Strengthening open disclosure in maternity services in the English NHS: the DISCERN realist evaluation study

Open disclosure: policy and research background

Since the mid-2000s, the term ‘open disclosure’ has been adopted internationally by policy-makers and researchers to capture the process of ongoing communication with patients and families after adverse events. 1,2 Open disclosure (OD) is defined by Manser (2011) as: ‘the open and timely communication about adverse events to keep patients and family members informed, to acknowledge their suffering and grief and help in reducing feelings of abandonment’. 3–5 OD communication involves providing accurate information about the critical incident, its immediate consequences and appropriate remedial action, as well as providing an expression of regret and information about what will be done to avoid harm reoccurrence. 5

Internationally, OD of adverse events to affected patients and families has been recognised as an important ethical and patient safety concern for several decades. 6–13 Additionally, evidence suggests that OD reduces long-term emotional distress and injury to families and healthcare staff. 5,14–22 As an added benefit, greater openness with families post incident might enhance their confidence in service transparency, thereby reducing their need for investigatory and legal processes and, when practised routinely and in the longer term, may enhance the public’s trust in services. 14,19,23–29

Yet, services consistently providing OD that meets the needs and expectations of patients, families and staff remains a challenge. Barriers to OD are deeply rooted and widely documented, and include factors like physician management fears over loss of reputation or litigation and second victim trauma,11,12,30–33 fear of reprisals in punitive workplaces,34 lack of clarity over accountability for injury,15,35,36 varying perspectives on when disclosure is appropriate12 and heavy-handed procedural systems that can constrain empathic communication. 12,23,37,38 Furthermore, since the inception of OD policy initiatives, researchers have noted a tension between what patients and families expect after injury and what healthcare providers (HCPs) think that they should provide. 8,39,40 For HCPs, post-incident communication with families encompasses a complex array of legislative, policy, professional, organisational, local workplace and interpersonal considerations. 12,17,37 In the English NHS, and particularly in NHS maternity care since 2013, the development and implementation of a raft of legislative, policy and safety improvement initiatives have given rise to complex and sometimes contradictory OD improvement contexts. The following sections outline the rationales, components and effects of the most significant of these.

Open disclosure initiatives in the English National Health Service

Study phase 1: scoping

Study phase 2: ethnography

Study phase 3: interpretation and output development

Finally, we facilitated five interpretive forums: one at each organisational case-study site (two online and one in-person), one national forum with the PAG and one online family forum (FF) (online due to COVID-19 travel concerns). The interpretive forum research process is described in Appendix 2. Appendix 2, Table 3 details participation at the interpretive forums.

Building trust in organisations

In 2014, the impetus for involving SUs in reviews and investigations was to build individual and public trust in organisations. 95,101 A shift from ‘paternalism’ to ‘partnership’ and the ability to have candid conversations were emphasised for improving trust. 101 Recommendations related to SU involvement were focused on organisational obligation to provide patients and families their rightful access to their health information101 and to notify, acknowledge and apologise for what happened. 95,101 This was described not only as a professional obligation, but also as ‘the right thing to do’; a key priority was creating an organisational culture of openness. At the time, SUs were one-directional recipients of information about their case. Care Quality Commission Guidance recommended a single point of contact in case of SU questions or queries about what happened, but the emphasis was on organisational communication rather than SU involvement. 95 The CQC recommended the provision of emotional support and other relevant assistance for SUs, but details on how this should be facilitated were not provided. 95

Improving systems of care and ensuring accountability

The Kirkup report, published in early 2015 on the independent investigation into Morecambe Bay NHS Trust, uncovered numerous failures in safety, care and respect for injured families within the Trust. 54 Recommendations addressed organisational failings in admission of the ‘extent and nature’ of injuries caused to families. The Trust was instructed to review family involvement practices, particularly with respect to complaint procedures. It was noted, for the NHS more widely, that there was a ‘strong case’ for standardising the internal investigation process to include input and feedback from families, with families offered the opportunity contribute evidence to their investigation. Shortly after the Kirkup report, strategy documents began to highlight the importance of reducing stillbirths and neonatal deaths. 105,107 These documents did not include SU involvement recommendations for when harm occurred, but did suggest that post-mortem examinations were important in the event of stillbirth or neonatal death so that counselling could be provided for future pregnancies. 105,107

Around the same time, the rationale for conducting investigations into incidents of injury shifted from building and restoring public trust in the NHS to improving systems of care while encouraging public accountability. 96,98,106 Although documents did not make explicit recommendations for family involvement, they still warned that the relationship between patients and organisations was in danger of remaining merely ‘transactional and contractual’ without changes and improvements in leadership. 104 When involvement was included, it was envisaged as more two directional than in previous documents, with a language shift from ‘apologising, acknowledging, providing and telling’42,99 in earlier documents to ‘engaging’,96,98 ‘hearing’106 or ‘involving families’97,98 as ‘active participants’96 ‘at the centre’ of investigation processes98 from ‘start to finish’. 106 Policy documents continued to advise on the provision of a single point of contact for families during the investigation process,96–99 along with the opportunity for SUs to ask questions. 98,99 For the first time, the importance of establishing family expectations and preferences for communication96,98 was described as part of active engagement. Although little detail was provided on how to achieve active engagement or involvement,106 the unique perspective SUs offer on their adverse events was acknowledged and it was noted that these perspectives should be shared and listened to as part of the review process. 96,98 The specifics on how these contributions were to be garnered and used remained unclear.

Notably, the serious incident framework (SI Framework), which involved some families in its production, was the first document to detail specific actions and resources to produce meaningful involvement (e.g. the provision of letter templates to facilitate the initial communication with families). 98 The SI Framework suggested the following actions: arranging meetings with families, ensuring that investigation teams have expertise in facilitating family involvement, keeping families informed of investigation progress, providing support and opportunity to express concerns and questions, enabling families to inform ToR and give evidence, sharing findings with families, giving media advice as appropriate and providing the opportunity for families to comment on findings and recommendations. 98 The SI Framework included expectations that investigation reports include a description of how families have been engaged in the investigation process and how they were supported following the incident. Finally, after the conclusion of the investigation, the framework recommended that families be offered opportunities for continued involvement should they wish. The rationales for family involvement were to ensure accountability and build service-user confidence that the investigation findings were ‘robust, meaningful and fairly presented’. 98

Improving the safety of maternity care and saying sorry

Most documents in 2016 were focused on improving the safety of maternity care, rather than improving the processes after harm has happened. 126–129 However, in the Better Births National Maternity Review, where recommendations for improving family involvement were made, the language used to describe this involvement matched the shift in late 2015. 108 Recommendations focused on improving communication with families, putting families at the centre of processes and enabling informed decision-making. In addition to informing families of what was involved in reviews and investigations, the importance of individualised, relational care in maternity care and in investigation practices was foregrounded. 108 The importance of a single point of contact, especially in the wake of trauma, was described as essential to communicating family involvement options. 108

Despite these recommendations, evidence from the CQC in 2016 suggested SU involvement in investigation processes was still inadequate. 49 SUs reported that they were not always told that an investigation was happening, what their rights were in this process, what an investigation involved or how to access support and advocacy. 49 Family involvement was perceived by families as ‘tokenistic’ rather than ‘meaningful’. 49 This suggests that, although the language had shifted towards promoting meaningful involvement after publication of the Kirkup report, there was a lag in translating these priorities into practice. To combat this, the CQC review recommended production of a complementary family engagement framework, developed in partnership with families, for learning from deaths.

Throughout this period, a central concern was how to manage NHS staff concerns with liability and organisational obligations to meet the DoC. In 2016, the NHSR ‘Saying Sorry’ poster was released for circulation through Trust services. Its purpose was to encourage more NHS staff to apologise to injured families. The poster sought to clarify the difference between an apology and an admission of liability. 100 It also alerted staff to their obligations, as employees and professionals, to ‘apologise’, ‘acknowledge’, ‘share information about what went wrong’, ‘provide truthful information’ and ‘inform’ families after an incident. 100 In addition to this reminder of obligations, NHS staff were encouraged to see that apologising had a wider moral basis, in that it is ‘the right thing to do’. The poster also touched on the challenges of standardising communication with injured families, suggesting the ‘tailoring the apology’ to individual patient’s needs. 100

Shifting to individualised, relational care

In 2017, the shift to individualised, relational care was even more pronounced. A new vocabulary was adopted: rather than ‘active participants’ as in previous publications, families were described as ‘equal partners’. 109,110 This was captured in the recommendations, which were extended to incorporate individual choice. In addition to communicating openly and honestly about what happened, services were advised to communicate investigation and complaint procedures with families55,109,110,112 and to involve families to whatever extent they wish in these processes. 55,109–112 This involvement was envisaged as providing the opportunity for families to ask questions and voice concerns during reviews,55,109,110 inviting families to contribute to evidence, inform ToR and comment on findings and recommendations,55,110 providing families a meaningful, plain English explanation of what happened and what could have prevented what happened,112 and finally, offering the option of involving families in service learning efforts after the conclusion of the review or investigation. 109,110 However, it was also noted that it should be made clear that SU feedback may not be included if was not considered ‘relevant or appropriate’. 110 Who determines what is ‘relevant or appropriate’ was not described in this guidance, which was produced for Trust Boards. However, this creates an interesting paradox: families cannot truly be considered equal partners if their contributions to reviews can be dismissed.

Additional new recommendations, such as providing a sympathetic environment for disclosure, also surfaced at this time. 109,110 Improved communication between healthcare professionals across services was also recommended, for instance, increasing communication with outside facilities during transfer and with the family’s general practitioner (GP), to improve continuity and handover. 109,110 Although these recommendations were more actionable than their predecessors, it was still suggested that more detailed guidance on family involvement was required. 109,111 In some cases, recommendations signposted to existing outside guidance, such as that by the stillbirth and neonatal death charity (Sands). 55,111,112 Concurrently, two consultation documents were produced to inform policy development in ways that aligned the interests of families and HCPs. 57,113 These stated that both parties value the opportunity for meaningful apology,57 increased family involvement in reviews and investigations,57 a single point of contact57 and continuity of carer. 113

Enhancing communication

Despite mounting recommendations in previous years for increasing family involvement, in 2018, a progress update on the Each Baby Counts (EBC) programme revealed that only 41% of parents were invited to be involved in reviews. This was an increase from 34% in the previous EBC report, but still startlingly far from involvement ambitions. 116 The EBC recommendations that followed these statistics were mainly focused on procedural compliance, but reiterated that families should be informed of any reviews and investigations taking place and be invited to contribute according to their wishes. 116 The Maternity Safety Training Fund (MSTF) Fund, one of the other mechanisms set out for achieving the national maternity ambitions, also released an evaluation report in 2018 which listed the training opportunities provided to staff around maternity safety. Disappointingly, none of the offered trainings were related to family involvement or disclosure, other than training on the DoC. 130 The content of this training was not described, and it was not one of the ‘popular’ courses selected by the Trust. The course was also not featured in the MSTF catalogue, but rather was included as a course funded by the programme in the ‘other’ category.

Also in this year, National Health Service Improvement (NHSI) launched their scheme for ‘Maternity Safety Champions’ operating at the front-line, regional and national levels. 114 Maternity Safety Champions are representatives who act as ambassadors for improving safety in maternity care by learning and sharing best practice. Notably, this programme does not include a SU representative; however, the guidance recommends that champions ‘work with service users to address their needs, particularly in the redesign of new services’ (p. 10). Although no advice on how to achieve this work is given, it is the first time that a co-design approach to family involvement in maternity safety/service improvement is recommended.

The year 2018 did not yield many new recommendations for SU involvement; however, NHSI produced a document on promoting effective spoken communication between clinicians and patients. 115 This highlighted factors like providing the right environment for communication, ensuring information is accurate and understood, listening, conveying an attitude of respect and aligning expectations as facilitators to effective communication. 115 Although not specifically about disclosure, these principles can be applied to the practice. However, as Iedema et al. pointed out the following year in their report on the findings of this initiative, translating these principles from work-as-imagined to work-as-done presents a number of challenges. 137 Namely, there are differences between factually accurate communication and cultural, emotional and situationally sensitive communication, between imagined calm contexts for communication and the reality of the hospital setting, and between structured, evidence-based communication templates and individual, flexible, situated judgement. As such, it is unsurprising that these recommendations were not carried into future policy documents.

Conceptualising families as active partners rather than passive recipients

In mid-2019, the publication of the new NHS Patient Safety Strategy (PSS) marked a significant shift in the potential for patient and family involvement. 120 Transparency, providing opportunity for families to raise concerns and the use of the Patient Safety Incident Response Framework (PSIRF) and ENS were emphasised, and family involvement was embedded in a wider national strategy. The first ENS progress report, published in 2019, highlighted that the programme would enable families to receive answers, support and compensation more quickly. The ENS report did not reflect any of the developments in family involvement recommendations since 2015, instead integrating the original DoC steps of apology, openness, candour and providing support. 121 The biggest addition to existing recommendations in the PSS was the explicit recommendation to approach patients not as ‘passive recipients’ but ‘active partners’. In other publications published that year, the language matched that of 2017, employing phrases like ‘meaningful engagement’ and family as ‘equal partners’. 118 The PSS differentiated itself terminologically, opting for the word ‘active’ instead of ‘equal’. This new status was to be embodied by the ‘Patient Safety Partner’ who was to represent patient interests in their work with innovators on co-production, safety, strategy and policy developments. 120 Other publications signposted to existing recommendations for family involvement from the SI Framework, Sands or HSIB rather than producing new recommendations. 117,119 A report of qualitative findings on the implementation of the learning from deaths national guidance was published, reporting that good practice involved engaging families. 118 The report revealed that environments where there is a culture of openness and staff are supported/trained help with achieving engagement, but did not give specific recommendations beyond that. Several reports published between 2019 and 2020 on maternity safety topics by NHS England, NHSI, NHSR and HSIB did not include any sections on family involvement. 66,125,131–136

Enabling families to guide the process

As in previous years, 2020 policy recommendations by professional, NHS and inspection bodies focused on reminding services of the importance of informing families that a review or investigation was taking place,93,94,122 offering an opportunity for families to share their perspective on their care and raise any questions that they have,93,94,122 enabling families to contribute to evidence,68 personalising care123 and providing a point of contact. 124 Interventions that include family notification, such as the PMRT93,122 and ENS,122 were also signposted. The impact of the gradual recommendation changes and specification on steps for boosting involvement was evidenced in the EBC progress reports published in 2020, which revealed that invitations for family involvement in reviews had increased from 40% to 51% from 2016 to 201793 and to 70% by 2018. 94 These results were approximated in PMRT annual review reports (see Report Supplementary Material 1).

The language shifted slightly in some documents from ‘partnership’ and ‘active engagement’ to allowing patients and families to ‘guide’ their own engagement122,124 with flexibility, inclusivity and transparency during the review process. 93,94 The transition of SUs from recipients to equal partners, to active partners, and finally to guides, demonstrates the changes in perception of family involvement over time. HSIB stressed that families offer unique perspective as the only individuals with insight into what transpires at all stages of the healthcare journey and that families should have continuous involvement. 68 Fittingly, new recommendations at this time included encouraging and instructing patients, families and carers on how to record and share information about patient safety incidents,124 proactively seeking feedback about service openness and transparency,124 having a discussion with families about incidents124 and helping families to understand reports and recommendations. 68 Although national guidance remained the same, HSIB published a diagram illustrating the different steps for family involvement throughout HSIB-led reviews and investigations, illustrating their process for continuous engagement. 68 The process spans the initial contact with the family through investigation completion, with opportunities for family engagement at each stage. Each step has a section with additional information on questions to ask families, considerations to keep in mind and advice for where, when and how meetings with families should take place. 68 For the first time, this process involves giving the family access to a draft report for their input in addition to shaping ToR. 68

Healthcare Safety Investigation Branch reports on their progress for involving families in their investigations indicates the success of organisational investment in an adequately staffed, systematic and flexible family involvement strategy. It also indicates the neglected interests of many families in involvement. In 2019–20, 88% of families were involved in HSIB maternity investigations, whereas, in that same year, 34% of families were involved in Trust investigations. 103

In late 2020, the preliminary findings of the Ockenden report were published with immediate actions and recommendations. 89 Disappointingly, as in Kirkup,54 few of these recommendations were related to processes of family involvement. While Ockenden reports that families want their questions answered and want systems to learn, the only recommendation related to families made is that family voices must be heard and that families must have an advocate. 89 This advocate is to provide an oversight of meetings that happen with families in services. The final Ockenden report, published in 2022, builds on these recommendations to some degree, adding that families should be the primary concern during incident investigations, that families must be actively involved, that feedback must be shared with families openly and transparently by senior team members, that governance teams must work with the Maternity Voices Partnership (MVP) and that bereavement care and other support should be provided. 90 In 2021, an inquiry report published by the House of Commons Health and Social Care Committee recommended that families should be involved in a ‘compassionate manner’ and acknowledged that investigations have often failed to involve families in a ‘meaningful’ way. 103 It was again stressed that families need to be heard and that lessons need to be learnt; HSIB’s family engagement pathway68 was cited as a programme that had boosted family involvement. The report concludes that: ‘… it is important that [HSIB] continue to pursue improvements in this area to ensure that all investigations are informed by the experience of families’ (p. 22). 103 What improvements were needed, or how these might be achieved, were not noted.

Initial programme theories hypothesised

The following sections describe the synthesised underlying factors for OD as a series of generalised hypotheses and explore how these are considered in the included papers. We summarise the realist data extraction findings in relation to the five identified programme theories in Appendix 5, Table 8.

The five identified programme theories are summarised as:

1. receiving a meaningful acknowledgement of the harm that has happened

2. being involved during the review/investigation process

3. making sense of what happened

4. receiving care from clinicians who are skilled and feel safe during post-incident communication

5. knowing that things have changed because of what has happened.

Strengths and limitations of the realist synthesis

The strength of this analysis is that the identified underlying factors for OD have been developed iteratively with input from discussions with stakeholders and recent policy interests in this field. The programme theories have been constructed based on their pragmatic relevance in guiding future ethnographic research within maternity services. 73 Following realist principles,173,174 a non-linear approach to data searching, along with the inclusion of heterogeneous evidence sources, allowed us to develop and refine our approach during the synthesis process. More traditional search strategies would have excluded many relevant sources that were not peer reviewed. At the same time, this approach included data that might not have the same rigor as data extracted in a traditional systematic review. While we have maintained a focus on international maternity settings, the most recent documents are from English health services, where there is a notable policy drive for maternity safety improvement. This may decrease the external validity of the results. These themes should be explored more widely in empirical research in both the English NHS and other health systems. In the following five chapters, which are drawn from the results of SP2 and SP3, we explore each programme theory from the perspectives of national and regional stakeholders and research participants from the three case-study services.

General reflections

Most participants highlighted the importance of communication by clinicians as soon as possible after the incident for injured families to begin to feel acknowledged. Several also noted the importance of honesty. Two national leads described that OD stood as the ‘litmus test’ of an open organisation and committed safety culture. Few QI leads felt that there has been significant improvement in OD in NHS maternity care. There was an overall sense that the reported OD improvement did not capture the quality of communication required by families. A senior charity lead commented:

Parents say the same thing today that I would have said 19 years ago or someone else would have told me 10 years ago. Parents are not being told.

SG2

Few participants felt the organisational DoC had encouraged meaningful discussion with or involvement of families but rather had ‘just become another process you’ve got to do’ (Qi17). Participants rarely discussed the practice of acknowledgement beyond spoken apologies and later explanations of what happened. The provision of tailored postnatal care or compensation was rarely mentioned. Rather, participants focused on the practical and ethical challenges of honest conversations with families, particularly in relation to the demands of corporate assurance measures. As significant variations between thresholds of clinical harm determined the application of particular review and/or investigation procedures and guidelines for family involvement in these procedures.

The ‘human’ and the ‘technical’

Many participants noted a difference between the ‘human’ and ‘technical’ aspects of candour. This captured the tensions between what constituted honesty in clinical relationships, who had authority to determine how harm is defined and categorised and what this authority conveyed about organisational, professional and family interests.

For most clinicians, candour was envisaged as honesty embedded in a taken-for-granted relationship of care. Some objected to the notion of the organisation of a conversation of candour as a series of procedural steps in a mandated time frame. Referring indirectly to the DoC, they stressed the importance of a situated and responsive approach to honesty within an ongoing clinical relationship.

A senior obstetrician commented:

Candour is the start of your relationship with a patient [and is] built upon many things: integrity of what you are saying, how you are saying it and the relationships that you have developed with that person, be it long or short term. We tend to make up for that by having policies and procedures. The other day at our clinical governance meeting a senior member said: ‘did you do candour?’ and I put my hand up and said: ‘It’s not a thing!’

Qi13

Many clinicians felt that the demands of organisational compliance had diverted attention away from ‘the human story’ (FS9), thus narrowing the care of injured families. Some charity leads and defence solicitors also noted the effects of the ‘legal imprint’ on conversations with families, with clinicians fearful of ‘getting it wrong’ (PB1) and with conversations reduced to ‘legal transactions’ (ML2). Overall, many clinicians and some others considered that the organisational DoC had undermined the relational care of injured families.

Apologising

All participants agreed that an apology in person after any incident was important to all families, irrespective of the circumstances of an event. However, many commented on the difference between interpersonal expressions of regret offered by those caring for a family and routine apology initiated by organisational directives, which were less likely to be attuned to the family.

For senior clinicians, apologising was seen as the ‘human thing to do’. Many described the long-term impact of the difficulties with expressing regret to families injured in their care. These descriptions differed from the more consumer-orientated, instrumental apologising described by stakeholders from governance and commissioning backgrounds. A former commissioner commented:

And it doesn’t matter [about your Band], you say ‘I’m really, really sorry, I’ve done this.’ The family anger is when ‘they didn’t even say sorry.’ To just apologise reduces the level of anger and not admitting that you’ve done something wrong is going to cost the Trust money.

Qi15

In the discussion between clinicians and families, apologising had value as a strategy for reducing conflict between a family and a service, with legal action implied. However, for clinicians anxious about litigation, a misinterpreted claim of individual accountability made apologising feel risky. For some, the blurring of personal and organisationally directed apology was complicated by evolving understanding of the ‘avoidability’ of an incident. The risk of misconstrued apologies was also discussed by participants from Trust legal teams. Their concern was that an apology for ‘distress caused’ could be an admission of liability. Overall, the emphasis of legal teams on ‘right words’ and ‘right time’ was observed to make already difficult conversations more challenging.

In summary, in their reflections of the conditions and effects of personal apology, participants reiterated the felt tension between organisationally prescribed behaviours and personal responses to injured families. For those in corporate management, technical processes and procedures were essential to the management of risk to the organisation.

Degrees of candour

Summary of stakeholder perspectives on acknowledgement

Apart from family representatives, national stakeholders felt that the regulation of candour had disabled the situated judgement of clinicians to shape family-centred OD. At the same time, situated judgements opened the possibility for OD to be avoided or delayed for reasons besides the family’s best interests. Maternity safety improvement schemes also contributed to a managerial/procedural stance of individual family entitlement.

Perspectives and experiences of families

For family participants, acknowledgement of harm had a considerable impact on their overall journey after injury. While immediate priorities varied between families depending on the context, the way their harm was disclosed mattered. The ‘human’ aspects of this communication were fundamental to their experience. Clarity about what was known at the time of the event made a positive difference to families who were overwhelmed and confused. They understood that full explanations were not always possible soon after an event, but nevertheless valued the sharing of available information. Honesty was important even if the news was difficult to hear. Sometimes this news came as a shock, but open communication was appreciated, especially if the message was delivered in a kind and caring way. For some families, communication that harm might have been avoidable occurred as events were unfolding. For others, it came later, during or after the arrival of the incident investigation report. Whatever the situation, empathic and responsive communication was key.

Few families identified formal DoC processes, but instead referred to meetings with staff in which they learnt about the harm experienced. Families often appreciated immediate appointments followed by further scheduled meetings, which provided time for processing information and sent a message to them that questions and discussion were welcome. Participants whose injuries were not identified as avoidable, or who had no interest in raising questions about avoidability, were more likely to feel that staff had been open with them:

I think genuinely everything was done that could have been done, so therefore they could be completely honest with everything.

SU1-5,7

Honest conversations led to trust in practitioners that was lacking if there were concerns that information was being withheld. Some participants described tight guarding of information and defensive behaviour from clinical staff during initial and later conversations. This generated a climate of suspicion and a concern that staff were ‘closing ranks’ (SU7) to control information and prevent access to answers. Yet, in some situations, families revisited these initial impressions. For example, a family who experienced a neonatal death in 2015 later acknowledged that it was more likely that staff were overwhelmed with work. However, their suspicion that clinicians were guarding information affected confidence in immediate care post injury (SU7).

Irrespective of questions of avoidability, sincere expressions of regret for the injury were important to participants, even if these were sometimes delayed or prompted by a family’s insistence on further conversations (e.g. after a complaint about how they were treated). Impersonal, written apologies were more defensively phrased and less welcome. ‘Non-apologies’ (‘I’m sorry you feel that way’ SU20) were distressing for families, particularly when they conveyed the impression that the family was somehow responsible for the situation. Other unhelpful responses include being told ‘it is one of those things’, especially when, as with one family whose baby required cooling following his birth in 2008, families were discouraged from asking questions and suspected that information was being withheld (SU14).

The physical setting where OD happened was remembered as important. Although families appreciated invitations from the service after the event to discuss what took place, checking with them about how and where they would like these conversations to take place did not always occur. Women reported being ‘being crammed in a tiny room and really hot’ (SU17), which was particularly hard if the space was close to where the harm happened. One family, whose baby was stillborn in 2020, was obliged to return to the place where the event happened to receive post-mortem results:

And we got a letter through the post … asking us to attend an appointment at that hospital, where we would have to go and sit in the waiting room amongst all the other pregnant women whilst we wait for the second half of the post-mortem results.

SU16

The right words used at the right time matter, but the location also needs to be considered for the message(s) to be delivered with sensitivity. Honest, prompt and in-person OD conversations were important to families. It was appreciated that a complete picture of what happened would not be immediately available.

Perspectives and experiences of in the case-study services

What is meant by family involvement

Across the case-study services, there were a range of definitions of family involvement. Clinicians understood involvement as the process of asking a family if they had any questions or concerns during a DoC conversation, offering a meeting to discuss family questions, requesting questions about the draft report and providing a final explanation when the report content was agreed by the Trust. By contrast, charity and HSIB leads envisaged a more open and reciprocal approach whereby family perspectives and concerns informed the scope of a review or investigation, the ToR, the emergent findings, draft reports and final action plans. Without this sort of involvement, these participants felt:

it’s not unusual for a Trust to do what they think is a good job investigating and to share it with the family at a very advanced stage when in fact it has already been signed off and can’t really be changed.

SG6

A widespread experience of national stakeholders was that there was a lack of family representation in internal SI investigations. Informal and less visible work on CG teams and clinical leads to represent family interests and perspectives in service-level reviews and Trust-level investigations was overlooked as involvement work.

In addition to holding individuals and services accountable, families needed to revisit the events, in their own terms, as they experienced them to make sense of what happened. One charity lead explained:

This gap between what clinicians know and what parents go away with, it is about the entire narrative of that day and the need to process a life-changing event for a family.

SG2

To meet this need, this participant noted, the incident must be approached in ways that were different to clinical or safety improvement frameworks and that had the inclusion of family personal experience as a starting point.

The Perinatal Mortality Review Tool

Nine participants considered the variable effects of the implementation of the financially incentivised (in England) PMRT, which includes prompts for disclosure and inviting questions from parents about their or their baby’s care. Several noted the importance of this tool to provide ‘a coherent answer to what happened’ (Qi6) because it was designed to include multidisciplinary and family perspectives. However, most participants agreed that PMRT implementation in Trusts was stymied by a widespread lack of service-level investment in skilled staff with time for ‘that type of conversation’ (FS5) to ‘understand how a family is feeling’ (ML2). Local service capacity for the review determined whether families were ‘invited into the investigation process or just … asked “can you send us some questions?”’ (FS5).

Some participants noted the increase in the percentage of parents reported to have been informed of a review. Some improvement leads also suggested that the online prompts for reporting family engagement that were built into the tool could act as an important ‘signalling exercise’ (Qi6). However, most also noted the practical and ethical difficulties of asking time-squeezed staff to initiate ‘people-related difficult conversations where you just don’t know what a family is going to need’ (Qi20). They worried that the simplification of this work from ‘actual engagement and the desire to change things for families to a tick-box’ (Qi11) was enhanced by the financial incentivisation of the tool through the Maternity Incentive Scheme. Participants cited examples of midwives responsible for tool completion limiting family engagement to e-mail correspondence or simply asking colleagues for any signs of family questions or concerns.

As with the discussions about the DoC, participants who were clinicians stressed their concerns that the implementation of national safety improvement programmes reduced the emotionally demanding, skilled, intricate and unpredictable conversations with families into an administrative procedure. Several participants noted that despite the provision of sensitive guidance on family engagement, the completion of the financially incentivised tool was directed by organisational rather than family needs.

Family representation in Healthcare Safety Investigation Branch investigations

Almost all participants who were also practising clinicians felt that family engagement in the HSIB maternity programme had been positively received by these families. They felt that this had encouraged a wider reconsideration of family engagement in Trust-based (or internal) investigations. Several factors contributing to the success of HSIB’s family engagement approach were identified. First, it was felt that more families would engage with HSIB because it was independent and perceived as trustworthy. A senior HSIB lead commented that before this programme, ‘families would have felt that nobody was even vaguely in their corner’ (Qi4). HSIB leads described family involvement as a ‘moral compass’ with guiding values of ‘openness, honesty and transparency throughout relationships with the family’ (Qi1). They felt that concerns about reputational and litigation risk for the Trust created a conflict of interest when applying these values. Second, the HSIB approach was underpinned by centralised strategic planning and funding. Senior HSIB operational and family engagement leads brought practical expertise from non-maternity healthcare settings, emphasising the need for interpersonal and organisational support systems for investigation teams and the ongoing development of family engagement support tools and facilities. Finally, responsiveness to family needs was central to the practice of the investigation teams. In some situations, senior leads modelled an approach to family engagement that included home visits, which created ongoing opportunities for families to ‘talk through their story’ (Qi8) and to develop family care directly from knowledge of their needs.

Some services, particularly senior midwives, worried that HSIB family engagement work undermined the relationship between families and the NHS. Some midwives also reflected on the impact of external investigations on the morale of service staff. Despite this, the contrast between the type of family representation promoted by some PMRT questions and the type of representation that occurred through a well-resourced national organisation with the capacity to develop strategic and operational plans for family engagement was stark.

The impact of national safety developments on staff and families

Many participants noted that the introduction of multiple review and investigation systems was a barrier to improving disclosure with families. This was a point of frustration for some clinicians. Participants felt that, particularly since 2018, both families and clinicians have been faced with an ever-expanding range of national safety improvement initiatives. Many identified a sense of disconnect between national ambitions of maternity safety improvement, the implementation approaches used (notably, incentivised schemes with regular assurance reporting), and actual, felt changes in services. This situation left clinical and service leads feeling ‘frustrated’ (FS11), ‘overburdened’ (Qi5) and ‘overwhelmed’ (SG2) with a ‘well-intentioned, brilliant, but complicated jigsaw’ (ML2). Misunderstandings over referral criteria or thresholds of severity led to situations where some injured families ‘slipped though the investigation net’ (SF5).

Attempts by national safety improvement leads to consolidate the reporting tasks of clinicians with the construction of single reporting portals were welcome. However, the need for clinicians to respond to multiple initiatives with different information requirements remained. The overall effect was that ‘people still feel there is just lots of different bits (of reporting activity) without specific purpose’ (Qi4). Participants also indicated that families required ongoing advice for understanding these schemes, yet this task often fell to families themselves. The ongoing efforts demanded of staff to stay abreast of complex and changing schemes could reduce the time spend with families and a sense that this time diverted attention away from systems designed to include them. The introduction and rapid development of multiple and sometimes overlapping programmes to improve maternity safety investigation and family involvement were complicated by a lack of capacity to keep up to date on changes, rendering clinicians ill-equipped to provide the expert reassurance that families required.

The organisation of post-incident family involvement

Data from the three case-study services were used to establish a diagrammatic representation of family involvement in reviews and investigations, focusing on Trust and service-level processes (see Appendix 6, Figure 5). All case-study services followed DoC procedures including identifying a named linked person for each family. Some clinicians also initiated responsive link arrangements with families already known to them, occasionally working across services and teams as an ongoing contact. Most clinical leads and CG teams felt that service resourcing, to develop more coherent structures for family representation post incident, was inadequate and were aware that ‘many families with questions [were] falling through the gaps’ (SDSi06). While some clinicians justified this as a difference in families’ interest in engagement, others highlighted a lack of time for family outreach and the poor commissioning of language services as contributory factors. Issues of service capacity were interconnected with distinctive patient-flow patterns in maternity services. The most prominent of these were the rapid throughput of families from hospital to community services and the transfer of babies requiring intensive care across hospital sites and services or between Trusts.

Communication with families was difficult and incomplete when conducted by e-mail or telephone. The situation was improved when families were visited in person at nearby neonatal units. For example, a senior neonatologist described that the ongoing updating of families on the emergent process and findings of reviews and investigations was more an aspect of day-to-day communication within the unit because, compared to maternity professionals, ‘it is more in our nature to be constantly explaining’ (SASi06).

Service capacity for encouraging family representation in review and investigations differed between the case-study services and in relation to the differing attitudes of the CG teams towards the risks and advantages of family involvement. In Site A, where there was greater reluctance to establish processes for involving injured families, CG teams were more likely to work within DoC procedures, where families were invited to contact a named link person if they had concerns or questions. At Sites B and C, a more proactive approach to involving these families had been introduced. One senior obstetrician explained:

Historically, we have situations where even when the woman had the letter, they didn’t know who to contact so they felt very alone so we’ve tried to build in gentle touch points to update them and say, ‘is there anything you want us to make sure we’ve covered?’ And some will send an e-mail and other families we’ll meet with them.

SDSi06

Nevertheless, even in services taking a more progressive and proactive approach to the involvement of injured families, initiating and sustaining this process required the staff to encourage first contact, respond to expressed needs and to incorporate family perspectives into panel presentations, discussions and reports. The value and undeveloped possibility of community midwives representing injured families during investigation processes was noted across all three services.

Determining facts with families

Across the case-study services, both CG and lead clinicians described several different rationales for ensuring that families were represented in review and investigation meetings and different ways that insights from families were elicited and included in investigation processes. Most often, family representation was understood to be about the exchange of factual details about the incident between CG teams and families. These exchanges happened in response to those families who had themselves contacted the CG team with a question about the incident. However, CG staff valued these family enquiries:

And when they come back to us with an e-mail full of questions, of this is what we want answered by the review panel. Which is great because it gives us a real steer and you feel like you are doing a good job for them [because] you can answer what is actually on their mind rather than what we think the priorities are.

Risk Midwife, SASi01c

Without ‘getting a steer’ on what a family might want to know, the CG teams felt less able to anticipate how a family might respond to report findings and might overlook what is important to a particular family.

In some situations, families were invited to draft their own factual accounts and chronologies of events for incorporation into the risk scenario, which was otherwise derived from clinical records. However, these contributions were most likely to be included where a family could produce ‘hard evidence’ (SASi1b), such as smartphone records or photographs where there can be ‘no debate’ (SASi1b). CG teams often noted the difficulties of dealing with unsubstantiated accounts from families, especially where these insights were missing from or contradicted the clinical notes. In these situations, it was almost always assumed that the family had forgotten or had misinterpreted events or conversations. CG teams agreed that the reason for building a factually detailed risk narrative of the event was to improve the depth of information available to panels and, from this, to improve the quality of a report sent to the service and parents. The inclusion of information from the family meant the CG team ‘don’t feel the need to revisit things after the report is published’ (SCSi04) and avoided a situation where ‘if there was some crucial bit of information that they had, and we didn’t know then the whole thing slightly unravels’ (SCSi06). There was an awareness amongst CG teams that without inclusion of a family perspective, reports produced by them could be dismissed by some families as ‘full of lies’ (SASi16). Additionally, CG teams felt that the inclusion of family perspective in the incident investigation was important for demonstrating that their service was not defensive or duplicitous.

Of the four case-study families (all from Site A), two had had their incidents investigated internally rather than through HSIB and all had contributed towards their investigations by adding factual knowledge of events that was absent from clinical records. One family had not been notified of the internal investigation of a serious near-miss medication error. They felt that ‘the service had shown itself as defensive’ (SAFi03) by not informing them of the planned investigation, and ‘unwilling to learn’ by not including their perspectives on the incident. A clinical risk and debrief lead met with them to re-examine the incident. Using this family’s factual account with time-stamped photographs, the clinical risk lead initiated changes to drug administration equipment in the department and across the Trust. However, this consultant described his meetings with this family as unusual. He felt that, unlike some families he met, this family was interested in system safety concerns as well as their isolated case (SASi20).

The significance of family experiences

In addition to determining the facts and timeline of the case, at least initially, most families wanted to determine their own frames of reference for their engagement with services. These were more often highly detailed, personalised narratives, related to feelings and memories surrounding the event. This immediate priority for families often eluded staff who were more engaged with evidence gathering for risk identification and learning. For example, a risk midwife commented that, when you start to ask a family what they want to know:

We ask ‘is there anything that you want us to raise on your behalf so it can be discussed?’ Nine times out of ten, we will get questions from the patient to say, ‘This is how I felt’ and you have to respond to that.

SASi01

Clinical governance teams and other clinicians, as this risk midwife indicates, engaged with injured families because the experience of these families demanded in-principal acknowledgement of their situation.

There were several ways that CG teams and risk midwives managed the tension between collecting facts from families and recognising other means of family representation. For example, in Site A, some risk midwives kept separate notebooks of family reflections on the impact of their injury. At least during the research period, the use of family reflection documents appeared to be for personal reference only. These documents were managed in addition to records of family descriptions of events for the risk record. Across all services, the stories of some families were remembered by CG teams and clinical leads as important and enduring reference points that informed their personal approaches to injured families. These also became integrated into in-service teaching sessions as it was hoped that they would enhance collective approaches to care.

As mentioned in the previous chapter, some clinicians developed close relationships with some injured families, that often extended beyond their core work responsibilities. For example, they sometimes called to check if a family was prepared for a meeting, accompanied them around the hospital and arranged or escalated various outpatient appointments. At Site A, risk midwives identified themselves as family representatives because of their wider interest in the experiences of families beyond the risk record. According to one risk midwife there:

So we say before at the end of the panel meeting, in summary, ‘if the parents were watching us now, are we agreed and assured that we have discussed everything you would want to know as a parent? that we’ve done everything that we possibly can’. Just to draw people’s attention back to the fact that this is a woman and baby that we’re talking about. That there’s a family hurt at the end of it.

SASi04

Our field-work observations were that the possibilities for improving panel meetings to better accommodate family perspectives were very limited at this site. For the most part, panel discussions focused on questions of informed choice about maternity care and on the reliability of a family’s memories about events. Despite CG team and some clinicians’ best intentions, most panel meetings observed pivoted on the construction and defence of risk and its management (see further discussion below).

Parent involvement in the Perinatal Mortality Review Tool

Five PMRT panels were observed in two services (Sites A and B) and 10 staff regularly involved in PMRT work were interviewed. One of the panels observed reviewed a case of serious harm that was also investigated internally. In all three case-study services, attitudes towards parent involvement in the PMRT were entangled with the wider implementation challenges of the PMRT.

There was a significant difference in the attitudes towards the PMRT at Site B (a PMRT pilot service), and the other services. At this former pilot site, a consultation obstetrician led on bereavement care and PMRT completion and, from their perspective, the PMRT carried significant value for learning across the service and the Trust:

We feed the findings up to the mortality surveillance group within the hospital [and] it’s increased the awareness of the importance of mortality review so now maternity had helped lead improvement of reviews of death in other specialities.

SDSi12

This PMRT lead also noted the ‘huge potential’ of the PMRT across her local network that was lost without ‘having the time and resource to be able to do everything as you would like it to be done’ (SDSi12). However, this PMRT lead noted, along with risk midwives and clinical leads at Sites A and B, that improvements to the design and content of the tool to avoid the need to complete many questions that appeared to her to be unnecessary would avoid some duplication.

It is quite laborious and quite repetitive, but I think you can get useful stuff out of it if you used it properly.

SDSi12

Without a clear understanding of the potential of the tool to generate learning for their own service or Trust, staff involved tended to feel that ‘the PMRT is a lot of work for not very much insight’ (Fieldnotes, PMRT meeting, Site C, 2 September 2021). This perspective was notable at Site A, where staff learning forums had been significantly diminished in recent years (see later discussion). Site A risk midwives and clinical leads were particularly frustrated with the prescriptive design of the tool and the poor functionality of the web pages.

Across the three case-study services, responsibility for tool completion was differentiated between maternity and neonatal cases. However, at Sites A and C there was an ongoing lack of clarity over who was responsible for family involvement and for completion of these sections of the tool. The senior risk midwife at Site A reflected:

There’s a tension between where responsibility for the PMRT sits, in clinical governance or in bereavement. It feels like the bereavement team feel like it’s their thing, but they don’t really want it. But they’re pissed off when the clinical governance team get involved.

SASi1c

At Site C, where a recently appointed obstetrics lead for maternity quality and safety was especially keen to engage staff, it had been possible for the lead bereavement midwife to negotiate the PMRT responsibilities that had originally fallen to her. They explained that until recently,

I would complete the PMRT summary of care and then complete all sections of PMRT [but] now we’ve got a wider team to grade the care as it’s always weighed heavily on me to ensure I’m grading someone’s care.

SCSi04

At the same time, and most notable in neonatal care settings, senior medical staff questioned the effects of the tool on undermining their own clinical expertise:

There is no space for judgement about what questions you should be asking. You can’t actually get to anything important.

SASi06

Besides the ongoing tensions evoked over clinical responsibilities and judgement, the implementation of the PMRT was particularly challenging during 2020. During this year of exceptional pressures on staffing and maternity care, staff worried about care quality and how to reach judgements of care. Additionally, time for discussion with colleagues was often surrendered to meeting reporting timelines into the Trust and so guarantee financial benefit.

Clinicians’ frustrations with the PMRT at this time impacted on staff attitudes towards the representation of parents during these reviews. Of the panel meetings observed, the time and space given to considering parents’ concerns and questions were limited because of scheduling and difficulties with accessing the website pages for uploading information. There were parent questions that were not mentioned at all because the panel agreed that a bereavement midwife had already answered them. There were also questions considered to be so unfocused that an answer during the time and space available was not possible:

Ginny (the obstetric risk lead) notes that one question asked by the parent. Jane (the risk midwife) says: ‘it’s a very vague question which is “what happened?”’ The panel quickly decide that they will explain to her about a Strep B infection as that is what they believe the case of death was.

Fieldnotes, PMR panel, 2 August 2021

Panels tended to reduce explanations to parents into single clinical causalities (e.g. ‘it was a failing of the placenta’ or ‘it was a severe infection’). Overall, an indifference of clinicians to the consideration of baby deaths during the PMRT process was noted by researchers as they were often considered to be the responsibility of the bereavement midwife:

At the close of two PMRT meetings, both completed in less than 30 minutes, the risk midwife chairing the panels exclaimed ‘that was record speed! We can fit a few more in next time!’

Fieldnotes, PMRT meetings, 5 July 2021

In all, the PMRT had most significance for staff as a means of including families when it was used alongside responsive family updates and in-person debriefings where ‘you have that meeting people and interacting on a face-to-face human basis’ (SDSi12). One approach for improving a more empathic approach to families during the PMRT was to include a family representative onto the panels.

Other approaches to the representation of families

Explanations

Several participants considered the importance of a family’s search for explanations. Participants from clinical backgrounds often commented that they were sometimes unable to offer completeness and certainty to a family, and this could generate suspicion of their motives in the light of wider public expectations of medical expertise:

To the outside world there is always a definitive answer that something did or didn’t happen. There is no grey area, and it is this idea of us being infallible or able to discover some deeper truths. You have got clinicians being as transparent as they can be, but that transparency comes across as fudging it.

Qi14

Many participants also reflected on how multiple programmes and perspectives contributed to a family’s attempts to ‘uncover what really happened’ (Qi11). They noted situations where families had received ‘two reports with conflicting information’ (Qi11) from services, leading families to struggle to determine ‘what bit of all of that do you believe?’ (Qi6). The need for families to receive ‘one coherent report’ (Qi6) from a service was emphasised. However, this approach did presume that one ‘coherent answer’ could be gleaned from reports based on different investigation methodologies or viewpoints where questions and answers were framed differently (e.g. to explain the action or decisions leading to an incident or the contexts in which these actions or decisions occurred).

Healthcare Safety Investigation Branch leads viewed the challenge of multiple perspectives differently. They felt that the solution was to remind families of the purpose of each report:

Reminding the family all the time that this is about shared learning for cases in the future and not particularly about them is hard to hear when the investigation is about you, your baby and your incident.

Qi5

Several charity leads felt that this approach failed individual families when the particularity of their incident and its outcomes were marginalised as a result:

From a family perspective, these reports are not focusing closely on what should have happened in an individual’s treatment.

SG3

Participants working outside the health system also noted the tension between explanations of events focused on systems-learning and explanations focused on the affected family. For example, a coroner commented:

We have SI reports looking at standards of care and learning lessons, so where is the test of a root cause? Where is the balance of probabilities? A family may have an expectation that we will reach the same conclusion at inquest, but there are so many types of investigation all with different approaches. It’s a bit of a minefield. And can you imagine a family trying to navigate all this?

Qi6

This coroner’s view summed up a wider feeling that multiple investigation and reports, while constructed for different purposes, frustrated families’ attempts to come to an acceptable understanding of what happened. The ‘evidence’ that families relied on differed between reports constructed as learning documents and investigations of a case initiated to determine legal causality.

In summary, participants identified incident reporting as a significant process that epitomises the tensions between family expectations of OD and service offerings. A central dilemma identified across the interviews was that many families lacked confidence in their report findings and Trusts and services may be defensive of these findings due to the perceived reputation and litigation risk posed by injured families. Family misinterpretation of the purpose of reports was expected to contribute to this risk, as families often desired a coherent, linear and sometimes reductionist account of what happened. Participants indicated that organisational learning frameworks clashed with adversarial assumptions and legislative frameworks. This tension was compounded by investigation approaches that sought to involve families as contributors of learning for the future. A perceived danger of the system approach to learning was the potential marginalisation of families’ needs for acknowledgement and understanding of their case.

Review and investigation reports

Senior clinicians widely agreed that most families had three main questions to ask of an incident review or investigation: ‘what went wrong, was anybody at fault and how can we stop this happening again?’ (SASi04b). Those experienced in bereavement or trauma care also felt that these questions overlay a more fundamental need for many families for services and clinicians to fully recognise the impact of their injuries. CG and clinical leads were aware that a family’s receipt and discussion of their investigation or review report were critical times for them. Readings of the report, and discussions with debrief leads, were times when the integrity of the service from the perspective of the family were at stake. The final drafting and design of reports received by families often conveyed a sense that their own interests and experiences were marginal to service interests in limiting the sharing of knowledge with them.

Clinical governance teams rarely shared draft investigation reports with families, even though some families had raised complaints about this. The demands of additional administrative processes, along with the personal risks of sharing knowledge that was unapproved by corporate leads, conspired to prevent the sharing of this information with families:

Bits [of the draft report] go to the clinical director and Trust safety people and then they say ‘maybe don’t write that’ or ‘that isn’t right’ so do we send the report to families before or after this? We could end up getting into some legal hot water.

SASi01c

Instead, CG teams tended to focus attention on ensuring that families received a signed-off report that was a ‘good, finished product, the distilled thing that is the most useful thing’ (SASi01). A widespread sense of risk that some families might report staff or seek legal advice of knowledge acquired during the investigation process hindered family inclusion in the construction of this knowledge.

Service and family interests in the report

Additionally, the open question of who the report was for complicated the design of the report and the satisfaction of families with this document.

Clinical governance teams and clinical leads highlighted the purpose of a report as a response to the injured family’s search for answers and some also noted that the purpose of the report was to contribute to ‘learning that would benefit other women’ (SDSi02). Less often mentioned, but observed, was the use of incident reports as evidence in business case submissions for investment in infrastructural improvement. A CG midwife at Site B reflected on how they managed the different objectives of investigation and reporting:

So it’s not about the family, it’s more about Trust learning, but in all these cases I’ve always got the families in the back of my mind with thinking ‘I want to do this on your behalf’.

SDSi03

At Site A, the CG team emphasised that incident reviews and investigations were conducted in the interest of the injured family rather than for organisational learning:

'So, you are telling them that actually,' what you are struggling with here, we actually think is really quite serious. It is serious enough that we are going to galvanise this process, and I have come to talk to you about it because I don’t want this going on without you knowing about it, and you being at the absolute centre of it.

Risk Midwife, SASi1c

At Sites B and C, service and clinical leads were more likely to explain the benefits of an investigation for future events or other families:

I always say upfront ‘look, really sadly, we can’t change the outcome of what happened here but what we really need to do is to do all the learning we can to make services safer'. Most parents are really engaged in that.

SCSi14

Healthcare Safety Investigation Branch investigations, which required family consent for data sharing, were explained to families in more immediate, relational terms:

I always explain ‘If you have another baby, we want things to be safer moving forward. If your baby girl has a baby, we want her maternity care to be better, and we’ve got to look at that wider picture’. And I think most people are very receptive of that.

SDSi11

Service leads and risk midwives assumed the complicated interpretive tasks of making reports relevant to families. Sometimes, a ‘systems approach’ (SASi04) meant that clinicians felt family questions and concerns were ‘just personal to them’ (SASi04) or ‘actually don’t have a bearing on the outcome of the investigation at all but have to be heard at some point’ (SASi13). In other cases, particularly at Site A, where clinical teams had less confidence that interventions would lead to tangible improvements, CG teams and some clinical leads emphasised the importance of responding to individual family concerns.

Crafting reports and balancing interests

Irrespective of the ultimate purpose of the report, across all three services, clinicians felt that a report that give a clear explanation of what had happened, would help a family ‘be able to close that incident off a little better’ (SCSi08). A risk midwife commented:

Normally it’s quite a lengthy document and it all looks quite official, so generally families are quite pleased with what you have done even if they might not be pleased with what we’ve said as the outcome.

SASi1b

The production of the report was, in itself, an indication of care by the service. Teams described how reports should also generate action plans, thus contributing to organisational learning and in some cases committing the service to the completion of improvement requirements.

A difficult aspect of investigation report drafting described by CG midwives was the need to balance both family and service priorities. Where possible, reports were drafted to include any previously shared findings with ‘everything tied together so that they resemble one another’ (SCSi04) or ‘woven in’ (SASi08). This was to minimise variation in perspectives on the incident shared with families and to create consistency. Reports were also crafted for accessibility to families with understanding of their emotional situation:

[It’s important to make] sure that the language used is weighted correctly so that a family understand as best as we understand what the truth is behind the events. There’s a lot of guilt families carry about so when things aren’t black and white like when there’s uncertainty, we need to make sure that we say it in a measured way.

SDSi06b

However, from the perspective of families, their own information and experience were often marginalised, for example, added into the report as an addendum or excluded from this document:

If that’s something important [to a family] but not significant, then we verbalise that in the debrief meeting [and] put in a letter so that the parents are very clear that we’ve acknowledged whatever bit the report hasn’t addressed.

SDSi06

Exclusion of family information from the report also happened when there was a discrepancy between a family’s view of events and the risk scenario drawn from clinical records. For example, a clinical lead at Site A reflected on why a recent incident report was poorly received by a family whose understanding of events had been excluded from the report:

So, they raised things like ‘the doctor in the night’. I can’t know what the doctor in the night may have said, so this is a flaw in the process isn’t it? People often write in the notes their own interpretation of what the woman feels rather than what the woman felt.

SASi16

Site C developed a way for the experience of parents to be incorporated alongside the clinical notes; this process was informed by the inclusion of parents in investigation panels (as described in the previous chapter). The governance lead explained that even when some families had not submitted detailed accounts of the event from their perspective, there was a different tone to reports:

The report is better balanced and rather than saying ‘and then this is what happened to the woman and that is what happened to the woman’ we put ‘well actually the woman’s perspective was that she came in’. When you look at the clinical documentation and staff statements it doesn’t always reflect what the woman felt. So the way we’ve made our actions, it’s woman-centred because it’s about experience. So, yes, perhaps we have less questions and it’s less contentious and I think it’s more digestible when people give their own perspective and it’s fed into the report.

SCSi08

This participant describes an approach to report drafting that does not necessarily include greater depth or breadth of event information but offers a different way of framing the events described. She indicates the benefits of this approach to families, staff and services because the report produced is accessible and prioritises the narrative of family experience.

More generally across the case-study services, the management of knowledge reported extended more often extended beyond questions of accessibility and recognition of the family. Reports were more often about the management of different sets of interests with an awareness that the more the report identified, the more likely it was for a family to ‘give it to their lawyers’ (SCSi05).

For the four Site A families, receiving their incident report was a critical moment that determined their view of the service. For them, a key aspect of reporting was that their views, experiences and questions were included in the body of the report. These families interpreted omissions as a sign of the unwillingness of the service to learn from the event:

I made the point that [my baby] should have been brought out earlier and I don’t think there was a satisfactory answer for why it shouldn’t have happened so, for me, I was satisfied with the process and the apology but there can’t be a change in practice.

SAFi01

Family debriefs

In all case-study services, families were invited to meet with members of the CG team to discuss the findings of their incident investigation or report. These meetings were expected to alleviate underlying tensions that were likely to surface in the disclosure process. Debrief leads noted the importance of these discussions to bring ‘cohesiveness’ to a family’s understanding of events and to mitigate any growing suspicions that a service was not being entirely honest. Another purpose was for the service to attempt to acknowledge the situation of the injured family.

Meetings with families to discuss report findings was an onerous task for all but the most experienced CG and clinical leads. They explained that planning for timing, tone and content of debrief meetings was difficult:

And they are all so different. You can have some awful cases that are the nicest debriefs to do and you can have some that are awful. They are unique, there’s never a common theme to them all.

SCSi14

One experienced lead in family debriefs explained the principles he had developed himself, over time, to mitigate this:

So, I tell people that it’s their time and whatever is going to be most useful. We can go through the timeline, if they would like, or if they would like us to hear their story then we will listen to that and draw out the points as to whether we can learn something or whether there’s been a misconception of why something was done, if it’s helpful.

SASi20

The accommodation of family needs was sometimes difficult:

Before Lola’s family meeting, the HoM and clinical lead met to plan for ‘getting the meeting right for the family’. Timing was a consideration for both of them. They agreed on an hour to discuss the HSIB report on Lola’s HIE injuries, but ‘who knows they might have four questions or their might have 47 questions’. At the meeting, Lola’s mother said that she had ‘loads of questions with no answers yet’ and they said that a further meeting was possible to ‘cover everything’. Lola’s mum became very distressed, saying ‘we’ve already waited four-and-a-half months and things really need to be wrapped up now’.

Fieldnotes, Report Debriefing with Family, 6 July 2021

Power dynamics at meetings became more tangible when the clinical leads neglected to translate terminology used in the report:

Lola’s mother asks ‘and what’s macrosomia? I don’t even know what macrosomia is and I know for a fact that wasn’t mentioned to me before birth’. The clinical lead explains ‘macrosomia is just a word for a big baby. You were informed of the possibility of a big baby several times’. Lola’s mother snapped ‘big words shouldn’t be used because most people don’t know what they mean’.

Fieldnotes, Report Debriefing with Family, 6 July 21

Investigations conducted using interpretations of human factors methodology could generate further tension over the language used and the report priorities implied, exacerbating suspicion by families and defensiveness by clinicians:

The neonatologist sought to respond to Lola’s mother’s question of what happened to Lola just after she was born. She explained ‘it’s all done by human factors that are meant to help with the messiness of humans by asking like, how can we maintain safety alongside the humanness of interactions. Like making an incorrect call’. Lola’s mother asked, ‘so is human factors the panicking that staff were doing?’ She replied ‘yes, we work hard to standardise what happens in emergencies, but we can’t guarantee what will happen when people are under stress. So, those kinds of incidents don’t happen very often and we’re trying to train staff’. Lola’s mother looked confused and moved on to another question.

Fieldnotes 7 June 2021

This case study illustrates the difficulties for families of report debrief meetings where matters of timing, the use of specialist terminology or explanatory frameworks and the difficulties of challenging organisationally determined frames of reference.

For Lola’s mum, as for the four Site A families interviewed, debriefing conversations opened the opportunity for them to discuss issues with their care that fell beyond the interests of investigation panel and scope of the report. During debrief meetings observed and described, families often felt that investigators’ classification of events and experiences that families felt were connected to their incident as ‘incidental’ was symptomatic of a myopic or disingenuous attempt at addressing service shortcomings. In some meetings, where families discussed a completely different version of events to those in the report document, CG and risk leads sought to ‘unify and iron things out, just to maintain trust really’ (SASi16); however, it is not clear how far this approach satisfied families themselves. For example, Lola’s mother was expected to return several times more to continue her conversations with the clinical lead and other staff involved in her care.

Three of the four families interviewed from Site A had decided not to seek legal action because of the quality of post-incident communication. It was particularly their experience of the attention given to them and responsiveness to their needs in person-to-person debrief meetings, which led them to conclude that they would not seek legal action.

The background conditions and costs of successful family debriefs

Most generally, for clinicians, the success of the family debriefs rested on the acceptance of responsibility of the service for the incident. Some senior clinicians described the most difficult meetings to be those where no explanation of what caused the injury can be offered but ‘the apology is probably really valuable’ (SASi13).

There are things we don’t know, we don’t necessarily have an explanation for. And as one of the patients put it ‘don’t tell me it was just one of these things’. But sometimes it is one of these things, it is bad luck, you happen to be 1 : 10,000 and that is very, very difficult to accept.

SASi16

In these situations, families were more likely to become suspicious that a clinician or the service was hiding things. Both CG teams and families felt more satisfied when the service directly and immediately accepted responsibility:

The last debrief I was at they said ‘absolutely, 100% said this is our fault. Your baby is dead because of a fault in our administration processes. We are truly sorry. We will try to do everything we can, but this is nothing that you have done. We have failed you’. Now these are the kind of debriefs I can get behind.

SASAi1b

Several clinical leads agreed that if responsibility for the incident is clearly assumed by the service, families are less likely to hold individual staff accountable. For example, Fay’s family’s meeting with a clinical lead left the family unconvinced that the service had really addressed what mattered to them. For them, ‘real point of things, that she should have been brought out earlier was missed’ (SAFi1) and without this clarification they felt that the obstetric team who delayed the C-section were both ‘off the hook’ and ‘unlikely to change their practice’. All case-study families felt frustrated with family meetings and clinical debriefings that appeared to avoid a clear identification of who was accountable for what happened. However, questions of responsibility acceptance were more complicated in situations where parents’ own decisions were implicated in poor outcomes.

Senior clinicians who were expert in family debriefings were often resigned to withholding clinical judgements when family choices might have contributed to an incident. This was particularly so in services that actively promoted the value of women-centred care over risk management (i.e. at Sites B and C). 179,180 However, this position was more challenging for junior clinicians to sustain, particularly where they were themselves directly involved in incidents are perceived as connected to decisional autonomy that is unsupported by the obstetric or maternity team.

Listening services

All case-study services offered injured families some access to reflections or after-thoughts services. These services were described by a Professional Midwifery Advocate (PMA) as ‘about families feeling like we understand how they are feeling’ (SDSi08). Depending on the approach of individual clinicians as well as the family, these one-to-one sessions with clinicians were described as providing families with a clearer explanation of what happened to them, often by a careful reading of their notes with them, or as a ‘another opportunity for families to be able to voice how they felt, their perspectives, their concerns’ (SCSi06).

The PMA started the session by explaining to the woman that this was a chance for ‘me to listen to your story and to see if there are any areas where I can offer a perspective, either by filling in any gaps by looking at the medical notes or to explain things so you have a better understanding of why things happened the way they did’. The woman was silent and the PMA suggested ‘why don’t you explain to me your birth. Start at the beginning’.

Fieldnotes, Reflections Clinic, 24 September 2021

While some matters raised by families were seen as a more straightforward need for ‘filling in the gaps’ of what happened, often to more immediately relieve families of feelings of confusion or guilt, others were more complicated by families sharing experiences of poor care or questioning documented events. A Practice Development Midwife (PDM) explained:

People are unhappy and have been treated badly. I never want read the notes in too much detail because I want their story; you don’t really get that from the notes and I don’t want to go in with preconceived ideas of what happened.

SASi30b

The legitimisation of family’s experiences of injury in their own terms was central to the service, as was an approach to understanding events that were different to those offered by clinical or incident explanatory frameworks. Service managers noted the success of these services in reducing informal complaints; however, percentage changes were difficult for them to calculate due to data collection methods in that department.

The organisation of the service differed across the case-study services (see Appendix 2) At Site A, the service was always oversubscribed with families arriving by self-referral or matron referral following a complaint. At Sites B and C, referral into this multidisciplinary pathway for injured families was more responsive, with therapeutic approaches determined by family need. As noted in Chapter 6, service managers felt that care through this pathway was important for families who wanted to attend their own incident investigation meetings. Across the services, according to staff there was high demand for the service even though these services were not always publicised to families themselves.

Unplanned opportunities to make sense of events

In home birth settings, where relationships extended beyond incident management or a hospital stay, some families and staff found less-formal ways of making sense of incidents.

A few days after the emergency admission of a baby following a home delivery, the mother contacted the preceptor midwife present at the birth with baby photos and updates. She felt a ‘special connection’ with this midwife. The midwife replied to the mother’s invitation to meet in the hospital cafe. The midwife described: ‘she did ask me “was there anything wrong with my birth?”. I did say to her ‘No I was as surprised as you were until the baby came out it all looked perfect and that I didn’t have a lot of experience but from my perspective, everything sounded perfect’. Then her husband said, ‘yes, everything was perfect, but she blamed herself’.

SDSi17

Most service and clinical managers from hospital settings were wary of these informal meetings during which families might search for explanation and some resolution. However, several staff recalled the importance to them of meeting a family months or years after an incident as both found comfort in hearing one another’s perspectives. Where families remained feeling unacknowledged and unheard by the service, staff often found these accidental meetings ‘very terrifying because it was just me and my face was the face that they remembered’ (SCSi08).

Disclosure training

Few senior stakeholders who also worked clinically had any formal communication skills training, despite several leading on debriefing work with families in their service. They described how they, like others, had ‘just learned on the job’ (FS1) and ‘got wheeled out because [I’m] quite good at talking to families’ (Qi13). Five participants were actively involved in developing pilot training for legal firms, national charities or professional colleges.

Eight participants commented on the lack of central investment in training for OD. Bursary and sponsorship opportunities were available in some services (FS5). Although attending an investigation panel and a family debrief is part of the obstetrician training curriculum, clinicians more often prioritised updates of technical skills and funding for midwives or allied maternity care professionals was also very limited (FS5; ML5). Interest in training approaches to prepare for less defensive OD varied between maternity services and between individual professionals. Most staff had a brief induction session by a Trust legal representative on meeting the statutory DoC with advice on minimisation of professional and organisational litigation risk:

You get told how you can handle this in a way that best protects yourself from unwanted attention from the CQC, litigation or other unwanted outcomes.

SG3

The lack of NHS training investment in communication with families was contrasted to the HSIB maternity safety programme that was designed around ‘what investigators on a daily basis cope with’ (Qi8). Investigator training and support ranged from introductory workshops on ‘steps’ to involvement over the investigation time frame (first phone call, home visits and closing relationships after investigation reporting) to bespoke sessions on resilience, bereavement and trauma care. Routine regional team meetings, ongoing peer support and expert telephone advice on family engagement issues were described as ‘a must for our wellbeing’ (Qi5):

There is a lot we take on like sometimes families who aren’t coping and the dynamics of the families. Lots of tears and support and open talking within our teams.

Qi5

The connection between HSIB’s investment in training of their investigation teams and the better experience of families was noted by many participants. Overall, discussions about disclosure training noted the comparative underinvestment in disclosure training in services and by professional bodies. Participants also emphasised the value of acquired practical knowledge over training programmes. Training offered by national charities was welcome but had not yet been evaluated for immediate and longer-term outcomes.

The care of staff post incident

Surprisingly, few national stakeholders identified post-incident support for staff as important to families, services or staff themselves. Those who identified this need were more often former or practising clinicians who drew on their own personal experiences of post-incident trauma and approaches to recovery. Two other participants who considered this were from mediation rather than clinical or safety improvement backgrounds. They noted the psychosocial challenges of disclosure ‘when you take it as personal failure’ (Qi1) and ‘need to start acknowledging the impact of incidents or you get stuck’ (Qi19). Five safety improvement leads discussed the importance of ‘fair’ or ‘open’ workplace culture for supporting OD; these discussions were of general principles and their implications for practice. It is significant that the support needs of clinicians conducting OD were widely overlooked. It may be related to a wider ethos amongst some staff in NHS maternity units where discussion of post-incident support needs is unsupported by many colleagues in this speciality. 181,182

Clinician care and disclosure: workplace features and variations

However, the clinical fellows recruited from our national study highlighted the relationship between local unit practices of staff support as well as ongoing mentoring and the quality of their post-incident communication with families. Significantly, they did not identify connections between formal systems of post-incident staff support and their behaviours towards families.

Several of these participants reflected on the impact of the first moments after the incident on their subsequent conversation with the affected family. The immediate reactions of colleagues, particularly of senior colleagues or managers, were described as critical for providing non-judgemental support to trainees. This, in turn, influenced how comfortable the trainee was to share what had happened with the service. One trainee described this:

If I do a Cat 1 section at 02:00 in the morning and that baby looks like they’re going to have brain damage, if I know that at 8 a.m. when I discuss that it’s going to be ‘Are you okay? Do you want to talk about it? Have you spoken to the mum about it?’. Whereas if you think ‘I know so-and-so is on in the morning and they’re going to be … Oh no, you’ve sent a term baby to the Neonatal Intensive Care Unit (NICU)’ then you’ve got that kind of reticence. You’re not going to go into it as much. You’re going to keep things a bit to yourself because you’re scared of being judged. If you can’t have conversations like ‘I feel like I’ve done something wrong’ or ‘maybe I should have done that quicker’ with your team, then you’re not going to have them with the patient.

Qi7

This description highlights the impact of the attitudes and behaviours of immediate colleagues on the ways that incidents are discussed, and initial OD might unfold.

Opportunities to observe differences in the behaviours and skills of clinical shift leads as they communicated with a family post injury was also important. Learning on the job by informal mentoring, often by senior colleagues and bereavement midwives, was envisaged as a gradual building of the necessary expertise for meeting with families after traumatic events and situations. One senior trainee described:

You see things that you think are wonderful, that person has just told somebody the worst imaginable thing that you can hear and in an incredible way. I want to be like that or equally, I never want to be like that.

Qi10

These learning experiences were recognised by trainees and clinical leads as essential to developing the embodied skills of reflexivity and being attuned to a family’s situation. However, clinical fellows agreed that there were declining opportunities to observe post-incident conversations with families due to lack of time or a reluctance of senior colleagues to invite them.

Clinical fellows also identified leadership in CG as a significant factor affecting the possibility of their post-incident communication with families. Where CG was more risk adverse, these relationships were actively discouraged. One experienced trainee described the effects of a CG approach that ‘disempowers patient-facing clinicians’ who feel ‘too junior to say sorry’ because it has become ‘only something matron does …’ (FS2). She continued:

We work intimately, with lots of honesty and openness with families then when something goes wrong it’s like everyone is on the defensive and then it’s something for governance. That’s quite harmful from both sides when someone says, ‘No you can’t go and see her and her baby’ or ‘if they ask questions just tell them to talk to so-and-so.’ It entrenches the thing. Then the parents will say ‘nobody spoke to us’.

FS2

This trainee, and other clinicians, indicated their interest in seeking some early resolution with families and the reluctance of many services to facilitate these opportunities. However, clinical fellows recruited for our national interview study described that while these meetings might be organised at the request of a family, service or clinical leads lacked the protected time or mediation skills to steer these events. Two clinical fellows described ‘disastrous meetings’ with families. One became a ‘shouting match’ (FS2) with a family when all attendees were ill prepared for the meeting and the Chair had been called away. Both fellows were now determined to revise attitudes towards incidents and OD.

In summary, the post-incident support experiences of clinical fellows (both obstetricians and midwives) were highly varied across services, units and shifts. In the absence of any post-incident support from immediate colleagues and managers, these staff looked to partners and friends, particularly their former student peers, for care and advice.

‘Putting out the flames’

During the case-study research (April–November 2021), there was a growing public awareness of the importance of staff well-being in NHS services. This was reflected in interviews with senior and specialist clinicians in the case-study services where, despite some recent investment in staff well-being interventions, many participants highlighted the inadequacies of the post-incident care of front-line clinicians. A series of factors contributed to this. Most immediately, there was limited time available for post-incident support and recovery:

When everybody’s go, go, go, you feel like you’re just putting out flames.

Matron for Quality and Safety, SASi17b

We just make them a cup of tea, we do a hot debrief, we rub them off and send them back off out into the battlefield, let’s be truthful.

Director of Midwifery, SCSi16

I think as a Trust we are not very good at it. Not good at it at all.

Clinical Lead, SDSi19

Particularly at Site A, senior clinicians felt that service attrition and sickness rates were directly related to the lack of post-incident support of clinicians:

Staff used to think that when if anything happened, they would be 100% supported and it would be fine. Now I believe that staff don’t feel that and that’s why they think twice about coming to work unless they feel completely right.

PMA, SASi22

Reflections on the inadequacy of post-incident support for staff were also shaped by a sense of wider workforce neglect and demoralisation. Some clinicians at all case-study services described feeling ‘disempowered’ or ‘defensive’ in the face of corporate systems that managed the complaints of families who assume poor clinical outcomes or experiences of care centred on the behaviours of single individuals (SASi36; SDSi19). Some front-line clinicians felt ‘forgotten’ or ‘neglected’ by their immediate managers or by their service more widely.

Several clinical leads noted the need for individualised post-incident support for clinicians that would include the possibility of ongoing care by the service and the commissioning of specialist support. This approach relied on the provision of collectively agreed guidance on post-incident support. At the same time, some clinical leads were wary that such provision was a salve for more fundamental issues of staff welfare, while post-incident workplace response was the more pressing issue to be addressed.

Workforce trust, flux and diversity

In the case-study services, we examined the formal resources and interventions for post-incident staff support with service, clinical and business managers. Our detailed interview and observational findings on the organisation and reception of post-incident support for staff are presented in Appendix 7, Table 9.

We note that all case-study services invested in formal post-incident staff support structures (ranging from weekend coverage of PMAs, who provided personal and educational for midwives and sometimes doctors; PMA group and individual reflection sessions, Trust counselling services and, at Site A, a dedicated service psychologist). Despite this, front-line staff described a lack of both formal and informal support resources during our fieldwork and interviews. This may have been because of a lack of available time for staff to access formal resources as well as various structural and cultural factors undermining the embedding of the resources in the case-study services. For senior clinicians, in particular, informal arrangements of ‘one-off’ informal peer support, occurred between team members sharing the same office space. For example, a senior clinical lead at Site A described the common practice of ‘talking to office colleagues and then literally closing the book and hoping the next time is easier’ (SASi12) as her way of coping post incident.

The puzzle for post-incident support specialists was the tailoring of post-incident support so that staff could reach it at critical times (which were frequently out of office hours). This was complicated by patterns of part-time working, night work and long shifts, as well as by a significant workforce turnover in Sites A and C. Several PMAs and Consultant Midwives were also aware that as long-standing, senior employees, they may not be immediately trusted by junior staff. Specialist leads and advisors at all case-study services reflected on the ongoing difficulties of engaging with many midwives who ‘get anxious because they think they are going to be blamed for something’ (SDSi08). At the same time, trainees interviewed and observed at the case-study services had highly varied experiences of support from educational supervisors.

Some senior medical staff were not keen to encourage junior colleagues to seek MDT post-incident support and preferred to have issues discussed only in trainee meetings. While PMAs were available for them at some services, obstetric trainees rarely made use of this resource. Indeed, 8 months after Trust commissioning of a full-time staff psychologist at Site A, there was widespread reluctance of medical staff to make use of scheduled group sessions (e.g. breaking bad news and self-care after events) or individual appointments (SASi35). Some managers also noted the shortcomings of commissioned post-incident support for diverse staff groups that have different attitudes to counselling and psychological services.

In all, the fluidity and diversity of clinical teams, workforce churn and protectiveness over professional identities contributed to the difficulties of embedding support interventions in practice. Additionally, access and use of these resources and interventions often relied on workplace histories where individuals had the time and opportunity to develop an ongoing trusting and confidential relationship with a clinical specialist. Our interviews with junior clinicians at the case-study services (n = 6) highlighted important parallels between the post-incident needs of families and of junior front-line staff.

The immediate impact of an incident on clinicians directly involved

When incidents were out of hours, and the mother was conscious, junior clinicians were often the first staff to engage with the family immediately after the event. In these situations, senior trainees led the initial conversation with the woman and her family. Although two junior staff had completed various training courses on conducting the DoC and its implications for staff, both still felt poorly prepared and anxious about these encounters with families. One trainee described this first conversation as ‘an aspect of the fear of a serious incident’ (SASi19).

As with clinical fellows recruited in our national study, front-line post-incident staff support in the case-study services was described as highly varied, and changeable from shift to shift depending on floor or shift leads. As also indicated in our national study, better experiences of post-incident support were described as having significant more immediate effects for the family, and longer-term effects for the development of clinician confidence and skills.

For example, Roberta, a senior trainee, described the effects of immediate and unjudgemental collegial support at Site B. Immediately after an intrapartum stillbirth on the unit where Roberta was shift lead, her senior colleague from another area came over. This colleague ‘stood right next to me’ (SDSi18) during the very difficult and initial conversation with the distraught father. This colleague was with Roberta again, the following night, when the baby’s mother wanted to meet only with her:

Like I know it wasn’t down to me but when I arrived the following night I was physically shaking. I was still really upset myself. So, I got the mother’s permission to talk to her with my colleague. Then we spent about 40, 45 minutes talking with the family. She was present for support for me and to answer any questions I couldn’t. Then the next four nights I had more conversations [with the family] to answer any questions they thought of.

SASi18

These examples highlight the importance and impact of senior support for junior members of staff, both for their own well-being, the development of post-incident care of a family and for the informal acquisition of OD skills.

However, in all case-study services, junior doctors considered that, because they lacked a formalised ‘risk or governance brief’, they struggled to negotiate time to adequately prepare, both emotionally and practically, for initial and subsequent family meetings. The importance of ‘being given at that very time the headspace for a considerate response to a family’ and ‘to start to “absorb”’ (SASi27) or ‘process’ (SASi19) the event was expected to be crucial for everyone concerned. This relied on, to some extent, an immediate staffing capacity during a shift but also on senior colleagues recognising this as important.

There were significant differences in post-incident leadership support between the case-study services. At Site A, it was often felt that ‘we don’t have that luxury’ (SASi01), whereas, at Sites B and C, senior service managers sometimes ‘ended up staying there and being with them or sending them home when they are safe to leave’ (SCSi16).

Sometimes, the fast-paced environment of the maternity unit was not conducive to discussing events with an injured family. In these instances, the regret for not meeting with that family lingered with junior clinicians, sometimes for years. For example, an experienced trainee senior described how 3 years after an event she still experienced enduring regret for being unable to return to speak to a woman whose delivery had led to severe harm to the baby and who was transferred, rapidly and unexpectedly, to another unit.

When junior clinicians were directly involved with a traumatic incident, they experienced initial and ongoing emotions of self-blame, regret, self-doubt and isolation. These initial feelings ‘weighed heavy’ (SASi28), irrespective of issues of avoidability. This was also the case where junior clinicians felt that families had opted for care that contradicted clinical advice. Being offered some time to begin some recovery was highly valued, often as a sign that their distress had been acknowledged by senior colleagues.

Ewa, a Band 7 midwife, was involved in a difficult neonatal resuscitation while covering a colleague’s lunch break. Two years later she remembered:

So, when I finished the notes then it hit me as a really big deal because it’s a baby’s life, like you know for a fact it’s going to be severely disabled. The second I walked out the room, the midwife-in-charge just said ‘Oh well, your other patient needs you’ so I just went back to my unit. It would have been nice for someone to say ‘just go and get a cup of tea and I’ll deal with your patient, just recognizing what had happened and that I might have needed five minutes to collect myself. It was that desensitized.

SASi28

From Ewa’s perspective, the failure of her immediate manager to acknowledge the impact of the incident on her came to summarise the attitude of the wider service. This experience, and the neglect of the service to update her on what happened to the baby, contributed to her decision to leave the Trust the following year.

Particularly for more junior clinicians, there was also an important interplay between the acquisition of disclosure skills and immediate post-incident support for junior clinicians by practising experts. For example, Roberta’s support by a senior colleague, described above, was also an opportunity to observe and then develop disclosure skills. Apprenticeship learning, involving graduations of shared practice, focuses on the needs of one particular family ‘as a locus for embodied work practice rather than a Site of contemplation’. 183 Several clinical leads described significant moments in their own careers where they had worked out, from memorable events of tangible success or disaster when interacting with a family, certain tacit points of convention or ‘rules of thumb’ for OD. For example, a clinical and debriefing lead at Site A described the impact of a SI and his acquisition of OD expertise:

Years ago, I had a hideous case where someone died, and I made an error. I went through a lot of emotions. The single most important thing I did was to immediately talk about it to the family so that I wasn’t going down to them the next day. It was a roller coaster. It might be why I approached risk job afterwards. I felt very strongly that it was the speed at which I told them the truth.

SASi20

This clinical lead occasionally shared this experiential learning with staff in his team. He implied that this might encourage better OD. Opportunities for this learning were under considerable strain during the period of fieldwork. Learning-in-practice and through story telling relied on the ‘in-the-moment’ availability and accessibility of expert clinicians. Also, senior clinicians and specialists who had not previously met a family were understandably cautious of exposing them to any unnecessary attention from additional, unknown staff. As often, however, expert clinicians tended to overlook OD conversations as clinical learning opportunities for a trainee or midwife already involved in that families’ care.

Notably, very few CG teams or lead clinicians from the case-study services had attended professional or in-service training on disclosure skills. As far as the research team could establish, in-service staff inductions into the DoC at the case-study services were conducted by representatives from the Trust legal team, rather than by clinicians who were expert in this practice.

Clinicians’ experiences of reviews and investigations of incidents involving them

Junior clinicians across all the case-study services often felt that their involvement in internal incident review and investigation processes was poorly considered or planned by CG teams. Many felt that after initial support from senior colleagues, they were treated with less sensitivity as time went on (e.g. being told of a HSIB investigation in front of colleagues, being texted for information about a family complaint on days off and having to complete time-consuming and emotional written reports in their own time). For some trainees, distress was compounded by requests for statements for inquests arriving years after an incident that had happened on a previous rotation. With variable support from senior staff in their previous service, they often completed these alone and without professional advice. While trainees worried about the consequences of investigations as well as potential legal action by a family because of emotional strain and impact on career, they also felt it a necessary recourse for families in need of financial support after injury. Not knowing how a review or investigation was progressing left them feeling confused and anxious. Lack of follow-up after an investigation has been reported, was also a problem, with one midwife at Site B describing how:

You just get an e-mail from management asking for a statement and nothing else comes of it.

Fieldnotes, shift observations, 21 September 2021

In situations where front-line staff felt ignored or misrepresented in internal investigation processes or report findings, they often felt frustration and resentment towards families whose perspective was (they assumed) better represented. A CG lead confirmed this sense of neglect of front-line clinicians post incident:

We go back to them as individuals ‘are you Ok, do you need anything?’ and then there’s the investigation and there’s learning and I don’t think we actually come back together with them.

SCSi06

Most interviewees had not received final reports of investigations involving them and those who did felt that an e-mailed report without opportunity for further discussion did not help their understanding of events or their personal recovery. The interplay between post-incident learning and the care of staff post incident was complex. It was approached differently across the case-study services.

Staff debriefs

Immediate and longer-term post-incident team debriefs were the most common way that services supported all staff directly involved in the incident to begin to make sense of events. However, as indicated in Appendix 7, Table 9, immediate (end-of-shift) debriefs were initiated at the discretion of shift leads and might be conducted for different purposes (e.g. reflecting on immediate feelings, composing a clinical picture and advising on investigation processes). No one mentioned immediate debriefs as an important aspect of their post-incident experience. However, some clinicians did discuss later post-event debriefs as important for their recovery of feelings of guilt and isolation. For example, a few weeks after the incident involving Ewa, a MDT meeting was organised by a risk midwife. Several senior clinicians involved in the neonatal resuscitation attended and the service psychologist facilitated the discussion. Ewa described the positive effects of this meeting on her self-esteem and personal recovery:

I realised I wasn’t the only one blown away by how unexpected it was. It reassured me that as a relatively junior member of such a team, I wasn’t the only one who didn’t know what to do. That you just do as much as you can do.

SASi28

Clinical leads in the case-study services (who were senior doctors) often remarked on the difficulties of organising these debriefs. Identifying and reassembling the team involved in the incident (including those who had witnessed events) was a ‘major headache’ (SASi13) when staff came from different service areas, had different work patterns and differing seniority. The impetus to organise these events could diminish as collective memory of the incident seemed to fade. However, several junior clinicians directly affected by incidents described these meetings as invaluable to their recovery post incident.

Meeting with families

As already discussed, in some circumstances, it was possible for injured families to contact clinicians involved in their incident. Some clinical leads identified the valuable potential of these encounters for clinicians and families. In hospital births, especially where families were quickly discharged or transferred, these meetings were rarely suggested or initiated. However, some service managers and midwives found ways of reconnecting with families, on the advice of colleagues, when families were re-admitted to the unit in later pregnancies. Pamela, a Band 6 midwife, described spotting the name of a woman on an admissions board ‘a name I’ll never forget’, as she felt responsible for this family’s intrauterine death (IUD) several years earlier:

We had an opportunity to chat then, and very interesting to see how she perceived the event. I certainly did feel a huge amount of guilt and sadness. She didn’t harbour any feelings of anger or sadness towards us. That was my closure, after all these years that was my closure.

SASi27

The possibility for facilitated meetings that might contribute towards resolution, for families as well as for staff, was considered by some service and governance managers. However, it was generally felt that the unpredictability of these meetings, with possibilities of further trauma for staff and for families, and risks for the service if legal action was progressed, meant that they were unlikely to be formally encouraged.

Personal and service investments in post-incident learning

The organisation and processes of post-incident learning in the case-study services are examined in detail in Chapter 9. However, from the perspective of more junior front-line trainees as well as midwives, this service work had limited reach to them. Interviewees had little knowledge of changes being implemented because of an investigation of an incident they had been involved in. However, most felt a personal responsibility to make changes to their own or others’ clinical practice, either by learning more about the clinical aspects of the adverse event (with three becoming recognised specialists and teachers in that area of practice) or by becoming involved in related QI programmes risk governance (three clinicians). A senior obstetric trainee described the aftermath of a disastrous delivery that resulted in the admission of a baby to NICU. A delayed investigation of the incident (that happened outside of the three case-study services), generated considerable personal concern about her own safety as a clinician.

It all makes you very apprehensive. It was such an awful thing and I wanted to leave medicine at that time. Then I did a lot of my own research. These skull fractures do happen with some vaginal deliveries and perhaps it’s a little bit like a lottery. But it’s very hard to change your practice when you don’t know what happened.

SASi29

The significance of post-incident learning for both future patient safety and ongoing personal recovery was highlighted by trainees. As one trainee expressed it: ‘when you look at it more clinically and less emotionally, you start to realise that you did your best’ (SASi19). This realisation was, for all participants, a necessary step towards the rebuilding of professional confidence.

The support needs of governance teams and clinical leads

The emotion work and support needs of clinicians less directly involved in incidents but communicating with injured families and distressed staff on a regular basis were rarely directly considered by participants.

Particularly for risk governance teams and HoMs, office spaces were the usual forum for listening, learning, encouragement and skills development between colleagues.

The researcher walked into the offices of the risk team just as there was a round of applause for the HoM by everyone present. They explained ‘she had just managed to talk a really upset family down from a formal complaint, it was so well done!’

Fieldnotes, office observations, 11 July 2021

Particularly at Sites B and C, mentoring support for challenging conversations extended beyond the HoM to senior managers, where the escalation of family concerns was identified as important for families as well as for service leads (senior midwives) themselves. A Director of Nursing explained:

These need to be dealt with at a senior level and when they go wrong, they usually haven’t been. I absolutely want sight on all of them. My Director of Midwifery just did a second debrief of a family that had already contacted me, and I said: ‘Look, I’m aware you’ve got a debrief meeting on Wednesday. Have that meeting and let’s catch up after to see how I can support,’ like follow-up.

SCSi14

These mentoring systems were reported to be highly valued by distressed families as well as by staff responsible for communicating with them.

Anticipating the needs of families

Most stakeholders commented on the overlapping interests of families and services as both wanted to avoid the reoccurrence of similar incidents. One charity lead identified the importance for everyone involved in the incident for services to:

Act on those moments where you go ‘Oh, if only something different had happened there then things would have been different’.

SG4

However, some participants were cautious of assuming an automatic alignment of interests between services, staff and injured families. For instance, clinicians who were trauma and bereavement specialists commented on the need to prioritise and give families time to grieve before involving some in service improvement agendas. It was also recognised that families’ interests in sharing their experience for the benefits of a service was one of many approaches they might take to make sense of what had happened to them.

Furthermore, several participants from clinical backgrounds were cautious that families might be promised changes in action plans that either were difficult to achieve or had marginal effects on underlying systems:

I’ve heard it all the time. They want to make changes so that it doesn’t happen again to someone else. Having a report that says we’ve done things might appease them so that feel that. Whether that is truly making a difference? It’s a fine balance between how do you do an investigation so that people get their answers at the same time as having a wider, deeper level investigation that is going to make a difference? Can you do both or just one?

FS1

Discussions like this centred on concerns about how the focus and scope of both report recommendations and action plans were negotiated to meet the interests of both services and families.

The positioning of families in service improvement work

For the most part, participants’ reflections on organisational learning and safety improvement after incidents rarely involved an expectation of families actively participating in informing action plans or holding services to account for implementing them.

The national stakeholders who reflected on the question of how service changes were demonstrated to families and staff identified several key considerations. First, participants noted the potential for alignment of family and service priorities for organisational learning. They described the importance of identifying and acting on ‘pinch points’, or the critical loci where learning can be translated into significant service improvement to impact outcomes. Finally, some national stakeholders suggested the need for a rethinking of improvement approaches, recognising the potential to involve families as active participants to drive improvement, rather than as passive recipients of improvement efforts.

Implementing actions plans for safety improvement (April–November 2020)

The period of fieldwork was a time when staff were ‘just about getting by, from shift to shift’ (Site C clinical lead, Fieldnotes, 23 July 2022). Across all case-study services, there was a sense that organisational learning post incident and communication of this learning to harmed families were particularly inadequate. Also, clinical and service leads discussed two longer standing difficulties with this work. These were, first, how to balance the desire for ‘quick wins’ with longer-term change in a service. The pace of change was a concern for some injured families as well as front-line staff and, second, how to communicate ongoing changes to staff.

Site A’s CG lead was especially outspoken about the undermining effects of extensive reporting requirements of Trust and external QI programmes.

They, along with many senior clinicians, felt that required action for learning from incidents had become an exercise in the pragmatics of demonstrable success organised around service risk registers rather than a response to clinical realities. 187

At Site A, staff were frustrated that learning from incidents simply bypassed some service areas and front-line staff. A risk midwife commented:

There’s lot of learning that needs to be known more widely. Like for junior doctors and labour ward. It’s really important for them to see the same issues come up time and time again.

SASi01

Improvement leads at Sites B and C were also tired of assurance and audit that did not seem to contribute to any new learning or service change. Site C’s CG lead remarked:

I feel like learning could be better. I feel like it all gets fed upwards, and you go to meetings, and you explain what actions you’re going to take and why but actually it needs to trickle back down to staff.

SCSi08

However, reasons for disappointment at the pace and spread of service changes post incident were significantly different across the case-study services.

At Site A, where safety improvement was less embedded in regular service business, many staff interviewed felt demoralised by the lack of investment in learning opportunities and in ongoing safety improvements. At Sites B and C, where clinical leads were more actively engaged in Local Maternity and Neonatal System (LMNS) and service-level improvement projects and ongoing initiatives, staff were more often frustrated by the technical and political complexity of securing changes to systems [e.g. revisions in care pathways or information technology (IT) systems]. At Sites B and C, many clinical leads were invested in professional and service development work (e.g. site selection for national pilots; leadership in LMNS projects, specialist training initiatives and secondments into national schemes). These investments were evident in service-level expertise, commitment to and ownership of quality and safety improvement processes along with knowledge of organisational change.

‘Quicker wins’ and system changes

Across the case-study sites, post-incident action plans were constructed pragmatically, with an eye to service capacity and ambitions, as well as to the interests of injured families who might want to know whether changes highlighted from their own incident investigations had happened.

‘Quicker wins’ or local changes that did not require inclusion on the risk register or a business case could sometimes be identified as a way of demonstrating safety improvements to families and front-line staff as well as meeting requirement for local audit. However, there was understandable caution that action plans drafted for more rapid completion were not always the most beneficial for managing ongoing safety concerns. A clinical lead commented:

We change the guideline, share message of the day, put individual training in place, do an e-learning module, we do all these things to share learning. We will get 10 to 20 people to alter their behaviour but that won’t alter the system. And that, I think, is the challenge.

SDSi06a

Changes to the physical environment were often expected to be the most straightforward to implement. However, a relatively straightforward change to the workplace could be difficult and time-consuming to secure, as the changes required the support of business cases and external recommendations before they were taken seriously:

Getting a second printer for Maternity Admissions triage should have been simple stuff so that the midwife did not have to leave triage and walk 30 paces to another printer. When we got an actual anonymised external report, then all the things we’d been saying, when it was in the report, the changes happened overnight because some external came in an explained it.

Clinical Lead, SDSi06a

Indeed, action plans established from incident report recommendations often involved complex as well as protracted revisions across a service.

So, we met with the family whose baby died 18 months ago and that was an HSIB investigation with recommendations for us. We’ve got traction on quite a few of them, but not all are completely completed. Some things are big: antenatal storage of CTGs requires a business case, capital investment, finance committee investment, then you have to tender, then you have to buy the equipment …

Director of Midwifery, SCSi16

Changes to care pathways were particularly complex and time-consuming to develop and embed in practice. Site C’s Clinical Director explained what was involved in the 2-year project of revising the ‘Induction of labour pathway’ that involved changes in guidelines and working practices, staff attitudes and ongoing improvement cycles. Senior clinicians at Sites B and C sometimes reflected on how best to balance the ambitions of service improvement and the completion of action plans. The clinical director explained:

We need to close on actions. Making sure these actions are caught when some people were not aware an action had their name on. But they all have to be realistic and with a reasonable timescale to deliver it but seeing impact on the shop floor.

SCSi03

This clinical lead was commenting on previous situations where clinicians had been overly ambitious in drafting action plans, identifying service changes that were out of reach for improvement teams. She felt that this lack of change would be difficult to justify to a family. She noted that overambitious post-investigation action plans could also alienate staff and showed in failing audits. In this service, there has been some initial discussions to involve some families with insights into the demands of this work in the prioritisation of identified safety improvement work streams.

Communicating changes to front-line clinicians

Many senior clinicians involved in safety improvement highlighted a tension between the demands of ‘upward reporting’, auditing and assurance to service and Trust committees and the felt need to communicate ‘down to the shop floor’ (SASi17a; SDSi12) and to injured families. While CG teams and clinical leads recognised the importance of involving front-line staff in learning and service change, they also stressed the difficulties of engaging highly fluid teams in multiple units and across geographically dispersed parts of a service.

Appendix 8, Table 10 documents service communication approaches in the three services and how they were received by different clinicians working in the three case-study services.

In brief, we found that none of the services were able to implement a post-incident communication strategy to reach front-line clinicians. Most notably, the effectiveness of newsletters, bulletins and e-mail updates for sharing learning was questionable, with messages lost within other correspondence; e-mail overload and difficulties with accessing IT systems for staff e-mail updates during working hours. Special learning events, dedicated safety weeks and invitations to reviews panels often missed staff working shifts on busy units or out of hours. We found that more organic informal networks of learning188,189 were more accessible and enjoyable for staff. Multidisciplinary team learning meetings and informal outreach activities were reported as the most effective approaches to knowledge sharing. As discussed below, these approaches were effective for influencing the attitudes of junior clinicians towards families post incident as well as for sustaining the work of CG teams and clinical leads.

Particularly at Site A, the CG team was acutely aware of the tense division between their team and front-line clinicians affected by incidents. Risk midwives tried various approaches to improve awareness of their work and foster better relationships. For example, they networked to ‘tempt people to risk meetings’ and invited Band 6 midwives to the PMRT and SI review meetings to ‘take up any learning and spread an understanding by word of mouth of what we do far more effectively than we can’ (SASi01). They composed letters of reassurance to staff before HSIB interviews, ran ‘tea and cake trolleys’ to update staff on learning from investigations and worked some clinical shifts ‘so that people don’t say how can you judge my clinical care’ (SASi04b). Overall, the clinical risk team were aware that without service-learning forums, their engagement with clinical teams was limited:

When I work clinically, I get asked all the time ‘Whatever happened to that woman who had so-and-so?’ and I’m like ‘it’s this, and this,’ and everybody’s interested. I just feel that we have got this huge gap here. Because you experience these things, you see all these things, but the circle doesn’t close for you because you don’t know.

SASi01a

Another aspect of this informal outreach work of risk midwives, some nurse consultants and PMAs was post-incident support of some midwives and, occasionally, junior doctors (see also Chapter 8).

Most senior clinicians at Sites B and C described the services where they worked as places of learning for quality and safety improvement. Weekly MDT learning meetings were the tangible expression of this organisational focus and ethos. Significantly, detailed case studies of SIs that had happened in the service, which were narrated from precipitating events to learning and change outcomes, were included amongst other learning topics. The inclusion of these case studies normalised discussion of SIs as events of learning and collective reflection:

During a weekly zoom learning meeting 30 staff discussed a recent SI and its investigation. The clinical lead and some staff from the unit where the incident happened were at the meeting. Discussion initially focused on flaws in tools and systems until the Chair asked who admitted the woman. The clinical lead replied it was midwifery triage, adding ‘we try our best under immense pressure. We manage the best we can. Maternity Assessment Services in all units are under such high risk at the moment.’ The Chair quickly added ‘yes, yes, there is city-wide action on awareness of these problems at the moment. It’s just not manageable.’

Fieldnotes, Weekly Multi-Disciplinary Learning Forum, 4 August 2021

As this example shows, case-study discussions took account of wider service conditions beyond individual clinical scenarios. Particularly where staff involved were present, there were frank discussions about service pressures and limitations. These discussions also drew attention away from more immediate assumptions about individual clinical error. In addition, during these meetings clinicians pondered ethical questions, for example, on the role of families in their own safety.

During this learning meeting, some participants reflected on the role of women themselves in drawing staff attention to a change in signs of illness. Some meeting participants suggested that if a woman felt any changes, she should be listened to. A senior clinician commented on the difficulty of this:

'We’ve got to think about how much onus we leave for families if we wait for this.’ There was a pause and some agreement before the meeting moved on.

Fieldnotes, Weekly Multi-Disciplinary Learning Forum, 4 August 2021

Multidisciplinary team learning meetings were also an important space where service and clinical leads could promote ongoing reflection and shared discussion. The examples given above illustrate how complex questions of accountability were explored through case-study examples. As often, ‘one-off’ remarks at these meetings could express and reiterate a wider service ethos of shared responsibility for events of harm.

At the start of a SI case presentation, about a baby who required admission to NICU after a home delivery, the presenter mentioned that a Band 5 midwife had been called for this delivery. The maternity safety lead chairing the meeting blurted out ‘how could we put her [the midwife] in this position?’ (Fieldnotes, Weekly Multi-Disciplinary Learning Forum, 2 August 2021).

At Sites B and C, there were multiple examples of senior clinicians reflecting, individually and in small groups, on the best ways to approach teams and families post incident to communicate subsequent learning. ‘Learning how to learn’ was a noticeable dimension of approaches to SIs and action plans and their implementation. Reflecting on the difficulties of including all staff in learning meetings, a senior clinical lead commented:

So it’s quite hard to sort of spread the message wider. I just need to make sure that whatever allowed a Band 6 midwife or doctor to drift into that error, I need to fix that, and not them.

SCSi05

The situation was very different in Site A where there was less capacity for learning by MDT discussion and learning; there was widespread resistance to systems-based approaches to events, and consequently staff sometimes felt scapegoated for incidents. At Site A, a senior clinical lead complained of the loss of this clinical space in recent years leading to a ‘denudation of our culture’ (SASi12). A ‘lack of manpower’ to ‘look into things’ (SASi35) or to ‘work out who was involved’ (SASi01; SASi04b) was repeatedly noted by senior clinicians. During the fieldwork period, Site A had also disengaged from many LMNS activities and had no regular forum for multidisciplinary engagement and learning. Senior leads talked of ‘the luxury of learning’. At Site A, CG and clinical teams more often described their exhaustion and cynicism towards the narrative of safety improvement. At this site, there was a widespread sense that service or systems-wide learning and improvement post incident were either ‘just not done very well’ (SASi1c) or ‘not disseminated effectively’ (SASi14). One long-established clinical lead noted the ‘disintegration’ of opportunities for learning and QI in recent years due to the sheer increase of patient numbers and pressures on clinical time. The gradual erosion of maternity teams to protect time for learning also appeared to correspond with a hiatus in service leadership (see also Appendix 2 on general and topic-relevant features of the case-study sites).

This meant that QI practices were not routinely embedded through learning forums and ongoing safety improvement work. Without this, staff felt, there were limited opportunities to share knowledge and make changes, resulting in demoralisation.

It all does nothing. Absolutely nothing. I end up just feeling like a monkey. It does nothing. Don’t pat yourselves on the back that we’ve spread a learning message because we haven’t.

Risk Midwife, SASi1b

As described above, for injured families wanting to be involved in service improvement, this lack of capacity was interpreted as a lack of interest in collaborative working to make improvements.

In contrast, at the same time, clinical leads at all case-study services were keen to use the voices of injured families to strengthen the case for front-line staff to attend to various practice and service improvement work. This inclusion depended on the availability of time between more pressing clinical learning priorities. Family stories or experiences were considered a ‘soft side of safety learning’ that encouraged staff to be ‘better, more empathic listeners’ (SDSi04). Also, when these stories impacted directly on clinical leads during debriefs, they might have a more significant impact on approaches to clinical care:

The clinical director at Site C described: ‘an incident that really hit me. [The woman] said she didn’t want forceps and the doctor just ignored her and put the forceps in and pulled the baby out. The debrief will always stay in my mind. She felt, at the time, raped. That’s what she said. Violated, is the word she used. Violated. “A piece of meat”. This really stayed with me all the way. We use that in teaching when the junior doctors join us, we bring up listening and not making a woman feel violated’.

SCSi05

The possibilities for injured family experiences to contribute to formal teaching events were also important to clinicians who contributed these stories. However, in contrast to situations where families shared their stories more directly, there were relatively few opportunities for families themselves to shape these narratives for different audiences or to have a sense of the impact of the experiences they had shared.

Communicating changes to families

As previously described, all case-study services routinely invited families to at least one in-person or virtual meeting to discuss the findings from their finalised incident report. Some of this discussion was to reassure families that ‘learning had happened’ (SASi08b). For more expert debrief leads, this aspect of the meeting was carefully considered in relation to the interests and expectations of that family.

Several clinicians who were expert in developing relationships with injured families (before, during and after debriefs) were cautious of assuming that all families are interested in knowing about service and systems changes. Some families were still dealing with the ‘raw emotions’ (SASi01b) of family injury, and for them, such promises of changes made were irrelevant or insulting. A risk midwife described:

An angry woman said to me ‘I don’t want my baby to be your f***ing learning curve. I want my baby alive’. At the end of the day, you’re going home to an empty cot and I don’t suppose you give two shits then as to whether we are taking learning forward.

SASi01b

For some clinicians, another consideration of the debrief meeting was how to respond to a family’s expectations of the outcomes of safety improvement work. For example, a specialist midwife commented:

Always have to say to a family ‘I can’t promise you that this will never happen again’ but I can say we will make improvements.

SDSi08

She felt that promising too much to families risked the further erosion of trust in the service. Clinicians involved in the longer-term support of injured families (CG teams or specialist clinicians) were also aware that families had different expectations for the pace and the scope of changes. The completion of safety improvement measures that were rapid and more tangible was sometimes expected to offer some families greater reassurance that changes were happening:

By the time we’ve done the debrief following the SI closure, we’ve normally completed most of the actions so that’s why we can sit with them and say to them that is what we’ve done, so not many families come back for more closure.

Risk Midwife, SDSi03

However, some senior clinicians felt that a more honest approach to what was possible for services was important to many families:

So I say there’s certain things that I cannot fix but at the same time there are things that I can influence straightaway, so this shows learning and improving on things all the time.

CG Lead, SCSi01

Debrief leads were also aware that the pace of some changes in services change left families with the impression that nothing had been done.

Families want that meeting quite early and actually it takes us time to work out what’s happened and then institute changes. So, like changes to our CTG monitoring off the back of a couple of cases with some really awful outcomes, we told the families we were going to do that, actually now that’s three or four years ago and they wouldn’t know that we’ve actually done that and are doing that.

SASi13

This senior clinician highlights the difficulty of demonstrating long-term changes to families without a Trust or service strategy for keeping injured families updated on long-term improvements and without public access to risk registers.

Indeed, all four families recruited from Site A highlighted that the lack of proactive communication from the service about changes made. This lack of public communication led families to conclude that the service, rather than all staff working there, was untrustworthy with respect to what they had committed to do for this family and others. For example, 3 years on, Fay’s mother remembered that the appointment of two specialist cardiotocography (CTG) midwives was identified as a priority action following Fay’s injuries. The family had not been informed that these appointments had happened. For Fay’s family, this situation reflected a wider disinterest of the service in their daughter and their avoidance of any responsibility for learning. She felt that the harm that had happened to her should be used as a case study for staff to demonstrate where things went wrong to promote learning (SAFi1). However, Fay’s mother remained unsure about the learning that had happened in clinical teams after her daughter’s avoidable injuries. She wanted assurance that the doctors who had dismissed the midwife’s concern about the CTG and who ‘didn’t even ask my name, just looked at the chart and walked off again’ knew of the outcome for her family (SAFi1).

These views illustrate the importance of ongoing service efforts to sustain relationships, provide acknowledgement and demonstrate learning to injured families. For some families, the attribution of events surrounding the incident as ‘incidental’ by investigation panels overlooked their lived impression of that time. They stressed the connection between their poor relational care on their incident and its immediate aftermath.

While CG leads and clinical leads tended to focus on providing reassurance of specific changes connected with the SI, many families were as concerned with improvements to the general care before and after their incident (e.g. the cleanliness of a room or how staff spoke to one another). These aspects of care quality indicated a wider interest in their experience and in making improvement. However, without Trust or service prioritisation of and support for updating injured families on service changes made, CG teams or clinical leads responded to family enquiries on a discretionary basis.

Involving families in ongoing learning

The involvement of some injured families in safety changes in more active ways was noticeable at Sites B and C, where experience-based learning and learning networks were already embedded in other aspects of maternity care provision.

At Site B, in-person invitations to injured families to become part of maternity quality and safety improvement work fostered a wider ethos of reconciliatory practice as well as in family engagement. Family engagement was envisaged as ‘the most powerful way of getting change’ (SDSi04). For example, small grant and service funding secured by clinical leads supported two women from injured families to join the ‘User Experts by Lived Experience’ panel that influenced the reorganisation and renaming of the maternal mental health service. Another woman, with perineal injuries was directly involved in the development of the ‘new pelvic floor programme’. Programme completion and promotion gave further social capital to improvement work with injured families.

Three (of four) families recruited from Site A expressed the desire to become directly involved in changes that would prevent the reoccurrence of similar injuries. For example, Hugh described that he had offered to contribute to teaching events for the service:

I offered, it think, on multiple occasions to speak to whoever wanted to speak to me about my experiences. I would have been happy to speak to a room full of midwives or obstetricians or anaesthetists or whatever. I would argue that would be a lot of use if you are going to change the system.

SAFi3

For these families, the apparent lack of interest of the service to learn from their own events, and to include them directly in their learning, led to disappointment and continued distrust. At the same time, some families were more ambitious than clinicians in driving service improvement. While some clinical teams were satisfied in framing successful change as that happening in their unit, hospital or LMNS, some families sought to address these issues on a national scale.

The development of working relationships with some injured families was described by several CG and clinical leads working in some services. For example, the CG lead described her relationship with one father where:

He actually stayed in contact and phoned me like every 3 or 4 months. He did chase me up on it. It actually made me realise [that] it’s good when [families] do that because it means you follow through on these actions, people get busy and actions slip off into the file on some computer somewhere.

SCSi01

This emphasis on giving time and personal attention to disappointed parents, understanding what exactly the family expected and voicing dissatisfaction to Trust leads encouraged some demonstration of trustworthiness to families despite the slow pace of service improvements during 2020–1 (due to the COVID-19 pandemic).

This service approach also opened the possibilities for a realignment between ‘work with injured families’ and ‘accounting upwards’ as families came to inform some of the priorities of senior committees or the Board. The Director of Midwifery (DoM) at Sites C and D took the view that families played a significant part in holding CG and clinical leads to account:

Families are quite searching, they’ll go through the Public Board minutes. Literally, I’ve had families go and look and try to identify their case. They’re anonymised but they’ll recognise them. If you haven’t presented things at Trust Board you wouldn’t get away with it.

SCSi16

This perspective was informed by the attitudes of Trust Board members towards taking family voice seriously and by an embedded ethos of family involvement in different aspects of the service. Where families were proactive and able to access systems for holding services to account for changes to be made in the longer term, families could effect changes in services that had a service ethos that welcomed their involvement and the associated development of new family involvement perspectives.

Initial programme theories

Table 2 summarises the if/then hypotheses that structured our five initial programme theories and the a priori propositions identified from our realist literature synthesis (see Chapter 4). These working hypotheses were used to guide data collection and analysis of our national interview and service case-study findings. These differ from the stages of the ‘Being Open’ guidance41 as they are focused on outcomes, particularly from the perspective of families and front-line clinicians.

Revised evidence-informed development of the programme theories: what this study adds to knowledge of open disclosure policy and practice improvement

Additional programme theories informing disclosure

In the above sections, we have presented our initial programme theories, discussed our deeper interpretation of the key elements of these programme theories as informed by our empirical research and have offered realist C–M–O configurations to illustrate what works, for who and in what contexts for successful OD. In addition to our five original programme theories, throughout our empirical research, we inductively identified three additional programme theories during thematic analysis and in discussion with our CIG and interpretive forums. We tentatively propose that these additional theories also encompass mechanisms that underpin effective disclosure in many circumstances for many families. These programmes theories are:

1. RECOGNITION: If attempts to understand the incident prioritise the family’s own experience and position, as well as their unfolding priorities for making sense of their incident, then some families might rebuild some sense of trust in those who recognise these priorities.

2. INCLUSION: If maternity services were proactive in the inclusion of injured families in ongoing safety and QI work, then family interests and experiences would be more effectively integrated into this work.

3. RECOVERING TRUST: If lead clinicians and services are both engaged in the recovery of a family’s trust (rather than the management of their complaint), then families are more likely to consider the service, rather than a few individuals working there, trustworthy.

Recognition of family experiences of event

We understand recognition as different from acknowledgement, which is the ‘non-denial’ of an injury and the effects of the injury on an individual, family or community. 205,206 Recognition is considering a family’s experiences of injury and its effects in their own terms. 207,208 Our national interview and service case-study findings identify and examine various situations where the priorities and perspectives of injured families are not aligned to the organisationally defined imperatives of safety improvement agendas. Misalignment is most immediately felt when a family’s incident of harm falls outside clinically recognised thresholds of significant injury as defined by the DoC or outside service-defined thresholds for maximum learning as defined by EBC and UK-MBRRACE criteria. 105,209,210 Second, we explore a range of situations where a family’s need to understand the details of what happened in their case becomes obscured by organisational priorities to generalise learning for ongoing and future-oriented systems improvement. Third, misalignment arises during incident investigation and reporting practices, for example, when a family’s sense-making of the event becomes categorised as ‘incidental’ or is excluded from recommendations and related action plans. In these situations, family concerns become irrelevant as they were filtered through limited definitions of safety with a short-sighted interpretation of the information seen as necessary for identifying risk. 211 Finally, in reviews and investigations, family testimonies of what happened, along with broad questions of ‘why?’, sat uncomfortably with a risk investigation focused on establishing ‘objective’ facts. 212

Families who offered experiential testimonies rather than objective evidence were disregarded because this knowledge did not fit neatly into review and investigation systems or wider safety improvement priorities. This marginalisation was sorely felt by families as further injustice. In these situations, families encountered what is described as a ‘double injury’: their experiences are silenced by both the difficulty of communicating aspects of their experience and by the assumed inaccessibility of understanding these personal experiences. 213 Birth Reflections clinics were positively experienced by many families for this reason; their experiences were legitimised.

In the absence of meaningful inclusion in reviews, investigations or responses to their concerns, some families initiate their own searches for explanation. Our 27 in-depth family interviews indicate that the injured families from Shrewsbury and Telford NHS Trust and from East Kent are not exceptional in this regard. Discussions during the SP3 FF and Project Advisory Forums (PAF) identified and considered a fundamental difference between the interests of families and the interests of services in disclosure improvements. Participants identified a ‘mismatch’ between family and service interests that underpinned our findings (PAG forum participant). This mismatch was identified in the tensions between the personalisation of acknowledgements of harm and the standardisation of these acknowledgements in ‘tick box’ procedures required for services. Sometimes, organisations’ ‘family involvement’ agendas differed from the at-that-time needs of families. For example, considering an example of a family feeling deluged by information and visits from different staff (a risk midwife, complaints team, clinical lead and HoM) soon after receiving the news that their baby was likely to have been seriously harmed after a traumatic labour and delivery, a forum participant reflected:

[Your data] suggests that the organisation is deploying resources to meet four or five different things it needs to be, but it’s not thinking about how you support the person who had suffered the problem. So I wonder if it’s not so much ‘is the family involved or not involved’ as ‘what is the mismatch between the needs of families and the things that drive the processes’.

PAG forum participant

This emphasis on the experience of families as the credible starting point for inclusive practices was elaborated in this same forum discussion:

[The best approach is not] ‘this is what it looks like, now ask questions’, but actually: ‘tell us how it was, what’s your part of this whole experience so that we can get all of the information together’.

PAG forum participant

Standardised review and investigation procedures, while important, have set in motion a practice where the stories, voices and concerns of families have been muted or shaped for standardisation to be achieved. 171 In improving future disclosure work, it is important not to lose sight of the importance of enabling practices that prioritise listening, discussing and integrating family voices during review and investigations. Previous research has highlighted that without an open conversation about what happened, many patients and families feel that their experience has been disrespected by providers. 13,214 Reflections Clinics and responsive post-investigation debrief meetings were some interventions that opened the possibility for families to rebuild confidence in a service, yet ensuring that families also contribute their experiences should still be a necessary care priority. Like Wu et al., we identified some conflict between patient safety theory and patient expectations. 11 However, we were unable to characterise these families as intent on focusing on individual clinician failings and involved in a ‘blame game’. Instead, we found the families wanted an assumption of responsibility from them by services and, more significant, time for a hearing of events in their own terms.

We are currently exploring the sociopolitical issues raised by our findings on family experiences of post-incident care through the paradigms of epistemic injustice215,216 and knowledge encounters. 217

Family inclusion in service change

In our PAG and FFs, proposals for co-production arising from the study findings were discussed by families as a means of achieving change and partnership. Family participants described a desire to help services lobby for resources and design better processes. Participants wanted to be apprised of any challenges with making service changes and be involved with overcoming any barriers to change. Also in these discussions, several families expressed an empathy with clinicians, who were identified as victims of ‘process’:

[If] you had just had something go really wrong and almost like you’re in a war zone, you’ve just got to go straight on to the next one. There is no time … for personalised care where you actually treat the people in front of you … There’s this feeling … [of] … process, process, process. And that sucks the humanity out of the staff.

FF participant

Families were cognisant of challenges faced by staff during OD and acknowledged the potential for incidents to deeply affect the healthcare team. They were keen to work with staff in improving OD and safety processes. 218–220

However, our findings show that the marginalisation of injured families in organisational systems and processes that promise to be inclusive is deeply rooted in health care. 219,221 The importance of a ‘whole systems’ and multilevelled approach to the development of more meaningful PPI and partnership working with patients within health care has been noted. This approach encourages cultural shifts that place patients at the centre of safety and address taken-for-granted paternalism. 218 Our case-study findings present new evidence on the wider value of ongoing organisational PPI and partnership working. It is the organisational foundation for listening to injured families. Also, we emphasise the need for this inclusion to extend beyond the imperatives of maternity safety improvement.

An issue not addressed in our forums were accountability structures to encourage ongoing family involvement in interventions to strengthen OD. In contrast to most of the NHS, the development of mechanisms for gathering maternity SU feedback and for promoting service-user involvement in the co-production of local services has been incentivised at Trust level. 58 However, the effects of this intervention on promoting empathic and meaningful engagement of injured families in OD work have yet to be explored.

During our data analysis, we surfaced the two critical factors for strengthening disclosure for a family’s perspective that underpinned others. These were, first, recognition of the incident in the family’s own terms and, second, the need for skilled clinicians who feel safe and have time to practice in ways that are responsive to family needs. Figure 1 shows how the programme theories map against the general steps of SI management within organisations.

FIGURE 1.

Programme theories mapped against the general steps of SI management within organisations.

However, we found that underlying service or clinician efforts to strengthen OD was a medicolegal background that undermined this, compromising staff and families. Figure 2 illustrates how the programme theories interconnect to explain how disclosure can be improved.

FIGURE 2.

How the programme theories interconnect to explain how disclosure can be improved for some families.

Patient and public involvement

As described in Chapter 2 and Appendix 2, public and family involvement were embedded in all stages of the research cycle in four ways. First, one experienced PPI representative and parent was an active member of our Study Steering Group (SSG). Their concerns with the practical outcomes and impact of the study were an important public contribution to this academic group. Second, three PPI co-investigators were included in the study team from the outset of the project. The entire study team met 6-weekly for meetings and ongoing correspondence. This was essential to reaching an agreement on the refinement of study methods, adjustments in research design, identification and prioritisation of study findings, report design and drafting, and research dissemination and impact priorities. The three PPI co-investigators supported study progress in practical ways, for example, in networking the researchers into new campaign groups promoting diversity, so extending the scope of our PAG and promoting the study across their charity and association networks from where most members of the national family cohort were recruited. However, the most positive outcome of the PPI co-investigators was their ongoing involvement in the discussions about study progress and research findings. They became a trusted and critical ‘sounding board’ for the researchers to ensure the relevance of work to injured families and the wider public. Third, our PAG included six families, who became the direct voice of SUs during these meetings. The outcome of this direct involvement within the PAG was to ensure that disclosure improvement for families remained the focus of discussion. The inclusion of these families was a central focus of the second PAG meeting where two researchers worked with the families to prepare and present their own stories of disclosure and their expectations for service and practice change. Finally, a second family group (recruited from SP1 and SP2 family research participants) gathered to contribute towards the interpretation of study findings and research dissemination and impact plans. In all, the study engaged in four ways with a variety of PPI representatives. This inclusion of different voices was important so that differences were not muted and critical engagement with the research team and study findings were possible. 223 Nevertheless, the study team recognises that the families who expressed participated in the study represent a population who are more proactively involved in post-incident service change.

Reflecting on the overall outcomes of multifaceted PPI for the study, the study process and outcomes have been enriched by a multifaceted PPI approach (e.g. the study team have been able to identify key issues that would otherwise have been overlooked by those without this experience). At the same time, particularly for the CIG teamwork, the project has benefited from exceptionally experienced PPI contributors who had worked on previous studies together or were aware of one another’s work. We feel that this level of expertise and collaboration is critical for the success of multifaceted PPI. Our PPI included lively debates on the value of immediate recommendations over theorisation and generalisation from findings.

Diversity and inclusion

Although the research was designed for relevance to all families and clinicians involved in disclosure, during year 1 of the study, it became clear that black and ethnic minority families, families with non-heteronormative sexual orientations or gender identities and families with disabilities had been less actively included in the study (as co-applicants, project advisors and research participants). We addressed this to some degree by identifying and directly inviting representatives of two black parenthood groups on to our PAG. We also purposively identified and approached four families from ethnic minority backgrounds as study participants, with three consenting to interview. The challenge with accessing and recruiting these families echoes our findings about systemic inequalities for families who wish to engage with improvement work. It is important that we recognise whose experiences our work captures and ensure that we work to amplify marginalised voices, thus accurately representing the experiences, needs and wants of a diverse population in our improvement interventions. This is an important and salient reminder that engaging in research, accessing networks and advocating for one’s own care requires privilege and resources which may not be available to all families. Breaking down these systemic barriers is an essential priority for all interventions related to family involvement.

Study strengths and weaknesses

A key strength of the study is generation and integration of in-depth and extensive qualitive research findings to develop a nuanced understanding of the critical factors shaping disclosure practices and outcomes, from the multiple perspectives of three stakeholder groups. These groups have also been directly involved in research design, ongoing study management, data collection, analysis and decisions over dissemination and impact. The research was structured around a realist evaluation methodology. A realist synthesis of international disclosure work in maternity settings underpinned the development of initial theories to guide investigation of the complex causal processes of disclosure improvement. This methodological strategy also allowed for the iterative development of previously unidentified theoretical insights from the research findings. Interviews with injured families and ethnographic observations of disclosure practices including them have generated new perspectives on the relationships and disconnections between the expectations of injured families and the imperatives of services. Given the significant variation in disclosure nationally, the participation of families recruited from the case-study services where disclosure improvements were ongoing strengthens the reliability of findings.

A limitation to the in-depth case-study research is its limitation to three services in two NHS Trusts due to difficulties of accommodating non-COVID-related research during the pandemic. These services were selected for their positive deviance in OD improvement. Therefore, the generalisability of enablers to OD improvement beyond these case-study services has not been examined. Additionally, families were recruited from only one case-study service and so the impact of OD improvement in two case-study services has been explored without direct communication with families affected. However, the research team was able to identify and explain considerable variation within and across these three services, as well as to explore the generalisability of findings in interpretive forums with national stakeholder groups, including a family group. A significant limitation to the study findings has been a lack of access to HSIB investigators and HSIB investigation practices, where many SIs in NHS maternity care are now investigated. As a qualitative study, inevitably influenced by relational aspects of data collection, researchers’ personal perspectives may have influenced the interpretation of study findings.

The inclusion of a co-investigator engaged with families who had experienced marginalisation during maternity care, whether because of sexual orientation, gender identification, ethnicity or disability, would have generated findings that better reflected population diversity. Our well-networked third-sector co-investigators enabled some mitigation of this during the study. Finally, while the study has followed a realist evaluation methodological strategy, co-investigators have found the conceptualisation of findings in realist evaluation terms particularly challenging in terms of jargon and causal modelling. Clear differentiations between programme interventions and mechanisms could not always be established, particularity where programme improvement was explicitly directed to issues of attitude or cultural change in the workplace. Our analysis of findings in terms of programme theories highlights the cumulative and interlinked quality of factors underpinning disclosure improvements.

The impact of coronavirus disease 2019 on the study findings

Our ethnographic case studies were conducted between April and October 2021, immediately after some decline in service pressures associated with the second wave of the COVID-19 pandemic. However, service leads in consenting case-study services as well as our site PIs were aware of more enduring declines in staff morale and critical staff shortages within midwifery and obstetrics. The pandemic magnified longer-standing issues of short staffing, escalating organisational demands and increasing barriers to the provision of relational care,224 along with anticipated increases in preventable incidents associated with care modifications. 225–227 Our research, while conducted in unusual circumstances, documents magnified rather than exceptional situations in which post-incident care and communication with families and staff were conducted.

Receiving a meaningful acknowledgement of the harm that has happened

We identified ongoing tensions over the ethical need and learning benefits of OD, particularly in relation to concerns with litigation and reputational risk. Our evidence from families suggests that for over a decade, post-incident communication has been limited, closed and suspicious. Incentivisation of reviews and reporting, which included a family inclusion element, was felt to have taken time and energy from relational care. While these views were recognised in the case-study services, staff managed these tensions personally. A few staff were experts in in-person communication for the purpose of recovering the trust of the injured family. This work was poorly recognised or resourced by services. Inconsistency in OD processes was managed by informal staff networks and against the churn of shifts, hospital discharges and transfers.

Being involved during the review/investigation process

We found widespread agreement that this aspect of OD required improvement. The implementation of family involvement as an aspect of the PMRT highlighted the lack of resourcing for necessary post-incident relational care. By contrast, external reviews such as those conducted by HSIB were perceived as well-resourced, integrating more structured family involvement and ultimately more successful. Many families felt that they, their views and their experiences were excluded from NHS organisational processes. In the case-study services, some staff attempted more proactive approaches to family involvement, and sought to incorporate family perspectives into panel presentations, discussions and reports. Nevertheless, the interests of families (to understand their case) and wider service interest (to extract learning) sometimes clashed, and the unclear status of uncertain knowledge discussed in report drafts or panels could undermine the intentions of staff to foster trust. Involvement of harmed families in service improvement was unusual. In the one example of this that we were made aware of, a senior Trust lead had independently championed this work.

Making sense of what happened

We identified widespread challenges of establishing explanations for an incident when they are examined through multiple review and investigation approaches and when there is a reliance on often inconsistent or patchy clinical records. In the case-study services, enabling participation in report crafting, providing in-person support and ensuring ongoing report debriefs were identified as crucial practices for the regaining of family confidence in clinicians and services and making sense of what happened. We also explore informal opportunities for injured families and affected staff to explore the consequences of an incident. From our national and case-study injured families, we briefly note the capacity of some to gather information and garner professional networks and expertise independently of a service. We identified the value of post-pregnancy support and Birth Reflections services when injured families were referred into them.

Receiving care from clinicians who are skilled and feel psychologically safe during post-incident communication

We identified a national underinvestment in OD training for clinicians, as well as a lack of awareness raising of the needs of injured families across services. We found that immediate workplace responses to an incident were crucial for the quality of OD, both in terms of the quality of the communication with the family and in terms of the experience of the clinicians, many of whom felt unsupported in conducting OD. While the case-study services were investing in post-incident staff support, we found that this was often inappropriate and underutilised. The value of facilitated, ‘cool’ post-incident staff debriefings, informal staff support systems and informal meetings with some families, as well as being informed of system changes and learning, were important for post-incident staff recovery.

Knowing that things have changed because of what has happened

We found that knowing that learning and service improvement had happened was a crucial aspect of OD for most families. Evidence of post-incident learning and ongoing communication, even after the review or investigation had concluded, was often critical for families to begin regaining trust in a service. However, services did not proactively communicate news of the progress of changes with families, and improvement leads often sought to demonstrate ‘quick wins’ rather than longer-term and more significant improvement investments, both of which were disappointing to families and created additional feelings of mistrust. Correspondingly, we found there was a lack of effective, post-incident, learning communication strategies across services. The non-judgemental discussion of an incident and its implications, in either ongoing MDT meetings or during outreach work by risk midwives, were found to be important in shaping staff attitudes towards incidents and injured families.