Abstract
Context
Over the past decade, there has been widespread adoption of end-of-life care pathways as a tool to better manage care of the dying in a variety of care settings. The adoption of various end-of-life care pathways has occurred despite lack of robust evidence for their use.
Objectives
This integrative review identified published studies evaluating the impact of an end-of-life care pathway in the acute and hospice care setting from January 1996 to April 2010.
Methods
A search of the electronic databases Scopus and Cumulative Index of Nursing and Allied Health Literature as well as Medline and the World Wide Web were undertaken. This search used Medical Subject Headings key words including “end-of-life care,” “dying,” “palliative care,” “pathways,” “acute care,” and “evaluation.” Articles were reviewed by two authors using a critical appraisal tool.
Results
The search revealed 638 articles. Of these, 26 articles met the inclusion criteria for this integrative review. No randomized controlled trials were reported. The majority of these articles reported baseline and post implementation pathway chart audit data, whereas a smaller number were local, national, or international benchmarking studies. Most of the studies emerged from the United Kingdom, with a smaller number from the United States, The Netherlands, and Australia.
Conclusion
Existing data demonstrate the utility of the end-of-life pathway in improving care of the dying. The absence of randomized controlled trial data, however, precludes definitive recommendations and underscores the importance of ongoing research.
Key Words
Introduction
The setting and manner in which people die is of interest to clinicians and policy makers as a way not only of meeting consumer choice, in relation to dying with dignity, but also of reducing costs and avoiding unnecessary and clinically futile interventions. Most deaths in the developed world now occur in the acute care setting,
1
, 2
and this is expected to increase as the population ages and more people die of chronic and complex conditions.3
Yet, few of these acute care deaths will be managed by specialist palliative care providers, with most being handled by health professionals with no formal training in diagnosing dying and for whom end-of-life care is not their primary area of expertise.4
In the acute care environment, where the focus has traditionally been on “curing and prolonging life,” there can be tensions and challenges in moving care toward a symptom management model, such as palliative care.
4
Supporting patients and families to make this transition in a curative culture can seem confronting to many acute care health professionals and, in part, may explain late recognition of dying and a delay in initiating end-of-life care in hospitals. Unfortunately, late recognition of dying is likely to result in poor symptom management and suboptimal psychosocial and spiritual care for dying patients and their families.5
, 6
Consumer concerns over inadequate end-of-life care have been identified in various international health service reform agendas7
and currently accounts for more than half of all complaints received by the National Health Service.5
The seminal work of Ellershaw and colleagues in developing integrated end-of-life care pathways emerged in response to the shortcomings of the health care system’s capacity to deliver evidence-based end-of-life care in the late 1990s.
8
, 9
, 10
Integrated care pathways detail the essential elements of care required to manage a specific clinical problem and ensure that the best available evidence is systematically integrated into care delivery while providing a framework for auditing and benchmarking care.11
The Liverpool Care Pathway (LCP) was designed to improve care of cancer patients in the last 48 hours of life and facilitate monitoring of the level and type of end-of-life care provided.10
A multidisciplinary team philosophy underpins the approach of this end-of-life care pathway, which is divided into three sections: initial assessment, ongoing care, and care after death, with 18 goals of care identified.12
The template provides an aid to decision making, with the pathway replacing all other documentation.10
Over the past decade, the LCP has undergone many revisions and is now used in more than 1000 care organizations in the United Kingdom.8
, 13
The LCP is a key element of the Gold Standards Framework promoted by the National Institutes of Clinical Excellences in the United Kingdom to facilitate the delivery of palliative care beyond specialist care settings.14
More recently, the pathway has been adapted for use in other population groups, such as people dying with advanced chronic kidney disease15
or from stroke and those dying in the emergency setting; these adaptations have been accomplished in parallel with translation of the LCP into other languages.16
, 17
The widespread diffusion of the end-of-life pathway has resulted in the development and implementation of end-of-life care pathways in the United States,
18
, 19
, 20
Europe,21
Australia,22
and China.23
These pathways are now used in hospices, acute care, community, and residential aged care sectors.8
, 24
Despite the end-of-life care pathways’ widespread adoption, recent public comments have called into question the promotion and adoption of these clinical tools because of fears that these pathways fail to be sensitive enough to accurately diagnose dying or meet the individual needs of dying people; there is concern that they may actually hasten death, particularly if all reversible causes for decline are not systematically investigated.25
, 26
A recent Cochrane review found 920 potentially relevant articles but no studies that met the inclusion criteria of a randomized controlled trial, confirming widespread uptake of the pathway in the absence of supporting evidence.
27
This review concluded that there was no evidence of harm that would prevent the use of the end-of-life pathway to manage the care of people actively dying.27
In an era of evidence-based medicine, parallels can be drawn between the rapid uptake of the end-of-life care pathway and laparoscopic cholestectomy,28
with both practices being adopted in a vacuum of evidence from randomized controlled trials.Given the global adoption of the end-of-life care pathway, emerging concerns over its use, and a lack of randomized controlled trials,
27
there is a need for further appraisal of the end-of-life care pathway.25
The purpose of this integrative review was to identify published studies describing the use of the end-of-life care pathway to determine its 1) impact on consumers (patients and families), health professionals, and the acute care and/or hospice systems and 2) barriers and facilitators to its implementation. The integrative review aimed to address the following questions: 1) In which population(s) has the end-of-life care pathway predominately been used to manage care of the dying? 2) Is there evidence to support the end-of-life care pathway’s use in acute care and/or hospice systems? 3) What are the implications of these findings for evidence-based care of the dying in the acute care and/or hospice setting? 4) What are the key elements underpinning effective implementation of the end-of-life care pathway? and 5) What are the gaps in the evidence and future research directions?
Methods
Definitions
An integrative review, using prespecified parameters, allows systematic appraisal of the literature and synthesizing of research findings on a focused topic. This method allows for the inclusion of varied research designs and methodological approaches to provide a comprehensive analysis of a topic and draw overall conclusions from primary studies.
29
An integrative review includes problem identification, a literature search, data evaluation, data analysis, and presentation.29
This structured and focused approach to conducting a literature review is appropriate in subject areas where a summation of effect and identifying issues of implementation are beneficial. Similar approaches to a systematic review are incorporated in this technique to ensure rigor through a standardized review protocol to enable replication.30
Setting and Sample
Both randomized and nonrandomized studies were eligible for inclusion in this review. Articles were selected if an end-of-life care pathway was used to manage the dying phase in the acute care and/or hospice setting and if care delivered to dying patients and/or their families was evaluated. Articles were excluded if they reported a single case study or described process measures only.
This integrative review was conducted over the period from November 1, 2009 to the final week of April 2010. The initial literature search was undertaken in Cumulative Index of Nursing and Allied Health Literature, Medline, Embase, PsycINFO, Scopus, and the World Wide Web, using Google Scholar and Mednar search engines, and was limited to articles published since 1996 in the English language, related to adult patients. The Medical Subject Headings (MeSH) terms were identified, with four key concepts explored: acute care, palliative care, end-of-life care, and pathways. MeSH key words including palliative care, terminal care, end-of-life care, dying, acute care, pathways, integrated care pathway, end-of-life care pathway, evaluation, and health services research were used in the search. Reference lists were checked for additional sources; hand searching of relevant journals also was undertaken.
The search strategy generated 638 articles. No randomized controlled trials or meta-analyses were identified. Five hundred sixty articles were excluded based on the inclusion criteria, leaving 78 potential studies to be included in the review. After reviewing the full article, 52 of these articles were rejected because they did not provide empirical data about the impact of the end-of-life care pathway for adults dying in the acute care or hospice setting. Articles that were rejected tended to describe single case studies, or the development or implementation process, without reporting outcomes. Three studies did not meet the inclusion criteria because these were not available in English, and a further four articles were rejected because they were editorials. At the end of this process, 26 studies matched the inclusion criteria.
Data Analysis
Appraising the quality of a study for inclusion in an integrative review is a complex undertaking.
29
To assist with this quality assessment process, two evidence evaluation tools developed by the Australian Palliative Residential Aged Care (APRAC) Project Guidelines were used to appraise the studies.31
The APRAC quantitative studies evaluation tool adopted the Australian National Health and Medical Research Councils level of evidence categories.32
This tool was modified to include verifiable quality or program evaluation data (Level V evidence).National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
33
This data extraction tool facilitated a systematic approach to appraising the strength of the evidence, and the quality of methods used, while determining the transferability of the results to make a recommendation about the pathway’s applicability.34
The APRAC qualitative studies level of evidence (Level QE) evaluation tool uses eight questions to appraise the aim of the study and appropriateness of the method.31
For each included study, data on methods, setting, population, findings, evidence levels, implementation features, and strengths and weaknesses of the approach were extracted onto the data collection tool and into a matrix by J. L. P. and P. M. D. In the event of disagreement, a third reviewer (E. J. H.) provided clarification. This process facilitated the summary of key points and synthesizing of information to inform organizational decision making around the implementation of an end-of-life care pathway. After completion of the matrix, it became clear that the heterogeneity of studies precluded formal meta-analysis. Therefore, thematic content analysis was undertaken to identify key issues.
Results
Table 1 summarizes the 26 studies included in this integrative review and data elements included in the data analysis process. The defining features of each study are summarized in Table 2. In accordance with the hierarchy of evidence,
32
most studies (n=11) were case series reporting pre- and post-test outcomes (Level IV), with the highest level of evidence (Level III-3) coming from a small number (n=2) of comparative studies without concurrent controls. Despite this limitation, the studies included in this review provided useful insights into the level of end-of-life care delivered and information relating to the pathway implementation process.National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
Table 1Summary of Included Studies
| Source | Country | Focus | Design | Participants | Outcomes | Level of Evidence | Features |
|---|---|---|---|---|---|---|---|
| van der Heide et al. 2010 6 | The Netherlands | Use of LCP for management of cancer deaths in acute, aged, and a home care setting. | Mixed method. Survey and retrospective baseline and prospective post-LCP chart audit. No control. |
| Pathway usage had no significant impact on the types of drugs used in the last 72 hours of life and reduced the use of drugs with a potentially life-shortening effect from 46% to 28%. | IV |
|
| Jackson et al. 2009 22 | Australia | Use of modified LCP in four general medical wards. | Retrospective baseline (limited data) and post-LCP chart audit. No control. | Baseline audit (n=15) and post-pathway audit (n=20). | Descriptive statistics only. Limited baseline data reported. Increased adherence to evidence-based care observed in 20 KPIs. | V |
|
| Mullick et al. 2009 41 | United Kingdom | Perception of carers of people who had died on LCP in acute care. | Prospective carer survey based on the LCP domains of care. No control. | Survey offered to 42 consecutive carers over a 14-week period. | Small sample (n=25) precludes statistical analysis. 59% response rate. 84% reported high levels of satisfaction. | V | Carer survey |
| Paterson et al. 2009 36 | United Kingdom | Use of modified LCP in a short-stay emergency ward. | Mixed method. Retrospective post-LCP chart audit. No control. Health care provider survey. |
| Descriptive statistics only. Survey themes: improved continuity of care, documentation of care, and communication—team and patient/family. | V |
|
| Lo et al. 2009 16 | China | Use of modified LCP for people dying from cancer in an inpatient palliative care unit. | Retrospective benchmarking LCP chart audit—three sites: Rotterdam, United Kingdom, and China. Unmatched sample. No control. | Post-LCP audit China (n=51); compared with United Kingdom (n=40) and Rotterdam (n=40). | Descriptive statistics only. 25% missing data—awareness of dying and religious or spiritual needs. | IV |
|
| The Marie Curie Palliative Care Institute, 2009 53 | United Kingdom | Impact of LCP on EOL care in acute hospitals. | Retrospective LCP chart audit. Repeat 2007 methodology. 54 No control. | Invitation sent to 214 hospitals in 140 acute care trusts. Audit of 30 (max) consecutive deaths on LCP in each participating hospital from October 2008–December 2008. 155 hospitals from 114 acute hospital trusts (n=3893 patients) participated. | 68% of sites using LCP. 21% of all patients who died were on LCP. 39% of sites had an LCP facilitator and 99% of sites have an SPC team. 61% of people on LCP had a diagnosis other than cancer. 33 hours median time on LCP. Drugs prescribed for agitation and restlessness were only given in 37% of cases. | IV |
|
| Veerbeek et al. 2008 42 | The Netherlands | Impact of LCP on communication and carers' bereavement. | Retrospective pre-post test design questionnaire administered three months after death. No control. |
| Similar patient and relative characteristics at both times. Evidence that the LCP moderately contributes to lower levels of bereavement in relatives. | IV | Pathway used across multiple settings |
| Veerbeek et al. 2008 40 | The Netherlands | Impact of LCP on symptom burden and communication in seven settings: four oncology units, two aged care facilities, and one home care organization. | Mixed methods. Prospective pretest (2003–05) and post-test (2005–06) chart audit. No control. Surveys: nurse and carer. |
| No statistical difference in patient characteristics. Improvement in aspects of care documented with LCP (P<0.001). Perception that symptom burden is less on LCP: nurse (P=0.008) and relatives (P=0.016). Nurses’ assessments not blinded. | IV |
|
| Hardy et al. 2007 46 | Australia | LCP audit tool used to identify EOL priority areas in three hospices, four hospitals, and one aged care facility in a defined geographical area. | Retrospective baseline LCP chart audit. No control. | Eight health care institutions completed the baseline audit on 20 consecutive charts (n=160). | Descriptive statistics only. Discussion focuses on differences between settings but data not provided. | V |
|
| The Marie Curie Palliative Care Institute, 2007 54 | United Kingdom | Impact of LCP on EOL care in U.K. acute care sector. | Retrospective LCP chart audit—30 (max) consecutive deaths on LCP in each participating hospital from September 2006–November 2006. | Invitation sent to 209 hospitals in 108 acute care trusts. 94 hospital trusts (n = 2672 patients) participated. | Descriptive statistics only. 40% of all hospitals and 60% of acute care trusts participated. Variations across setting. 15% of patients on LCP at time of death. Median time on LCP 33 hours. 55% of patients had a non-cancer diagnosis. 47% of sites had an LCP facilitator. SPC team present at 95% of sites. | IV |
|
| Lhussier et al. 2007 39 | United Kingdom | Perception of health professionals and bereaved carers familiar with the LCP in two primary care trusts. | Qualitative study. Semistructured interviews. | Pathway facilitators (n=22), nurses (n=10), and bereaved carers (n=10). | Multiple challenges identified related to variability across care settings, communication with patients, and the provision of emotional or spiritual support. Perception that ICP promoted greater care consistency, improved continuity, and facilitated proactive care delivery. | QE |
|
| Hugel et al. 2006 50 | United Kingdom | Management of RTS in people dying on the LCP in a specialist palliative care unit. | Retrospective pretest and prospective post-test chart audit. No control. | Matched sample. Prospective data—RTS patients on glycopyrronium (n=36) compared with those on hyoscine (n=36) (retrospective data). | Glycopyrronium group more likely to have a response than the hyoscine group (P<0.01). No statistical difference in the levels of agitation observed in the glycopyrronium and hyoscine groups. | IV | Evaluation symptom management |
| Veerbeek et al. 2006 21 | The Netherlands | Comparison of EOL care in an inpatient palliative care unit in The Netherlands and a hospice in the United Kingdom. | Retrospective benchmarking LCP audit comparing two sites. No control. |
| Descriptive statistics only. 9/14 goals of care achieved for >80% of deceased at both sites. | IV |
|
| Gambles et al. 2006 38 | United Kingdom | LCP perceptions of hospice clinicians. | Qualitative study. Semistructured interviews. |
| Perception that the LCP has a role in the hospice setting; improves documentation, promotes continuity of care, and enhances communication and care of relatives. | QE | Ongoing LCP education and feedback |
| Hinton and Fish, 2006 35 | Ireland | Use of a modified LCP for people dying in an acute renal ward. | Retrospective baseline and post-LCP chart audit. No control. | Baseline sample (n=10) and post-LCP (n=10). | Small sample precludes statistical analysis. | V |
|
| Main et al. 2006 44 | United Kingdom | Use of an SCP for older people dying in acute care. | Retrospective baseline and post-SCP chart audit. No control. | Baseline sample (n=34) and post-SCP sample (n=35). | Preliminary data Improvements in all SCP domains. Formal evaluation to be undertaken. | V |
|
| Bailey et al. 2005 18 | United States | Use of inpatient comfort care order sets for people dying in acute care. | Retrospective baseline and postcomfort care order sets audit of last seven days of care. No control. | Baseline sample (n=108) and postcomfort care order set plan sample (n=95). | Intervention improved EOL care with an increase in the mean number of symptoms documented, comfort care order sets implemented, and opioid and DNR orders (P<0.001). | IV |
|
| Luhrs et al. 2005 20 | United States | Use of PCAD pathway for people dying of cancer in acute oncology ward. | Controlled retrospective baseline and post-PCAD chart audit. Noncomparable sampling. |
| No significant characteristic difference between patient groups. Evidence that introduction of PCAD improved EOL care with better goals of care documentation (P = 0.0001), fewer interventions and more symptoms assessed (P = 0.004) and symptoms managed in accordance to guidelines (P = 0.02). | III–3 |
|
| Bookbinder et al. 2005 19 | United States | Impact of PCAD pathway on patient’s access to resources, communication, and care delivery in a geriatric or oncology unit (study units), palliative care unit (benchmark), or two general medical units (comparison). | Retrospective baseline and post-PCAD chart audit. |
| 33% of patients on the geriatric/oncology units and 100% on the palliative care unit died on the PCAD. Despite positive results, unable to confirm if the PCAD drove improvements in the care of the dying. | III–3 |
|
| Mirando et al. 2005 45 | United Kingdom | Impact of LCP in 18 clinical areas in six acute hospitals. | Retrospective baseline and post-LCP chart audit. No control. | Baseline (n=50) and post-LCP (n=50) chart audit. | Small sample precludes statistical analysis. 20% improvement in 9/16 areas after LCP implementation. | IV |
|
| Grogan et al. 2005 49 | United Kingdom | Impact of ICP on EOL prescribing in a specialist palliative care unit. | Retrospective audit. No control. | Retrospective chart audit (n=68). | Descriptive statistics only. Agitation most unstable symptom. Median dose of drugs in syringe driver did not change between LCP commencement and death for all symptoms and drugs. | V | LCP used to evaluate symptom management |
| Kass and Ellershaw, 2003 51 | United Kingdom | People who died on the LCP in a specialist palliative care unit with RTS. | Noncontrolled retrospective trial. | Sample (n=202). | Risk factors for RTS: male (P=0.034), lung cancer (P=0.003), and prolonged dying (P=0.001). 59% people with RTS given hyoscine. 35.5% died with RTS and 18.6% had no RTS-free period. | IV | LCP used to evaluate symptom management |
| Jack et al. 2004 43 | United Kingdom | Use of LCP for people who die in an acute stroke unit. | Retrospective noncontrolled baseline and post-LCP chart audit. | Baseline 2001–02 (n=20) and post-LCP 2002–03 (n=20) on a U.K. stroke unit. | Descriptive statistics. Small sample sizes. Improvements observed in 6/7 key areas assessed. | V | Audit undertaken by research team |
| Jack et al. 2003 37 | United Kingdom | Acute care nurses perceptions of the LCP. | Focus groups (n=2) and semistructured interview (n=1). | Sample (n=15) | Perception that LCP enhances EOL care by improving symptom control and communication with relatives/carers; increases EOL knowledge and confidence; and promotes more streamlined documentation of care needs and delivery. | QE | |
| Fowell et al. 2002 52 | United Kingdom | Use of LCP died in four distinct care settings: acute care, hospices, specialist inpatient units, and community care. | Retrospective benchmarking of LCP. | First 500 variances comparing four sites: acute care (n=133), hospices (n=185), specialist inpatient units (n=104), and community care (n=78). | Descriptive statistics only. No recorded variance for 50%, implying symptoms controlled. Differences in variance rates across care settings. | V |
|
| Ellershaw et al. 2001 12 | United Kingdom | Impact of LCP on EOL symptom management in an inpatient hospice unit in a hospital. | Baseline LCP audit. | Inpatients (n=168) | Increased length of time on ICP associated with better pain and agitation control (P<0.01). | V | |
LCP=Liverpool Care Pathway; EOL=end-of-life; ICP=integrated care pathway; KPIs=key performance indicators; PCAD=Palliative Care for Advanced Disease Pathway; RTS=respiratory tract secretions; SCP=Supportive Care Pathway.
Evidence levels: I,
32
systematic review of all relevant randomized controlled trials; II,National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
32
at least one properly designed randomized controlled trial; III-1,National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
32
well-designed pseudo-randomized controlled trials; III-2,National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
32
comparative studies with concurrent controls and allocation not randomized, case-control studies or interrupted time series with a control group; III-3,National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
32
comparative studies with historical control, two or more single-arm studies, or interrupted time series without a parallel control group; IV,National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
32
case series, either post-test or pretest and post-test; V,National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
33
case report or systematically obtained verifiable quality or program evaluation data; QE,31
qualitative evidence.Table 2Summary of Study Design, Populations, Settings, and Implementation Features
| Features | Source |
|---|---|
| Design | |
| Prepathway chart audit | 12 , 46 |
| Pre- and post-test pathway chart audit | 6 , 18 , 19 , 20 , 22 , 35 , 40 , 43 , 44 , 45 |
| Post-test pathway chart audit | 22 , 36 |
| Benchmarking | 16 , 21 , 52 , 53 , 54 |
| Symptom control audits | 49 , 50 , 51 |
| Survey—health professionals | 6 , 36 , 39 , 40 |
| Survey—carers | 6 , 39 , 41 , 42 |
| Prospective consent | 40 |
| Statistics reported | 6 , 18 , 19 , 20 , 40 , 42 , 50 |
| Matched sample | 21 , 50 |
| Controlled | 20 |
| Qualitative studies | 37 , 38 |
| Populations | |
| Malignant | 6 , 16 , 20 , 51 |
| Nonmalignant | 35 |
| Geriatric | 43 , 44 |
| Mixed (malignant, nonmalignant, and geriatric) | 12 , 18 , 21 , 22 , 36 , 38 , 39 , 40 , 42 , 45 , 46 , 49 , 50 , 52 , 53 , 54 |
| Settings | |
| Acute care | 18 , 19 , 20 , 22 , 35 , 37 , 41 , 43 , 44 , 45 , 53 |
| Palliative care unit or hospice | 12 , 16 , 21 , 38 , 49 , 50 , 51 |
| Mixed (acute, hospice, and community) | 6 , 39 , 40 , 42 , 52 , 54 |
| Implementation features | |
| Clinical champion | 18 , 19 , 35 , 45 |
| Facilitator | 35 , 39 , 45 , 53 , 54 |
| Education | 18 , 19 , 20 , 35 , 45 , 53 , 54 |
| Defined performance indicators | 22 |
| Quality Improvement approach | 19 , 20 , 53 , 54 |
| Modified pathway | 18 , 35 , 36 |
| In-house audit process (nonresearch team audit) | 46 , 53 , 54 |
Populations and Settings
Fifteen studies were conducted in the United Kingdom, four in The Netherlands, three in the United States, two in Australia, and one each in Ireland and China. Although end-of-life care pathways were originally intended for cancer populations, only four studies focused exclusively on people dying with cancer. Three studies examined the impact of the pathway on care for people dying from other causes, such as stroke, renal failure, postcardiac arrest, or an intracranial hemorrhage. Eleven studies reported on pathway data related to the care of people dying in a range of acute care settings, and seven studies were undertaken in a palliative care or hospice environment. Four studies outlined pathway modifications, including tailoring end-of-life care for specific populations,
35
, 36
increasing the observation period to manage workforce issues,16
expanding the time period that death was predicted to occur within seven days,18
and maintaining active treatment.18
Qualitative Studies
Four qualitative studies focusing on end-of-life care pathway domains reported findings from focus group
37
and semistructured interviews.38
, 39
These studies all tended to be undertaken early in the pathways evolution. The key themes emerging from these qualitative studies revealed a perception that the end-of-life care pathway had impacted positively on care of the dying through better coordination of care, enhanced communication, and greater adherence to evidence-based clinical guidelines (Level QE).37
, 39
Health Professional and/or Carer Perceptions
Six studies reported health professional
6
, 36
, 40
or carer survey data6
, 40
, 41
, 42
on various aspects of end-of-life care pathway usage. There was some evidence (Level IV) that the use of an end-of-life care pathway contributed to lower levels of bereavement in relatives42
and lower levels of perceived patient symptom burden in the last days of life.40
Pre- and/or Post-Pathway Audits
Nearly all studies identified in this end-of-life care pathway integrated review were nonexperimental studies that reported baseline and post-pathway implementation chart audits.
6
, 18
, 19
, 20
, 22
, 35
, 40
, 43
, 44
, 45
Few of these pre- and post-test chart audit studies included a control group19
, 20
or collected contemporaneous data,6
and only one study matched the groups.21
A subset of chart audit studies reported just baseline12
, 46
or post-pathway chart audit data.22
, 36
The baseline pathway chart audit data confirmed previous reports that people dying in the acute care sector experience varying levels of care depending on who was responsible for managing the dying process.5
, 47
Inadequate symptom management and poor communication within the multidisciplinary team and with patients and their families were noted.48
The majority of these chart audits involved small sample sizes, reflecting a quality improvement approach, which limited data analysis to descriptive statistics only. A pre-post test pathway study with a larger sample noted some positive changes in end-of-life care delivery but, because of inability to control for confounders, was unable to confirm whether the pathway actually drove improvements in care of the dying.
19
Another large pre-post test pathway study found that the intervention improved end-of-life care, with an increase in the mean number of symptoms documented, comfort care plans implemented, and opioid and “do not resuscitate” (DNR) orders (P<0.001).18
However, this same study also reported an increase in the use of restraints (P<0.001), which is not reflective of best practice end-of-life care.18
A mixed methods study using pre-post pathway chart audits and carer follow-up obtained patient consent demonstrating that prospective end-of-life care pathway studies are indeed possible.40
Most studies provided details of the chart audit selection process, but few defined the criteria by which “expected to die” was determined.
18
It is estimated that approximately 50% of deaths in acute care cannot be predicted, and it can be assumed that only half of all retrospective baseline chart audit records would contain information suggesting that the patient’s death was expected.4
These end-of-life prognostication challenges may account for some of the suboptimal pre-pathway implementation results.6
Retrospective Symptom Management
Three studies reported retrospective symptom management pathway data.
49
, 50
, 51
A matched sample study without a control compared retrospective and prospective pathway data to determine the effectiveness of various medications to manage respiratory tract secretions.50
One other study retrospectively examined the impact of the pathway on physician prescribing,49
whereas another described the profile of people who experienced excessive respiratory tract secretions at the end of life.51
These studies suggest that end-of-life care pathway clinical data could be used to determine feasibility and design larger prospective multicenter controlled trials.Benchmarking
The widespread uptake of the end-of-life care pathway has enabled benchmarking studies involving local,
52
national,53
, 54
or international16
, 21
comparisons to be undertaken. There may be scope for further larger benchmarking studies that would allow for more robust evaluation of the impact of end-of-life care pathways on care outcomes at the patient, health professional, and system level, across jurisdictions.Implementation Issues
Some of the key strengths of the end-of-life care pathway (Table 3) include a perception that it increased the accessibility of palliative care outside of specialist services or units
36
, 40
, 43
and promoted the adoption of best evidence-based end-of-life care regardless of care setting or competencies.12
, 36
, 40
, 43
, 45
The capacity of the pathways to act as a template to guide the delivery of care in the last hours or days of life, regardless of setting, appears to have propelled the pathways’ widespread adoption. The appointment of an experienced nurse as the pathway facilitator for 12–18 months,45
, 48
selecting acute care wards with a significant burden of death, and piloting the pathway in units with high-level clinical support18
, 36
also were singled out as important success factors.48
In the large U.K. benchmarking studies, hospitals with a pathway facilitator were noted to have higher pathway compliance levels.53
, 54
Table 3Strengths and Weaknesses of Using an End-of-Life Care Pathway in the Acute Care or Hospice Setting
| Strengths | Weaknesses |
|---|---|
|
|
Discussion
This integrative review did not identify any randomized controlled trials, and most studies provided low levels of evidence for the pathway’s effectiveness, despite its widespread adoption. Notwithstanding this limitation, the end-of-life care pathway has been used for a range of dying populations in a variety of acute and hospice care settings across much of the developed world. However, the lack of evidence from pragmatic clinical trials means that only low-level evidence is available to support the use of the end-of-life care pathway in the acute care and/or hospice settings.
It is currently estimated that only 50% of deaths in acute care can be accurately predicted.
4
, 21
, 44
The lack of universally accepted and validated criteria by which dying can be accurately predicted in different populations may impact on health professionals’ preparedness to initiate an end-of-life care pathway. Implementing an end-of-life care pathway is predicated on recognition of dying and multidisciplinary team agreement that there are no potentially reversible causes for the patient’s decline that need to be addressed.18
Alternatively, the ability to predict death also may reflect the fluctuating trajectories of people dying from organ failure and the slow dwindling trajectories associated with dementia and frailty, which make diagnosing dying in these populations a complex process and an inexact science.55
, 56
Although the end-of-life care pathway may be applicable in the final days of life, it does not preclude advance care planning early in the illness trajectory and may in fact demonstrate the significance of this issue in priming the patient, clinician, and health care system for reform.7
Several studies provide insights into strategies that supported the implementation of the pathway and drive practice change. Clinical education sessions, strong clinical leadership,
36
, 45
and the presence of a pathway facilitator45
were all identified as critical success factors, as well as health professionals having the necessary competencies to initiate and use the pathway.18
, 48
One implementation study tapped into a larger external physician education program on end-of-life care and provided education on a continuous cycle to manage junior staff rotations.18
All learning sessions were repeated frequently to increase clinicians’ competencies and assist with embedding change.18
Managing perceptions that the introduction of the pathway would be an additional documentation burden and workload for clinicians is an important aspect of the implementation process.48
The palliative care capabilities of the pathway facilitator appear to be central to ensuring that the dying patient’s transition onto the pathway is appropriately negotiated and safely managed. The pathway facilitator also plays a key role in building the palliative care capacity of health professionals. Despite this positive relationship, the number of U.K. pathway facilitators actually decreased over time, reflecting a trend to use facilitators for a defined period during the pathway implementation phase.
45
Yet, what is not known is whether appropriate pathway usage can be sustained over time, especially in the absence of a designated facilitator. This integrative review revealed that the settings effectively using the pathway had evidence of clinical leadership, access to a pathway facilitator, availability of a specialist palliative care team, a structured education program and high-level organizational support. These factors appear to be important considerations if the end-of-life care pathway is to be safely and effectively used by nonspecialist palliative care to provide best evidence-based care to dying people.There was limited acknowledgment in these studies of the need for ongoing evaluation of the pathway’s effectiveness. Few studies explicitly stressed the critical importance of integrating individual clinical expertise and critical decision making into end-of-life care delivery. Strengthening this linkage is critical if the end-of-life care pathway is to be safely and effectively implemented in the context and manner in which it was intended to be used. This is particularly important given the global focus on extending best evidence-based care to dying people, regardless of care setting.
57
Limitations
Only studies in English were accessible for this review, and issues relating to publication bias need to be considered. Furthermore, given the resource limitations of this review, no attempts were made to access research reported in the gray literature. These methodological limitations make it difficult to fully evaluate the impact of the end-of-life care pathway on care of the dying. However, the barriers to undertaking randomized controlled trials in palliative care are well noted,
58
and this does not mean that the lessons learned from these studies are not valuable in improving future care.Implications for Future Research
The challenge of undertaking end-of-life care research is well documented.
58
This underscores the importance of using pragmatic clinical trial considerations. To date, few studies have been able to accurately define imminent dying, describe the perceptions and experiences of the families and carers with the end-of-life care pathway, or explain the impact of care pathways on long-term bereavement outcomes. Greater understanding of the facilitator role also is required to gain insights into the degree to which this position bridges the gap between health professionals’ palliative care capabilities and their ability to effectively determine when the use of an end-of-life care pathway is indicated. Further research also is needed to determine whether safe and effective pathway usage can be sustained over time, especially in the absence of a designated facilitator. It also is important to consider the cultural aspects of death and, therefore, international comparisons of outcomes would be of particular interest.16
Because the use of end-of-life care pathways has become more widespread, several benchmarking studies have emerged; yet, the process for setting these benchmarks is unclear. Using systematic approaches for developing benchmarks, deriving consensus, and endorsement by policy and professional bodies is warranted.59
, 60
Conclusion
End-of-life care pathways have been widely implemented over recent decades in the absence of evidence from randomized controlled trials to support their use. In spite of this limitation, this review has identified some favorable outcomes in adopting this approach, as well as barriers to implementation. Furthermore, assessment of the impact of the end-of-life care pathways within pragmatic clinical trials is recommended.
Disclosures and Acknowledgments
No funding was received for this study, and the authors declare no conflicts of interest.
References
- Place of death: U.S. trends since 1980: fewer Americans died in the hospital in 1998 than in 1980, but some racial disparities raise troubling questions.Health Aff. 2004; 23: 194-200
- Where people die (1974-2030): past trends, future projections and implications for care.Palliat Med. 2008; 22: 33-41
- Goals of care toward the end of life: a structured literature review.Am J Hosp Palliat Med. 2009; 25: 501-511
- Diagnosing dying in the acute hospital setting—are we too late?.Clin Med. 2009; 9: 116-119
- Dying in an acute hospital setting: the challenges and solutions.Int J Clin Pract. 2009; 63: 508-515
- End-of-life decision making for cancer patients in different clinical settings and the impact of the LCP.J Pain Symptom Manage. 2010; 39: 33-43
- A healthier future for all Australians.Australian Government, Canberra, Australia2009
- Care of the dying: what a difference an LCP makes!.Palliat Med. 2007; 21: 365-368
- Developing an integrated care pathway for the dying patient.Eur J Palliat Care. 1997; 4: 203-207
- Care of the dying patient: the last hours or days of life.Br Med. 2003; 326: 30-34
- Integrated care pathways.Br Med J. 1998; 316: 133-137
- Care of the dying: setting standards for symptom control in the last 48 hours of life.J Pain Symptom Manage. 2001; 21: 12-17
- Benchmarking: a useful tool for informing and improving care of the dying?.Support Care Cancer. 2008; 16: 813-819
- Improving end-of-life care: a critical review of the gold standards framework in primary care.Palliat Med. 2010; 24: 317-329
- Symptom management for the adult patient dying with advanced chronic kidney disease: a review of the literature and development of evidence-based guidelines by a United Kingdom Expert Consensus Group.Palliat Med. 2009; 23: 103-110
- The implementation of an end-of-life integrated care pathway in a Chinese population.Int J Palliat Nurs. 2009; 15: 384-388
- Dutch experiences with the Liverpool Care Pathway.Eur J Palliat Care. 2006; 13: 156-159
- Improving processes of hospital care during the last hours of life.Arch Intern Med. 2005; 165: 1722-1727
- Improving end-of-life care: development and pilot-test of a clinical pathway.J Pain Symptom Manage. 2005; 29: 529-543
- Pilot of a pathway to improve the care of imminently dying oncology inpatients in a Veterans Affairs Medical Center.J Pain Symptom Manage. 2005; 29: 544-551
- Audit of the Liverpool Care Pathway for the Dying Patient in a Dutch cancer hospital.J Palliat Care. 2006; 22: 305-308
- The development and implementation of the pathway for improving the care of the dying in general medical wards.Intern Med J. 2009; 39: 695-699
- Palliative care in old age.Rev Clin Gerontol. 2001; 11: 149-157
- End-of-life care pathways.Curr Opin Support Palliat Care. 2007; 1: 198-201
- The Liverpool Care Pathway: its impact on improving the care of the dying.Age Ageing. 2005; 34 ([letter].) (author reply 198–199.): 197-199
- Continuous deep sedation: Dutch research reflects problems with the Liverpool care pathway.BMJ. 2008; 336: 905
- End-of-life care pathways for improving outcomes in caring for the dying.Cochrane Database Syst Rev. 2010; (CD008006)
- Clinical trials and the development of laparoscopic surgery.Surg Endosc. 2001; 15: 1-3
- The integrative review: updated methodology.J Adv Nurs. 2005; 52: 546-553
- Integrative reviews of nursing research.Res Nurs Health. 1987; 10: 1-11
- Guidelines for a palliative approach in residential aged care—Enhanced version.NHMRC, Canberra, Australia2006
National Health and Medical Research Council. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra, Australia: Commonwealth of Australia. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/publications/synopses/cp30.pdf. Accessed February 3, 2011.
- Utilization-focused integrative reviews in nursing service.Appl Nurs Res. 1998; 11: 195-206
- NHMRC additional levels of evidence and grades for recommendations for developers of guidelines: Stage 2 consultation.NHMRC, Canberra, Australia2009
- A care pathway for the end of life in a renal setting.J Ren Care. 2006; 32: 160-163
- Introduction of the Liverpool Care Pathway for end of life care to emergency medicine.Emerg Med J. 2009; 26: 777-779
- Nurses’ perceptions of the Liverpool Care Pathway for the dying patient in the acute hospital setting.Int J Palliat Nurs. 2003; 9: 375-381
- The Liverpool Care Pathway in hospices: an exploratory study of doctor and nurse perceptions.Int J Palliat Nurs. 2006; 12: 414-421
- The evaluation of an end-of-life integrated care pathway.Int J Palliat Nurs. 2007; 13: 74-81
- The effect of the Liverpool Care Pathway for the dying: a multicenter study.Palliat Med. 2008; 22: 145-151
- Liverpool Care Pathway carers survey.Palliat Med. 2009; 23: 571-572
- Using the LCP: bereaved relatives’ assessments of communication and bereavement.Am J Hosp Palliat Med. 2008; 25: 207-214
- Towards a good death: the impact of the care of the dying pathway in an acute stroke unit.Age Ageing. 2004; 33: 625-626
- The development of an Integrated Care Pathway for all patients with advanced life-limiting illness—the Supportive Care Pathway.J Nurs Manag. 2006; 14: 521-528
- Introducing an integrated care pathway for the last days of life.Palliat Med. 2005; 19: 33-39
- Audit of the care of the dying in a network of hospitals and institutions in Queensland.Intern Med J. 2007; 37: 315-319
- Dying for attention: palliative care in the acute setting.Aust J Adv Nurs. 2006; 24: 21-25
- Role of a clinical facilitator in introducing an integrated care pathway for the care of the dying.Int J Palliat Nurs. 2004; 10: 497-501
- Drugs at the end of life: does an integrated care pathway simplify prescribing?.J Integr Care Pathw. 2005; 9: 78-80
- Respiratory tract secretions in the dying patient: a comparison between glycopyrronium and hyoscine hydrobromide.J Palliat Med. 2006; 9: 279-284
- Respiratory tract secretions in the dying patient: a retrospective study.J Pain Symptom Manage. 2003; 26: 897-902
- An integrated care pathway for the last two days of life: Wales-wide benchmarking in palliative care.Int J Palliat Nurs. 2002; 8: 566-573
- National care of the dying audit—hospitals: Generic report 2008/2009.The Marie Curie Palliative Care Institute Liverpool, Liverpool, UK2009
- National care of the dying audit—hospitals (NCDAH)—Summary report.The Marie Curie Palliative Care Institute Liverpool, Liverpool, UK2007
- Living well at the end of life: Adapting health care to serious chronic illness in old age.RAND Health, Santa Monica, CA2003
- Illness trajectories and palliative care.Br Med J. 2005; 330: 1007-1011
- The gold standards framework for care homes.(Available from) (Accessed February 3, 2011)
- Challenging the framework for evidence in palliative care research.Palliat Med. 2005; 19: 461-465