Background

Antipsychotic medications are regularly prescribed in care home residents for the management of behavioural and psychological symptoms of dementia (BPSD) despite questionable efficacy, important adverse effects, and available non-pharmacological interventions. Prescription rates are related to organisational factors, staff training and job satisfaction, patient characteristics, and specific interventions. Psychosocial intervention programmes aimed at reducing the prescription of antipsychotic drugs are available. These programmes may target care home residents (e.g. improving communication and interpersonal relationships) or target staff (e.g. by providing skills for caring for people with BPSD). Therefore, this review aimed to assess the effectiveness of these interventions, updating our earlier review published in 2012.

Objectives

To evaluate the benefits and harms of psychosocial interventions to reduce antipsychotic medication use in care home residents compared to regular care, optimised regular care, or a different psychosocial intervention.

Search methods

We used standard, extensive Cochrane search methods. The latest search date was 14 July 2022.

Selection criteria

We included individual or cluster-randomised controlled trials comparing a psychosocial intervention aimed primarily at reducing the use of antipsychotic medication with regular care, optimised regular care, or a different psychosocial intervention. Psychosocial interventions were defined as non-pharmacological intervention with psychosocial components. We excluded medication withdrawal or substitution interventions, interventions without direct interpersonal contact and communication, and interventions solely addressing policy changes or structural interventions.

Data collection and analysis

We used standard Cochrane methods. Critical appraisal of studies addressed risks of selection, performance, attrition and detection bias, as well as criteria related to cluster randomisation. We retrieved data on the complex interventions on the basis of the TIDieR (Template for Intervention Description and Replication) checklist. Our primary outcomes were 1. use of regularly prescribed antipsychotic medication and 2. adverse events. Our secondary outcomes were 3. mortality; 4. BPSD; 5. quality of life; 6. prescribing of regularly psychotropic medication; 7. regimen of regularly prescribed antipsychotic medication; 8. antipsychotic medication administered ‘as needed’; 9. physical restraints; 10. cognitive status; 11. depression; 12. activities of daily living; and 13. costs. We used GRADE to assess certainty of evidence.

Main results

We included five cluster-randomised controlled studies (120 clusters, 8342 participants). We found pronounced clinical heterogeneity and therefore decided to present study results narratively. All studies investigated complex interventions comprising, among other components, educational approaches.

Because of the heterogeneity of the results, including the direction of effects, we are uncertain about the effects of psychosocial interventions on the prescription of antipsychotic medication. One study investigating an educational intervention for care home staff assessed the use of antipsychotic medication in days of use per 100 resident-days, and found this to be lower in the intervention group (mean difference 6.30 days, 95% confidence interval (CI) 6.05 to 6.66; 1152 participants). The other four studies reported the proportion of participants with a regular antipsychotic prescription. Of two studies implementing an intervention to promote person-centred care, one found a difference in favour of the intervention group (between-group difference 19.1%, 95% CI 0.5 to 37.7%; 338 participants), while the other found a difference in favour of the control group (between-group difference 11.4%, 95% CI 0.9 to 21.9%; 862 participants). One study investigating an educational programme described as ‘academic detailing’ found no difference between groups (odds ratio 1.06, 95% CI 0.93 to 1.20; 5363 participants). The fifth study used a factorial design to compare different combinations of interventions to supplement person-centred care. Results showed a positive effect of medication review, and no clear effect of social interaction or exercise. We considered that, overall, the evidence about this outcome was of low certainty.

We found high-certainty evidence that psychosocial interventions intended primarily to reduce antipsychotic use resulted in little to no difference in the number of falls, non-elective hospitalisations, or unplanned emergency department visits.

Psychosocial interventions intended primarily to reduce antipsychotic use also resulted in little to no difference in quality of life (moderate-certainty evidence), and BPSD, regular prescribing of psychotropic medication, use of physical restraints, depression, or activities of daily living (all low-certainty evidence). We also found low-certainty evidence that, in the context of these interventions, social interaction and medication review may reduce mortality, but exercise does not.

Authors’ conclusions

All included interventions were complex and the components of the interventions differed considerably between studies. Interventions and intervention components were mostly not described in sufficient detail. Two studies found evidence that the complex psychosocial interventions may reduce antipsychotic medication use. In addition, one study showed that medication review might have some impact on antipsychotic prescribing rates. There were no important adverse events. Overall, the available evidence does not allow for clear generalisable recommendations.