A tenant lives in a flat with persistent damp. The damp aggravates a respiratory condition. The condition produces an A&E admission, then a second one six months later. By the third admission the consultant flags the housing situation in the discharge summary. The summary lands in primary care. Primary care has no formal route by which to escalate the housing concern, because the tenant's housing provider sits in a different governance system, with its own regulator, statutory data flows, and no shared definition of what either organisation owes the other.
The episode is not unusual. It is the modal pathway for a particular kind of regulatory and clinical failure, and it sits inside a data architecture which makes the failure invisible to both boards until it produces a regulatory finding, an ombudsman case, or a civil claim.
The boundary is institutional, not logical
I have spent twenty-five years inside the systems on the NHS side of this boundary. Clinical coding, terminology standards, Master Person Service, the structures by which the NHS captures and reconciles data about populations. I now sit on the housing side, building products which touch the same people from the other direction. The view from across the boundary is harder to defend than either sector wants to admit.
Housing data and healthcare data are kept apart for institutional reasons. The two sectors have separate primary legislation, separate regulators, separate funding flows, separate professional cultures, and separate technology estates. The separation is not the result of a deliberate governance decision. It is the result of two parallel histories which never had a reason to converge.
The logic does not support the separation. A person's housing conditions, clinical history, and social circumstances form a single causal system, and the data about it has been split into two halves which neither sector has a clean way of joining. A respiratory consultant treating a tenant does not see the housing record. A repairs operative attending the flat does not see the clinical record. The boards governing each system see only their own half, and the joint accountability sits in a gap neither board is currently equipped to manage.
For most of the last thirty years, the cost of this gap has been carried by the tenant. The position has now changed.
What the recent legislation has done
Awaab's Law, the Social Housing (Regulation) Act 2023, and the maturing Integrated Care System architecture have together altered the regulatory weight applied to the boundary.
Housing providers are now required to investigate and remediate damp and mould within prescribed timescales, with an audit trail per case. The regulatory test is not whether the provider acted reasonably given what it knew. It is whether the provider acted on the information which was available, and increasingly, whether the provider should have sought information it did not hold. A tenant whose A&E admissions are clinically linked to their housing conditions is no longer a fact known only to the NHS. It is a fact a regulator will expect a provider to have considered.
Integrated Care Systems are simultaneously being asked to demonstrate population health outcomes which depend on housing conditions. ICS leadership cannot show measurable improvement in respiratory admissions, falls in older adults, or mental health outcomes without engaging with the housing stock those populations live in. The pooled budget conversations now happening in most ICS regions are the early signal of an institutional rearrangement which will eventually require shared data.
The implication for boards is direct. Governance arrangements designed for the old separation are no longer fit for the new joint accountability. A housing board which cannot evidence what it knew about the clinical context of its damp cases is exposed under Awaab's Law. An ICS board which cannot evidence what it knew about the housing context of its respiratory admissions is exposed against its own outcome targets. Both exposures are live, and most boards now subject to them are not yet managing them.
Why this is harder than it looks
A board which decides to address the gap will find the problem is not technical. The technical mechanisms exist. NHS Number, UPRN, secure data environments, established terminology standards, mature consent frameworks. The infrastructure for a shared data view between a housing provider and a local NHS partner is broadly available, and a board willing to fund it has a working integration within twelve to eighteen months.
The harder problem is governance. Once the two halves of the record are connected, the question of who is accountable for what becomes immediate, complicated, and uncomfortable. If the housing provider knows a tenant's clinical condition is being aggravated by a remediable hazard, and the housing provider does not act, the legal exposure is unambiguous. If both organisations know and the matter falls between them, the joint exposure is shared, and neither board has a contractual basis on which to manage the other's behaviour.
The technical project to connect the two estates is straightforward. The governance project to define the joint accountability is not. It requires both boards to accept they are now operating in a shared regulatory perimeter, and to put in place the agreements, escalation routes, and data sharing arrangements which make the perimeter manageable.
The reason most boards have not started this conversation is the legislation has moved faster than the institutional habit. The Social Housing (Regulation) Act 2023 received royal assent three years ago. Awaab's Law commenced in 2025. The ICS statutory framework has been in place since 2022. The regulatory implications of the convergence have only begun to surface in inspection findings and ombudsman judgements in the last twelve months, and the next round will surface considerably more.
The board-level question worth opening
For a housing board, the question is whether the organisation has any formal data sharing arrangement with its local NHS partners which goes beyond ad hoc safeguarding referrals. If the answer is no, the next regulatory inspection will be entitled to ask why, and the answer "we have not historically held NHS data" is no longer sufficient.
For an NHS board, the question is whether the organisation has any formal data sharing arrangement with the registered providers serving the populations whose health outcomes the system is being measured against. If the answer is no, the same question applies in reverse, framed against population health accountability rather than housing regulation.
For a NED considering an appointment on either side, the question is whether the organisation's risk register has named the joint exposure or has carried it forward by default. Most risk registers I have seen on the housing side have not. The NHS side has begun to address it but inconsistently across regions.
The conversation between the two boards has not started in most localities. It needs to. The data which would link a damp flat to a respiratory admission has been there all along. The institutional permission to draw the link has not. The legislation has now provided the permission, and is steadily turning it into a requirement.
The boards which open the conversation now will manage the joint exposure deliberately. The boards which wait will manage it under conditions set by whichever regulator gets there first.