At the end of life, people wish to be treated with dignity, have relief from pain and, as much as possible, have some control over what happens to them. Advance care planning (ACP) can help a person’s wishes and preferences be considered by others when they are not able to communicate or to make decisions. [1-8] With an advance care directive (ACD) in place, people are more likely to have care which respects their wishes. [3,7,9,10] People with an ACD are also more likely to experience fewer unwanted medical interventions, less likely to be moved from their home or community care to a hospital, and less likely to die in a hospital. [4,9]
The legal and policy framework for ACP varies across countries  and across states in Australia. [5,7]
Overall, the quality of the included reviews is good with systematic reviews being of acceptable to high quality.
Four systematic reviews relate to ACP for older people in a palliative care or end-of-life context. [4,10,12,13] Nine systematic reviews relate to ACP for older people, [11,14-16] residents in an aged care facility [2,3,17] and frail residents. [9,10] Eight systematic reviews relate to ACP for people living with cognitive impairment and dementia, [1,18-23] advanced dementia  and for their carers. Two systematic reviews relate to end-of-life decision-making  or surrogate decision-making  in a general adult population. Three papers included meta-analysis. [14,15,24]
No guidelines for ACP specific for older people receiving palliative care were retrieved. The Australian Therapeutic Guidelines include a section on ACP.  In a Position Statement, Palliative Care Australia  calls for national guidelines to promote good practice in ACP and for initiatives to increase community discussion of death and dying.
Other sources are included to provide context to this topic. [5-7,27,28]
Most research has been conducted in USA. [2,3,9,10,12,14,17,20,24,29]
The ACP process may include interactive and perhaps multistage discussions between a person and his/her family members, his/her carers [1,2,8,19,20] and relevant health professionals. [1,2] A person can create an ACD without talking to others. Cultural beliefs and practices will shape a person’s preferences for future care and for end-of-life and readiness to engage with ACP.  Cultural and religious beliefs may influence decisions made by professionals, carers and family. 
Successfully implementing ACP in residential aged care (RAC) may be complex. [3, 19] Health professionals and RAC organisations are seen to play an important role in the successful implementation of advance care plans. [3,19] Having identified the numerous preconditions related to successful implementation of ACP in RAC, Gilissen et al.  recommend a whole-system approach. Beck et al.  recognise the complexity of ACP in the care of older adults with dementia.
Benefits of ACP are cited in the literature including end-of-life care in line with a person’s wishes, care and/or death in a person’s preferred place, reduced unwanted treatment or interventions at the end of life and reduced regret experienced by carers or substitute decision-makers. [1,9,14-16]
Prevalence of ACP
In practice, discussions around end-of-life issues may often not take place or start too late.  Research shows that only 14% of the Australian population has an ACD  and only a minority of older people actively engage in advance care planning. Internationally, there is a low prevalence of advance care planning in residential aged care [3,8,19,20] despite a majority of residents welcoming the opportunity for such a discussion. 
Decision-making in ACP and substitute decision-makers (SDM) or surrogates
Advance care planning (ACP) is a voluntary process of reflecting, discussing and recording preferences and plans for future care should an adult lose capacity or the ability to communicate. [1-7,11,19,20] Séchaud et al.  encourage the participation of many actors (person, family, relevant medical and health professionals) in ACP.
An Advance Care Directive (ACD) may result from ACP. An ACD is a document which provides a legal means for a competent adult to nominate and instruct one or more Substitute Decision-Maker(s) (SDM) and/or to record preferences for future health and personal care. [2-6,9] Forms of ACD include living will, [2,5,6,9,20] advance directive (AD), [2,5,7,9] advance health directive, advance personal plan, medical direction,  do-not-resuscitate (DNR) order and do-not-hospitalise (DNH) order, [4,9] physician orders for life-sustaining treatments (POLST) or refusal of treatment certificate. 
Statements of values are also used to describe a person’s wishes. These can describe the person’s values, his/her conception of existence and wishes, hopes and expectations in the context of health and disease, what the person means by ‘quality of life’ or ‘dignified end-of-life’.  These may help guide decision-makers.
Kelly et al.  reviewed individuals’ preferences in ACP and decisions around the choice of SDM. Half the included studies (but only 20% participants) focussed on elderly people. In a family with close ties and free of conflict, it is very common for people to choose close family members to make treatment decisions for them during periods of decisional incapacity with a preference for a spouse or adult child(ren). These choices and preferences are based on trust yet were made after discussion with family members in sometimes less than 50% cases. People find it difficult to judge how much “leeway” or “freedom” they are happy to give to the SDM. In certain cases, people indicate clear, inflexible wishes or ask family or SDM to adhere to their wishes to avoid possible feelings of guilt about a “wrong decision”. Other people are satisfied that general preferences are enough to guide family and SDM in decision-making. 
ACP and aged care residents and frail elderly people in the community and in RAC
Older people with multiple comorbidities, frailty, cognitive impairment or dementia are often cared for in residential aged care. [9,16,17] Frailty and comorbidities are often associated with hospital admissions for residents at a rate up to three times higher than for older people living in the community.  Martin et al.  investigated ACP interventions such as ACP educational programs or the introduction and evaluation of a new ACP in residential aged care facilities (RACF). This review finds that these interventions have beneficial effects for residents: ACP interventions can reduce the number of hospital admissions of residents, reduce unwanted life-sustaining treatments and improve the alignment of preferences for care and delivered care.  ACP interventions are associated with death in preferred place and increased and earlier referrals to palliative care.  The authors caution that these results, although promising, are from low-quality studies.
ACP is important for the frail elderly.  Frail elderly people often have multiple comorbidities, functional and/or cognitive impairment and commonly frail elderly people are cared for in RAC. [9,16] Prognosis is difficult and the dying trajectory unpredictable. [9,6] An acute deterioration in health, commonly an infection, fall-related injuries or an acute exacerbation of a chronic illness, may trigger a transfer to ED or hospital.  For frail elderly people, hospitalisation can be associated with delirium, pressure ulcers, hospital-acquired infections, and functional decline post-admission.  For a certain number of these residents, a transfer to hospital may disrupt or inhibit appropriate palliative care. 
The completion of an ACD, DNR or DNH order may reduce the probability of transfer to hospital for acute treatment, therefore avoiding the effects of hospitalisation, and reducing the risk of dying in hospital rather than in the RACF. [9,16] However, the type and prevalence of ACP documents completed by residents and frail elderly residents is variable. For example a) 2-29% of frail older people had discussed some form of end-of-life care plans with a healthcare professional, b) 15-66% of older people have end-of-life care plans documented in their records, c) 54-73% and 2.1-49% of residents have DNR and DNH orders, and d) 3-37% of residents presenting to ED have a DNR order and up to 7.9% have a DNH order. [9,16] Dwyer at al.  note that RACF residents are less likely to be transferred to hospital as age approaches the very extremes of old age. These authors propose that if a RACF has a high percentage of residents with completed ACD, DNR, and/or DNH order, this can be an indicator of quality of care for acute illness and at end of life for the residents. 
Sharp et al.  note that a majority of older adults (61-91%) want to discuss their end-of-life care early and in an honest and straightforward way. Older adults see ACP as an opportunity to plan for their future care whilst they have the capacity to do so, and a way of enhancing the possibility of their future being in line with their wishes and reducing the burden of decision-making on their family.  Often older adults see a doctor, GP or specialist as the person who should initiate end-of-life discussions.  Sharp et al. find discrepancies in the literature in that 28% older adults had discussed terminal care with a family member yet only 2% had had a similar discussion with their doctor. Information related to this topic can be found in: Communication at End-of-Life.
Efficacy of ACP interventions
Houben et al.  investigated the efficacy of ACP interventions in different adult patient populations. Whilst not specific to older adults receiving palliative care, this review relates to adults and older adults in RAC or with a life-limiting disease. Included studies are 55 RCTs of interventions focussed on the completion of ACDs and or communication about end-of-life care and related choices. Both types of interventions increase the completion of ACDs and the stimulate end-of-life care discussions between patients and healthcare professionals. The authors note that the optimal time for these discussions is outside of an acute episode of ill-health, when the person’s health is stable.
The interventions that include both communication about ACP and ACDs, improve alignment of preferences for care and delivered care. It seems that these interventions are not associated with participants’ anxiety or depression. It is not clear if ACP interventions have significant positive impact on satisfaction with health care, decisional conflict or the use of healthcare services. 
The cost-effectiveness of ACP
One systematic review focussed its attention to the cost-effectiveness of ACP and found limited evidence.  Fourteen of the 18 included studies were conducted in USA. No evidence suggests that ACP is more costly and some well-conducted studies indicate that ACP is associated with healthcare savings.  The savings seem to be for older people in RAC, people living with dementia who are cared for at home, and people with high-care needs and low income.  Although savings are recognised, the broader costs and benefits are not understood. Examples of such costs are those borne by RAC, community services or families and carers with the avoidance of hospital admissions. Examples of benefits may be the reduction of stress and improved quality of life with care aligned with the dying person’s wishes.
ACP and people with dementia
A palliative approach for people with dementia, particularly in the advanced stages, is appropriate [1,13] but not always put in place.  In residential aged care, a palliative approach supports advance care planning, decision-making and goal setting for people with dementia.  The involvement of the resident (when possible) and key family members and/or substitute decision-maker(s) is important to an integrated approach to care.  Case conferencing, a useful tool for integrated care, can improve palliative care outcomes for residents with dementia particularly in the areas of advance care planning, physical symptom management, psychological support, family support and terminal care. 
ACP is a beneficial process for adults living with dementia. [1,23] Ideally, ACDs should be completed with input from a treating doctor and a discussion of prognosis, possible complications and treatment options. In case of loss of capacity, illness or injury, whether expected or not, ACP assists family, friends and health professionals to consider options and make decisions in line with a person’s wishes. [1, 19, 20, 23] ACP can reduce inappropriate hospital admissions and health-care costs for people with cognitive impairment and dementia.  People with dementia with an ACD in place are less likely to die in hospital or in ICU. 
The progressive decline associated with dementia can be associated with under-treatment for pain and other symptoms, numerous emergency hospital admissions and/or non-beneficial medical treatments.  Hospitalisations and invasive treatments can exacerbate the behavioural and psychological symptoms of dementia (BSPD). 
Dixon et al.  note that the majority of the 18 studies included in their review indicate ACP to be associated with positive end-of-life outcomes. These outcomes were diverse but mostly in relation to healthcare utilisation including place of death, hospitalisation and specific treatments. It was unclear if patient and carer outcomes also improved. Most of the included studies were conducted in residential aged care and therefore focussed on late(r)-stage dementia. Some of this evidence supports ACP reducing burdensome transitions in the last three months of life.
Like other authors, Dixon et al.  note that the ACP interventions described in the literature are diverse. These can range from ACP discussions, the completion of an ACD, an ACD directly guiding care, to multi-component programs involving staff education and training, education for patients and carers, and documented ACP on charts as a means of checking that care respects stated preferences. Dixon et al.  note that ACP programmes appear to influence end-of-life outcomes by influencing the policy and culture of the residential aged care facility and by improving staff skills and knowledge. They also note that education and training for health and care staff can increase acknowledgement of resident and carer preferences. This may reduce the number of transfers to hospital from a RACF and create a higher threshold for transfer to hospital for residents without an advance care directive.
Implementing ACP may be more challenging in dementia, but Dixon et al.  find it important as it allows people with dementia to have a say in their future care and because of the currently poor outcomes for people with dementia at end of life, which ACP may help to improve. However, Dening et al.  find limited literature and little evidence in support of the effectiveness of ACP in changing outcomes for people with dementia.
Dening et al.  find limited evidence on the timeliness and initiation of ACP discussions for people with dementia. Some evidence suggests the important role of health professionals in the initiating timely and informed discussions about planning for future care with both the person with dementia and their carer, the success of which is based on background ACP education and supportive programs.  Dening et al. suggest that a key ACP facilitator, a dedicated professional available to educate both families and health professionals about ACP, can help promote ACP and improve outcomes. 
Family carers require emotional support, information and education about dementia and its trajectory and this can include a better understanding of palliative and end-of-life care.  Education is also needed to enable health professionals to engage with people with dementia and their families in the process of ACP, acknowledging the role of the family within this context. [19,20]
Initiating ACP in dementia
There is general agreement that planning for end-of-life can be beneficial in dementia. Anticipating the patient’s cognitive decline, starting ACP early can be beneficial.  In practice, however, these discussions may not take place or may start too late. The time of diagnosis may be seen as a good time to start planning or, on the contrary, may be so emotionally-charged and information-intensive that ACP may be too difficult.  Van der Steen at al.  describe the complexity of factors associated with initiation of ACP for end-of-life care in dementia. Noted commonly in the literature are the family’s initiative or lack of it, and willingness or reluctance to begin ACP discussions. Terms such as “unwillingness,” “active avoidance,” “not being ready”, “passive avoidance” or “blind trust”, are used. Previous experience with ill-health or dying can prime families’ readiness to initiate ACP early. Noted in the literature is families’ regret on having been too late to discuss ACP or having failed to anticipate problems in the dying phase.  Initiative of health professionals and carers are also noted as important factors in the initiation of ACP. 
Communication tools and decision aids (DAs) and ACP
The communication process in ACP is important and involves addressing fundamental values, goals and preferences.  The involvement of family, carers [1,2,8,19,20] and the relevant health professionals may enrich and strengthen the process. [1,2] Time constraints due to the advancement of disease or the availability of people to participate in this process are cited in the literature. [1-3,9,12,15-17,19-21,23,24]
Communication tools (including decision aids, structured meeting plans, and educational interventions) can assist patients and clinicians with end-of-life (EoL) decision-making.  A review by Oczkowski et al.  provides low-quality evidence that the use of structured communication tools to assist with EoL decision-making may increase the completion of ACP (discussions or ACD documentation). They may also increase alignment of desired care and actual care at EoL. The authors believe that the mechanisms behind these positive outcomes are improved patient and/or SDM knowledge and/or improved communication skills. This evidence is from a broad range of outpatient populations with approximately 40% of participants being elderly people.
Decision aids (DAs) are tools that provide high-quality, synthesised information to assist patients and clinicians compare the risks and benefits of treatment options, clarify what matters most to the patient and make a shared decision about what is the best option for the person. [12,14,27] DAs are most useful for situations when there is not one clear option and when each option has benefits and harms that patients may value differently. [12,14,27]
Low-level evidence suggests that video decision aids lead to greater knowledge of ACP and preferences for less aggressive treatment at end-of-life.  What needs to be better understood is whether DAs help confidence in decision-making, the actual use of aggressive treatments at end-of-life, or end-of-life care in alignment with a person’s wishes. 
There are few DAs available relative to end-of-life care of older people.  These seem to be generally accepted by older people and help by increasing knowledge of care options and reducing decisional conflict.  DAs should be an adjunct to and not a replacement of discussions with health professionals.  As death approaches, symptoms, care needs and care costs change and therefore DAs can be used regularly to assist with the evolving decisions of care. 
Page updated 22 October 2018
- The evidence base for ACP for older Australians and aged care residents is weak. Further high-quality studies are required to support the reported benefits and to help identify the types of ACP interventions that are most effective and beneficial for older Australians at the end of life.
- The value of decision aids in assisting advance care planning for or at end of life is not yet clear.
- Future work may find strategies to align the receptiveness, knowledge, and attitudes of patients, families and health professionals to the initiation of ACP.
- The knowledge, attitudes, experience and confidence of aged care staff, primary care providers, specialist geriatric and palliative care staff may have a positive influence in the process of ACP. Future research may help us better understand the interrelationship.
- There is a lack of evidence around the costs associated with implementing ACP and the subsequent benefits to patients and families and to the health system.