Me & My Wishes: Concordance of End-of-Life Preferences between Residents with Dementia,Family,and Staff

Introduction

More than two million people in the United States reside in long-term care (LTC) settings, such as assisted living communities and nursing homes. ¹ More than half of LTC residents are living with Alzheimer's disease or a related dementia and one-third die from it.^2,3 Despite the importance of understanding care and end-of-life (EOL) preferences for LTC residents living with dementia, routine approaches for eliciting and documenting such information are underdeveloped.^4,5 Some resident preferences and advance directives are collected through standardized assessments at admission, but few LTC settings personalize the approach to ascertain preferences of residents with dementia while they are able to convey them. ⁶

Ascertaining resident preferences and sharing them with informal (family) or formal (staff) caregivers is important so that care preferences of persons with dementia are honored, even when the caregivers have different opinions of what care should include or entail. ⁷ Because family, in particular, may serve as proxy decision makers for persons living with dementia, ensuring they understand residents' preferences is vitally important to assure that care is concordant with the residents' wishes. ⁵

Previous research has shown that persons living with mild-to-moderate dementia can provide information about care preferences in advance care planning conversations.^8–13 When wishes are shared and documented before cognition declines, residents feel involved, family caregivers' feel prepared, and providers' decision making may be better informed to provide resident care consistent with those wishes.^5,14 Although care preferences can be documented in advance directives, more personalized methods, such as videos, can complement such documentation and also include information about a residents' values and preferences that foster person-centered care. Increased understanding of values and preferences in everyday and EOL care can help inform future decision making, comprehensive care planning, and delivery of concordant care. ¹⁵

We previously established the feasibility and acceptability of the Me & My Wishes intervention—a video recording approach for viewing, listening, and discussing daily and EOL care preferences of residents living with dementia. ¹⁶ Our protocolized structured Me & My Wishes conversation guide ensures the video interviews elicit more detailed resident preferences for EOL care that are often missing from other assessment processes ¹⁷ and which may improve the alignment of care with those preferences. In this Stage II ¹⁸ efficacy trial, we examined the concordance of the EOL preferences expressed by residents in their Me & My Wishes videos with family and staff understanding of residents' EOL preferences.

Methods

We employed a randomized wait-list control design to examine the concordance of EOL psychosocial and treatment preferences among LTC residents living with dementia, families, and staff. Assisted living communities and nursing homes were assigned to either an intervention group (resident videos recorded and shared) or a wait-list group (resident videos recorded and shared after all data from the intervention group were collected). The study was approved by the University of Utah Institutional Review Board and registered at ClinicalTrials.gov.

Results

Figure 1 illustrates the flow of enrollment and Table 1 shows participant characteristics. We enrolled 50 residents; our attrition rate was 28%. The 36 residents who completed the intervention on average were 78.4 (standard deviation [SD] = 9.4) years of age and were more likely to be female, widowed, non-Hispanic, and white. Almost half (47.2%) had some college and 27.8% had graduated college or had postgraduate education. Family members (N = 50) on average were 60.7 (SD = 11.5) years of age. Most were female, married/partnered, non-Hispanic, and white. Nearly half of the family caregivers were adult children followed by siblings and the category “Other.” Staff (N = 38) were mainly female, non-Hispanic, and white. Sixty-four percent of staff in this sample had a college education and 13.2% reported having a postgraduate or professional education.

OPEN IN VIEWER

FIG. 1.

Me & My Wishes resident enrollment flow.

Discussion

We found the Me & My Wishes intervention of sharing resident videos of care preferences in care plan meetings efficacious in improving family and staff concordance of understanding the EOL psychosocial and treatment preferences for assisted living community and nursing home residents living with dementia. Our findings align with previous research, including ours, that has shown that persons living with mild-to-moderate dementia are able to express their preferences for everyday and EOL care, and that advance care planning conversations have value.^9–13,20 In addition, older adults living with mild-to-moderate dementia can respond to questions about values and preferences related to quality of life, and treatment preferences such as a breathing tube, and can identify a surrogate decision maker.^10,13

In our descriptive analysis, we found discordance in all areas with the highest discordance related to the preference of receiving life support or having a pet at the bedside. Family and staff response of “I don't know” to a preference before video sharing accounted for nearly half of the discordance in treatment preferences for family and more than half of the discordance in treatment preferences for staff. The percentage of “I don't know” responses by family and staff for EOL and psychosocial preferences were less frequent at time of sharing and follow-up, suggesting family and staff caregivers were more informed about residents' preferences after viewing their videos. However, concordance, especially for family members, was not maintained at follow-up and possibly demonstrates that family know some preferences but not others. ²¹

Staff discordance may be explained in part if staff members completed study surveys without accessing the medical record. We recognize that many staff appropriately look in the medical record to identify the residents' preferences—a good practice as staff care for many residents simultaneously and are not expected to remember all the nuances of their preferences. Me & My Wishes provides residents an opportunity to convey their preferences, although sometimes their responses are ambiguous. Prompting further conversation with the resident to clarify any discrepancies and documenting the conversation in the resident's chart would be a best practice when the resident, care team, or family members report “I don't know” about EOL preferences.

Our findings demonstrate that Me & My Wishes can improve concordance and highlight the need for continued conversations among residents living with dementia, their family, and staff caregivers. ¹⁶ Reviewing care plans with residents quarterly is already a part of regulation, but may not always happen consistently or include the nuances of psychosocial preferences near EOL. To ensure person-centered care is delivered in concordance with the preferences of residents with dementia, their preferences must be routinely assessed and documented to identify changing needs.^20,22 Me & My Wishes videos are designed for ongoing conversation. The four modules do not need to be viewed at one time; one could be viewed during each quarterly care plan meeting. In the event of a significant change, staff could review the most relevant module. Using Me & My Wishes videos routinely could facilitate conversation among residents, family, and staff and identify changing preferences.

Barriers to conversations include the assumption that caregivers know the resident or the resident assumes their caregivers know them, ²³ but our findings suggest that knowing may not be enough as preferences are nuanced. It is possible that residents' decisions of EOL treatment preferences were influenced by acceptable characteristics of the treatments such as if they were temporary, would relieve suffering, ensure a positive outcome, or enabled a return to baseline health. ⁷ Organizational barriers such as the structure of the care plan meeting (e.g., time allotted for each meeting) and team members focused on completing the checkboxes (or current care needs) may influence participants' ability to engage in meaningful conversation. Furthermore, care teams may lack communication skills to engage residents and families in ongoing conversations about preferences. ⁵

Improved understanding of the nuances that accompany preferences potentially improves family and staff knowledge of preferences, which increases the likelihood of providing care and that supports both resident preferences and family caregivers who are concerned about decision making. ²⁴ Although many residents with dementia may already attend their care plan meeting, our findings suggest that sharing their video during the meeting provides an opportunity to ask residents about their preferences or to clarify preferences, for example when residents convey limitations or misunderstanding about treatments.

Concordance may further improve if parts of the video are shown routinely to spark conversation about preferences. Participation in the care plan meeting may be particularly important for staff and family caregivers who may not have had informal or interspersed remarks within casual conversations with residents over time. The coronavirus disease 2019 (COVID-19) pandemic has increased the use of, and comfort with, telehealth, including use by family members to participate in care discussions. Therefore, we hope that family participation in attending care plan meetings where the Me & My Wishes videos are shared and discussed will increase and improve care concordance.

Limitations and Future Research

The results from this analysis should be interpreted in light of the limitations. Randomization occurred at site level and despite randomizing in blocks our sites differed. For example, each site has their own culture and buy-in, which may influence interpretation of preferences.^5,15 The small sample size, the majority of participants being of one racial group (white), and one geographic region limits generalizability. Despite a small sample size, we did detect statistical significance. Because nursing home champions provided a list of eligible residents, rather than researchers conducting a chart review, there may be recruitment bias.

Participant bias may have occurred among all groups. For example, residents and staff who participated in the study may have wanted to engage in conversations. The definition of life support and breathing machine was left open to interpretation by the residents, family, and staff, which may have led to discordance between the groups. We did not examine dialysis or hospitalization and considering these preferences in the future may be helpful.

Future research recommendations include greater representativeness of the study population, examining the documentation surrounding nuanced EOL treatment and psychosocial preferences, and if engaging residents in repeated conversations maintains concordance over time. More specifically, studies will be designed to address lack of concordance (e.g., interviews when staff/family's understanding is not aligned with resident's video; staff training to lead conversations and address “I don't know” responses). Interviews with staff and family may help explain why the understanding or knowledge of preferences was not maintained at follow-up or why conversation was not documented. A larger and/or longer study could also examine actions taken by staff or family compared with their perceived understanding. Finally, establishing a specific place in the medical record to document care conversations would allow information to be queried and compared with regulatory guidelines.

Conclusion

Sharing Me & My Wishes videos in care plan meetings improved family and staff concordance of EOL psychosocial and treatment preferences for assisted living community and nursing home residents living with dementia. These findings build on prior Me & My Wishes research demonstrating that engaging residents in discussion about their preferences promotes person-centered care, can be accomplished by sharing videos during care plan meetings, and is feasible within the established workflow. Because informal caregivers, particularly family, may serve as proxy decision makers for persons living with dementia when they are no longer able to communicate, ensuring they understand residents' preferences is vitally important. As the COVID-19 pandemic heightened the importance of advance care planning, this may be a promising intervention to facilitate conversations and improve goal-concordant care.