Secondary Mental Healthcare Use Shifts Around Cognitive Diagnosis in a Large South London Cohort
A large retrospective cohort from a major UK mental health provider examines how secondary mental healthcare use and costs change before and after a diagnosis of mild cognitive impairment or dementia. The study drew on routinely collected electronic health records from 16,081 adults identified through the Clinical Record Interactive Search platform within South London and Maudsley NHS Foundation Trust.
The design is straightforward, but the clinical question is not. A diagnosis of cognitive decline often marks the point at which care becomes more coordinated, more resource-intensive, and more visible to specialist services. In practical terms, that means inpatient episodes, outpatient contact, and memory clinic activity can all shift as clinicians move from diagnostic uncertainty to ongoing management.
By examining the same individuals during the 12 months before and after diagnosis, the analysis avoids the weakness of purely cross-sectional comparisons. It captures each patient as their own comparator, which is especially useful when the question concerns changing service needs over time rather than a single static snapshot.
A Cohort Built From Routine Care Records
The size of the cohort gives the study unusual reach for a service-based analysis. Using routinely collected electronic health records allowed the investigators to study real-world care delivered in everyday practice rather than in a tightly controlled trial setting. That matters in cognitive disorders, where patients often move across services, present with complex psychiatric or behavioural symptoms, and receive care from multiple teams over long periods.
The platform also enabled assessment of utilisation across inpatient, outpatient, and memory clinic services. Those settings do not carry the same clinical meaning. Inpatient care can reflect acute deterioration, behavioural disturbance, delirium, or caregiver strain. Outpatient follow-up may signal surveillance, medication review, or psychiatric comorbidity. Memory clinic contact often sits closer to diagnosis and early planning, when cognitive symptoms are being formally investigated and explained to patients and families.
This approach makes the study relevant to service planners as well as front-line clinicians. A rising burden of cognitive disorders rarely shows up in one service alone; it redistributes work across the system, often without clean boundaries between diagnosis, psychiatric support, and long-term monitoring.
Why the 12-Month Window Matters Clinically
The 12 months before diagnosis can be just as instructive as the period that follows it. In many patients, the lead-up to a diagnosis of cognitive impairment involves escalating anxiety, depression, agitation, behavioural change, confusion, or repeated contact with mental health teams before the underlying syndrome is formally recognized. Once the diagnosis is made, service use may change again as care pathways are reorganized and follow-up becomes more structured.
Looking at pre- and post-diagnosis care within the same person helps isolate those transitions. It also reflects the bedside reality of cognitive disorders: diagnosis is rarely an isolated event. It is often a turning point that reshapes referral patterns, risk management, and the intensity of ongoing support.
What the Data Can and Cannot Prove
The study has important strengths, but its boundaries are equally important. It was restricted to secondary mental healthcare within South London and Maudsley NHS Foundation Trust, so it does not capture primary care, acute hospital care, social care, medication costs, or informal care. Those missing elements are not minor. For patients with cognitive impairment or dementia, a large share of the real-world burden often sits outside specialist psychiatric services, especially in community support, caregiver time, and general medical admissions.
Because of that narrow scope, the findings may not generalise beyond the SLaM catchment area. Service configuration, referral thresholds, case mix, and local practice patterns can all influence utilisation. A mental health provider in another region, or a health system with different memory clinic pathways, may see a very different pattern of contact and cost.
Routine data bring their own hazards as well. Electronic health records are powerful for scale, but they are not pristine research instruments. Missing, incomplete, or inconsistent entries can distort utilisation estimates, obscure timing, or leave uncertainty around how diagnoses were recorded in practice. A record may exist without conveying the full clinical context behind it, and a coding label may not fully capture the patient’s cognitive trajectory.
Service Use as a Proxy for Disease Burden
Secondary care utilisation is more than an accounting exercise. In dementia and mild cognitive impairment, rising service contact can reflect worsening symptoms, caregiver distress, comorbid psychiatric illness, or the consequences of delayed diagnosis. Costs track that burden in a concrete way, translating clinical complexity into pressure on staffing, clinic capacity, and inpatient beds.
The study framework is therefore useful even before the detailed results are considered. It provides a way to map how diagnosis changes the flow of patients through specialist mental health care, and it creates a basis for later work that could link service use to functional decline, crisis presentations, or specific care pathways.
The main limitation is simple and substantial: the analysis only covers what happened inside one mental health trust. Everything else that shapes the lived experience and full economic burden of cognitive decline remains outside the frame.
Why This Kind of Cohort Still Matters
Retrospective observational work like this cannot establish causation, and it does not answer whether any particular intervention reduces costs or service use. It does, however, show how a large clinical system can be studied with precision using existing records. For conditions such as mild cognitive impairment and dementia, where trajectories are prolonged and care is fragmented across settings, that kind of service mapping is often the first step toward smarter planning.
The most useful next questions are operational rather than abstract: which patients drive the steepest rise in specialist use, which parts of care absorb the greatest cost, and how early recognition changes the trajectory of contact with mental health services. Those answers require the detailed results from this cohort, but the study design already makes one point unmistakable — diagnosis is a pivotal moment for specialist mental healthcare demand, and the burden does not sit in a single clinic.