We would like to commend Sampson et al ^{Reference Sampson, White, Leurent, Scott, Lord and Round1} on undertaking the difficult task of identifying and monitoring behavioural and psychological symptoms of dementia (BPSD) in people admitted to acute medical wards. The authors have also done their best to untangle the BPSD syndrome from similar clinical symptomatology seen in delirium, which still remains an ongoing conundrum for many of us working in liaison psychiatry. The study not only provides a wealth of information, but also raises a number of issues about how BPSD presentation might differ in older people when admitted to acute medical settings and how it influences their outcomes.

The authors used the Behavioral Pathology in Alzheimer’s Disease scale (BEHAVE-AD), ^{Reference Reisberg, Auer and Monteiro2} which has been widely used to detect behavioural problems in people with dementia, in particular Alzheimer’s disease. However, this scale has a number of shortcomings, neglecting some important symptoms of dementia, such as apathy, irritability and/or disinhibition, ^{Reference Robert, Ferris, Gauthier, Ihl, Winblad and Tennigkeit3} that are frequently present in dementia. The low-medium BPSD scores (2.6–4.4 mean BEHAVE-AD severity) are somewhat surprising, since the majority of the enrolled participants had higher Functional Assessment Staging Test (FAST) ^{Reference Reisberg4} staging, corresponding to more advanced stages of dementia. Such low BPSD scores are usually associated with mild cognitive impairment. ^{Reference Reisberg, Auer and Monteiro2} Only five individuals coming from an ‘other’ place of residence (continuing NHS care?) had substantially higher BPSD severity scores. Similarly, the severity of the dementia (as measured via the FAST) did not influence BPSD scores, suggesting that either the medical problems modified the BPSD presentation, or the BPSD were pharmacologically managed. The only factor to have a significant effect on BPSD was presence of delirium, thus highlighting the difficulties in routine clinical settings of differentiating between BPSD and clinical symptoms of delirium.

Nevertheless, Sampson et al’s work undoubtedly indicates that hospitals make people with dementia worse, trebling their paranoia and delusional beliefs, causing hallucinations, making them more aggressive and disturbed, and substantially worsening their moods and anxieties. These findings support the public’s widespread beliefs that hospitals are dangerous places, not only filled with sick people and germs, but with a wide potential for something to go amiss in lieu of wrong. And this ‘wrong’ ranges from having newly acquired diagnosis of dementia when physically unwell to worsening BPSD, ^{Reference Sampson, White, Leurent, Scott, Lord and Round1} further complicating their polypharmacy and making them more frail and with poorer functional outcomes, ^{Reference Runganga, Peel and Hubbard5} as well as increasing their likelihood of death. ^{Reference Sampson, White, Leurent, Scott, Lord and Round1} Not surprisingly, this also affects their formal and informal caregivers. There is a striking discord between the severity of the recorded BPSD and the caregiver’s distress, arguing that the problems around the escalating in-hospital behavioural changes are much more serious than the physical illness itself.

Many of the highlighted BPSD could be easily regulated with non-pharmacological approaches, including better orientation, information and knowing our patients. What is happening to make our hospitals more dementia-friendly? A number of hospitals have already introduced dementia-friendly wards that should be fully equipped with the professional experts in dementia care. However, surprisingly, there is a void of research evidence regarding how the newly introduced dementia-friendly policy in acute medical settings influences the behaviour of people with dementia. Since most of the people with BPSD come from residential and nursing care, ^{Reference Sampson, White, Leurent, Scott, Lord and Round1} one wonders whether there should be another way of introducing dementia-friendly management. Perhaps an ‘in-home health dementia care’ approach would avoid (unnecessary) hospital admissions, and involve medical care professionals treating people with dementia in their own home whenever possible. This would reduce substantially not only the distress that both people with dementia and their caregivers face when in the acute medical setting, but also reduce hospital admissions, and thus result in substantial financial savings. The data from a recent study ^{Reference Treloar, Crugel and Adamis6} on people with advanced dementia stages prove that this can be successfully done, stressing that managing the distress, rather than behaviours that challenge, is central when treating people with dementia. We now need to take these lessons on board and implement them not only within our acute medical wards, but also adapt them to use in the community.