An assessment for an underlying cause of behavior is needed before prescribing antipsychotic medication for symptoms of dementia .  Antipsychotics in old age dementia showed a modest benefit compared to placebo in managing aggression or psychosis, but this is combined with a fairly large increase in serious adverse events. Thus, antipsychotics should not be used routinely to treat dementia with aggression or psychosis, but may be an option in a few cases where there is severe distress or risk of physical harm to others.  Psychosocial interventions may reduce the need for antipsychotics. 
By the time medicinal products are granted a marketing authorisation, knowledge about their safety is incomplete because the clinical trials investigating the safety profile and efficacy are usually performed with a relatively small number of subjects. Furthermore these subjects are selected with regard to special aspects and in general do not represent the average patient in medical practices. Therefore, rare or very rare adverse drug reactions, interactions or other hazards related to the use of a medicinal product cannot be observed during clinical trials. Especially the rarely occurring serious adverse reactions are very important for the overall assessment of a new medicinal product. New findings regarding the safety of a medicinal product may occur even long time after its authorisation and depend on new developments in medical science.