Potentially Low-Value Medical Interventions in Frail Older Adults Near the End of Life: A Narrative Review

Frail older adults approaching the end of life frequently receive multiple medications, diagnostic tests, hospital-based treatments, and procedures despite limited physiological reserve, shortened time-to-benefit, and changing goals of care. Low-value care in this setting is not defined solely by treatment cost or intensity. It refers to an intervention that is unlikely to provide meaningful benefit, has burdens that outweigh its expected benefit, offers benefit only beyond the patient's anticipated survival, or conflicts with the patient's values and priorities. This narrative review examines low-value medical interventions in frail older adults near the end of life and considers the evidence supporting deprescribing, de-escalation, and goal-concordant care. The literature was organised into six domains: preventive medications, diagnostic testing and screening, hospital and intensive care interventions, nutrition and hydration, disease-directed treatment, and antimicrobial or symptom-directed therapy. Evidence is strongest for identifying potentially inappropriate preventive medications and reducing medication burden, although effects on survival, quality of life, hospitalisation, and function remain heterogeneous. Tools such as STOPPFrail can support structured medication review, but most deprescribing instruments have been developed largely from expert consensus and have limited clinical validation. Evidence concerning statin discontinuation suggests that stopping preventive statin therapy may be reasonable for selected patients with limited prognosis when cardiovascular benefit is unlikely to be realised within the patient's remaining lifespan. Enteral tube feeding in severe dementia has not demonstrated clear survival or quality-of-life benefits, but available evidence is predominantly observational and affected by confounding. Intensive care, emergency hospitalisation, routine investigations, cancer screening, and invasive procedures may become of low value when they impose substantial burden without a realistic prospect of improving outcomes that matter to the patient. However, these interventions should not be discontinued solely because of age, frailty, dementia, or palliative care involvement. Decisions should incorporate prognosis, treatment indications, time to benefit, symptom burden, reversibility, risks of withdrawal, patient preferences, family perspectives, and available alternatives. The central clinical task is therefore not simply to provide fewer treatments, but to provide treatments that are proportionate, evidence-informed, symptom-responsive, and consistent with the older person's goals. Future research should prioritise patient-reported outcomes, caregiver outcomes, equity, treatment-related harms, and implementation strategies across primary care, hospitals, nursing homes, and hospice services.