PLoS Medicine: Packages of Care for Dementia in Low- and Middle …

1276030812 88 PLoS Medicine: Packages of Care for Dementia in Low and Middle ... The Evidence on the Management of Dementia Top

The principal goals of management of dementia are: early diagnosis; optimization of physical health, cognition, activity, and wellbeing; detection and treatment of BPSD; and the provision of information and long-term support to carers. The evidence base for dementia care comes, overwhelmingly, from HICs (Table 1). All the studies discussed below refer to HICs, unless otherwise specified.

Detection and Diagnosis of Dementia

Many cases of dementia, particularly in LMICs, go undetected,in part because of lack of awareness. Awareness of this disorder can be boosted by dissemination of information from governments, health care providers, and media. Help-seeking can be encouraged by improved case-finding. In India and Brazil, for example, community health care workers could, with a few hours training, identify dementia in the community with a positive predictive value of 66% [16],[17]. Population screening for dementia is not considered cost-effective even in HICs [18]. Selective screening can be accomplished by cognitive testing or by informant report of cognitive and functional decline. The Mini-Mental State Examination [19] is widely used in HICs, and adapted versions have been developed for use in many LMICs [20]–[22]. However, this assessment takes 10 min to administer and is prone to educational and cultural bias [23],[24]. A systematic review of brief screening assessments identified three tests (the General Practitioner Assessment of Cognition, Mini-Cog, and Memory Impairment Screen) that took less than 5 min to administer and were considered suitable and valid for routine use in general practice [25]; none of these tests has been validated in LMICs.

Dementia diagnosis requires cognitive testing, clinical interview, informant interview, and physical examination. The 10/66 Dementia Research Group’s culture- and education-fair diagnostic protocol is validated in many LMICs [26],[27], but requires adaptation for clinical use. Practice guidelines in HICs [18] advocate a routine “dementia screen” to exclude treatable causes, which includes haematology, biochemistry, thyroid function tests, vitamin B12 and folate levels. The feasibility and cost-effectiveness of this approach needs to be tested in LMICs.

Finally, a recent systematic review based on evidence collected in HICs identified good practice for disclosing dementia diagnosis. This practice should include: preparation; integrating family members; exploring the patient’s perspective; disclosing the diagnosis; responding to patient reactions; focusing on quality of life and wellbeing; planning for the future; and communicating effectively [28]. The person assessed should be asked if they, and/or others wish to be told the diagnosis. If so, they should be given information about the signs, symptoms, and course of dementia, available treatments, care and support services.

Physical Assessment

To avoid missing underlying conditions, a physical assessment is recommended before specific treatments for BPSD are initiated [29], and should be a regular part of care for all patients. BPSD may sometimes be caused by pain, constipation, and urinary tract infection, although such associations are not always observed [30]. Pain is common and poorly controlled in severe dementia [31]. Hearing and visual impairment impede communication and exacerbate disorientation, and deafness predicts rapid cognitive decline [32],[33]. Visual and auditory impairment can be associated with hallucinations and delusions [34],[35]. Studies indicate that all these impairments are overrepresented in people with cognitive impairment [36]–[38].

There have been very few trials of the effects of physical assessments and interventions on the course of dementia. In a randomised controlled trial (RCT), pain assessment among nursing home residents with dementia was associated with increased analgesic use, reduced pain, and improvements in staff morale [39]. Uncontrolled studies show that audiological assessment is feasible, that hearing aids can be beneficial [40], and that referral to an optician to improve visual acuity may reduce visual hallucinosis [34]. Nutrition is often impaired because of apathy, aversive feeding behaviours, poor dental health, and dysphagia. Although difficult to sustain, nutritional supplementation improved nutrition among nursing home residents [41]; nutritional education for carers had the same effect in the community [42]. A “vascular care” secondary preventive intervention for people with dementia and cerebrovascular disease (part of current good practice guidelines [18]) that addressed hypercholesterolemia, hypertension, smoking, obesity, exercise, and micronutrient deficiency had no impact on subsequent cognition, disability, or institutionalisation [43].

Psychological Treatments

A well-conducted RCT of cognitive stimulation (reality orientation, games, discussions based on information processing rather than knowledge) conducted in the United Kingdom as a group intervention [44], and a small pilot trial from Brazil [45], suggest that cognitive benefits from this intervention are similar to those for cholinesterase inhibitors (ChEIs). More specific cognitive training produced no benefits [46]. Cognitive rehabilitation, an individualised therapy designed to enhance residual cognitive skills and cope with deficits, showed promise in uncontrolled case series undertaken in HICs [46]. A meta-analysis of four trials of reminiscence therapy (the discussion of past activities, events, and experiences) [47] provides evidence for short-term improvement in cognition, mood, and carer strain, but the quality of these trials was poor.

Pharmacological Treatments

Targets for pharmacological treatment include cognitive impairment, behavioural symptoms (agitation and aggression), and psychological symptoms (depression, anxiety, and psychosis).

There is a strong evidence base for the efficacy of ChEIs (donepezil [48], rivastigmine [49], and galantamine [50]). The use of each of these drugs is associated with modest and comparable improvements in cognitive function, global clinical state, and activities of daily living [51]. The evidence base for ChEIs from LMICs is limited to one small RCT of donepezil in Brazil [52] and open-label trials of galantamine in Brazil [53] and China [54]. The efficacy of this class of drugs in severe dementia is unclear, although useful cognitive benefits were identified for galantamine [55]. A fourth drug for the treatment of cognitive impairment, memantine, has a different mode of action, and is well tolerated, but evidence for its efficacy is limited to people with moderate to severe dementia [56]. ChEIs and memantine are less efficacious in vascular dementia than in other forms of dementia [57]. Their efficacy for the treatment of disturbed behaviour is not established; manufacturer-sponsored licensing trials [51] and post hoc analyses [58] indicate small improvements that have not been confirmed in independent trials [59] and meta-analyses [56].

Meta-analyses of RCTs of haloperidol [60] and atypical antipsychotic drugs for the treatment of agitation [61] and BPSD [62] indicate small treatment effects, most evident for aggression [62],[63]. Atypical antipsychotic drugs have also been widely prescribed for psychosis in dementia, but a meta-analysis of their efficacy indicated that only aripiprazole had a statistically and clinically significant effect [62]. Use of these drugs in dementia is associated with an increased risk of death and cerebrovascular adverse events [62],[64]–[68].

The benefits of antidepressant treatment in older people are clear [69], but a meta-analysis of their efficacy in people with dementia was inconclusive [70]. Only two small trials were included in this meta-analysis, one of which suggested a benefit of sertraline for some depression outcomes [71]. Antidepressants have also been proposed for the treatment of BPSD. A meta-analysis of two small RCTs of trazodone was inconclusive [72]. Citalopram showed efficacy over placebo for the treatment of agitation in one small RCT [73], while sertraline showed no benefit on any primary behavioural endpoint [74].

A systematic review of trials of anticonvulsants to treat BPSD found sodium valproate to be ineffective [75]. Carbamezepine may be more promising with large benefits noted for global clinical outcomes and agitation in one small parallel group trial [76] and more marginal effects in a small pilot trial [77].

Sensory Therapy

Various sensory therapies have been proposed as treatments for BPSD but the evidence base for this approach is small and limited by the poor quality of the trials. Current evidence does not support the use of bright light therapy [78] or multisensory stimulation [79]. One small RCT of aromatherapy suggested considerable benefit across a range of behavioural outcomes [80]. Another small but well-conducted trial suggested that hand massage may be effective in reducing agitation [81],[82]. More evidence is required to confirm efficacy, exclude the possibility of harm, and define the optimal content and mode of administration of sensory therapies in both HICs and LMICs, where the approach is untested.

Carer-Focused Interventions

A large literature attests to the benefits of carer interventions in dementia [83]. These include: psychoeducational interventions, often including carer training; psychological therapies such as cognitive behavioural therapy (CBT) and counselling; carer support; and respite care. Many interventions combine several of these elements. There are several systematic reviews and meta-analyses of these interventions [84]–[88]; all of the constituent trials in these studies were conducted in HICs, and many were nonrandomised [84]. Outcomes studied include carer strain, depression, and subjective wellbeing; behaviour disturbance and mood in the care recipient; and institutionalisation. Most carer-focused interventions seemed to reduce carer strain and depression, CBT having the largest impact on depression [84]. Psychoeducational interventions required the active participation of the carer (for example, in role-playing activities) to be effective [84]. Carer support increased carer wellbeing but no other outcomes [84]. For respite care, three methodologically flawed RCTs showed no benefit on any outcome [87]. However, nonrandomised studies suggest that respite care significantly reduces carer strain and psychological morbidity [84]. Interventions targeting the carer may also have small but significant beneficial effects upon the behaviour of the person with dementia [84]. A systematic review of ten RCTs indicated a 40% reduction in the pooled odds of institutionalisation [88]; the effective interventions were structured, intensive, and multicomponent, offering a choice of services and supports [84],[88]. Two small trials in LMICs of a brief carer education and training intervention, one from India [89] and one from Russia [90] indicated much larger treatment effects on carer psychological morbidity [89] and strain [90] than typically seen for such interventions in HICs.