Drug therapy optimization at the end of life.

Older people reaching end-of-life status are particularly at risk of adverse effects of drug therapy. Polypharmacy, declining organ function, co-morbidity, malnutrition, cachexia and changes in body composition all sum up to increase the risk of many drug-related problems in individuals who receive end-of-life care. End of life is defined by a limited lifespan or advanced disability. Optimal prescribing for end-of-life patients with multimorbidity, especially in those dying from non-cancer conditions, remains mostly unexplored, despite the increasing recognition that the management goals for patients with chronic diseases should be redefined in the setting of reduced life expectancy. Most drugs used for symptom palliation in end-of-life care of older patients are used without solid evidence of their benefits and risks in this particularly frail population. Appropriate dosing or optimal administration routes are in most cases unknown. Avoiding or discontinuing drugs that aim to prolong life or prevent disability is usually common sense in end-of-life care, particularly when the time needed to obtain the expected benefits from the drug is longer than the life expectancy of a particular individual. However, discontinuation of drugs is not standard practice, and prescriptions are usually not adapted to changes in the course of advanced diseases. Careful consideration of remaining life expectancy, time until benefit, goals of care and treatment targets for each drug seems to be a sensible framework for decision making. In this article, some key issues on drug therapy at the end of life are discussed, including principles of decision making about drug treatments, specific aspects of drug therapy in some common geriatric conditions (heart failure and dementia), treatment of acute concurrent problems such as infections, evidence to guide the choice and use of drugs to treat symptoms in palliative care, and avoidance of some long-term therapies in end-of-life care. Solid evidence is lacking to guide optimal pharmacotherapy in most end-of-life settings, especially in non-cancer diseases and very old patients. Some open questions for research are suggested.