Minimally enhanced usual care participants are provided with print educational materials that include a 44-page brochure developed by the National Institute on Aging entitled "A Guide for Older People: Talking with your Doctor" and a blank easy-to-complete advance directive.

Documentation of advance directive completion in the electronic health record is defined as having a durable power of attorney or a living will documented in the primary care electronic health record. The Medical Order for Life Sustaining Treatment (MOLST) will not be included for this outcome as the completion of a Maryland MOLST is mandatory in certain situations, such as on transfer between settings of care, and is not indicative of having had an advance care planning discussion or naming a durable power of attorney.

Caregivers of patients who die are asked to assess their Decisional Conflict 2-3 months after patient's death. The Decisional Conflict Scale is a 16-item instrument scored on a 5-point Likert scale ranging from 0 ("strongly agree") to 4 ("strongly disagree").The 16 items are summed, divided by 16, then multiplied by 25 to yield scores that range from 0 (no decisional conflict) to 100 (extremely high decisional conflict). Higher scores indicate higher decisional conflict.

Caregivers of patients who die are asked to assess their Decisional Regret 2-3 months after patient's death. This outcome assesses bereaved family-reported decisional regret using a 5-item instrument that assesses the extent to which decision-makers experience regret about care. Response options are assessed using a 5-item Likert scale in which scores of 1 indicate the least regret and 5 the most regret. Scores are then reduced by 1 point and multiplied by 25 for a scale that ranges in value from 0 to 100. Prior studies have categorized scores of 0 as no regret, 1 to 25 as mild regret, and more than 25 as heightened regret. Higher scores indicate stronger feelings of regret.

Caregivers of patients who die are asked to respond to questions about Symptoms of Anxiety 2-3 months after patient's death. This outcome assesses bereaved family symptoms of anxiety using the Generalized Anxiety Disorder 7-item questionnaire (GAD-7). The GAD-7 is a validated instrument that asks about symptoms of anxiety using a two-week recall period with response categories that vary from 0 ("not at all") to 3 ("nearly every day"). Item responses are summed to construct composite scores. Higher scores indicate more symptoms of anxiety.

Caregivers of patients who die are asked to assess their Satisfaction with Care at the End-of-life in Dementia 2-3 months after patient's death using the Satisfaction with Care at the End-of-Life in Dementia (SWC-EOLD). This is a 10-item instrument measured on a 4-point Likert scale that ranges from 1 ("strongly disagree") to 4 ("strongly agree") with a summary score ranging from 10-40 in which higher values indicate higher satisfaction. Higher scores indicate better satisfaction with care at the end-of-life in dementia.

For patients who die while enrolled in the study by 24 months, this outcome assesses the proportion of patients who received burdensome care near the end of life. Burdensome care is defined as any intensive care unit use or life prolonging care (cardiopulmonary resuscitation, mechanical ventilation, tracheostomy, dialysis, artificial nutrition, chemotherapy) within 30 days of death using dates and services abstracted from medical records and the Chesapeake Regional Information System (CRISP), Maryland's Health Information Exchange, which includes all hospital encounters.

Of 273 patient-family dyads enrolled, 128 patients and 128 caregivers were assigned to the control group and received minimally enhanced usual care encompassing an introductory letter, the guide for communicating with clinicians and blank advance directive. A total of 256 participants were randomized to the control group.

Total of all reporting groups

The row population for each row (patient and caregiver) differs from the overall number of baseline participants because each row specifies the number of participants analyzed separately by patient or caregiver.

The row population for each row (patient and caregiver) differs from the overall number of baseline participants because each row specifies the number of participants analyzed separately by patient or caregiver.

The row population for each row (patient and caregiver) differs from the overall number of baseline participants because each row specifies the number of participants analyzed separately by patient or caregiver.

The row population for each row (patient and caregiver) differs from the overall number of baseline participants because each row specifies the number of participants analyzed separately by patient or caregiver.

Patients' and Caregivers' measures were assessed separately.

Patients only

We use the Modified Telephone Interview for Cognitive Status (TICS-m) to categorize normal cognitive functioning (scores of 32-50), mild cognitive impairment (scores of 28-31) or severe (scores of less than 27) cognitive impairment.

Patients only.

Patient only

This measure was only assessed for the caregivers.

The row population differs from the overall number of participants because this measure was only assessed for the caregiver population.

Quality of Life Alzheimer's Disease (QoL-AD) scale is dementia specific and brief and uses the patient's own responses. It has 13 items covering the domains of physical health, energy, mood, living situation, memory, family, marriage, friends, chores, fun, money, self, and life as a whole. These are scored on a 4-point Likert scale, ranging from 1 (poor) to 4 (excellent), with a possible total range from 13 to 52. Higher scores indicate better quality of life.

Patient only

Caregivers were asked to rate their physical health using a 4-item Likert scale (poor, fair, good or excellent). The number of caregivers indicating poor or fair physical health is reported.

Caregivers only

Caregivers were asked to estimate the number of hours in a typical week spent helping the enrolled patient.

Caregivers only

Caregivers were asked to estimate the number of years accompanying the enrolled patient to primary care doctor visits.

Caregivers only

Caregiver burden measured with 13-item Zarit Burden Inventory +1 global item as reported. Each item is scored 0-4 for a range of 0-52,higher scores indicating greater burden.

Caregivers only

Caregiver anxiety was measured using the Generalized Anxiety Disorder (GAD-7) which is a 7-item validated instrument that ask about symptoms of anxiety using a two-week recall period with response categories that vary from 0 ("not at all") to 3 ("nearly every day"). Item responses are summed to construct composite scores; range 0-21. Higher scores indicates more symptoms of anxiety.

Caregivers only

Caregivers were asked to reflect on their interactions with the enrolled patient's primary care doctor and their team in the past 6 months using the CollaboRATE-10 scale. CollaboRATE-10 is a 3-item measure of the shared-decision making process on a 10-point anchored scale, ranging from 0 (no effort was made) to 10 (every effort was made). Participant's scores were summed and multiplied by 3.704 to transform the CollaboRATE-10 scale to 0 to 100. Higher scores indicate better interactions on shared-decision making process between caregiver and patient's primary care doctor.

Caregivers only

Caregivers were asked to reflect on the extent to which they felt a sense of mutual understanding, caring and trust with the enrolled patient's doctor. The 16-item Human Connection Scale (THC) was used to assess therapeutic alliance between patient and physician on a 4-point Likert scale ranging from 1-4 with various response categories depending on the relevance to question. The individual scores were summed; score range from 16-64. Higher scores indicate better sense of mutual understanding, caring and trust with patient's doctor.

Caregivers only

Minimally enhanced usual care participants are provided with print educational materials that include a 44-page brochure developed by the National Institute on Aging entitled "A Guide for Older People: Talking with your Doctor" and a blank easy-to-complete advance directive.

Minimally Enhanced Usual Care: Minimally enhanced usual care participants are provided with print educational materials that include a 44-page brochure developed by the National Institute on Aging entitled "A Guide for Older People: Talking with your Doctor" and a blank easy-to-complete advance directive.

The Advance Care Planning Engagement Survey is a validated patient-reported questionnaire that assesses advance care planning process measures on a 5-point Likert scale. The outcome assesses a 6-item subscale of the Advance Care Planning Engagement Survey that includes parallel items for patient and family participants. The scale for each item is from 1 ("I have never thought about it") to 5 ("I have already done it"), with a range of 6-30 with higher scores indicating higher perceived readiness to engage in advance care planning.

216 represents the total number participants (patients, caregivers) assessed in the intervention group for this outcome at 6 months, and 194 represents the total number of participants (patients, caregivers) assessed in the control group for this outcome at 6 months.

Documentation of advance directive completion in the electronic health record is defined as having a durable power of attorney or a living will documented in the primary care electronic health record. The Medical Order for Life Sustaining Treatment (MOLST) will not be included for this outcome as the completion of a Maryland MOLST is mandatory in certain situations, such as on transfer between settings of care, and is not indicative of having had an advance care planning discussion or naming a durable power of attorney.

Data was not collected.

Caregivers of patients who die are asked to assess their Decisional Conflict 2-3 months after patient's death. The Decisional Conflict Scale is a 16-item instrument scored on a 5-point Likert scale ranging from 0 ("strongly agree") to 4 ("strongly disagree").The 16 items are summed, divided by 16, then multiplied by 25 to yield scores that range from 0 (no decisional conflict) to 100 (extremely high decisional conflict). Higher scores indicate higher decisional conflict.

Caregivers of patients who die are asked to assess their Decisional Regret 2-3 months after patient's death. This outcome assesses bereaved family-reported decisional regret using a 5-item instrument that assesses the extent to which decision-makers experience regret about care. Response options are assessed using a 5-item Likert scale in which scores of 1 indicate the least regret and 5 the most regret. Scores are then reduced by 1 point and multiplied by 25 for a scale that ranges in value from 0 to 100. Prior studies have categorized scores of 0 as no regret, 1 to 25 as mild regret, and more than 25 as heightened regret. Higher scores indicate stronger feelings of regret.

Caregivers of patients who die are asked to respond to questions about Symptoms of Anxiety 2-3 months after patient's death. This outcome assesses bereaved family symptoms of anxiety using the Generalized Anxiety Disorder 7-item questionnaire (GAD-7). The GAD-7 is a validated instrument that asks about symptoms of anxiety using a two-week recall period with response categories that vary from 0 ("not at all") to 3 ("nearly every day"). Item responses are summed to construct composite scores. Higher scores indicate more symptoms of anxiety.

Caregivers of patients who die are asked to assess their Satisfaction with Care at the End-of-life in Dementia 2-3 months after patient's death using the Satisfaction with Care at the End-of-Life in Dementia (SWC-EOLD). This is a 10-item instrument measured on a 4-point Likert scale that ranges from 1 ("strongly disagree") to 4 ("strongly agree") with a summary score ranging from 10-40 in which higher values indicate higher satisfaction. Higher scores indicate better satisfaction with care at the end-of-life in dementia.

For patients who die while enrolled in the study by 24 months, this outcome assesses the proportion of patients who received burdensome care near the end of life. Burdensome care is defined as any intensive care unit use or life prolonging care (cardiopulmonary resuscitation, mechanical ventilation, tracheostomy, dialysis, artificial nutrition, chemotherapy) within 30 days of death using dates and services abstracted from medical records and the Chesapeake Regional Information System (CRISP), Maryland's Health Information Exchange, which includes all hospital encounters.

The control group received minimally enhanced usual care encompassing an introductory letter, the guide for communicating with clinicians and blank advance directive.

The SHARE intervention group received a packet from the primary care practice including a person-family agenda setting checklist, blank advance directive, guide for communication with clinicians, instructions for completing proxy registration for the patient portal and an invitation to a facilitator led ACP conversation.

The control group received minimally enhanced usual care encompassing an introductory letter, the guide for communicating with clinicians and blank advance directive.