The Poutasi report found that, at various points, the views of other agencies as well as those of Malachi’s and community should have been sought or shared by agencies so they could be considered in assessing and responding to Malachi’s needs. This might have resulted in a decision by Oranga Tamariki to go and see Malachi.
Critical gap 2
The process for assessing the risk of harm to a child is too narrow and one-dimensional
The Poutasi report made three recommendations to try and close this critical gap. Those recommendations were focused on enabling decision making at a community level to address harm and the risk of harm before it escalates to needing a statutory response, on giving health practitioners a wider view of interactions a child has had with the health system and on using health practitioners to help assess harm and support training under the Child Protection Protocol (CPP).28
In its report following the death of Malachi, the Ministry of Health made a similar recommendation to the Poutasi report about joining up medical records. In addition, the report by the Chief Social Worker following Malachi’s death noted a need to involve community agencies in decisionmaking processes.
There is still no consistent, structured process for considering perspectives from other agencies or community organisations when assessing the risk of harm to a child.
While there are models of multi-agency teams in place, these primarily focus either on responding to an incident of family violence where police have attended or where Oranga Tamariki has assessed a report of concern as not requiring its involvement. The issue of inconsistent risk assessment for family violence and sexual violence victim-survivors, including and , is a recognised gap within family violence and sexual violence responses.
Reports of concern tables triaging notifications that Oranga Tamariki has assessed as requiring no further action are promising, but they are not in every community and there is no consistent approach to how they operate. We could not identify any multi-agency models where professionals discuss concerns or worries about tamariki and rangatahi before the point of making a report of concern. If they did, and had resources to act, it would be a useful preventive measure – to provide help before harm occurs or escalates.
The Centre for Family Violence and Sexual Violence Prevention told us that, while this is mostly accurate, practice and capacity vary widely across and some multi-agency responses consider risks to tamariki prior to a formal report of concern being made. It told us that some multi-agency responses will bring to the table even if a family harm incident or report of concern has not occurred. This is because community providers sitting around the table often have knowledge of whānau within the community who are experiencing harm but are not known to Police or other statutory agencies.
The Centre for Family Violence and Sexual Violence Prevention also told us it has identified the need to strengthen the visibility of tamariki within multi-agency responses and ensure a more consistent approach and will be progressing this as part of work to respond to the recommendations of the Poutasi report. It told us that the Executive Board for the Elimination of Family Violence and Sexual Violence approved a target operating model in August 2025.
Models that have greater visibility of tamariki and can discuss concerns prior to lodging a report of concern could help to determine what whānau may need and how they can be supported by /Māori or community agencies before harm occurs.
Work in Health NZ to give health practitioners a wider view of a child’s interactions with the health system has not advanced since our 2024 review. Implementation of the Shared Digital Health Record is at such an early stage, it is not possible to measure its impact. Approval of a business case is still needed to progress work to deliver on the full intent of this work.
A decision has been reached for Health NZ to join the CPP, but this is yet to be implemented. Until we can see how this is working in practice, we cannot measure whether and how this will contribute to closing this critical gap. Child death reviews suggest that, even when tamariki are known to Oranga Tamariki and Police, they can still fall through gaps in the safety net. Since Malachi’s death, at least five further tamariki who were known to Oranga Tamariki and/or Police have also been killed by someone who was meant to be caring for them.
Widening the agencies that are involved in assessing risk to tamariki and rangatahi with a view to supporting earlier interventions to prevent harm may help to close this critical gap. More needs to be done to achieve this. Later in this review, we look at the ability of the child protection system to respond to reports of concern and why they are not necessarily acted upon.
The Poutasi report recommended that multi-agency teams should be working in communities in partnership with /Māori and other , resourced and supported throughout the country to prevent and respond to harm.
It noted that there are examples of this happening already across the country, but implementation in all communities must be a priority so relevant local teams can help assess, respond to the risks to and and provide support.
This recommendation was made with a view to widening how risk is assessed and supporting earlier interventions to prevent harm occurring. While assessing the risk of harm is an inexact science and assessments may not always identify this risk, bringing in more perspectives to help assess risk, including earlier assessments and community responses to address harm, would go some way towards closing this gap.
Our 2024 review noted that, while there were several multi-agency programmes in place across the country, they were not in place in every community. The multiagency programmes identified in our 2024 review were focused on responding to incidents of family violence, but they did not all operate or assess risk in the same way. We concluded that the collective impact these models were having on child safety was unknown.
There are reports of concern tables and family violence tables operating across
Oranga Tamariki told us that, since our 2024 review, it has worked with several community partners through its Enabling Communities approach to establish community responses to reports of concern. These tables receive reports of concern from Oranga Tamariki where it has already made a decision that the concerns raised do not require further action on its part. The term tables is used to describe the group of organisations that sit around the same table to discuss the response.
In communities where this approach has been set up, all reports of concern where Oranga Tamariki determines that a statutory response is not required will be referred to these tables. We heard about two examples of these tables during our recent community monitoring visit to Te Tai Tokerau – Te Kahu Oranga in Kaitaia and Te Tēpu in Whangārei.
In addition, we again heard about the network of multi-agency teams that respond to incidents of family violence. These teams are also often referred to as tables.
Most referrals to the family violence tables come from Police following a family harm incident that police officers have attended. The table collectively determines the appropriate response for whānau, which can include making a report of concern to Oranga Tamariki to assess whether further action is needed to support tamariki and rangatahi. The family violence tables do not usually consider cases where tamariki and rangatahi have been directly harmed, as allegations of abuse against tamariki and rangatahi are managed under the CPP between Oranga Tamariki and Police and require a direct report of concern to Oranga Tamariki.
To help inform this review, we met with two examples of these multi-agency teams – Integrated Safety Response (ISR) hosted by Police and multi-agency tables called Safety Assessment Meetings (SAMs).
While these initiatives are making a difference in the communities where they are operating, they are not yet closing the critical gap identified in the Poutasi report. Neither reports of concern tables nor family violence tables are present in all communities, although the family violence tables do cover every Police district. Oranga Tamariki further told us that it was not feasible or desirable to set these up in all communities.
There is no consistent operating model or approach to assessing risk for either reports of concern tables or family violence tables. This is because they have been developed to respond to local needs and conditions. It means that responses to harm vary across communities – what might be assessed in one community as needing a response might be assessed in another as not requiring a response.
For the most part, reports of concern tables and family violence tables respond to address harm that has already occurred. There is no consistent forum across communities for agencies to collaborate at an earlier stage, share information and prevent harm occurring. The vision of the Poutasi report was that multiagency teams would be set up to assess risk identified by agencies before needing to make a report of concern in order to help prevent harm from occurring.
Oranga Tamariki is piloting working more closely with strategic and community partners to respond to reports of concern
Oranga Tamariki told us Ngāti Kahungunu Iwi Incorporated is leading two multi-agency initiatives – Te Ara Hanganga and Te Kura. Te Ara Hanganga is a new programme operating in two Hawke’s Bay locations, with four community organisations and Oranga Tamariki collaboratively assessing reports of concern. Te Kura co-ordinates the assessment and response to family harm incidents across Napier and Hastings, involving approximately 20 government agencies and NGOs.
Te Kahu Oranga Whānau and Te Tēpu in Te Tai Tokerau are also multi-agency response tables that triage reports of concern, referring those that need early support to community organisations and escalating those that require statutory intervention back to Oranga Tamariki. Some community kaimahi felt that all reports of concern should be sent to the tables for triage as they saw inconsistency in the threshold for action at Oranga Tamariki.
Many kaimahi from these tables, from community agencies and Oranga Tamariki, emphasised that communication between them enables good outcomes for tamariki, rangatahi and whānau referred to the table. However, we heard that there is not consistent feedback between Oranga Tamariki and the tables about whether reports of concern have been closed or actioned.
Neither Oranga Tamariki nor the community agencies from these tables routinely hear what has happened once whānau have been referred to a community organisation for community response or back to Oranga Tamariki for a statutory response. This leads to concern about whether anyone has visited and if tamariki are safe.
The Whangārei community table, Te Tēpu, told us about several areas of concern in their interaction with Oranga Tamariki. They said there can be weeklong delays in referrals being sent to their table from sites, due to delays in sites receiving the reports of concern from the NCC.
Police and Te Tēpu kaimahi were also concerned that multiple reports of concern about the same child are not always linked in CYRAS, the Oranga Tamariki administrative database. They said that Oranga Tamariki does not consider the safety of siblings and whānau when responding to a report of concern, instead focusing only on the child the concerns were about.
We also heard that, once a report of concern is referred to a table for community response, it is closed in CYRAS as no further action. If the table determines further intervention is needed such as a statutory response from Oranga Tamariki, it needs to make a new report of concern.
In addition, we heard that, when cases are referred to some of the tables for follow-up, whānau may not consent to working with the community agency. In those situations, a new report of concern must be made as the community agency does not have a mandate to work with the whānau.
We further heard that Oranga Tamariki does not advise the table of the action it has taken on reports of concern referred to it by the table. The practice of prematurely closing the report of concern creates additional bureaucracy and potential delay. Tamariki would be better served if Oranga Tamariki kept the report of concern open until the community table advises the appropriate action has been taken and that tamariki are safe.
In Greater Wellington, we heard about the Hapori Community Intake and Assessment table. One regional leader said that Hapori went further than previous initiatives at Oranga Tamariki because Oranga Tamariki is “not at the head of the table”. The five community agencies29 that make up the table are given reports of concern received by Oranga Tamariki and collectively triage reports of concern and make intake decisions.
“Without the Hapori response, this [Oranga Tamariki site] would fall over. We don’t do it by ourselves, we do it all together.” ORANGA TAMARIKI KAIMAHI
We heard from Oranga Tamariki kaimahi that whānau are more willing to engage with community organisations, and at one site, this has reduced entries into care. However, we understand that not having secure funding makes it difficult to recruit kaimahi and that the table does not have all the resources it needs. At the time of writing, Oranga Tamariki is contracting the Hapori model on a rolling six-month basis as a pilot initiative.
Practice and resourcing across the family violence tables is also variable
Family harm tables are more common than reports of concern tables and exist in most communities. They are commonly made up of Police, Oranga Tamariki and NGOs, with some including iwi/Māori agencies.
The tables meet at varying levels of regularity to discuss recent incidents of family violence. Mostly, the representatives jointly decide the appropriate action based on the nature of the incident but we were told there is significant variation across the tables, and in some cases, the risk and action is decided by a single person in one organisation.
We heard from kaimahi from both SAM and ISR family harm tables about the importance of building relationships and the information sharing that happens in these meetings.
“I get phone calls [from professionals at the table] all the time. It’s always an open door, always an open line between agencies. They don’t hesitate to ask. We don’t have all the answers, we [as differing agencies] do think differently, we do have good relationships with other agencies.” POLICE KAIMAHI
However, we heard levels of resourcing created challenges across the different tables. For example, we heard that ISR tables were better resourced than SAM tables, and in practice, this can limit which iwi/Māori and NGO agencies the SAM tables can afford to contract.
“[The SAM table is] totally different to ISR where they get tonnes of money. We sort of run on faith and hope, which is a challenge.” COMMUNITY AGENCY KAIMAHI
We also heard that, at the SAM tables, each government agency is responsible for funding its own representation, and this can have an impact on which agencies attend and how frequently. We heard that Police put more funding and resourcing into the SAM tables than other agencies in terms of the kaimahi allocated to the SAM tables.
The Centre for Family Violence and Sexual Violence Prevention explained that ISR is an example of a multi-agency response model whereas a SAM table is a meeting and they are funded differently, with ISR receiving the most dedicated funding. ISR directly contracts NGOs and iwi organisations to pick up cases from its SAM table and work with whānau.
All other models do not receive direct funding for NGOs or iwi organisations to pick up cases. However, NGOs receive funding from to respond to family violence in their community more generally.
These different funding models lead to a variation in responses and services available in communities. There is also a lack of visibility of what happens with referrals, as agencies are not funded to enter outcome data across systems.
Kaimahi also told us about the lack of thorough induction for new agency professionals at the table and that they were often left to teach themselves new processes and systems. In some cases, agency representatives had received their own agency-specific induction, but this did not always cover the work of the table.
“The induction programme is not well distributed. It most definitely has not been made available to the SAM agencies. I have heard mixed things in the community about it.” COMMUNITY AGENCY KAIMAHI
The Centre for Family Violence and Sexual Violence Prevention explained that induction practices vary across the sites and that some sites have training and induction as a pre-requisite for participation.
We also heard that sometimes iwi and NGO agencies did not have the capacity to support the tables.
“… It was about capacity. [The iwi] had so many different contracts, and a lot of being in this space is tedious, it can be very ‘same shit different day’ … [It’s] not for a lack of trying and not to say they weren’t willing.” COMMUNITY AGENCY KAIMAHI
Work is happening at a national level to strengthen family violence responses
We heard from the Centre for Family Violence and Sexual Violence Prevention about the work it is doing under the second Te Aorerekura Action Plan30 to strengthen existing multi-agency responses to family violence. In its initial report,31 it summarised key insights, including enablers and barriers to multiagency responses. The themes we heard from the multi-agency tables we spoke with align with these insights such as how the membership of the table impacts its success but that membership varies across the tables, in part due to funding and available resourcing.
The Centre for Family Violence and Sexual Violence Prevention is working with tables in 12 localities to develop system improvement plans that will identify actions to improve multi-agency responses to family violence at the national, regional and local level.
As well as setting up multi-agency teams in communities, the Poutasi report recommended that medical records held in different parts of the health sector should be linked to enable health professionals to view a complete picture of a child’s medical history. This would further widen the view of the risk of harm to and and could support medical practitioners to make reports of concern at an earlier stage and with more information to support an assessment of the risk of harm.
Our 2024 review noted the linking of medical records was expected in 2026. We reported that Health NZ advised the Hira programme will give approved and health providers a comprehensive view of a child’s medical history and health system interactions. We were told the new system will help health providers monitor wellbeing indicators over time regardless of where healthcare is accessed and will give them secure, easy access to a child’s real-time information when needed but that this was still some years away from being realised.
We further noted that development of the business case for Hira began in early 2021. At the time of our last review, funding had only been confirmed for tranche one of the business case, which would make patient summaries available to individuals and their healthcare providers via My Health Record by mid-2024. Tranches two and three of the business case would enable consistent nationwide access to a child’s primary care medical records, but funding still needed to be confirmed for those tranches.
Limited progress has been made towards linking medical records since our last review
The Ministry of Health and Health NZ have been working to link medical records across the health sector since early 2021. This was also recommended in the Poutasi report and in the Ministry of Health’s own report following Malachi’s death. The intent of this work is to enable health professionals to view a complete picture of a child’s medical history, which in turn will allow for a greater assessment of risk.
We asked Health NZ about progress on this work. In response, it told us that this initiative is part of a multi-year programme to link medical records across relevant healthcare settings. Some regions already have record sharing between general practitioner and hospital providers, and functionality. Over the next year, Health NZ will progressively deliver integrated access to hospital and primary care records through existing systems, connected nationally.
The Poutasi report recommended that the health sector should be added as a partner to the CPP between Police and Oranga Tamariki to enable access to health professionals experienced in the identification of child abuse and to facilitate regular joint training.
Our 2024 review noted that the CPP was under review and that a decision had not been reached on health sector involvement in it. Options included full operational membership of the CPP, partial membership in areas such as governance, participation in review and training, and not joining but adopting other measures to enable access to health expertise and services in the context of the CPP.
We were told that a phased approach will be taken to Health NZ joining the CPP. Health NZ told us that its Board agreed in mid-October 2025 to it joining the CPP in a leadership and governance capacity as an initial step.
Health NZ also advised that an updated draft CPP that includes Health NZ is being developed by Oranga Tamariki, Police and Health NZ. However, it cannot confirm when or if full partnership will happen until it fully understands the likely resourcing implications for its frontline staff as part of the next phase.
It noted that, in some regions, additional clinical staff are likely to be required. In this next phase, Health NZ would also consult with clinicians about which and require health involvement and how to implement this practically. Subject to the outcomes of the scoping for this next phase and approval to proceed to full partnership, Health NZ anticipates fully joining the CPP from 1 July 2026.
In his evidence to the Coroner’s inquest into the death of Malachi Subecz, Dr Patrick Kelly, consultant paediatrician at Te Puaruruhau,32 said he supported Health NZ joining the CPP, stating that “in my view, it is the only way we have to work to eliminate the variability in practice shown by both the police and Oranga Tamariki, when it comes to deciding which children should be referred for a health assessment”. However, he raised concerns about the current ability of Health NZ to respond to the need for child abuse assessments.
Dr Kelly noted that there are no dedicated resources for these assessments, despite requests for regional centres with child protection expertise. Starship is the only hospital to have a dedicated, multi-disciplinary child protection team in the country. In areas outside Auckland, when acute assessments of injuries are sought by Oranga Tamariki or Police, they must be “somehow squeezed into the busy inpatient and outpatient clinical workload of general paediatricians or join the queue in equally busy emergency departments”.
Dr Kelly raised concerns about the limited training many frontline health professionals receive in interpreting childhood injuries. He noted that the only national standardised training is provided through the Violence Intervention Programme. However, he also noted that, while this training is mandatory for all Health NZ kaimahi, almost no doctors attend. Similar training for GPs was almost entirely voluntary.
Within his evidence, Dr Kelly raised concerns about Oranga Tamariki practice in relation to the identification of physical injury. Dr Kelly described Oranga Tamariki social workers receiving photographs and making decisions about the safety of tamariki, without any input from experienced health professionals. He also noted that this practice is embedded within the culture of Police.
The Poutasi report noted that a cousin of Malachi emailed Oranga Tamariki and included a photograph of what was thought to be bruising around Malachi’s eye. This photograph was only reviewed by Oranga Tamariki social workers, and the report of concern was closed.
“Oranga Tamariki and the police routinely make their own decisions about the significance of injuries observed in children. Decisions about which children are referred for a medical opinion are arbitrary and widely variable and largely made by statutory officers with little or no training in injury interpretation … One simple and achievable change to information sharing that might make a difference right now, is this. That every time Oranga Tamariki or the police receive a notification which involves an allegation that a child has visible physical injuries, that information (including any photographs of the injuries or possible injuries that the police or Oranga Tamariki receive) must be shared with a health professional with expertise in the assessment of injuries in children”. DR PATRICK KELLY
To achieve this, Health NZ would need to put considerably more resource into training of health professionals and providing adequate child protection expertise in all parts of . It was for this reason that Dr Kelly asked the Coroner to broaden the terms of reference for their inquest into Malachi’s death to ask “whether actions taken by the health system … are sufficient to reduce the likelihood of further deaths occurring in similar circumstances in the future”.
Until Health NZ provides child protection expertise in all parts of Aotearoa, the risks associated with Oranga Tamariki and the Police making health decisions in isolation remains.
The importance of working closely with health professionals was commented on in an Oranga Tamariki child death review.
“While Health is not currently a partner in the CPP, inviting Health to CPP consultations with Police about [the child] could have provided an opportunity to develop a more holistic understanding of the safety risks and wellbeing needs of [the child] at each point in the site’s work with [the child], his family … and to make more informed decisions about [the child’s] future safety.” ORANGA TAMARIKI CHILD DEATH REVIEW
Understanding of current CPP requirements is mixed, with inconsistent responses from Oranga Tamariki kaimahi and stretched Police
We heard concerns from some Oranga Tamariki and Police kaimahi about how the current CPP requirements are functioning on the ground.
Police kaimahi were concerned that some Oranga Tamariki social workers lack understanding of the CPP process and their role and there is inconsistency in responses from the sites. For example, we heard initial joint investigation plans are not always recorded in CYRAS or actioned by Oranga Tamariki kaimahi while the police are conducting their investigations under the protocol. Police kaimahi said that this can result in delays and unaddressed risk to tamariki and rangatahi, particularly when Oranga Tamariki closes cases during delays.
Police kaimahi in one region we visited also noted their own delays, telling us that they too are “swamped” by many CPP investigations.
“The timeliness and the delay in investigation causes kids stress. Kids have disclosed, and six months later, we are still having conversations.” POLICE KAIMAHI
We were told by a few Police kaimahi that individual Oranga Tamariki sites appear to have shifting priorities, with different demands, funding and resourcing that affect how they prioritise serious allegations that require a joint CPP response. We heard that some sites have more understanding of what the Police Child Protection Team does, and some Oranga Tamariki site leaders are more willing to attend meetings and work together than others.
“Some site managers have no idea what CPT is – and they’re ultimately in charge. We have quarterly meetings where we want the site managers and district managers [to attend]. We’ll be there but getting [Oranga Tamariki site managers] there is a struggle.” POLICE KAIMAHI
There is a similar picture in some child death reviews that shows interaction between Police and Oranga Tamariki regarding the CPP.
Five of the 24 who have died since Malachi because of confirmed or suspected abuse by a person who was supposed to be caring for them had reports of concern made prior to the incident that led to their death and resulted in a referral to the Police under the CPP. One of these cases did not meet the threshold for an investigation under the CPP.
For another child aged under 2 who was killed only a few months after Malachi, the CPP was not followed by Police. The PFVDR found that “a report of concern should have been submitted by the Detective or [Child Protection Team and] this would likely have triggered a follow-up response from Oranga Tamariki”. Nonetheless, this death review had no specific recommendations for Police.
“While there is clear evidence of working under the CPP, there were opportunities to work more closely with the Police and Health at key decision points. There was a lack of clarity about Oranga Tamariki’s [sic] role within the CPP which may have influenced our decision about ongoing involvement with [the child] based on the initial Police decision to close their investigation.” ORANGA TAMARIKI CHILD DEATH REVIEW
28 The CPP is the agreement between Police and Oranga Tamariki to work together where abuse or neglect is expected
29 Ngāti Toa, Wesley Community Centre, Taeaomanino Trust, Porirua Centre and ASK – A Safe Kapiti.
30 Centre for Family Violence and Sexual Violence Prevention. (2024). Te Aorerekura | Action Plan 2025–2030:
Breaking the cycle of violence. preventfvsv.govt.nz/assets/Uploads/Second-Te-Aorerekura-Action-Plan.pdf
31 Te Puna Aonui Business Unit. (2024). Understanding the current state of family violence: Multi-agency responses.
Centre for Family Violence and Sexual Violence Prevention. preventfvsv.govt.nz/assets/Resources/Data-and-Insights/Te-Puna-Aonui-Understanding-the-current-state-of-family-violence-multi-agency-responses-2024.pdf