the UK at a national BSOTS dissemination event, also attended by Patient and Public Contributors, adding further credibility and weight to the evaluation. The findings also resonated with midwives employed who visited sites using BSOTS to support and advise on implementation. Future research could explore the clinical impact, financial cost, and any unintended consequences of BSOTS.

Conclusion

In our study, successful normalisation and sustainment of BSOTS was facilitated and cemented by contextual factors, particularly organisational buy-in and leadership support, which mediated the mobilisation of conditions and resources required for implementation. Although strategic intentions to prioritise triage were determined by mandatory regulatory requirements by the CQC in some cases, our evaluation highlighted the importance of organisational support at every level from the seniors to the frontline in terms of legitimising and integrating BSOTS within their settings to be implemented with fidelity. Being underpinned by NPT, a comprehensive theoretical framework, allowed us to understand and explain how an evidence-based maternity triage system could become normalised in practice. Findings from this evaluation study have been used to inform an implementation toolkit for clinical triage teams wanting to implement or reinvigorate existing implementation [30].