Advance Care Planning Discussions with Adolescent and Young Adult Cancer Patients Admitted to a Community Palliative Care Service: A Retrospective Case-Note Audit

Introduction

Adolescents and young adults (AYA) with cancer require specialized healthcare models to address unique cognitive, developmental, and physical challenges.^1–4 They have poorer physical and psychosocial outcomes compared with pediatric or older adult populations and poorer survival for a number of malignancies, including sarcomas and leukemias.^1–4 Although a relatively uncommon diagnosis in the age group, cancer remains the leading cause of disease-related death in AYA in Australia. ⁵

The age range definition of AYA is 15 to 25 years, reflecting the unique neurodevelopmental profile of the age cohort.^6–8 Attainment of adult prefrontal cortex functioning, necessary for the higher cognitive tasks of self-navigated care planning, consequential thought, and complex decision-making is generally not achieved until the mid-20's.^6–8 In recognition, provision of care to AYA in adult healthcare settings in Australia is now augmented through a national network of specialist AYA cancer services, which aim to provide individualized, patient-centered care coordination, and supportive care. ⁹

Palliative care provision and end-of-life decision-making for AYA also necessitate cognitively appropriate approaches to facilitate conversations around goals of care and advance care planning (ACP). ¹ ACP is defined as, “a process of communication between individuals and their healthcare agents to understand, reflect on, discuss, and plan for future healthcare decisions” ¹⁰ It has been shown to improve end-of-life care in the adult population ¹¹ and may provide guidance for an individual's end-of-life care based on preferences regarding the following:

• General goals and priorities of care, including specific preferences for life prolonging or symptomatic care. ¹²

In general, healthcare decision-making by AYA is likely to be influenced by their immediate family, social media, advertising, peers, personal characteristics of their healthcare providers, financial costs, a lack of experience navigating systems and concerns regarding confidentiality.^15–18

International research supports the preference of AYA to be actively involved in decision-making.^14,19,20 However there are barriers to their open involvement, including healthcare professionals' concerns of causing distress, fear of extinguishing hope in the young person, ²⁰ and legislative frameworks around consent and capacity that may restrict binding decision-making for this group. ¹⁵ In Western Australia, advanced care planning legislation allows patients aged 18 years and over to appoint a legally authorized substitute decision maker (Enduring Power of Guardianship [EPG]) or document their wishes in a legally binding Advance Health Directive (AHD). ¹⁰

A significant further issue in Australia is the lack of local research on current practices within palliative care services, which cares for AYA cancer patients, to identify gaps, encourage development of specifically tailored tools, and inform best practice for this cohort.

Methods

Results

There were 51 records identified for 15- to 25-year olds. Sixteen did not have a diagnosis of malignancy, and eight were excluded as they were inappropriately identified, discharged from the service, or lost to follow-up. All patients included in the audit died while under the service. Twenty-seven AYA records and 37 adult medical records were reviewed.

Discussion

This retrospective audit provides a snapshot of the documentation of end-of-life care for patients aged between 15 and 25 years, under the care of a metropolitan community-based palliative care service in Western Australia. It focused on AYA's documented consent to palliative care service involvement, ACP, goals of care, place of death, and actual care received. It has demonstrated significant differences with adult palliative care patients, particularly in terms of the increased focus on potentially life-prolonging therapies and level of medical intervention received; still achieving high levels of death at home as per their wishes.

Documented ACP was very limited, only including preferred place of death and a low rate of completed DNR orders. The low number of DNR orders may indicate discussions not taking place, discussions taking place but not being documented, or discussions taking place and AYA patients choosing to receive CPR. There were two DNR orders (7.4%) documented for the AYA cohort and eight (21.6%) for the adult comparison group. In both cases where AYA patients actually received CPR, an ambulance was called when the patient had arrested at home, and CPR was commenced by paramedics on arrival at parental request. A study by Keim-Malpass et al. found 18% of young adult patients who died in a hospital setting received CPR. ²⁴

AYA patients with a malignancy are likely to be under the care of several different services, and specified communication pathways might be beneficial in assisting interagency communication of important documents such as DNR orders.

This audit examined documentation as it is important that these conversations are documented to provide healthcare professionals with guidance regarding a patient's wishes for care. The proposed role of a DNR order is to promote patient autonomy and prevent futile intervention. ²⁵ When caring for patients at home, a documented DNR may be beneficial if emergency health professionals attend the patient. Noticeably, despite no DNR order being in place, patients did die at home without receiving CPR. It is possible that close collaboration among the palliative care team, patient, and family meant that changes suggesting a person was closer to death may be discussed on a day by day basis, reflecting the patient's wishes and condition, with plans put in place and an agreed end-of-life care plan negotiated, negating the need for a formally documented DNR order. More work would need to be done to explore this further.

No documented AHDs for either the adult or the AYA group is reflective of the overall suggested low rates of uptake of AHDs in Australia, particularly in Western Australia. ²⁶ To try to increase the uptake of AHDs across the population, consideration could be given to tailoring education to cohorts with unique needs, including culturally and linguistically diverse and specific age groups.

There were documented conversations discussing goals of care with 96.3% of the AYA group. As described in the introduction, goals of care discussions incorporate preferences for treatments and procedures, including life-prolonging treatments or symptomatic care. ¹² The discussions were often initiated by an acute event requiring a decision to be made to guide further care. The discussions were documented within the body of the patient health record as opposed to on a separate care plan making them less accessible to health professionals requiring the documentation to guide care.

Although there were documented goals of care discussions, the intention of an intervention was not clearly documented, and it was not possible from the documentation to comment on patients' reasoning for their choices. For example, a patient may pursue a phase I clinical trial with the goal of helping other patients in the future. This highlights the complexity of trying to restrict goals of care into black and white categories, where many variables and potential outcomes preclude a straightforward yes or no response. These categories also do not allow for changes in preferences over time.

Communication with AYA patients needs to address their priorities: truthfulness, preservation of hope, and participation in decision-making. ²⁷ There is evidence for the preference of young people to be actively involved in decision-making, particularly in collaboration with healthcare professionals regarding their healthcare, including end-of-life care considerations.^14,19,20,28 The majority of discussions were documented by palliative care consultants, however, it was not documented who participated in the discussions. There was no documentation regarding AYA preference for degree of involvement in decision-making, however, there were documented discussions regarding parents' preferences for limiting information provided to their children. The usual practice for the palliative care service is to allow patients older than 16 years of age to provide consent to the service being involved in their care if they are physically and cognitively able.

Overall, the pattern of care received at the end-of-life portrayed a different approach in the AYA patient group when compared with the adult group with a higher proportion of patients receiving chemotherapy in the last 30 days of life. In the AYA cohort, late chemotherapy may be in the form of a phase 1 trial or compassionate/self-funded access to novel agents. It is only possible to speculate from this audit whether this was due to patient preferences, surrogate decision-making, or a default position from health professionals to heighten intensity of care in the context of the young age of patients. The results are consistent with research, which suggests that AYA patients may wish to receive more aggressive management toward the end-of-life. ²⁹

Notably, the two patients who received CPR had been documented to be in the terminal phase of their illness in their health records. This indicates that ACP processes and discussions may require adaptation for this unique cohort whose preferences may diverge from those of the adult population but are crucially important and need to be captured and understood. There is also a need to revisit and review advance care plans as patients transition through stages of illness.

These results show that although this is a unique cohort of patients, they can be cared for by an ambulatory specialist palliative care service and supported to die at home. A study by Cohen-Gogo et al. found that 13.4% of the AYA patients in their study died at home. ³⁰ The AYA cohort who had a documented preference for place of death all chose home. Checklists and care plans as part of the admission process to this palliative care service, which specifically record preferred place of death, likely foster these discussions and increase the probability of preferences being documented. Just over two-thirds of the AYA patients included in this audit died at home, indicating that if patient preferences are known and documented, healthcare professionals will try to honor these wishes.

To die at home, patients require a carer who is both physically and psychologically able to provide care for them in a conducive environment. Fourteen (37.8%) of the adult comparison group died in an inpatient hospice. Speculating, this could have been due to the unavailability of a carer as spouses may have their own health concerns, and other family members may not be able to provide care at home. Financial costs must also be considered for patients wishing to die at home, including medications costs and the costs of maintaining a comfortable environment. ³¹ AYA patients may be financially disadvantaged.^15,32

It is not possible from this audit to comment on why a third of the AYA patients did not die at home, whether this was due to a change in their preferences, financial concerns, or lack of a carer or if any of these deaths away from home may have been preventable with improved communication between healthcare professionals or shared documentation.

There were limitations to this audit. The AYA patient numbers were small, the audit was developed in house and conducted by one person, and restricted by the quality of documentation, which has been shown to be poor in other ACP studies, ³³ and the data sources may not have captured all possible interventions received outside of the community service. However, any interventions received outside of the service are likely to have involved tertiary hospital settings and a more aggressive level of intervention is likely, therefore underestimating the level of intervention received by these patients. As ACP is a fluid ongoing process, it is difficult to capture conversations unless they are documented. Documentation rates of AHDs are not an accurate indicator of whether or not ACP is occurring.

Conclusion

The results indicated limited documentation of ACP to guide health professionals providing care for AYA patients with a progressive, terminal malignancy in the community setting. An age appropriate ACP document incorporating a tailored discussion regarding goals of care, preferred location of care, including preference for place of death, consent to future intervention, including CPR, and prompts for review may be preferable to a DNR order in isolation in the community setting.

Further studies need to be carried out to evaluate AYA goals of care and pattern of end-of-life care received in more detail, and to specifically identify whether end-of-life decision-making discussions impact upon the pattern of care received at the end-of-life.