Many of the elderly in nursing homes are very ill and have a reduced quality of life. Life expectancy is often hard to predict. Decisions about life-prolonging treatment should be based on a professional assessment of the patient’s best interest, assessment of capacity to consent, and on the patient’s own wishes. The purpose of this study was to investigate and compare how these types of decisions were made in nursing homes and in hospital wards. Using a questionnaire, we studied the (...) decision-making process for 299 nursing home patients who were treated for dehydration using intravenous fluids, or for bacterial infections using intravenous antibiotics. We compared the 215 patients treated in nursing homes to the 84 nursing home patients treated in the hospital. The patients’ capacity to consent was considered prior to treatment in 197 of the patients treated in nursing homes and 56 of the patients treated in hospitals. The answers indicate that capacity to consent can be difficult to assess. Patients that were considered capable to consent, were more often involved in the decision-making in nursing homes than in hospital. Next of kin and other health personnel were also more rarely involved when the nursing home patient was treated in hospital. Whether advance care planning had been carried out, was more often unknown in the hospital. Hospital doctors expressed more doubt about the decision to admit the patient to the hospital than about the treatment itself. This study indicates a potential for improvement in decision-making processes in general, and in particular when nursing home patients are treated in a hospital ward. The findings corroborate that nursing home patients should be treated locally if adequate health care and treatment is available. The communication between the different levels of health care when hospitalization is necessary, must be better. ClinicalTrials. gov NCT01023763 [The registration was delayed one month after study onset due to practical reasons]. (shrink)