Dreamss Limited v Ofsted

We came to the following conclusions:

We set out further reasoning below. What follows is not designed to cover all the matters touched on in documents before us or within the Hearing. It is, however, designed to be sufficient to clarify the core reasons why we reached our decision.

Bearing in mind the principles set out further above, and without going as far as to find facts, we reached the views set out below on the evidence before us. Stepping back and trying to look at things in the round, we are satisfied that, as of today, there are reasonable grounds to believe that the provision of the relevant service by the Appellant may expose children to a material risk of significant harm. We rely on the following in support of this view.

The Appellant is registered as a supported accommodation undertaking, registered to provide accommodation with support in “category 1” and “category 2”. Category 1 relates to supported accommodation in a self-contained unit, where accommodation is for the sole user of the young person placed there, or for the young person and others that may live there are part of their family unit (e.g. partner, sibling, children). Category 2 relates to supported accommodation in a shared or group living situation, where the young person has their own bedroom and shares communal areas (e.g. living room, kitchen).

Both the Appellant and the Respondent must have regard to the Guide. It provides: “Where a young person has complex needs and/or requires a greater level of ongoing care and supervision, we do not expect that supported accommodation would be appropriate. However, supported accommodation should be flexible enough to accommodate temporary increases in support for young people who would otherwise manage well in this type of provision, enabling placement stability where appropriate.”

The Guide also provides (among other things): “Looked after children and care leavers are often some of the most vulnerable children and young people in society, and we must work together to do all that we can to ensure that they have access to suitable accommodation that can meet their needs and keep them safe. For most children who are not yet ready for greater levels of independence, and especially those who require increased care due to high needs, or who have additional needs, this is best achieved through a placement in foster care or a children’s home, for which there are already robust approaches to approving, registering and quality-assuring provision.”

The evidence indicates, in our view, that most (perhaps all) of the children referred to in the evidence probably have had a level of need/complexity/risk which means they have not been suitable for supported accommodation. While we do not consider it necessary to refer to all relevant information, we shall provide a summary.

It seems likely that Child A was being provided by the Appellant with one-to-one support and, at least at times (due to a propensity to act aggressively towards others), two-to-one support. Child A seems to have had safeguarding and public protection needs and vulnerabilities that would not typically be expected for a child living in supported accommodation. The evidence appears to indicate that Child A may have been the victim of an unreported stabbing a few years ago, and may well have had or still have (or be at risk of) links to gangs and the supply of street drugs. Child A’s risk assessment referred to “high risk” regarding emotional wellbeing, with references to sudden mood shifts, a history of potential self-harm, speech and language assessment, among other things. Similarly, it referred to a “high risk” in relation to experience of abuse and violence. It included a reference to a recent arrest for affray and possession of a bladed article, where it is said Child A was “chased” by “armed males” in what may have been a gang-related matter. The risk assessment put the risk regarding “associations and locations” at “high”: with references to Child A maintaining “secretive, high risk associations”. It put the risk “to others” at “high”, including an alleged assault on a pregnant woman, and recent “knife threats” (and to a bag containing a dangerous knife and/or a previous attempt to order a knife). It put the risk regarding “substance misuse” at “high”, perhaps rising, with a refusal to engage with drug/alcohol services. It put the risk relating to going missing at “medium”, with a reference to three recent missing episodes (from about a year ago), in the context of a wider history of going missing.

It seems likely that Child B was being provided with one-to-one support and, as such, could not be said to be living alone. Child B appears to have been admitted to solo accommodation after making threats to stab a staff member and/or peer in a previous placement. It is reported that Child B experiences, or has in the past experienced, paranoia arising from substance misuse (believing, for example, that people are coming for him during the night, or similar). Child B’s risk assessment identified “substance misuse” as “high”, with references to reports of cocaine use (as a coping mechanism). There are also reports, among other things, of Child B self-harming, of recent property damage (when frustrated), and of being aggressive/threatening.

It seems likely that Child C was being provided with one-to-one support. Child C appears to have initially been provided with two-to-one support by the Appellant, before it was reduced to one-to-one. Child C’s risk assessment was “high” for violence/aggression towards others. There are also reports, among other things, of historical self-harm.

It seems likely that Child D was being provided with one-to-one support. There are references to problematic peer influences and of relatively high risk of exploitation (with a reference to a vulnerability to “grooming” and “county lines”). Child D was assessed, in formal risk assessment documentation, as being at “high” risk in relation to substance misuse, with references to a history of drug use and risk of relapse. There are references to allegedly engaging in theft, vandalism, and other anti-social acts, especially in response to peer pressure. The risk of going missing was put at “medium”. Child D appears to have recently – and perhaps still – been subject to bail conditions. There is a reference to a knuckleduster being found in Child D’s room, relatively recently, and to being found in public with a dangerous knife this calendar year. There is reference to an upcoming court hearing.

It seems likely that Child E was being provided with one-to-one support. The evidence indicates that Child E previously lived in the Appellant’s group living premises where Child E had engaged in several incidents of verbal and physical aggression, substance misuse, and self-harm. The move into solo accommodation was a response to apparently escalating needs/risk, due to perceived issues and risks relating to cohabiting with others. There is also evidence to suggest that Child E had been found in possession of a bladed weapon, involved in selling street drugs, and at risk of child criminal exploitation.

It seems likely that Child F had been provided with one-to-one support and, at times (due to an apparent propensity to make false/serious allegations against staff), two-to-one support. We were told that the Appellant had recently (on 16 March 2026) given notice to end the placement. The relevant email set out why Child F’s needs were (by then) judged to have surpassed the scope of supported accommodation. The Respondent’s evidence is that Child F’s needs, as recorded before the notice to terminate, had appeared to be the lowest of the children reviewed at the Visit.

It seems likely that Child G had been provided with one-to-one support. Child G’s risk assessment documentation has put the risk regarding criminal exploitation / weapons as “high”. It is recorded that not long before being placed at the Appellant, Child G had been involved in a robbery and was caught with a highly-dangerous knife. He appears to have had a GPS tag and a curfew, among other things, as a result. The Appellant told us that Child G had been an emergency referral and that the child had generally engaged well with them during the placement. We note, of course, the evidence referred to further above about recent events connected to the apparent shooting. We are told that the Appellant has recently given notice on Child G’s placement, after it came to a view that, in light of recent matters, it could no longer keep Child G sufficiently safe and that Child G would be better placed elsewhere. The Appellant maintained the position, however, that, prior to the shooting incident, there was no material risk or significant concern relating to Child G, as any earlier such risk and concern had reduced over time in the placement.

In addition, there is documentary evidence in the relevant risk assessments which indicates that: (i) one of these children had a live “referral order” in relation to a sexual offence involving another child in 2023; (ii) one was assessed at “medium” risk of being exploited sexually, referring to ongoing/recent police investigations; (iii) one has been assessed as posing, or to have recently posed, a “high” risk of sexual harm towards children under 16, with associated bail conditions (after, it seems, indecent images of children had been found) prohibiting contact with other children under 16; and (iv) one may have been the alleged perpetrator or victim of a serious sexual offence in relation to a previous placement.

The Appellant submits that some of the risk assessment documents relating to these children contain incorrect information and refer to steps being taken which did not reflect practice on the ground. It is said this is due to poor managerial oversight and/or the misuse of AI tools or similar. Some of these risk assessments, for example, refer to children being made subject to highly-restrictive levels of surveillance and other supervisions or interventions, such as the monitoring of internet use on phones or supervision in the community, which, according to the Appellant, are not, and have never been, in fact in place. While it may be correct that such documentation includes some inaccurate information, there was sufficient evidence, in our judgment, to indicate that many if not all of the children referred to above had needs which, in summary, were too complex and/or gave rise to risks which were too serious to be suitable for supported accommodation. Put another way: many if not all of these children would appear unlikely to have had the necessary level of independence required for category 1 supported accommodation.

We were further bolstered in forming such a view by the Respondent’s evidence on the nature of the support and/or “care” the Appellant was providing for these children and the associated costs of the same. Ms Forster’s evidence, which we accept as likely to be broadly accurate, is that the average cost, nationally, of a child placement in supported accommodation is between £600 and £1,500 per week. The average weekly cost of the same at the Appellant is, we are told, between £2,000 and £5,000, with evidence of a placing authority paying over £9,000 per week. That is striking. The amount a placing authority is paying for a child placement does, in our view, provide a further useful indication of the level of the child’s need. In Ms Forster’s view: a child with very low needs may cost around £300 per week; a child with more complex needs but still in scope of supported accommodation may move into a group home (category 2) with a staff member on duty 24 hours a day at a cost of around £1,000 per week. A child with complex needs, requiring intensive support/supervision will cost about £5,000 per week. A child living in category 1 solo supported accommodation, with remote support, would typically require no more than 20 hours of support a week, with the average being 2-10 per week.

There appears to be a pattern of the Appellant taking on children with relatively high needs/complexity/risk, and providing higher than typical levels of “support” (or in some cases, given the extent of supervision provided, “care”) on an extended basis. It does not appear to be a situation where the Appellant has generally tended to take on children who at the point of admission already had high or increasingly levels of independence and who then, for some temporary period of time, require a higher level of support – or some targeted intervention – to deal with some particular issue which has arisen. The evidence suggests a pattern of routinely taking on individuals with associated high risks. This pattern may well have emerged from a lack of understanding, at the Appellant, about the limits of what supported accommodation is designed for and/or what is actually permitted under the relevant regulations and guidance.

One might have some sympathy for the Appellant as, on its case, it has always, since registration, operated in the way is has been doing (i.e. the core “model” has remained consistent) and the Respondent raised no serious objection at the point of registration or, moreover, at the first inspection (when, indeed, it judged the Appellant to be operating in a manner which merited the most positive outcome). However, such things appear to us to indicate that the Respondent was simply not adequately aware of the apparent pattern of admissions and ongoing practice at earlier points. Had it been aware earlier of the apparent reality on the ground, it would have likely taken some form of action earlier. The inspection appears to have only focused, in any detail, on two cases (i.e. a small sample) relevant to the area of the Appellant’s operations we are concerned with in these proceedings (and neither case, when considered in context, gave rise to significant concerns). The Respondent then looked at a wider range of cases in the Visit and, by the time of the Decision, had a more reliable evidence base and a better understanding of the likely reality. By then, it was increasingly clear to the Respondent that the Appellant was providing very extensive support (and/or care), for extensive periods of time, for a large proportion of the children within this service.

The Appellant also submitted that the position, in the Guide and/or elsewhere, as to the solo accommodation requirement (category 1) was not clear, or at least not clear to it, particularly with regard to the fact that the prohibition on other people routinely being at the accommodation alongside the child extended to staff. Ms Russell said that recent changes have now clarified the position, so the Appellant is fully aware, now, that staff providing ongoing one-to-one support, including overnight, would cause the Appellant to act out of scope of the supported accommodation category. The Appellant is further aware now, it submits, that any previous understanding, held by it, that it had effectively changed its category was misplaced.

In our judgment, the Respondent had reasonable grounds to have believed, and to have sustained such a belief, that admitting additional children, in this current period of time, would put them or other children at a material risk of harm. In essence, supported accommodation is not designed for children with the type and level of risk that children the Appellant has been admitting and continuing to support have. The Appellant may well not have staff (or sufficient staff) with the right level of qualification, skill, knowledge or experience to manage such risks adequately (with the requirements regarding qualifications being less stringent than, for example, in children’s homes). The Appellant may well not have the right tools or powers required to manage such risks adequately (when compared, again, for example, to a children’s home). More widely, the regulatory framework is not designed to manage such risks within supported accommodation arrangements.

Further or in the alternative:

Being satisfied that the risk threshold is met, we turn to this second question about proportionality. In summary: in all the circumstances, we are satisfied that the Restriction is objectively justified as being both reasonable and proportionate.

First, the harm contemplated is more than merely trivial or minor: it is, in our view, sufficiently significant/serious. It could, realistically, include a significant level of physical, sexual, emotional, psychological, or developmental harm. There is very recent evidence of a child apparently being shot. While the circumstances relating to that incident are, as yet unclear, and while any police/criminal procedure will follow its course (we are told there is an ongoing attempted murder investigation), this is a solid indicator of a high level of potential harm. We think that the likelihood of the risk of harm materialising, to one or more child, is not yet low enough. Whilst there is some evidence of developing insight, whilst one or two children have now moved out, and whist the Appellant’s witnesses gave oral evidence of the Appellant doing some good and positive work more widely, the balance of the evidence suggests a likely pattern of problematic admissions to this supported accommodation provision over an extended period of time. At this time, in light of the other evidence, those positive things are not sufficient to reduce the overall risk profile enough.

Second, there is a legitimate aim engaged in relation to the measure adopted. It relates, ultimately, to the protection of vulnerable children from a risk of significant harm. The Restriction relates directly to the pursuance or achievement of that aim. The high importance of the aim means that the Restriction is more readily justified as proportionate. Its immediate purpose is to keep children safe during a (limited) period of time in which the Respondent can conduct further assessment and/or the Respondent/Appellant can take further steps necessary to reduce the relevant risk to an acceptable level. Importantly: the Respondent is planning another monitoring visit (or similar interaction) soon, as early as next week, to further assess and/or to manage any relevant ongoing risk. The Restriction is justified, at present, while the Respondent carries out further assessment and/or mitigation. Further steps may need to be taken, by the Respondent or the Appellant, to manage any relevant risk. Alternatively, it may be possible for the Respondent to lift the Restriction, if satisfied at the further upcoming visit that the risk (particularly that arising from the policy and practice regarding admissions) is by then reduced. The Respondent is under a duty to keep the Restriction under review.

Third, we have tried to have careful regard to the adverse impact of the Restriction on the Appellant and any other relevant people. We can see that, generally, it may well have an adverse impact. It may well reduce its income, for example. That said, there is a lack of direct or persuasive evidence from the Appellant about the nature and degree of any such impact. There is no sufficient evidence before us that the Appellant, or any relevant children or others, are or will be, as a result of the Restriction, disproportionately impacted. There is no financial evidence before us. The Restriction is limited in both scope and time. Importantly, the Respondent decided to limit the intervention to a time-limited restriction on additional service users. The Appellant submitted that the Restriction would have an adverse impact on the availability of suitable placements, in the wider context of a sector where there’s a problem with such availability. The Respondent, however, told us there was ample provision of the specific type of supported accommodation placements relevant to this case (with the key shortfall relating to children’s home beds). We preferred the Respondent’s evidence on this point. Overall, we are content, on the evidence available, that the importance of the legitimate aim being pursued outweighs the adverse impact at present. In forming this view, we have in mind the limited period of restriction and the obligation to actively keep it under review.

Fourth, we are satisfied that the legitimate aim cannot realistically be achieved adequately by a less-restrictive measure. It therefore appears appropriate and reasonably necessary to continue the Restriction for at least some additional time. Our decision is based, as it must be, at the time of the Hearing and on the basis the Restriction is due (in principle at least) to end relatively shortly.

In our judgment, the Decision was rational, proportionate, and lawful; and it continues, for now at least, to be so.

The Appeal is dismissed. The Decision is confirmed.