Physicians need to take the lead in advance care planning.

Abstract

Our population is ageing, and older people are living longer with chronic illness. Technology-laden end-of-life care is often associated with poorer quality of life, lower satisfaction with care and increased anxiety and depression for carers and families.1 When given appropriate information about different care options and encouraged to express their preferences, frail older patients often choose comfort measures only or significantly limited interventions, rather than invasive or aggressive treatments.2 Randomised trials have shown that timely and nuanced discussions around end-of-life care options involving patients suffering advanced or end-stage disease can improve quality of life, prolong survival, reduce use of aggressive care and lower healthcare costs.3, 4 Advanced care planning (ACP) is a process for iteratively identifying and facilitating what patients consider important for a ‘good death’ (such as managing symptoms, avoiding prolonged dying, achieving a sense of control, relieving burdens placed on the family and strengthening relationships) and for informally communicating their wishes in case they are rendered at some future time unable to make or voice decisions for themselves.5 The processes involved in ACP are well described and comprise a systematic, yet personalised, approach.6 At its heart are structured, iterative conversations about goals, values and preferences for end-of-life (EOL) care involving trained and trusted healthcare professionals (who may include nurses and social workers as well as doctors). Such conversations occur over several encounters, actively involve properly informed substitute decision makers (SDM) and are supported by oral, written and even video-recorded information. Randomised trials have shown that ACP improves concordance between expressed preferences for care and delivered care, provides greater patient and carer satisfaction with care and lessens decisional conflict.7 Other studies have shown decreased use of hospital beds and more and earlier use of palliative care and advanced health directives.8-10 In recognition of these benefits, patients in hospital with serious illness and their family members place better EOL care, communication and decision-making as high priority targets for quality improvement.11 Unfortunately, despite this, few patients undergo ACP12 or complete an advance health directive (as noted by White et al. in this issue of the Journal).13 Hospital-based healthcare providers rarely engage patients and families in ACP conversations.14 Pleasingly, two other articles in this issue of the Journal describe efforts to correct this omission. Brimblecombe et al. introduced a routine Goals of Patient Care (GOPC) summary into a general hospital in Melbourne, targeting frail, older medical patients with multi-morbidity.15 The goals of care and limitations of treatment were discussed with patients and/or SDM during the initial assessment at time of admission or subsequently on consultant ward rounds, and recorded on a GOPC template. Additional details about discussions, including patient wishes, were recorded in the progress notes. Goals of care could be re-evaluated as the admission progressed in response to change in clinical status and did not necessarily carry over at discharge. The summaries were intended to guide actions, especially at times of clinical deterioration. A one-off survey of 101 inpatients conducted 3 months after roll-out of GOPC found 81 patients had a documented GOPC – such patients were significantly older than those without a GOPC (mean age 75 vs 60 years) – and for whom 48 (59%) indicated varying degrees of treatment limitation. In 42 cases (52%), there was evidence of discussions with patients or SDM, for whom 33 (79%) expressed treatment limitations. However, over a subsequent 7-day audit, a GOPC summary was documented prior to only nine of 23 (39%) emergency medical response calls, suggesting either a failure to identify at-risk patients who required an explicit management plan prior to deterioration, or too little time to complete a GOPC summary between admission and the call (mean 1.2 days for those without vs 6.8 days for those with GOPC, P = 0.04). De-escalation of care occurred in three instances. The authors acknowledge there is room for improvement with only one in 10 GOPC summaries containing substantial details of decision making, and only 20 (25%) clearly documenting involvement of a consultant despite all patients being reviewed by such a person within 24 h of admission. In another study, Cheang et al. undertook a retrospective 3-day survey of 100 consecutive patients aged 80 years or older admitted for 48 h or longer to a tertiary hospital in Sydney, aimed at determining prevalence of advanced care directives (ACD), resuscitation/end-of-life care intervention orders (REOL) and documented SDM.16 Patients who were not documented to have delirium, dementia, aphasia or drowsiness and could speak English were invited to participate in an ACP screening interview, with the aim of identifying those willing to participate in ACP discussions. At 3 months follow up, the authors then determined if a record of EOL/ACP discussions was contained within discharge summaries, and for those patients who had died, whether an ACD or REOL had been completed and, if so, the extent to which care given was consistent with this. Disappointingly, none of the 100 patients had an ACD filed in their chart (although four had completed them but did not think to bring them to hospital), only 17 had REOL forms completed by staff, and only eight had clear documentation of a SDM. On a more positive note, of 33 patients who completed the 10 to 20 min ACP screening interview, 32 (97%) were able to nominate an SDM, and 23 of 29 patients (79%) with no previously documented ACD were comfortable in having further discussions about ACP with a staff member, with 11 making positive comments about the interview. Of the 30 patients interviewed who did not have REOL, eight (27%) stated they did not want invasive life-prolonging measures if these were likely to be futile. Among 95 discharged patients, five died, of whom four had documented REOL and whose care was consistent with these orders. Of 14 patients discharged with REOL, only three had a record of EOL/ACP discussions, and all three were discharged to nursing homes. No mention was made of EOL/ACP discussions in any of the remaining 81 patients, including those with REOL orders. There are several implications of these two studies for physician practice. First, we need to be more involved in EOL and ACP discussions and assist patients and their carers to articulate what they feel are important goals of care, and ensure clinical interventions we may offer are aligned with those goals – while guarding against any compulsion to offer what is clearly futile and inappropriate care. Second, we should consider ACP in any patient with chronic progressive disease and all those who are being discharged to residential care facilities. The ‘surprise’ question: ‘knowing all I know about this patient, would I be surprised if he/she were to die within the next 12 months?’ is a suitable screening tool, if the answer is ‘no’, for identifying patients eligible for ACP.16 Third, most patients are comfortable about discussing ACP and, indeed, many welcome the opportunity to do so, and this can be done in hospital as well as in outpatient settings. Fourth, if discussions are held, they need to be clearly and fully documented in progress notes, and referred to in discharge summaries, so that the views and preferences contained within them are revisited, and further modified if necessary, during visits to general practitioners or at subsequent readmission to hospital. Finally, institutions that implement programmes aimed at systematising ACP among older medical patients with multi-morbidity should attempt to evaluate formally effects of such programmes, not only on ACP documentation and concordance of delivered and desired care but also on clinical processes and outcomes and utilisation of care resources. Such analyses are needed in growing the evidence base for ACP and further legitimising its inclusion as a duty of care.