Effects of Discontinuation and Tapering of Antipsychotics in Older Adults Living with Dementia Using the Target Trial Emulation Framework

Study type
Protocol
Date of Approval
Study reference ID
24_004238
Lay Summary

Dementia covers multiple diseases that reduce brain function beyond what is typically expected with ageing, affecting memory and the ability to perform daily tasks. Most people living with dementia experience behavioural and psychological symptoms of dementia (BPSD) such as distress, hallucinations, agitation, and aggression. Antipsychotics are often used to treat hallucinations, delusions, or disordered thinking and therefore, are often used to ease BPSD.

However, while there is some evidence that the use of antipsychotics can lead to modest improvements in BPSD, many studies have raised safety concerns. Studies have found that antipsychotic use by those with dementia has been associated with increase of adverse outcomes such as cardiovascular events, hospitalisation, stroke, and pneumonia. Therefore, guidance states that antipsychotics should only be used by those with dementia in cases of extreme distress, should be short term with regular medication reviews, and should be tapered when a decision is made to discontinue.

Among those who use antipsychotics for mental health conditions, many have reported withdrawal effects, some severe, when discontinuing. The impact of discontinuing antipsychotics use in those living with dementia—especially on safety outcomes such as stroke, death, and hospitalisation—is unknown. This study aims to investigate the safety implications of tapering, abruptly discontinuing, or continuing antipsychotics after varying lengths of treatment duration on risk of death, hospitalisation, stroke, fracture, and pneumonia. As antipsychotic use remains high in the population with dementia, this study may give insights into the safety implications of current guidance and may help better inform clinicians when considering discontinuation.

Technical Summary

Despite safety concerns, antipsychotic use by those living with dementia remains high. Many studies have concluded that often the harms associated with antipsychotics outweigh the benefits for those living with dementia. Clinical guidance states that antipsychotics should only be used in cases of extreme distress, for short periods of time and discontinued within 12 weeks. When discontinuing, it is recommended that the dose is reduced 25-50% every 1-2 weeks, depending on the initial dose. It is unknown what the safety implications of discontinuing antipsychotics are for those living with dementia. This study will provide insights into the safety concerns associated with current guidance to taper and discontinue antipsychotics.

We will investigate the risk of death, hospitalisation, stroke, fracture, and pneumonia after tapering, abruptly discontinuing, or continuing use after 12, 24, and 52 weeks of continuous antipsychotic treatment. Patients will have a first diagnosis of dementia between 1st January 1998 and 31st March 2021 and aged 65+. Comparisons will be made between three treatment strategies: continuing treatment at a stable dose, tapering (reduction of at least 50% from baseline dose followed by discontinuation), and abrupt discontinuation (no treatment for >=28 days). The study has been designed using the target trial framework and separate trials will be run for each of the outcomes and each of the treatment durations. We will use the clone-censor-weight approach in the primary analysis and intention-to-treat in the secondary analysis. As patients those with dementia are often treated with antipsychotics multiple times, the main analysis will be a per-protocol analysis using marginal structural hazards models, approximated with pooled logistic regression. Weights will control for confounding and clones that do not adhere to the treatment strategy will be artificially censored. The models will output adjusted hazard ratios for each trial. Bootstrapping will be used to obtain confidence intervals.

Health Outcomes to be Measured

1. All-cause mortality – identified using primary care records and Office for National Statistics death records. The study will consider the risk of death in the year following treatment allocation.
2. Stroke – identified using Hospital Episode Statistics and primary care records.
3. Osteoporotic fracture (hip, limbs, spine, pelvis) – identified using Hospital Episodes Statistics and primary care records.
4. Hospitalisation – identified using Hospital Episodes Statistics. The study will be considering risk of hospitalisation in the year following treatment allocation.
5. Pneumonia – identified using Hospital Episode Statistics.

These trials will be repeated to look at the risk of these outcomes after 12 weeks, 24 weeks, and 1 year of continuous antipsychotics use.

Collaborators

Wallis Lau - Chief Investigator - University College London ( UCL )
Olivia Bryant - Corresponding Applicant - University College London ( UCL )
Jan Chobanov - Collaborator - University College London ( UCL )
Kenneth Man - Collaborator - University College London ( UCL )

Linkages

HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation