Antipsychotics and risk of falls in people with Alzheimer's Disease and agitation. A self-controlled case series study and cohort study

Date of Approval
Application Number
Technical Summary

We aim to provide evidence of the risk of falls leading to hospitalisation from the use of antipsychotics (AP) to manage agitation in people with Alzheimer´s disease (AD). We will also calculate the excess risk of first falls leading to hospitalisation due to the use of any APs in patients with AD and agitation.

A Self-Controlled Cases Series (SCCS) will be conducted to estimate the incidence rate ratio (IRR) using a conditional Poisson regression model. Only first fall during the study period will be considered in the primary analysis. Time variant variables such as, age, time from AD diagnosis, other AP use, stroke and living situation will need to be accounted for in the analysis.

For each eligible participant, we would classify:
• Antipsychotics as: atypical APs (AAP), AAP in combination with benzodiazepines, typical AP and any AP.
• Exposed periods will be categorised into fully exposed periods followed by a sequence of five 35-day partially exposed periods up to a maximum of 175 days. This will allow to represent a gradual shift from full exposure, to a partially exposed and to an unexposed state (1). Any time in hospital (Immortal time) will be excluded from the exposure period.

A cohort design will be used to calculate the relative excess rate of falls by taking the ratio of the incidence rate (IR) of falls in patients exposed to AP and the IR of fall in those unexposed to AP.

Health Outcomes to be Measured

Primary objective
Given that a fall is an event that could affect the likelihood of a future fall, first fall leading to hospitalisation during the study period is the primary outcome. All subsequent falls will be excluded from the primary analysis and will instead be considered in the secondary analysis. History of falls in the 12 months before the index date will be included as covariate.

Secondary objective
• All falls: first and subsequent falls leading to hospitalisations of more than 1 day during the study period that are separated from a previous hospitalisation by more than 15 days from discharge to new admission.


Nawab Qizilbash - Chief Investigator - OXON Epidemiology - Spain
Bélène Podmore - Corresponding Applicant - OXON Epidemiology - Spain
Agueda Azpeitia - Collaborator - OXON Epidemiology - Spain
Ignacio Mendez - Collaborator - OXON Epidemiology - Spain
Itziar Ubillos - Collaborator - OXON Epidemiology - Spain
Jeff Schein - Collaborator - Otsuka America Pharmaceutical Inc
Julia Agúndez - Collaborator - OXON Epidemiology - Spain
Kristian Tore Jørgensen - Collaborator - Lundbeck Limited
Madhusudan Kabra - Collaborator - Otsuka Europe
Michael Frank Mørup - Collaborator - Lundbeck A/S
MIGUEL DESCALZO - Collaborator - OXON Epidemiology - Spain
Ruth Owen - Collaborator - OXON Epidemiology - Spain


HES Admitted Patient Care