bs_bs_banner

Aust. J. Rural Health (2015) 23, 195–200

Original Research Advance care planning and end-of-life care in a network of rural Western Australian hospitals Kirsten Auret, MBBS, FRACP,1 Craig Sinclair, BA/BSc, (Hons Psych), PhD (Psych),1 Barbara Averill, RN,2 and Sharon Evans, Msc, PhD, AStat1 1

Rural Clinical School of Western Australia, University of Western Australia and 2Albany Community Hospice, Albany, Western Australia, Australia

Abstract Objective: To provide a current perspective on end-oflife (EOL) care in regional Western Australia, with a particular focus on the final admission prior to death and the presence of documented advance care planning (ACP). Design: Retrospective medical notes audit. Setting: One regional hospital (including colocated hospice) and four small rural hospitals in the Great Southern region of Western Australia. Participants: Ninety recently deceased patients, who died in hospitals in the region. Fifty consecutive patients from the regional hospital and 10 consecutive patients from each of the four rural hospitals were included in the audit. Interventions: A retrospective medical notes audit was undertaken. Main outcome measures: A 94-item audit tool assessed patient demographics, primary diagnosis, family support, status on admission and presence of documented ACP. Detailed items described the clinical care delivered during the final admission, including communication with family, referral to palliative care, transfers, medical investigations, medical treatments and use of EOL care pathways. Results: Fifty-two per cent were women; median age was 82 years old. Forty per cent died of malignancy. Median length of stay was 7 days. Thirty-nine per cent had formal or informal ACP documented. Rural hospitals performed comparably with the regional hospital on all measures. Conclusions: This study provides benchmarking information that can assist other rural hospitals and suggests

Correspondence: Professor Kirsten Auret, Rural Clinical School of Western Australia (Albany), University of Western Australia, M701, 35 Stirling Highway, Crawley, Western Australia, 6009, Australia. Email: [email protected] Accepted for publication 25 November 2014. © 2015 National Rural Health Alliance Inc.

ongoing work on optimal methods of measuring quality in EOL care. KEY WORDS: advance care planning, palliative care, regional health planning, rural health services delivery, terminal care.

Introduction Of all deaths in Australia, just over half occur as an inpatient1; hence, looking after the dying is a common responsibility in Australian hospitals. There is growing recognition of the requirement to deliver this care well,2 but there is currently a lack of agreed standards of care for people dying in the hospital setting and little available evidence against which to benchmark end-of-life (EOL) care quality.3 Existing audits of hospital-based EOL care have been predominantly metropolitan and focused on measures including documentation of ‘not for resuscitation’ (NFR) orders, symptom assessment and management, and use of EOL care pathways.4,5 One comparative study of NFR documentation suggested the need for standardised policies across hospitals and clearer communication about decisions to withhold resuscitation.4 A retrospective notes audit of 159 patients who died in a network of Queensland hospitals suggested additional areas for improvement, including medication withdrawal, proactive prescribing, documentation of psycho-social and spiritual issues, and communication with family members after death.5 Australian Institute of Health and Welfare data for the 143 472 deaths in Australia in 2008–2009 confirmed previous reports that care at the EOL differs for those in regional or remote settings, particularly in access to palliative care.1,6,7 Despite this apparent disparity, there is just one published study of nonmetropolitan hospital-based EOL care in Australia.8 This retrospective notes audit assessed 73 adult deaths over 2006–2007 from Alice Springs Hospital. This centre has 176 beds with an Intensive Care Unit and doi: 10.1111/ajr.12167

196

K. AURET ET AL.

What is already known on this subject: • Of all deaths in Australia, just over half occur in hospital. • Measuring the quality of end-of-life care is difficult. • There is a lack of evidence to enable benchmarking of end of life care, particularly in rural settings.

What this study adds: • This study provides data that will enable comparison of end of life care in rural hospitals over time. • Advance care planning was not prevalent in the medical notes and was not strongly associated with any different type of end-of-life care.

dialysis service and is classified as ‘remote’ (RA 4).9 Deceased patients were relatively young (mean age 61.4 years); 64% were Aboriginal; and only 25% of deaths were from malignant causes. Diagnosis of dying was often made late (33% within 1 day of death); 68% had palliative care referral; 88% had a NFR order documented; and 48% had a high care admission during the last hospitalisation. The authors recommended more work on assessing patients’ experiences of EOL care, preferred place of death and the role of palliative are teams. Alongside the increasing focus on the quality of EOL care in hospitals, there is also wide community and policy support for increased use of advance care planning (ACP) as a way of ensuring people’s wishes are known and respected when they are dying.10 In 2010, Western Australian legislation supporting ACP was enacted, with provision for the Advance Health Directive (AHD), Enduring Power of Guardianship (EPG) and a hierarchy of proxy decision-makers.11,12 The Respecting Patient Choices program has demonstrated that ACP is safe and effective13; however the limited research on ACP prevalence in Australia suggests low uptake,14,15 with no information on uptake among rural communities. There is a lack of available evidence to benchmark EOL care quality and ACP uptake in rural hospitals. This study aims to measure key indices of EOL care quality, documented ACP and the extent to which ACP is associated with the type of EOL care received, in a network of rural Western Australian hospitals.

approval (University of Western Australia HREC RA/4/ 1/6895).

Methods This study was a retrospective medical records audit to describe existing practices with the goal of improving service quality. The protocols were institutionally approved by the WA Country Health Service Great Southern Executive. As non-clinical members of the research team did not have access to any identifying data, the study was judged to be exempt from ethics

Setting The Great Southern region of WA is serviced by a network of four small rural hospitals (10 beds or less) and one larger regional hospital (132 beds). These facilities are all between 200–500 km from the nearest tertiary hospital and classified as ‘outer regional’ (RA3).9 The region’s population is approximately 55 000, and 3% are Aboriginal people.16 The regional hospital is co-located with a four-bed hospice, while the small rural hospitals each have a designated palliative care bed. A palliative care team (nurse manager, clinical nurse and part-time social worker) services the region. A general practitioner (GP) with palliative care experience and a palliative care specialist provide medical consultancy on a sessional basis. The smaller rural hospitals and hospice are staffed by GPs, while district medical officers, GPs and specialists staff the regional hospital. Phone access to palliative care specialists is available 24 hours a day.

Audit tool Audit items reflecting best practice EOL care were developed after literature review and consultation among the team, comprising a senior palliative care nurse, a palliative care physician, health researcher and biostatistician. The tool was trialled on five medical records in the pre-audit period and further refined based on feedback. The final version comprised 94 questions, assessing demographics, medical condition and family support, as well as detailed items relating to medical treatments, communication and emotional support during the terminal phase, both 3 days and 24 hours prior to death. Documentation of ACP was defined as ‘formal’ (presence of legal documents such as an AHD, living will or EPG) or ‘informal’ (discussions about future medical care wishes documented by staff). The tool is available on request. © 2015 National Rural Health Alliance Inc.

197

END OF LIFE CARE IN RURAL WA HOSPITALS

Sample

mented. Thirty-nine (43%) patients died in designated palliative care (non-acute) beds, and 25 (28%) deaths occurred in the regional hospice. Seventy-eight (87%) patients had opioid medication charted for pain, and 44 (49%) patients had six or more pain assessments in the final 24 hours prior to death. Six patients (7%) had medical emergency response team calls activated, and nine (10%) patients were admitted to the High Dependency Unit. Forty-five (50%) patients died on the Liverpool EOL care pathway. Patients typically received one or more family meetings over the last admission (median = 1.0, interquartile range: 1.0, 2.0). Of the 78 NFR orders, the majority were documented by either the patient’s GP (33%) or senior admitting doctor (50%) with fewer completed by junior doctors (3%) or on call/overnight doctors (14%).

The final admission prior to death was assessed for the 50 most recent adult deaths at the regional hospital and the 10 most recent adult deaths at each of four small rural hospitals (majority occurred in 2012). Cases of traumatic death, suicide or maternal death were excluded.

Data analysis B. A. audited the medical records and K. A. independently re-audited five random medical records. Interrater reliability between the two raters was high (Cohen’s kappa on 63 items with two or three levels of response = 0.903). Subgroups of patients were analysed with Kruskal–Wallis and t-tests used to detect differences between subgroups for continuous variables based on normality conditions, and chi-square analysis for categorical variables. Alpha levels were set at P = 0.05, and two-tailed tests were used.

Diagnosis and impact on care Fifty-eight per cent of patients were admitted with a non-malignant diagnosis (cardiovascular – 17%, neurological – 10%, infection – 10%, renal – 4%, other – 17%). Those admitted with a malignant diagnosis, in comparison, were less likely to be admitted via the Emergency Department (ED) (32% versus 77%, P < 0.001), more likely to be referred to the palliative care team (if available) (76% versus 56%, P = 0.044), more likely to be directly admitted to hospice (if available) (53% versus 3%, P < 0.001) and less likely to die in an acute bed (34% versus 73%, P < 0.001). However, they were less likely (26% versus 48%, P = 0.04) to have some form of ACP (formal or informal) documented.

Results Demographic statistics Table 1 shows demographic statistics and admission characteristics for the sample of deceased patients by location.

EOL care For this cohort of 90 deceased patients, 66 (73%) were admitted for 3 days or more prior to death. Twentyeight (31%) patients had a clear diagnosis of dying (health-care professionals mentions dying/terminal) in the notes during the final admission, 58 (64%) patients were referred to palliative care during or prior to the final admission and 78 (87%) had an NFR order docuTABLE 1:

Presence of documented ACP Table 2 shows the number of patients with different types of ACP documented. Although there was a significant difference in formal versus informal ACP rates

Demographic statistics and admission characteristics for the sample of decedents by location

Characteristic

Regional hospital (n = 50)

Rural hospitals (n = 40)

Total (n = 90)

n (%) female Median (IQR) age in years n (%) with malignant cause of death n (%) admitted via emergency department n (%) admitted less than 3 days prior to death Median (IQR) length of final admission in days n (%) with no decision making capacity at admission Mean (SD) Karnofsky functional status in %

26 83 18 34 12 7.0 10 37

21 80.5 18 18 11 8.0 10 34

47 82 36 52 23 7.0 20 35

IQR, interquartile range; SD, standard deviation. © 2015 National Rural Health Alliance Inc.

(52.0%) (75, 90.2) (36.0%) (68.0%) (24.0%) (2.8, 16.5) (20.0%) (12.5)

(52.5%) (72.0, 86.8) (45.0%) (45.0%) (27.5%) (2.0, 18.0) (25.0%) (13.5)

(52.2%) (74.8, 87.0) (40.0%) (57.8%) (25.6%) (2.0, 16.5) (22.2%) (13)

198 TABLE 2:

K. AURET ET AL.

Number (%) of patients who had evidence of different types of ACP documented in medical record, by location

Type of ACP

Regional hospital (n = 50)

Rural hospitals (n = 40)

Total (n = 90)

n n n n n

49 44 11 19 1

40 34 16 7 8

89 78 27 26 9

(%) with nominated family member documented (%) with NFR order documented (%) with preferred place of death documented (%) with informal ACP documented (%) with formal ACP documented

(98.0%) (88.0%) (22.0%) (38.0%) (2.0%)

(100%) (85.0%) (40.0%) (17.5%) (20.0%)

(98.9%) (86.7%) (30.0%) (28.9%) (10.0%)

ACP, advance care planning.

across the two hospital types, the association between the presence of any type of ACP (combined formal or informal) and hospital type was not significant (P = 0.81). Of the nine patients with formal ACP, one had an AHD, seven had a living will and one had an EPG. Five of the nine patients with formal ACP documents received care in one small rural hospital site.

Impact of ACP on type of EOL care There was no association between presence of formal or informal ACP in the medical record and palliative care team referral (57% with ACP, 69% without ACP, P = 0.25), commencement of Liverpool EOL care pathway (43% with ACP, 55% without ACP, P = 0.28), likelihood of receiving four-hourly pain assessments in the final 24 hours of life (43% with ACP, 53% without ACP, P = 0.36), number of family meetings (1.80 with ACP, 1.56 without ACP, P = 0.99) or dying in a nonacute/palliative bed in hospital (43% with ACP, 44% without ACP, P = 0.94).

Impact of EOL care pathway The presence of a Liverpool EOL care pathway was associated with a greater likelihood of referral to the palliative care team (82% versus 47%, P < 0.001) and a greater likelihood of receiving six or more pain assessments in the final 24 hours of life (76% versus 22%, P < 0.001). Patients on the EOL care pathway had a greater number of family meetings (2.0 versus 1.0, P < 0.001) than those who were not on the pathway and were less likely to die in an acute hospital bed (38% versus 76%, P < 0.001).

Discussion This retrospective notes audit in a network of rural hospitals focused on the type of EOL care received and evidence of ACP documented in the notes. The patients in this study were very different from those in the Alice Springs Hospital audit8 (with a higher proportion of

elderly, palliative care patients). These patient demographics might contribute somewhat to the observed patterns of ACP, use of EOL pathways and referral for specialist palliative care. This patient case mix will be more generalisable to those seen in other less remote services in regional Australia, adding to the descriptive data available for regional services to benchmark the care they offer locally and the outcomes achieved. Just over one in three deceased patients had formal or informal ACP documented. This reflects the low rates of ACP within the wider community,14,15,17 but the awareness that those whose deaths were reported in this study were generally so close to death by the time of admission, and often not competent to make complex decisions, reinforces the value of early engagement with ACP in primary care.18 International NICE guidelines reinforce the key importance of aligning care with the needs and preferences of the dying person,19 which is only possible if they are known, and the clinicians and systems concerned are able to respond appropriately. One small rural hospital site accounted for five of the nine formal ACP documents in this audit, suggesting that institutional culture is important in determining clinical practice of documenting EOL care preferences. However, even for those with ACP, this audit did not detect meaningful differences in care compared with those without ACP. The literature suggests most ACP tends to request less aggressive treatment and more involvement with palliative care and family,13,20,21 so care should be different if ACPs are being responded to in a meaningful way. Previous studies have demonstrated some of the challenges in translating ACP into clinical care in the hospital setting.22 The lack of impact of ACP contrasts with the Liverpool EOL care pathway, which promotes a systematic response to care provision for the dying, based on a series of goals, clinician checklists and clear documentation. People on the pathway did receive a different type of care. This is of particular interest given the recent controversy surrounding the Liverpool EOL care pathway in the United Kingdom23 and calls for detailed evidence to inform policy decisions regarding such pathways in Australia.24 © 2015 National Rural Health Alliance Inc.

END OF LIFE CARE IN RURAL WA HOSPITALS

A limitation of this work, however, especially in commenting on the impact of ACP, is that it did not capture the experience of those who died in their own home or care facility; such deaths might have been impacted by ACP. As with a number of other studies, this audit shows that the diagnosis of cancer continues to be associated with a different pattern of care1,6,25 and remains a challenge to equitable delivery of high quality EOL care for all. Hardy et al. comment on the difficulties collecting data about important aspects of EOL care from medical records review,5 and we would reiterate their views and encourage ongoing work on tools for quantifying EOL care quality and the impact of ACP on the care being delivered.3

Conclusion This study provides benchmarking information that might assist other rural hospitals and suggests ongoing work to optimize measurement of the quality of end of life care.

Acknowledgements The research team acknowledges the support of the Western Australian Country Health Service – Great Southern. This audit was supported by a Rural Clinical School of Western Australia Small Project Grant 2012. K. A. and B. A. are providers of palliative care in the region under study, and both were involved in the provision of care to a small number of the audited cases in the regional centre.

Author contributions K. A. developed the research question, performed literature review, designed the audit tool, performed some of the medical record reviews, participated in data analysis and wrote the majority of the text. C. S. also developed the research question, performed literature review, significantly designed the audit tool, led the data analysis and wrote some of the text. B. A. performed literature review, contributed to the audit tool design, performed the medical record review, discussed the results and contributed to editing the text. S. E. contributed to the audit tool design, provided statistical leadership and second review, discussed the results and contributed to editing the text.

References 1 Australian Institute of Health and Welfare. Trends in palliative care in Australian hospitals. Cat. no. HWI 112. Canberra: AIHW, 2011.

© 2015 National Rural Health Alliance Inc.

199 2 Australian Council on Healthcare Standards. EquiP National Guidelines Standard 12. Sydney: ACHS, 2012. 3 Clark K, Byfieldt N, Green M, Saul P, Lack J, Philips JL. Dying in two acute hospitals: would usual care meet Australian national clinical standards? Australian Health Review 2014; 38: 223–229. 4 Levinson M, Mills A, Hutchinson AM, Heriot G, Stephenson G, Gellie A. Comparison of not for resuscitation (NFR) forms across five Victorian health services. Internal Medicine Journal 2014; 44: 671–675. 5 Hardy JR, Haberecht J, Maresco-Pennisi D, Yates P, Queensland AN. Audit of the care of the dying in a network of hospitals and institutions in Queensland. Internal Medicine Journal 2007; 37: 315–319. 6 Rosenwax LK, McNamara BA. Who receives specialist palliative care in Western Australia – and who misses out. Palliative Medicine 2006; 20: 439–445. 7 Evans R, Stone D, Elwyn G. Organizing palliative care for rural populations: a systematic review of the evidence. Family Practice 2003; 20: 304–310. 8 Nadimi F, Currow DC. As death approaches: a retrospective survey of the care of adults dying in Alice Springs Hospital. The Australian Journal of Rural Health 2011; 19: 4–8. 9 Australian Standard Geographical Classification – Remoteness Area (ASGC-RA). 2013. [Cited 27 Jun 2013]. Available from URL: http://www.health.gov.au/internet/ otd/Publishing.nsf/Content/RA-intro 10 Australian Health Minister’s Advisory Council. A National Framework for Advance Care Directives, 2011. 11 Department of Health WA. Preparing an Advance Health Directive. East Perth: Department of Health (WA), 2010. 12 Office of the Public Advocate WA. A Guide to Enduring Power of Guardianship in Western Australia. East Perth: Office of the Public Advocate (WA), 2010. 13 Detering KM, Hancock AD, Reade MC, Silvester W. The impact of advance care planning on end of life care in elderly patients: randomised controlled trial. British Medical Journal 2010; 340: c1345. doi:10.1136/ bmj.c1345. Epub 23 March 2010. 14 Bezzina AJ. Prevalence of advance care directives in aged care facilities of the Northern Illawarra. Emergency Medicine Australasia 2009; 21: 379–385. 15 Taylor DM, Ugoni AM, Cameron PA, McNeil JJ. Advance directives and emergency department patients: ownership rates and perceptions of use. Internal Medicine Journal 2003; 33: 586–592. Epub 2003/12/06. 16 Department of Health WA. WA Country Health Services – Great Southern. 2014. [Cited 5 Sept 2014]. Available from URL: http://www.health.wa.gov.au/services/detail.cfm? Unit_ID=111 17 Brown M, Grbich C, Maddocks I, Parker D, Roe P, Willis E. Documenting end of life decisions in residential aged care facilities in South Australia. Australian and New Zealand Journal of Public Health 2005; 29: 85–90. 18 Royal Australian College of General Practitioners. Position statement: advance care planning should be incorporated into routine general practice, 2012 [28/2/2014].

200 19 National Institute for Health and Clinical Excellence. Quality standard for end of life care for adults. 2011. [Cited 1 Jan 2013]. Available from URL: http:// publications.nice.org.uk/quality-standard-for-end-of-life -care-for-adults-qs13 20 Singer PA, Martin DK, Kelner M. Quality end-of-life care – patients’ perspectives. Jama-Journal of the American Medical Association 1999; 281: 163–168. 21 Emanuel LL, Barry MJ, Stoeckle JD, Ettelson LM, Emanuel EJ. Advance directives for medical care: a case for greater use. The New England Journal of Medicine 1991; 324: 889–895. 22 Burkle CM, Mueller PS, Swetz KM, Hook CC, Keegan MT. Physician perspectives and compliance with patient

K. AURET ET AL.

advance directives: the role external factors play on physician decision making. BMC Medical Ethics 2012; 13: 31. 23 Neuberger J, Aaronovitch D, Bonser T et al. More Care, Less Pathway: A Review of the Liverpool Care Pathway. London: Department of Health, 2013. 24 Chan RJ, Webster J, Phillips J, Currow DC. The withdrawal of the Liverpool Care Pathway in the United Kingdom: what are the implications for Australia? The Medical Journal of Australia 2014; 200: 572–573. 25 Billingham MJ, Billingham S-J. Congruence between preferred and actual place of death according to the presence of malignant or non-malignant disease: a systematic review and meta-analysis. BMJ Supportive & Palliative Care 2013; 3: 144–154.

© 2015 National Rural Health Alliance Inc.

Advance care planning and end-of-life care in a network of rural Western Australian hospitals.

To provide a current perspective on end-of-life (EOL) care in regional Western Australia, with a particular focus on the final admission prior to deat...
92KB Sizes 0 Downloads 8 Views