Advance care planning: empowering older frail people to document their end of life wishes

Introduction

Advance care planning (ACP) enables people to make decisions about their health and end of life care should they become seriously ill and unable to communicate those decisions.¹ It provides an opportunity for people to consider and discuss their preferences, values and beliefs with caregivers and health professionals to guide future medical decision-making.² It minimises conflict and confusion around treatment options, reducing stress and anxiety for families.³ While ACP should be encouraged for everyone, it is particularly pertinent for the older population given the increased chronic illness burden⁴ and cognitive decline.⁵

Research suggests that uptake of ACP in Australia is low.⁶ This is likely to be a result of inadequate awareness of ACP, societal reluctance or cultural sensitivities to discuss end of life issues and lack of involvement of health professionals in ACP.⁶

The frailty trajectory is well recognised as a general decline with episodes of deterioration and recovery, but with each recovery the person does not return to baseline (Fig. 1).⁷ The timeframe for this general decline is individual, but it is important that, in view of the fluctuations, ACP is considered.

Fig. 1.

Expected trajectory of decline in older people with frailty at the end of life.

Improvement is needed during various stages of ACP, including increasing community awareness, building confidence from health professionals to engage in ACP, and system-wide improvements to support the recording and accessing of ACP documents.⁸

The concept of empowerment involves supporting individuals to reach their full potential and to examine interventions aimed at changing the balance of power in society.⁹ Empowerment in healthcare systems, including ACP, has been limited because people dealing with decisions about end of life are usually facing challenging and distressing times and rely on health professionals to guide treatment options or initiate ACP.⁵

The primary objective of this research was to improve healthcare provider communication in the ACP process and empower patients to complete advance care directives (ACDs) that align with their values and wishes. The secondary objective was to assess the experiences of the patients involved in a systematic empowerment-based approach to ACD development.

Methods

Patients were recruited from March 2021 to December 2022 from outpatient geriatric clinics and community outreach services in the south of Sydney. Doctors, nurses and allied health workers referred the patient to the field researcher. Participants needed to be 70 years or older, have Mini-Mental State Examination (Folstein) score greater than 24/30, no existing ACD, and be English speaking.

Participants were contacted by the researcher to explain the study and gain consent. A meeting was held at the participant’s preferred location, and the concepts of ACP and its benefits were explained.

A person-centred consultation model was then used to assist participants in the completion of an ACD. More than one home visit/sessions were offered.

Upon completion, the field researcher organised the ACD to be signed by their geriatrician or general practitioner (GP), uploaded onto the participants’ e-health records and returned to the participant.

The field researcher assessed the participant’s current health status and frailty using the Supportive and Palliative Care Indicators Tool (SPICT)¹⁰ and Rockwood Clinical Frailty Scale.¹¹ Results were then shared in an empathetic way to enable understanding of potential health trajectories and likelihood of mortality.

The Patient Activation Measure (PAM)¹² questionnaire was conducted to assess patient engagement.

After completion of an ACD, feedback was sought on participant experience through a series of semi-structured interview questions. Referrers were also interviewed.

Contact within 6–12 months after initial contact was attempted via phone and a re-application of PAM was completed if possible.

The NSW Ministry of Health ACD template was used as the ACD.

Results

A total of 170 referrals were received during the time period of the study. Mean participant age was 83.1 years, and 60% were female.

A total of 47% of participants (n = 80) completed ACDs. This group had a mean age of 83.5 years and 65% were female. The group of 90 who did not complete ACS had a slightly lower age, 82.9 years.

There were 20 deaths in the total referral group and 40% (n = 8) of those had completed ACDs

Fifteen participants (18%) entered a residential aged care facility during the time period of the study and of these, only 60% (n = 9) completed ACDs

Of 27 couples, 56% (n = 15) completed ACDs; if one member of a couple was hesitant, the other tended not to participate. On only one occasion, one member of the couple completed the ACD and the other did not.

In 82.5% of ACDs the participant did not agree to cardiopulmonary resuscitation (CPR).

Of the 80 patients who completed ACDs, within 6–12 months after completion, 57% (n = 46) had at least one presentation to emergency and 47% had at least one overnight hospital admission.

Many participants expressed support for ACP. Despite this, most participants required 2–3 visits. This first visit was usually around 60 min. Only a small proportion of participants were happy to complete the ACD at this visit but they were aware of ACP and had already considered their wishes and only needed to document them. At the following visit, the duration was often shorter to complete the ACD and to complete any tools that were not completed previously. The purpose of the third and final visit was to return the signed ACD to the participant. In these subsequent visits, participants were more relaxed and willing to share life stories, which paved the way for more in-depth and insightful ACP discussions. This highlights the importance of rapport and trust with health workers to enable meaningful ACP, and this is developed over time.

Table 1 lists the Frailty scores for those who completed ACDs.

Of 90 participants who did not complete an ACD, Frailty scores were obtained for 32 participants (Table 2).

Fig. 2 shows a comparison of the Rockwood Clinical Frailty Scale score between these two groups.

Fig. 2.

Comparison of Frailty scores between those completing and not completing an ACD.

In the completed ACD group, 46/80 (57.5%) had one or more general indicators of serious illnesses or health conditions. The most common indicator was unplanned hospital admissions, followed by ‘carer needs more support’, followed by weight loss.

In the not completed ACD group, SPICT scores were obtained for 25 participants. A total of 13 (52%) had one or more general indicators of serious illnesses or health conditions, the most common being ‘needs more care’ and unplanned hospital admissions.

In the group of 20 patients who had died on follow-up, if a SPICT was completed, all but one patient had at least one general indicator.

In the completed ACD group, 75 PAM scores were obtained at baseline, and the mean PAM score was 55.96.

In the non-completed ACD group 18 PAM scores were obtained at baseline, and the mean PAM score was 61.55.

Of the 80 patients who completed ACDs, 40% (n = 32) were successfully followed up, and reapplication of PAM was successful for 26% (n = 21). Reapplication of PAM was low due to decline in cognitive ability or hearing difficulties. The PAM level reduced during the time period in 57%, remained the same in 38% and improved for only one participant. For patients with successful follow up, many were appreciative of the follow up and happy to re-engage with the researcher and provide an update of life events since participating in the research project.

Almost all participants who completed the semi-structured interviews reflected on the importance of ACP. Common responses included the ability to document own wishes while still capable and reduce the burden on families and/or adult children to make those decisions. Many patients listed being kind and non-judgemental as the attributes they want in a health professional initiating ACP conversations. To improve the experience of discussing ACP, patients suggested that the process should not be rushed and that the health professional should listen to and understand their wishes, rather than focus on completing a form. Finally, patients gave suggestions on how to increase ACD completion: advertising on traditional media channels, local newspaper, radio, and tapping into organisations that already have a large following of older people, for example Probus and neighbourhood aid services, or holding information sessions in residential aged care facilities or targeting residents in independent living.

Discussion

The original research method outlined that the field researcher complete the SPICT, Rockwood Clinical Frailty Scale and PAM to empower the participants to complete an ACD. In practice, the tools were often completed after the ACD was completed due to established rapport. Completing the tools first sometimes felt like the participant was being tested.

The Rockwood Clinical Frailty Scale was the easiest tool to use with patients. Most, if not all patients were very quickly able to identify themselves through the visuals and brief text.

Frail and unwell patients were difficult to engage. Often they deferred decision-making to their families: ‘it’s more important that my daughter knows what to do’ despite not being sure themselves of what they want at end of life. There was a sense that these participants already felt disempowered and they focused only on ‘living today’.

Participants and non-participants were mild to moderately frail. Although numbers are small, it is apparent that the group who completed ACDs were slightly less frail than the group who did not. This suggests that it may be important to target a less frail population. It seems from this study that the more frail a person is, the less likely they are to want to fill in an ACD, despite using an empowerment approach. Significant decision-making can be just too much to think about as deterioration in function, independence and cognition are occurring.

The SPICT appeared to be overwhelming for patients with its large volume of text. Most patients required guidance in identifying their indicators. For some participants, the field researcher determined the indicators. Most engaged conversations came from participants with minimal SPICT indicators. The SPICT scores, the subsequent deaths and aged care facility admissions in the studied group confirm the vulnerability of this population.

The PAM was a challenging tool to implement. Patients commented on its length, repetitive questioning and lack of ‘middle option between agree and disagree’. If the patient agreed to some but not all aspects of the question they would be inclined to simply ‘agree’. If patients appeared to disagree with a statement, many would avoid overtly selecting ‘disagree’ but rather offering a positive answer e.g. ‘I only know what some of my prescribed medication do’. When pressed to select a response, they were more likely to select ‘agree’. The PAM levels generally reduced in the study period, suggesting that their overall health, independence and cognition deteriorated during this time.

Many patients reported previous encounters with health professionals around ACP. Some had even been given the NSW Health ACD template and encouraged to complete it. Some patients admit to being wary of documents that require signing. A facilitated discussion and assistance in completing the form with an experienced health worker was integral to the outcomes obtained in this project. One participant mentioned that she was in the ‘control group’ of a previous ACP research study⁸ whereby she was given only information regarding ACP but no practical assistance to complete it. She did not see the urgency and did not want to go to the bother of gaining a GP’s signature.

Families provided feedback in the challenging nature of discussing ACP without a facilitator. The difficulty of navigating through complex emotions discussing end of life is a constant theme. The convenience of ACP being undertaken in the patient’s homes reduced carer burden.

Although there was an overwhelming negativity to CPR, many other aspects of the ACD template were discussed and documented. In this population it is essential that broad discussions take place, not only CPR status. These include other medical treatments, organ donation, and preference for place of death.

All families were observed to be encouraging of patients when completing their ACD. None was seen to ‘override’ or exert pressure or duress in any decisions. More often, families preferred ACD questions to remain blank or ‘unsure’ if patients did not indicate a strong preference.

Feedback from referrers to the project was that even when a patient would benefit greatly from ACP discussions, there were more pressing medical or social issues that needed to take priority, thus the patient was not referred. This suggests that ACP may need to be mentioned and encouraged on several occasions before patients are willing to engage. None of the patients who declined to participate changed their minds when the service was offered at a later time.

Referrers also gave feedback that they would refer patients who they knew would probably complete an ACD, as opposed to referring all patients who met eligibility criteria.

Conclusion

Patients who score as ‘less frail’ on the Rockwood Clinical Frailty Scale are more likely to participate and be more engaged in the ACP process. Therefore, we should target those people, given that the more frail are less likely to want to think about completing an ACD, even though they are the group who would benefit from it the most.

Patients benefit from a coordinated and facilitated ACP discussion in a place of their choosing, most likely their own home, and are more likely to complete an ACD through an empowered approach.

The timing of the discussion is important, as is the amount of time given to people to understand the ACD, the place of the discussion, and having an experienced health worker take them through the process and follow up with medical practitioners. Family support is also important and it is apparent that if one of a couple is not supportive, it is unlikely that the other will be supportive.