Restorative philosophy appreciates that humans are inherently relational beings, and that relationships can heal and harm us (17). A discussion of the rich and diverse roots is beyond the scope of this paper, which attempts to surface key challenges in patient safety, where it has more recently been combined with safety science (11, 18). Different terms are used in our respective countries, often interchangeably, to describe approaches underpinned by restorative principles and goals.

A restorative approach may be broadly conceived as a set of relational principles that finds expression in common practices that promote human agency, dignity, respect, voluntariness, responsibility, equity and safety. In the complex health environment, a restorative initiative appreciates the holistic, responsive and dynamic human contribution to safety and wellbeing. The restorative triangle (Figure 1) serves as a visual reminder as to how human relationships contribute, and how relational “slack”¹ influences system resilience capacity. Ideally, restorative systems focus their efforts on proactively promoting safety through strong relationships, anticipating that harm is inevitable in a complex system, and less time in a reactive state responding to harm. Developing relational capacity nurtures conditions in which individuals feel more able to have difficult conversations and resolve conflict. Doing so means that when harm inevitably occurs, it is understood as an event worthy of learning and a human experience that creates needs and obligations (4).

A restorative response involves listening to understand what happened, the impacts and needs involved, and clarifying the responsibilities for repair (20). NZ research concludes that the following principles are important considerations in complex health systems (4, 21).

A common myth is that a restorative response requires a face-to-face meeting, in alignment with the restorative justice conferences used in criminal settings. Whereas healthcare research identifies that offering a range of trauma-informed documentary and dialogical options is valued by harmed patients and families (21). Both can accommodate the use of art, poetry, or other forms of storytelling and stand in contrast to the investigative “interview” or “statement”. Dialogical practices associated with restorative practice encompass a continuum of affective questions and statements, facilitated meetings and Circle practices. Circles involve a structured and intentional conversation in which people, sitting in a circle, respond sequentially to questions posed by the facilitator. Community building Circles create foundations that nurture wellbeing, safety, connection and trust by encouraging collaboration, equitable decision making and cognitive diversity (22).

In health systems, restorative initiatives have found roots in communities with lived experience of inequity in all its forms and settings where the relational contribution to safety and well-being is incorporated in cultural norms, worldviews, or everyday practices. The key areas of development underway are mental health, women's health, and paediatrics. It is notable that in all of our countries, harmed patients and families, and advocacy organisations are amplifying lived experience voices to inform a community driven approach [e.g., (5, 23)]. In Canada and NZ, initiatives were developed in the aftermath of government inquiries that highlighted health system racism, bias or inequity (24, 25). In these settings, policy development is occurring in partnership with Indigenous communities. In NZ, the sovereignty of Māori over the rights and practices of hohou te rongo (peace-making) as a distinct Indigenous restorative approach is protected by Te Tiriti o Waitangi (The Treaty of Waitangi).

Restorative efforts are focussing on co-designing, implementing or evaluating restorative responses to adverse events or identifying enabling conditions. Table 1 provides an overview of policy and practice initiatives in our countries, and the evidence supporting development. Australian and NZ adverse event policies and restorative guides focus on supporting healing, learning and improvement equitably (42–44). NZ has the most established national approach, being co-designed with a diverse range of stakeholders in the aftermath of a Ministry of Health inquiry (22). As well as providing information about surgical mesh harm and what to do about it, the restorative inquiry identified that compounded harm was widespread and contributed to mental distress, PTSS and suicidal ideation. A descriptive evaluation determined that the restorative response met most participants psychological and procedural needs, captured information crucial for learning, and recommended embedment within systems that mitigate and respond to harm (21).

The NZ approach has been informed by Indigenous worldviews and Western research that identifies the relational contribution to safety (see Table 1). To mitigate the risk of marginalising voices or communities, co-design has been shaped within a collaborative framework that is guiding development (44). The expectation to offer restorative responses (restorative practices or hohou te rongo) is embedded within the Adverse Events Policy 2023 and Health and Disability Standards 2022 (42, 45). Section 17(d) of the Mental Health Bill 2024 includes the expectation that hui whaiora (wellbeing meetings) “support restorative practice to uphold the mana (power and authority) of all parties following the use of coercive practices”. Capability building has been underway for two years, initially focussing on workers in mental health settings.

In Australia, mental health has been at the forefront of development, with New South Wales piloting the approach under the umbrella of a restorative just and learning culture (RJLC) (43). A development in safety culture thinking, RJLC encourages organisational justice, and ‘forward looking’ accountability, rather than blaming individual clinicians (18). In Queensland, the approach gained traction as part of a Zero Suicide framework at Gold Coast Hospital (46). Their healing, learning, and improving model includes a peer response for clinicians, demonstrated benefits to a range of stakeholders, and enhanced the quality of recommendations (29, 47). A mixed methods evaluation of a restorative practice initiative that balanced proactive and reactive elements in a secure adult community mental health service concluded that the model was beneficial for worker and client relationships. Evaluation participants almost unanimously indicated that there was no downside to introducing restorative practices, identifying that most of the benefits were gained from alignment with the mental health recovery model and everyday use of the proactive elements (30).

The Patient Safety Incident Response Framework in England is based on a programme of participatory research with harmed patients and families that included interviews, ethnography and documentary analysis (12, 14, 48). The ‘Learn Together’ programme incorporates five principles that support a systems based approach to ‘restorative learning’ (49). In British Columbia, a research programme is underway to investigate feasibility and inform contextually relevant, evidence based restorative approaches (50). In the US, an interdisciplinary network is raising awareness about the requirement to enable ‘restorative competency’ within Communication and Resolution Programmes. Academic institutions in the US and Canada have also applied restorative principles and practices to address workplace harms in medical, nursing and dental settings (41, 51).

In our unique and collective journeys, several key factors are enabling the development of restorative potential within our distinctive contexts. Interdisciplinary collaboratives, co-facilitated by individuals with restorative knowledge and skills, have guided development in NZ and BC (44, 52). Indigenous worldview, leadership and voice is ensured and has been integral to advancement. Indigenous communities have many approaches to addressing harm, which can differ by place, be dynamic (shift and change over time), and hold competing perspectives. In the US a similar network connects Western restorative expertise with safety scientists. Opportunities that connect and explore structural and relational interdependencies between institutions, those with lived experience of the system, and communities is essential to build a mutual understanding about what works (or not). Developing relational infrastructure (i.e., collaboratives) is pivotal to ameliorate adversarial relationships and enables the co-creation of systems, key concepts (e.g., safety, harm, justice, responsibility), and supports those involved in the emotional work of change.

Mental health is proving to be a fertile area for policy and practice development and conditions may be conducive for numerous reasons. Firstly, priority populations are overrepresented and there is a strong focus on social justice. Mental health teams work within an interdisciplinary model, have transferable skills, and are familiar with trauma informed dialogical therapies which have some alignment with restorative practices (4, 30). In mental health settings, lived experience is increasingly viewed as a credible form of evidence that has been structuralised into peer worker roles in the UK (53), Australia and NZ. Furthermore, restorative justice is embedded in disciplines used to navigating complex legal and safety matters e.g., Forensics (54). Importantly, most adverse events are suicides, can affect over a hundred people, and thus require a community response (55). Existing suicide postvention services afford a structural opportunity to incorporate restorative principles.

An important first step in a restorative response is to acknowledge that harm has occurred, affects the community providing and receiving care in different ways, and creates unique needs and obligations. Doing so creates opportunities to shape holistic responses and attend to the range of physical, psychosocial, cultural and other needs involved, whilst also learning to improve system safety. It is more challenging to offer a restorative response in the context of adversarial systems, when responses privilege the rights or wishes of institutions or providers, focus on minimising reputational or liability exposure, or are inequitable (9). However, restorative potential can be enabled, and compounded harm mitigated, if approaches explicitly acknowledge and respond to the human experience of harm and participants can make an informed choice about the potential benefits and risks of stepping into procedural responses (4).

Adverse event policy and practice that explicitly acknowledges the complex human experience; enacts an equitable focus on healing, learning and improvement; and expects a range of diverse outcomes is advantageous [e.g., (42–44, 46)]. In Australia and the US, apology protections enable restorative dialogue during open communication and resolution practices (39, 56). An influential institute in England recently proposed that no fault no blame legislation may be advantageous (57). It is important to note that tax payer funded compensation and consumer protections have supported restorative potential in NZ, but can also generate compounded harm if the needs of harmed patients and families are minimised or dismissed (4, 58).

Use of restorative inquiry as a triage tool supports a focus on who is affected and the impacts, needs and obligations involved from the outset of an investigation (29). Furthermore, restorative practices are proving useful tools to enable shared understandings, equitable and safe conditions, honest (often courageous) conversations or the restoration of dignity and trust (21, 30). Early adopters should be aware that restorative responses can result in compounded harm when efficiency is prioritised over quality.