By Richard Devine, Social Worker and Author of Messy Social Work: Learning from Frontline Practice with Children and Families
There is something uneasy about this review.
This national review was commissioned following the death of baby Victoria Marten in 2023, whose parents, Constance Marten and Mark Gordon, were later convicted of gross negligence manslaughter. The case was marked by years of involvement from multiple agencies, repeated care proceedings, and the removal of four previous children. Victoria’s birth and short life, however, were concealed from professionals entirely. The review does not set out to re-examine the criminal case, but to identify broader lessons for safeguarding practice.
I don’t think the themes identified in the review are wrong. Few would argue against the importance of understanding trauma, engaging parents, or thinking more systemically about families who experience repeated child removal.
These are serious and longstanding issues in child protection practice.
Nor is there anything inherently incompatible between trauma-informed practice and decisive child protection intervention; in principle, both are necessary.
But reading this report, it is hard to escape the feeling that the conclusions sit awkwardly on the case they are supposed to illuminate.
The review states clearly that it will not speculate on whether different actions would have changed the outcome. That is an understandable position. Hindsight bias is real, and there are limits to what any review can credibly claim. Responsibility for Victoria’s death, as the report rightly says, lies with her parents.
And yet the report is threaded with an implicit sense that things might have been different if the system had operated differently.
Practitioners are described as struggling to “think themselves into the parents’ shoes”, opportunities for deeper reflection are said to have been missed, and the system is portrayed as insufficiently trauma-informed. The difficulty is not that these observations are wrong, but that they are rarely grounded in a clear account of what could realistically have been done to work with this family.
This also marks a noticeable shift from the approach taken in most serious cases and national reviews. Traditionally, such reports have sought, however cautiously, to explore what might have been done differently. That counterfactual analysis has been central to how the safeguarding system learns.
The defining feature of the family’s history was persistent and determined non-engagement. Across multiple pregnancies, local authorities, and interventions, information was withheld, identities obscured, and opportunities for assessment repeatedly avoided. Even the review acknowledges that, as a result, practitioners had only “limited insight” into the couple. Against that backdrop, it is not clear what a more “trauma-informed” or “relational” approach would have looked like in practice, still less how it might have been sustained.
It is here that the review begins to feel as though the broader framework increasingly takes precedence over the specifics of Victoria’s case.
The emphasis on trauma-informed practice is particularly striking.
The report sets out, at length, the impact of trauma on parental behaviour and the need for systems to respond differently. In principle, this is sound.
Many practitioners would recognise that trauma-informed approaches can help build trust, make sense of avoidance, and reduce adversarial dynamics. But the connection to this specific family remains largely theoretical. At no point does the review establish that such an approach would have led to greater engagement, clearer assessment, or a different outcome for Victoria.
There is also a further difficulty in how “trauma-informed practice” is presented.
Throughout the review, it is presented as a clear, operational framework that practitioners can straightforwardly adopt and be measured against. In reality, it is much less settled than that. While the term “trauma informed” is widely used, there is no single agreed model that sets out clearly what practitioners should do in complex situations, how decisions should be made, or how empathy should be balanced with the need to act decisively to protect children.
It is, in that sense, easier to invoke than it is to operationalise. To argue that practice should be more “trauma-informed” risks sounding both self-evident and imprecise, particularly in cases like this where engagement was minimal, and opportunities for relationship-building were severely constrained. Without greater clarity about what this would have meant in practice, it becomes difficult to assess the weight of the critique.
That distinction matters. A recommendation may be valuable in general terms without being meaningfully evidenced by the facts of a particular case. At times, the review appears to move too quickly from “this is good practice” to implying that it therefore had explanatory relevance.
Instead, the argument rests on a general proposition: that repeated child removals can exacerbate trauma, that trauma can drive disengagement, and that disengagement increases risk. That may be true in many cases. But here, it is applied without sufficient attention to the particularities of this family: the sustained avoidance, the escalation in concealment, and the difficulty professionals faced in gaining a reliable foothold.
A related tension runs through the treatment of non-engagement itself.
On the one hand, it is rightly understood as a significant safeguarding risk. On the other, it is reframed as a possible expression of trauma, mistrust, or fear.
A genuinely trauma-informed approach would need to do both simultaneously: understand the meaning of avoidance while recognising the dangers it creates for children, and in particular, unborn or newborn babies.
At times, however, the report seems to slide between these positions, creating the impression that practitioners are being asked both to treat non-engagement as a critical risk and to view it as something that could have been mitigated through a different relational approach – without a clear account of how.
Perhaps the most unusual feature of the review is its use of Constance Marten’s own reflections. Her voice is given space, particularly in relation to her perceptions of social work intervention: that services were coercive rather than supportive, that parents fear being judged as unfit, and that engagement with health services risks triggering removal.
In principle, the inclusion of parental perspectives is welcome. Understanding how safeguarding systems are experienced by those subject to them, including those who struggle to engage, can offer important insight.
It is also true that, in many reviews, parental voice is either absent or heavily mediated. Seen in that light, the decision to include Marten’s reflections could be understood as a genuine attempt to broaden the lens.
However, the way those reflections are used here raises questions.
Her account is presented largely at face value, with limited critical engagement or contextualisation alongside it. There is little attempt to explore how her perceptions sit alongside the known history of the case, the practitioners’ legal duties, or the repeated findings that led to the removal of her children. The result is that her narrative carries weight without being meaningfully tested.
A parent who has lost multiple children and whose actions resulted in the death of an infant is positioned, in part, as a key informant on the shortcomings of safeguarding systems without sustained interrogation of the limits of that perspective. It is difficult to see what is gained by this, particularly when it risks reinforcing a narrative that professionals acted on arbitrary assumptions rather than in response to clear and escalating risks.
There are also questions of balance. The review does not appear to include equivalent reflections from Mark Gordon. If both parents were approached but only one contributed, that is important context. If not, the asymmetry becomes harder to justify, particularly given the central role of both parents in the events in question.
This, in turn, raises a broader issue about how voice is selected and represented. Marten is articulate, educated and able to express her views clearly. That inevitably makes her account easier to present and arguably more compelling to read. But it also risks giving disproportionate influence to a particular kind of voice. Without clearer framing, there is a danger that what is presented as “lived experience” may, in practice, reflect whose perspective is most accessible.
None of this is to suggest that the system worked perfectly.
The review identifies areas where multi-agency reflection could have been stronger, where links between offender management and child protection were not fully realised, and where post-removal support for parents was limited. These are important points and align with wider concerns across the sector.
But there is a difference between recognising systemic issues and attributing them to a specific case. In Victoria’s situation, the overwhelming challenge was not a lack of professional effort. If anything, the review acknowledges the persistence and commitment of those involved. The challenge was that, despite those efforts, the parents remained largely beyond reach.
That raises a more uncomfortable question, one that, in my view, the review does not fully engage with:
What are the limits of safeguarding when parents actively and effectively evade intervention?
And what should we reasonably expect practitioners to do in those circumstances?
This may be the central unresolved issue.
Serious case reviews often proceed on the assumption that failures are identifiable and remediable. But some tragedies reveal not only the limits of practice, but the limits of state reach itself.
By focusing so heavily on trauma-informed practice and parental engagement, the review risks obscuring that question. It shifts attention away from the structural and legal constraints within which professionals were operating, and towards a model of practice that, while valuable elsewhere, may have limited traction where engagement is actively resisted.
More broadly, this review reflects a wider pattern in recent national and high-profile safeguarding reviews. Increasingly, individual cases are used as a platform to advance broader system-level agendas, whether around trauma-informed practice, multi-agency reform, or parental engagement. Similar dynamics can be seen in other recent reviews, where the lessons drawn extend well beyond the specific circumstances of the child’s life and death.
There is a clear rationale for this. National reviews are intended to consider systemic change, not simply individual error.
But there is also a risk.
When learning becomes too detached from the particulars of the case, it can feel imposed rather than derived. The danger is not that the recommendations are wrong, as I have already said, I strongly agree with them all, but that they lose explanatory power. They tell us how the system should evolve, but less about what actually happened here, and why.
In the end, the report offers broadly familiar recommendations: more integrated working, better support for parents, and greater emphasis on relational practice. These are not new ideas, and in many ways they are welcome. But the link between those recommendations and the death of Victoria Marten remains, at best, indirect.
If we are to learn meaningfully from tragedies like this, we need to be clear not only about what systems should aspire to, but also about what was actually possible – and what was not.
Without that clarity, there is a risk that practitioners are left with a familiar but unhelpful message: that more could have been done, without a clear account of what that “more” would have been.
By Richard Devine (15.05.2026)
If you have found this interesting/useful, you may wish to join a growing community of 1200+ others by signing up for free blogs to be sent directly to your inbox.