Mental healthcare use changes around the first diagnosis of mild cognitive impairment and dementia
A diagnosis of mild cognitive impairment or dementia rarely arrives in isolation. For many patients, the pathway to diagnosis runs through secondary mental healthcare, where repeated assessments, crisis presentations, medication reviews, carer concerns, and safeguarding issues often accumulate long before a formal label is attached. This retrospective observational cohort study from a South London mental health service examines a question with direct service-planning consequences: how often people use secondary mental healthcare before and after diagnosis, and how those patterns translate into cost.
The clinical logic behind the study is straightforward. Cognitive decline does not begin on the date of diagnosis, and service use often reflects that lag. Patients may present with agitation, depression, anxiety, psychosis-like symptoms, sleep disruption, wandering, loss of insight, or functional collapse before anyone formally names the syndrome driving the change. Once the diagnosis is made, the care model often shifts again, with more structured follow-up, memory assessment, family counselling, coordination with primary care, and decisions about ongoing psychiatric treatment. A cohort design is well suited to capturing those transitions because it follows real patients through the same service over time rather than relying on a single encounter.
The study focus also speaks to a persistent systems problem. Cognitive disorders sit at the boundary between psychiatry, neurology, geriatrics, and social care, and that boundary can blur responsibility. Secondary mental healthcare services absorb much of the complexity when symptoms are behavioural, when risk is high, or when dementia is suspected but not yet confirmed. The costs generated in that setting are not abstract accounting exercises; they represent clinician time, specialist appointments, crisis contacts, and multidisciplinary work that can expand quickly when cognitive symptoms become disruptive. Tracking utilisation before and after diagnosis offers a practical view of where pressure lands on the service and where earlier recognition might shift the trajectory of care.
The study’s setting in South London adds another important dimension. Urban mental health services often care for diverse populations with wide variation in language, family structure, deprivation, comorbidity, and access to informal support. Those factors influence when patients present, how quickly diagnosis is made, and how intensively services are used once concerns emerge. In that context, a retrospective cohort drawn from routine care can illuminate the real burden of cognitive disorders as they move through an actual service rather than an idealised trial pathway.
One reason this topic matters is the biological and clinical overlap between psychiatric symptoms and neurodegeneration. Early cognitive decline can masquerade as depression, anxiety, apathy, personality change, or general functional deterioration. In some patients, behavioural symptoms dominate the picture, especially when insight is impaired. Secondary mental healthcare becomes the place where diagnostic uncertainty is managed, often over months rather than days. That uncertainty carries cost, but it also creates an opportunity: repeated contact can reveal evolving cognitive patterns, support family observers who notice the decline first, and trigger diagnostic work-up before avoidable crises occur.
The financial side of the question is equally consequential. If utilisation rises sharply before diagnosis, that pattern may indicate delayed recognition, recurrent acute contacts, and fragmented escalation through the system. If costs remain elevated after diagnosis, the data may reflect ongoing complexity, incomplete care coordination, or the need for longer-term specialist input. Either way, the cost curve helps mental health services understand whether current pathways are containing need efficiently or simply absorbing it later and at greater intensity.
A retrospective design cannot prove causation, and routine-service data rarely capture every determinant of use. Severity of symptoms, informal caregiver support, physical comorbidity, and social instability can all shape demand. Even so, observational work of this kind is often the most relevant for service planning because it measures what actually happens in daily practice, not what ideally should happen. In cognitive disorders, that distinction is not theoretical. It affects clinic capacity, referral thresholds, crisis planning, and the timing of memory assessment.
For clinicians, the practical implication is familiar but important: a change in behaviour, function, or engagement with services may be the earliest operational sign of emerging cognitive disorder. For managers, the question becomes where the system spends most heavily—before diagnosis, after diagnosis, or across both phases—and which contacts might be preventable with faster recognition and better coordination. For families, the study context reflects a common experience: by the time a diagnosis is formalised, the service burden has often already been building for some time.
The most valuable data from this cohort will be the shape of the utilisation curve itself, especially whether costs cluster before diagnosis, after diagnosis, or across both periods in different ways. Those patterns determine whether the next service intervention should focus on earlier case-finding, smoother transition into dementia care, or better management of behavioural and psychological symptoms once the diagnosis is established.