588007 research-article2015
PMJ0010.1177/0269216315588007Palliative MedicineBrännström et al.
Original Article
Effectiveness of the Liverpool care pathway for the dying in residential care homes: An exploratory, controlled before-and-after study
Palliative Medicine 1–10 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269216315588007 pmj.sagepub.com
Margareta Brännström1, Carl Johan Fürst2, Carol Tishelman3, Max Petzold4 and Olav Lindqvist3,5
Abstract Background: Clinical pathways aim to ensure that individuals receive appropriate evidence-based care and interventions, with the Liverpool Care Pathway for the Dying Patient focusing on end of life. However, controlled studies of the Liverpool Care Pathway for the Dying Patient, particularly outside of cancer settings, are lacking. Aim: To compare the effects of the Liverpool Care Pathway for the Dying Patient and usual care on patients’ symptom distress and well-being during the last days of life, in residential care homes. Design: Exploratory, controlled before-and-after study. During a 15-month baseline, usual care was carried out in two areas. During the following 15-months, usual care continued in the control area, while residential care home staff implemented Liverpool Care Pathway for the Dying Patient use in the intervention area. The intervention was evaluated by family members completing retrospective symptom assessments after the patient’s death, using the Edmonton Symptom Assessment System and Views of Informal Carers – Evaluation of Services. Settings/participants: Patients who died at all 19 residential care homes in one municipality in Sweden. Results: Shortness of breath (estimate = −2.46; 95% confidence interval = −4.43 to −0.49) and nausea (estimate = −1.83; 95% confidence interval = −3.12 to −0.54) were significantly reduced in Edmonton Symptom Assessment System in patients in the intervention compared to the control area. A statistically significant improvement in shortness of breath was also found on the Views of Informal Carers – Evaluation of Services item (estimate = −0.47; 95% confidence interval = −0.85 to −0.08). Conclusion: When implemented with adequate staff training and support, the Liverpool Care Pathway for the Dying Patient may be a useful tool for providing end-of-life care of elderly people at the end of life in non-cancer settings.
Keywords Clinical pathway, end-of-life care, homes for the aged
What is already known about this topic? •• Symptom alleviation is a fundamental component in high-quality end-of-life care. •• Despite increased clinical use, systematic reviews show that empirical evidence is lacking regarding the effects of end-of-life care pathways. •• End-of-life care pathways are not always implemented adequately.
1Department
of Nursing, Umeå University Campus Skellefteå, Umeå, Sweden 2The Institute for Palliative Care, Lund University and Region Skåne, Lund, Sweden 3Medical Management Centre (MMC), Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm, Sweden
4Akademistatistik
– Centre for Applied Biostatistics, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden 5Department of Nursing, Umeå University, Umeå, Sweden Corresponding author: Margareta Brännström, Department of Nursing, Umeå University Campus Skellefteå, S-93187 Umeå, Sweden. Email: [email protected]
Downloaded from pmj.sagepub.com at UNIV OF LETHBRIDGE on October 22, 2015
2
Palliative Medicine
What this paper adds? •• This exploratory study suggests that the use of the Liverpool Care Pathway for the Dying Patient (LCP) has potential to improve several aspects of the quality of end-of-life care for dying elderly people at residential care homes. Implications for practice, theory or policy •• The experiences gained during the process of conducting this exploratory study led to a policy decision to implement the LCP in end-of-life care in residential care homes in the study setting. •• When implemented with adequate staff training and support, the LCP may be a useful tool for providing end-of-life care of elderly people at the end of life in non-cancer settings.
Introduction The plea for palliative care as a human right is unfortunately based on the suffering and distress of all-too-many dying patients and their families.1 Despite the increasing evidence base for effective interventions to relieve pain and other symptoms at the end of life (EOL), the implementation of best practice care is still a challenge in most clinical settings. Integrated care pathways have been increasingly developed and used2 as one means to achieve high-quality EOL care. The use of integrated care pathways based on best practice guidelines aims to ensure that the individual patient receives appropriate evidence-based care and interventions. The Liverpool Care Pathway for the Dying Patient (LCP) is one such structured tool for assessment and continual evaluation of the needs of the dying patient and family, aiming to transfer best practice care derived from hospice to other care settings. It includes strategies and decision-making support for identifying imminent death, providing pharmacological and non-pharmacological symptom control and support for communication with patients and families.3 However, there is a scarcity of research on the effect of clinical pathways in general and of EOL care pathways specifically.4 The LCP has been the subject of recent debate and evaluation,5 particularly after the Neuberger report from the UK Department of Health.6 Although the report appreciates the positive effects of using LCP as it is intended, the authors criticize what they cite as inappropriate implementation of LCP, with many of the cited cases of malpractice in older non-cancer patients. Therefore, they recommend that LCP use be phased out in the United Kingdom. The report also emphasizes that robust scientific evidence for the effectiveness of LCP use is still lacking.6 Uncontrolled before-and-after studies7,8 and qualitative research9–13 support the value of using LCP in some types of services providing care for dying patients. Although much of this research focuses on patients with cancer, one exception is a recent qualitative mixed methods study of staff perceptions which found that LCP was described as supporting, guiding and promoting care delivery in residential care for the elderly in New Zealand.14 However, a
recent cluster randomized trial did not show a significant difference in the primary outcome overall quality of care for patients with cancer in hospitals. A significant improvement for the secondary outcomes of respect, kindness and dignity in care provision was observed, and an improvement in control of breathlessness was also noted.15 To the best of our knowledge, most studies on LCP to date have been carried out in hospital and hospice settings, with a lack of clarity as to whether findings can be generalized to other EOL care settings. In summary, the LCP research to date shows a need for more robustly designed (i.e. randomized and controlled before and after) studies in noncancer and non-hospital settings (i.e. residential care homes (RCHs)).15 The aim of this exploratory study was to begin to address these knowledge gaps, by comparing the effects of LCP and standard care on symptom distress and well-being during the last days of life in RCHs.
Methods Study design This exploratory controlled before-and-after study included all 19 RCHs with approximately 1000 residents in one Swedish municipality. All elderly people (called ‘patients’ hereafter for simplicity) who died at an RCH between 1 June 2009 and 31 October 2011 were included in the study population.
Context The municipality where the study was conducted is situated in the north of Sweden. The total population of approximately 72,000 inhabitants is spread over 6800 km2 in both urban and rural areas.16 Patients in RCHs are generally older with dementia diseases, multiple diseases and/or extensive disabilities. RCH staff consists primarily of assistant nurses with most responsibility for hands-on care and registered nurses (RNs) on a consultative basis, both employed by the municipality. General practitioners (GPs) are employed by the county council as consultants
Downloaded from pmj.sagepub.com at UNIV OF LETHBRIDGE on October 22, 2015
3
Brännström et al. to assess patient’s status, make diagnoses and prescribe medications.
Intervention All RCHs were organized under one of two equally large administrative areas (see Figure 1); one area included 10 RCHs with a total of 511 beds, and the other area had 9 RCHs with 506 beds. To avoid contamination between areas, one administrative area was randomly allocated to the LCP intervention and the other to control. During a 15-month baseline period (1 June 2009 to 31 August 2010), usual care (UC) was carried out in both areas. In the following 14-month period (1 September 2010 to 31 October 2011), UC continued in the control area (CA), while RCH staff introduced and used the LCP according to its guidelines in the intervention area (IA). This study was conducted during a period of ongoing development to improve EOL care in the municipality. This included the introduction of the Swedish Palliative Registry17 and a 2 × 3.5-h course in EOL care for all staff working in RCHs in the municipality during baseline in both areas, unrelated to this study. One contact RN was appointed at each RCH in both areas. The contact nurse was responsible for registering all patients who died in research protocols; this included information about the family member(s) usually present during the last days of life. The project principal investigator (PI; first author M.B.), the chief nurse for the municipality and the RN responsible for care development held hour-long meetings every third month with the contact nurses to reflect on issues about EOL care; these were held separately for the IA and CA. The Swedish LCP version 11 was introduced in the IA in collaboration with the Swedish LCP coordination centre according to standard criteria at the end of the baseline period. This included a 3-h session about the LCP and EOL care for all assistant nurses, RNs and GPs working at the RCHs in the IA. The IA contact nurses participated in an approximately 35-h web-based LCP train-the-trainer course during March–April 2010. Each contact nurse then taught staff at their respective workplace and acted as a resource person for LCP implementation.
Evaluation and outcomes To evaluate the intervention, a set of questionnaires was sent to a family member who had been in close contact with the dying patient, as determined from research protocol data, 1 month after the death of the patient. No reminder was sent out. Inclusion of family members to the evaluation ended somewhat earlier than planned, on 31 October 2011, due to organizational changes involving overlap of administrative areas which could increase risk of contamination.
The individual items evaluating symptom distress and well-being of patients, as rated by family members after the death of the patient and measured by the Edmonton Symptom Assessment System (ESAS), comprise the primary endpoints. In this study, we also analysed the symptoms pain and breathlessness during the last 3 days of life using the Views of Informal Carers – Evaluation of Services (VOICES) questionnaire. The ESAS was used to measure symptom burden during the last 3 days in life, assessing nine commonly experienced symptoms, that is, pain, tiredness, nausea, depression, anxiety, drowsiness, loss of appetite, deteriorated well-being and shortness of breath. A further item allows assessment of ‘Any other problem’. The severity of each symptom at the time of assessment is rated on a 10-step numerical scale, with higher scores indicating greater symptom intensity.18,19 In the analyses presented here, the symptoms pain and breathlessness during the last 3 days of life were further assessed using the VOICES questionnaire. VOICES is intended to retrospectively assess the quality of palliative and EOL care from the perspective of the bereaved person.20 Medical records and the National Board of Health and Welfare statistics on causes of death in Sweden were used to collect data about background characteristics. The cause of death for the deceased is classified according to International Statistical Classification of Diseases and Related Health Problems–Tenth Revision (ICD-10).21
Statistical analysis Descriptive statistics and linear regression analysis have been used. Comparisons between the IA and CA were made using the chi-square test for proportions and Student’s t-test for continuous variables. In the regression analyses, the unit of analysis was the RCH for which proportions or mean values were calculated and then used for comparison. The effect of the intervention was extracted as an interaction effect of being an intervention unit and post-intervention, after excluding patients in the IA who were not on LCP during the intervention period. This corresponds to calculating the difference between the changes from baseline to follow-up for intervention units and control units, respectively. A p-value
PMJ0010.1177/0269216315588007Palliative MedicineBrännström et al.
Original Article
Effectiveness of the Liverpool care pathway for the dying in residential care homes: An exploratory, controlled before-and-after study
Palliative Medicine 1–10 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269216315588007 pmj.sagepub.com
Margareta Brännström1, Carl Johan Fürst2, Carol Tishelman3, Max Petzold4 and Olav Lindqvist3,5
Abstract Background: Clinical pathways aim to ensure that individuals receive appropriate evidence-based care and interventions, with the Liverpool Care Pathway for the Dying Patient focusing on end of life. However, controlled studies of the Liverpool Care Pathway for the Dying Patient, particularly outside of cancer settings, are lacking. Aim: To compare the effects of the Liverpool Care Pathway for the Dying Patient and usual care on patients’ symptom distress and well-being during the last days of life, in residential care homes. Design: Exploratory, controlled before-and-after study. During a 15-month baseline, usual care was carried out in two areas. During the following 15-months, usual care continued in the control area, while residential care home staff implemented Liverpool Care Pathway for the Dying Patient use in the intervention area. The intervention was evaluated by family members completing retrospective symptom assessments after the patient’s death, using the Edmonton Symptom Assessment System and Views of Informal Carers – Evaluation of Services. Settings/participants: Patients who died at all 19 residential care homes in one municipality in Sweden. Results: Shortness of breath (estimate = −2.46; 95% confidence interval = −4.43 to −0.49) and nausea (estimate = −1.83; 95% confidence interval = −3.12 to −0.54) were significantly reduced in Edmonton Symptom Assessment System in patients in the intervention compared to the control area. A statistically significant improvement in shortness of breath was also found on the Views of Informal Carers – Evaluation of Services item (estimate = −0.47; 95% confidence interval = −0.85 to −0.08). Conclusion: When implemented with adequate staff training and support, the Liverpool Care Pathway for the Dying Patient may be a useful tool for providing end-of-life care of elderly people at the end of life in non-cancer settings.
Keywords Clinical pathway, end-of-life care, homes for the aged
What is already known about this topic? •• Symptom alleviation is a fundamental component in high-quality end-of-life care. •• Despite increased clinical use, systematic reviews show that empirical evidence is lacking regarding the effects of end-of-life care pathways. •• End-of-life care pathways are not always implemented adequately.
1Department
of Nursing, Umeå University Campus Skellefteå, Umeå, Sweden 2The Institute for Palliative Care, Lund University and Region Skåne, Lund, Sweden 3Medical Management Centre (MMC), Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm, Sweden
4Akademistatistik
– Centre for Applied Biostatistics, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden 5Department of Nursing, Umeå University, Umeå, Sweden Corresponding author: Margareta Brännström, Department of Nursing, Umeå University Campus Skellefteå, S-93187 Umeå, Sweden. Email: [email protected]
Downloaded from pmj.sagepub.com at UNIV OF LETHBRIDGE on October 22, 2015
2
Palliative Medicine
What this paper adds? •• This exploratory study suggests that the use of the Liverpool Care Pathway for the Dying Patient (LCP) has potential to improve several aspects of the quality of end-of-life care for dying elderly people at residential care homes. Implications for practice, theory or policy •• The experiences gained during the process of conducting this exploratory study led to a policy decision to implement the LCP in end-of-life care in residential care homes in the study setting. •• When implemented with adequate staff training and support, the LCP may be a useful tool for providing end-of-life care of elderly people at the end of life in non-cancer settings.
Introduction The plea for palliative care as a human right is unfortunately based on the suffering and distress of all-too-many dying patients and their families.1 Despite the increasing evidence base for effective interventions to relieve pain and other symptoms at the end of life (EOL), the implementation of best practice care is still a challenge in most clinical settings. Integrated care pathways have been increasingly developed and used2 as one means to achieve high-quality EOL care. The use of integrated care pathways based on best practice guidelines aims to ensure that the individual patient receives appropriate evidence-based care and interventions. The Liverpool Care Pathway for the Dying Patient (LCP) is one such structured tool for assessment and continual evaluation of the needs of the dying patient and family, aiming to transfer best practice care derived from hospice to other care settings. It includes strategies and decision-making support for identifying imminent death, providing pharmacological and non-pharmacological symptom control and support for communication with patients and families.3 However, there is a scarcity of research on the effect of clinical pathways in general and of EOL care pathways specifically.4 The LCP has been the subject of recent debate and evaluation,5 particularly after the Neuberger report from the UK Department of Health.6 Although the report appreciates the positive effects of using LCP as it is intended, the authors criticize what they cite as inappropriate implementation of LCP, with many of the cited cases of malpractice in older non-cancer patients. Therefore, they recommend that LCP use be phased out in the United Kingdom. The report also emphasizes that robust scientific evidence for the effectiveness of LCP use is still lacking.6 Uncontrolled before-and-after studies7,8 and qualitative research9–13 support the value of using LCP in some types of services providing care for dying patients. Although much of this research focuses on patients with cancer, one exception is a recent qualitative mixed methods study of staff perceptions which found that LCP was described as supporting, guiding and promoting care delivery in residential care for the elderly in New Zealand.14 However, a
recent cluster randomized trial did not show a significant difference in the primary outcome overall quality of care for patients with cancer in hospitals. A significant improvement for the secondary outcomes of respect, kindness and dignity in care provision was observed, and an improvement in control of breathlessness was also noted.15 To the best of our knowledge, most studies on LCP to date have been carried out in hospital and hospice settings, with a lack of clarity as to whether findings can be generalized to other EOL care settings. In summary, the LCP research to date shows a need for more robustly designed (i.e. randomized and controlled before and after) studies in noncancer and non-hospital settings (i.e. residential care homes (RCHs)).15 The aim of this exploratory study was to begin to address these knowledge gaps, by comparing the effects of LCP and standard care on symptom distress and well-being during the last days of life in RCHs.
Methods Study design This exploratory controlled before-and-after study included all 19 RCHs with approximately 1000 residents in one Swedish municipality. All elderly people (called ‘patients’ hereafter for simplicity) who died at an RCH between 1 June 2009 and 31 October 2011 were included in the study population.
Context The municipality where the study was conducted is situated in the north of Sweden. The total population of approximately 72,000 inhabitants is spread over 6800 km2 in both urban and rural areas.16 Patients in RCHs are generally older with dementia diseases, multiple diseases and/or extensive disabilities. RCH staff consists primarily of assistant nurses with most responsibility for hands-on care and registered nurses (RNs) on a consultative basis, both employed by the municipality. General practitioners (GPs) are employed by the county council as consultants
Downloaded from pmj.sagepub.com at UNIV OF LETHBRIDGE on October 22, 2015
3
Brännström et al. to assess patient’s status, make diagnoses and prescribe medications.
Intervention All RCHs were organized under one of two equally large administrative areas (see Figure 1); one area included 10 RCHs with a total of 511 beds, and the other area had 9 RCHs with 506 beds. To avoid contamination between areas, one administrative area was randomly allocated to the LCP intervention and the other to control. During a 15-month baseline period (1 June 2009 to 31 August 2010), usual care (UC) was carried out in both areas. In the following 14-month period (1 September 2010 to 31 October 2011), UC continued in the control area (CA), while RCH staff introduced and used the LCP according to its guidelines in the intervention area (IA). This study was conducted during a period of ongoing development to improve EOL care in the municipality. This included the introduction of the Swedish Palliative Registry17 and a 2 × 3.5-h course in EOL care for all staff working in RCHs in the municipality during baseline in both areas, unrelated to this study. One contact RN was appointed at each RCH in both areas. The contact nurse was responsible for registering all patients who died in research protocols; this included information about the family member(s) usually present during the last days of life. The project principal investigator (PI; first author M.B.), the chief nurse for the municipality and the RN responsible for care development held hour-long meetings every third month with the contact nurses to reflect on issues about EOL care; these were held separately for the IA and CA. The Swedish LCP version 11 was introduced in the IA in collaboration with the Swedish LCP coordination centre according to standard criteria at the end of the baseline period. This included a 3-h session about the LCP and EOL care for all assistant nurses, RNs and GPs working at the RCHs in the IA. The IA contact nurses participated in an approximately 35-h web-based LCP train-the-trainer course during March–April 2010. Each contact nurse then taught staff at their respective workplace and acted as a resource person for LCP implementation.
Evaluation and outcomes To evaluate the intervention, a set of questionnaires was sent to a family member who had been in close contact with the dying patient, as determined from research protocol data, 1 month after the death of the patient. No reminder was sent out. Inclusion of family members to the evaluation ended somewhat earlier than planned, on 31 October 2011, due to organizational changes involving overlap of administrative areas which could increase risk of contamination.
The individual items evaluating symptom distress and well-being of patients, as rated by family members after the death of the patient and measured by the Edmonton Symptom Assessment System (ESAS), comprise the primary endpoints. In this study, we also analysed the symptoms pain and breathlessness during the last 3 days of life using the Views of Informal Carers – Evaluation of Services (VOICES) questionnaire. The ESAS was used to measure symptom burden during the last 3 days in life, assessing nine commonly experienced symptoms, that is, pain, tiredness, nausea, depression, anxiety, drowsiness, loss of appetite, deteriorated well-being and shortness of breath. A further item allows assessment of ‘Any other problem’. The severity of each symptom at the time of assessment is rated on a 10-step numerical scale, with higher scores indicating greater symptom intensity.18,19 In the analyses presented here, the symptoms pain and breathlessness during the last 3 days of life were further assessed using the VOICES questionnaire. VOICES is intended to retrospectively assess the quality of palliative and EOL care from the perspective of the bereaved person.20 Medical records and the National Board of Health and Welfare statistics on causes of death in Sweden were used to collect data about background characteristics. The cause of death for the deceased is classified according to International Statistical Classification of Diseases and Related Health Problems–Tenth Revision (ICD-10).21
Statistical analysis Descriptive statistics and linear regression analysis have been used. Comparisons between the IA and CA were made using the chi-square test for proportions and Student’s t-test for continuous variables. In the regression analyses, the unit of analysis was the RCH for which proportions or mean values were calculated and then used for comparison. The effect of the intervention was extracted as an interaction effect of being an intervention unit and post-intervention, after excluding patients in the IA who were not on LCP during the intervention period. This corresponds to calculating the difference between the changes from baseline to follow-up for intervention units and control units, respectively. A p-value
