The timing varied by setting
Advance care planning, or ACP, meaning documented preferences about future treatment and care, was often recorded only as death approached in people with Parkinson's disease or atypical parkinsonism, a review of Dutch medical records found. Nursing homes tended to document treatment limits earlier and more completely than hospitals or general practices. Hospitals made their largest updates in the final six weeks, while general-practice records contained less information.
The review covered 189 records: 91 from nursing homes, 58 from hospitals and 40 from primary care. The records came from the Netherlands and represented people who died with Parkinson's disease or atypical parkinsonism. Researchers checked what was documented during the final year of life and again during the final six weeks, with the shorter period treated as a second observation point within the final year. All eligible records were considered; records missing more than 80% of required fields were excluded.
Those represented in the records had a mean age at death of 80.2 years, and 68.1% were male. Parkinson's disease represented more than 70% of cases, while mean disease duration was 6.4 years and the average medication count was 10.9.
The biggest changes came near death
The clearest difference involved do-not-resuscitate decisions. In nursing-home records, DNR documentation was already present for 95% of people a year before death and 100% in the final six weeks. In hospitals, it rose from 53% to 96%, a change with a reported p-value below 0.001. General-practice records rose from 50% to 60%, but entries marked unknown still accounted for 40% at one year and 37% in the final six weeks.
The same late increase appeared in other recorded treatment preferences, including whether someone should be admitted to intensive care or hospital, receive antibiotics, or be given artificial nutrition. The pattern shows that the chart held more treatment decisions near death, but it cannot tell whether those choices were planned in advance or made during an acute deterioration. The researchers used descriptive statistics and McNemar's tests, with Bonferroni-adjusted thresholds for the care-decision and care-discussion comparisons.
Documentation of disease-specific, end-of-life and sedation discussions became more common in nursing-home and hospital records near death. In nursing homes, disease-specific discussions rose from 33.0% at one year to 52.7% in the final six weeks. Discussions about sedation at the end of life rose from 3.3% to 17.6%, while end-of-life discussions rose from 23.1% to 62.6%. In hospitals, the corresponding figures were 32.8% to 65.5%, 1.7% to 15.5%, and 20.7% to 50.0%. These figures describe chart entries rather than confirmed conversations.
General practice looked different. The only discussion category to increase significantly there was preferred place of death, from 17.5% to 35.0%. Other listed discussion topics did not show significant changes. Euthanasia documentation was present for 35.0% at one year and 40.0% in the final six weeks.
What the records can and cannot show
The interventions recorded at the end of life also varied by setting. Palliative sedation was the most frequent intervention overall and in nursing homes and general practice. Hospitals most often recorded antibiotics and fluids. Nursing homes had the broadest documentation of physical, psychological, social and existential palliative-care goals, while hospitals emphasized physical goals. Records for Parkinson's disease and atypical parkinsonism were largely similar, with only small differences.
The results are about documentation, not proof that one setting caused a particular choice. Because the study looked back at records and compared two timepoints, it could not establish whether earlier planning would improve outcomes or whether a late chart entry reflected a newly made decision, a crisis response or a conversation that had simply been recorded late. The general-practice results also show why missing information matters: unknown DNR entries remained common at both timepoints.
Paper data and sources
Original title: Charting the future: A multi-setting medical chart study of advance care planning in Parkinson's disease and atypical Parkinsonism
Authors: Lennaerts-Kats H, Dijkstra B, Elbers A et al.
Journal/Repository: Journal of Parkinson's disease : Duplicate, marked for deletion
Status: Peer-reviewed
First online: 2026-10-13
DOI: Not available
Original paper · Full text