Dementia and End-of-Life Shared Decision-Making Among Older US Adults

Using nationally representative data from 2010 to 2022, this study reveals that while older US adults with dementia are more likely to have advance directives and involve family in care decisions, they face significantly higher decision-making demands near death, yet both groups achieve high concordance between expressed preferences and care received, highlighting the critical need for early planning and caregiver support.

The Gist

The Big Picture: Navigating the Final Journey

Imagine life as a long road trip. As we get older, the road can get bumpier, and sometimes we need a co-pilot. For people with dementia, the "map" (their memory and ability to make choices) starts to fade. Eventually, they can no longer steer the car, and a family member has to take the wheel.

This study looked at the "final stretch" of the road trip for over 5,000 older Americans who passed away between 2010 and 2022. The researchers wanted to see how the experience of saying goodbye differs for those with dementia compared to those without it. They used data from a massive, national survey called the Health and Retirement Study (HRS), which is like a giant, detailed logbook of American seniors' lives.

The Main Characters

• The Decedents: The people who passed away.
• The Surrogates: The family members (usually children or spouses) who had to make the tough calls when the person with dementia could no longer speak for themselves.
• The "Advance Directive": Think of this as a pre-written travel itinerary. It's a document where a person writes down, "If I get sick, I want X," or "I don't want Y."

What the Study Found (The Plot Twist)

1. The "Pre-Planning" Paradox

You might think that because people with dementia lose their memory, they wouldn't have written down their wishes. Surprisingly, the opposite was true.

• The Finding: People with dementia were more likely to have a written "travel itinerary" (Advance Directive) than those without dementia (81% vs. 69%).
• The Analogy: It's like realizing you're going on a dangerous hike, so you write a very detailed map before you get lost. Families often realize the dementia is coming and rush to get the paperwork done early.

2. The Heavy Burden on the Co-Pilot

Even though the "map" (the advance directive) was often there, the journey was still much harder for the family co-pilots of dementia patients.

• The Finding: Families of dementia patients faced more urgent decisions in the final days of life.
• The Analogy: Imagine you have a map, but the road is changing every minute. A person without dementia might have a clear path. A person with dementia has a foggy road where the terrain shifts constantly (infections, confusion, feeding tubes). The family has to make split-second decisions about whether to turn left (hospital) or right (hospice), even if they have the map. The map doesn't tell you what to do right now when the fog is thickest.
• Who took the wheel? For dementia patients, children or grandchildren were the ones making the calls 64% of the time, compared to only 46% for those without dementia.

3. The "Comfort" Destination

Here is the good news. Despite the chaos and the heavy burden on families, the destination was usually the same.

• The Finding: Most people, whether they had dementia or not, wanted comfort care (pain relief, being at peace) rather than aggressive life-prolonging treatments.
• The Result: In both groups, over 90% of the time, the care they actually received matched what they wanted.
• The Analogy: Even though the drive was bumpy and the driver was confused, the family successfully got the passenger to the "Comfort Zone" they wanted. The system worked well enough to honor the general wish, even if the specific turns were hard to navigate.

The Inequality on the Road

The study also highlighted that not everyone has the same access to maps or co-pilots.

• The Finding: Black and Hispanic older adults were significantly less likely to have written advance directives compared to White adults.
• Why? This isn't because they don't care. It's often due to mistrust in the medical system or cultural beliefs that say, "We discuss these things as a family at the table, not on a piece of paper."
• The Analogy: Some communities feel the "official map" is a trap set by the government or doctors, so they rely on their own family compass instead. The study suggests we need to build trust and offer better ways to talk about these plans that respect different cultures.

The Takeaway: What Should We Do?

The authors conclude that having a piece of paper (an advance directive) is great, but it's not a magic wand.

1. It's a Conversation, Not a Form: You can't just fill out a form once and forget it. Because dementia changes the landscape, families need ongoing support to interpret those wishes when the fog rolls in.
2. Support the Co-Pilots: The children and spouses making these decisions are under immense stress. They need better tools, training, and emotional support to make the right calls when the patient can't speak.
3. Start Early: Don't wait until the fog is thick. Have the "travel itinerary" conversation while the driver can still help steer.

In a nutshell: Dementia turns the end-of-life journey into a complex navigation challenge. While families are doing a great job getting their loved ones to a peaceful place, they are carrying a heavier load than necessary. We need to give them better maps, better compasses, and more support along the way.

Technical Summary

1. Problem Statement

End-of-life (EOL) shared decision-making (SDM) faces unique challenges for individuals with dementia. As cognitive decline progresses, decision-making authority shifts from the patient to surrogates (family caregivers). While advance care planning (ACP) is intended to guide these decisions, existing frameworks often fail to address the specific complexities of dementia, such as:

• The prolonged trajectory of illness requiring repeated decisions on life-sustaining treatments.
• The gap between documented preferences (advance directives) and the ability of surrogates to interpret them during acute clinical events.
• A lack of population-level data comparing the SDM processes of decedents with dementia against those without dementia.

The study aims to fill this gap by examining how dementia status reshapes the structure, intensity, and outcomes of EOL decision-making in the US population.

2. Methodology

• Data Source: The study utilized the Health and Retirement Study (HRS), a nationally representative longitudinal survey of US adults aged 50+. Specifically, it analyzed exit interviews collected between 2010 and 2022.
• Study Population:
  • Initial pool: 9,518 decedents.
  • Final analytic sample: 5,389 decedents (representing a weighted population of ~22.2 million US decedents).
  • Inclusion criteria: Complete data on dementia status, EOL decisional needs, advance directives, and demographic characteristics.
  • Exclusion: Individuals without positive sampling weights or missing key variables.
• Variables:
  • Primary Exposure: Dementia status (Yes/No), based on proxy reports of Alzheimer's or other dementia prior to death.
  • Primary Outcomes:
    1. Advance Directive (AD) Completion: Presence of living wills or durable power of attorney.
    2. Decision-Making Needs: Whether important medical decisions were required in the final days of life.
  • Secondary Outcomes: Capacity to participate in decisions, primary decision-maker identity (spouse, child, etc.), and concordance between preferred and received comfort-focused care.
  • Covariates: Age, sex, race/ethnicity, education, marital status, income, cause of death, and expectedness of death.
• Statistical Analysis:
  • Used survey-weighted descriptive statistics to generate nationally representative estimates.
  • Employed design-adjusted Wald tests for group comparisons.
  • Utilized survey-weighted modified Poisson regression with robust variance estimation to calculate Adjusted Risk Ratios (aRR) and 95% Confidence Intervals (CI), controlling for sociodemographic factors.
  • Software: SAS 9.4 and Stata 19.

3. Key Results

• Sample Demographics:
  • 17.7% (weighted) of decedents had dementia.
  • Decedents with dementia were significantly older (49.4% were ≥85 years vs. 32.0% without dementia), more likely to be female, non-Hispanic Black, widowed, and have lower educational attainment.
• Advance Directive (AD) Completion:
  • Higher prevalence in dementia: 81.3% of decedents with dementia had completed ADs compared to 69.1% without dementia ($p < .001$).
  • Adjusted Association: Dementia remained significantly associated with higher AD completion (aRR = 1.11, 95% CI: 1.07–1.15).
  • Disparities: Non-Hispanic Black (aRR = 0.70) and Hispanic (aRR = 0.73) decedents were significantly less likely to have ADs compared to non-Hispanic Whites.
• Decision-Making Needs and Surrogate Involvement:
  • Capacity Loss: 78.9% of decedents with dementia lacked decisional capacity in their final days, compared to 48.0% without dementia.
  • Increased Needs: Decedents with dementia had significantly higher decision-making needs in the final days (54.3% vs. 47.2%, $p < .001$).
  • Surrogate Shift: Children or grandchildren were the primary decision-makers for 63.9% of dementia decedents, compared to 45.6% for non-dementia decedents.
• Preference-Care Concordance:
  • Despite the differences in process, concordance was high and statistically similar between groups.
  • Both groups expressed a preference for comfort-focused care (90%) and received it (93-94%).
  • No significant difference was found in the alignment of stated preferences and actual care received between the two groups.

4. Key Contributions

• Paradox Identification: The study highlights a critical paradox: individuals with dementia are more likely to have completed advance directives, yet they experience higher levels of decision-making burden and reliance on surrogates in their final days. This suggests that documentation alone does not eliminate the complexity of real-time decision-making.
• Surrogate Dynamics: It quantifies the shift in decision-making authority, demonstrating that dementia specifically drives a reliance on children/grandchildren as primary surrogates, increasing the emotional and cognitive load on these family members.
• Population-Level Evidence: Unlike previous studies focused on specific clinical settings (e.g., nursing homes), this research provides nationally representative data, offering a broader view of how dementia impacts EOL care across the US healthcare continuum.
• Disparity Mapping: It reinforces existing racial and ethnic disparities in advance care planning, showing that Black and Hispanic populations remain underrepresented in AD completion regardless of dementia status.

5. Significance and Implications

• Clinical Practice: The findings suggest that advance care planning for dementia must be viewed not as a one-time documentation event but as a continuous, iterative process. Clinicians need to support surrogates in interpreting general preferences for specific, acute clinical scenarios (e.g., infections, feeding tubes) which are common in advanced dementia.
• Policy and Intervention:
  • Early Palliative Integration: Support systems should be integrated earlier in the dementia trajectory to reduce the decisional burden on families.
  • Targeted Interventions: Culturally tailored interventions are urgently needed to address the lower rates of AD completion among Black and Hispanic communities, focusing on trust-building and family-centered decision-making models.
• Future Research: The study calls for further investigation into the temporal evolution of SDM in dementia and the development of frameworks that specifically support surrogates navigating uncertainty when patient capacity is lost.

Limitations Noted: The study relies on proxy-reported data (potential recall bias), uses proxy-defined dementia status (potential misclassification), and utilizes cross-sectional exit data (limiting causal inference regarding the timing of ACP). Despite this, the large, weighted sample provides robust population-level insights.