Skip to main content

Table 6 Integrated palliative care

From: A scoping review of initiatives to reduce inappropriate or non-beneficial hospital admissions and bed days in people nearing the end of their life: much innovation, but limited supporting evidence

Author, date and journal Study design Sample and setting Research focus Relevant results Wider implications Limitations
Ferrell et al.; 2015; Journal of Pain and Symptom Management [60] Prospective, quasi-experimental. 544 American patients with Stage I–IV non-small cell lung cancer were enrolled in the study. The intervention group (n = 272) were reviewed at integrated care meetings. The control group received standard care (n = 219). To test the effectiveness of an integrated palliative programme for cancer patients. Outcome measures included QoL, symptoms, distress and hospital utilisation. There were no statistically significant differences in unscheduled hospital admissions between the groups in the last 2 weeks of life. Integrated palliative may not reduce hospital admissions The sequential design may have resulted in temporal bias.
This design did not allow for establishing the exact element of the intervention that was successful.
Benthien et al.; 2018; Journal of Pain and Symptom Management [61] Randomised controlled trial with a 1:1 ratio. 340 Danish in- and out- cancer patients were assigned to either an intervention or control group. The intervention group received standard care plus specialised palliative care enriched with a standardised psychological intervention for patients and caregivers at home, and the control group received standard care only. To investigate whether a transition process from oncological treatment to specialist palliative care at home for patients with incurable cancer results in more patients reaching their preferred place of care and death. The intervention group had more hospital admissions due to health deterioration (22% vs. 16%, p = 0.04) or an unmanageable home situation (8% vs. 4%, p = 0.0119). Hospital admissions were most often caused by symptoms without progression in the intervention group (11% vs. 7%, p = 0.0493). There was no significant difference in overall potentially avoidable admissions. Both groups felt generally safe regarding their place of care. This type of intervention may not decrease hospital admissions in incurable cancer patients. It may be more successful for those with a slower declining rate of health and with better diagnostics at home to rule out emergencies in need of hospital admission. Two-thirds of the control group received specialist palliative care, albeit considerably later than the intervention group.
Lack of a consistent definition of potentially avoidable admissions.
Blackhall et al.; 2016; Journal of Palliative Medicine [62] Retrospective electronic records review. 207 US patients with advanced cancer were referred to the Comprehensive Assessment with Rapid Evaluation and Treatment (CARE) Track palliative care intervention was compared with 198 deceased patients with similar diagnoses who were not. To measure the time of referral to outpatient palliative and impact on end-of-life care for patients with cancer. CARE Track patients had fewer end of life hospitalisations, were less likely to die in hospital and had increased hospice utilisation and decreased costs of care. Referral to outpatient palliative care within 3 months of death improved end of life care and may reduce costs. However, many patients were not referred, and systematic methods of referrals are needed. Non-randomised and single centre with a predominantly white population.
Selection bias.
Significant differences between control and intervention groups such as age and gender.
Jang et al.; 2015; Journal of the National Cancer Institute [63] Retrospective cohort study. 5381 patients who died of advanced pancreatic cancer between January 2005 and March 2011 in Ontario, Canada. Approximately half had received a palliative care consultation. To evaluate the impact of palliative care, including intensity, on the aggressiveness of care near death in patients with advanced pancreatic cancer. Patients who received a palliative care consultation had lower frequencies of intensive care unit visits near death (1.1% vs. 7.8%), multiple emergency department visits (7.4% vs. 28.5%) and multiple hospitalisations (3.8% vs. 12.8%). Palliative care is associated with less aggressive care near death. Selection bias and lack of control of confounding variables due to observational design.
Morita et al.; 2013; Lancet Oncology [64] Mixed-methods study. A multifaceted programme of interventions for patients with cancer was introduced in 4 regions of Japan. Place of death was compared before (5146 decedents) and after the interventions (5546 decedents). To assess the impact of these interventions on numbers of home deaths, coverage of specialist services and patient-reported and family-reported quality of care. Proportions of home deaths increased significantly, from 6.76% before the intervention compared with 10.48% after intervention. A regional programme of interventions could improve the quality of palliative care. The data used may not be fully representative of the region.
Youens et al.; 2017; Journal of Palliative Medicine [65] Retrospective, observational cohort study. 28,561 West Australian cancer decedents from 2001 to 2011, of which 16,530 accessed a community-based palliative care intervention. To compare place of death and acute care hospital use in the last year of life between cancer decedents who did and did not access a community-based palliative service. Intervention users had 3 times greater odds of dying outside of hospital than non-users.
Intervention users had fewer unplanned hospital admissions and emergency department admissions in the last year/last week of life.
Those who accessed the service had significantly shorter hospital stays than the control group.
The palliative care community service supported people to die outside of hospital and was associated with reduced acute care admissions, bed days and costs over the last year of life. Limited ability to establish causation due to it being an observational study.
Lack of data on symptom severity and functional status, which may have differed between groups.
May et al.; 2017; Palliative Medicine [66] Prospective cohort study. 863 adults with advanced cancer admitted to 3 US hospitals. Usual care (n = 637), early palliative care (n = 177) and late palliative care (n = 49) were compared. To establish the association between early palliative care consultation team intervention and 1) intensity of service and length of stay compared with usual care; and 2) day-to-day costs compared with a later intervention. Early palliative care patients had shorter hospital stays than late palliative care patients (mean length of stay 6.7 vs. 13.6 days). Early palliative care patients had shorter stays than usual care patients (mean length of stay 6.7 vs. 7.8 days). Reducing the length of stay is the biggest cost saver in early consultation for patients with advanced cancer. Lack of control over confounding variables due to its observational design.
Rogers et al.; 2017; Journal of the American College of Cardiology [67] Randomised controlled trial. 150 patients with advanced heart failure received standard care (n = 75) or standard care plus a palliative care intervention (n = 75) at a single centre. To investigate whether an integrated palliative care intervention improves certain outcomes in heart failure care. Randomisation to the intervention did not affect hospitalisation or mortality.   The single centre and intervention care was implemented by a single nurse. The control group may have received palliative care not representative of standard care.
Brannstrom et al.; 2014; European Journal of Heart Failure [68] Randomised controlled study. 36 Swedish patients with chronic heart failure were randomised to PREFER and compared with 36 patients in a control group. To compare the outcomes of PREFER with regard to patient symptoms, quality of life and hospitalisations with those of usual care. The mean number of hospitalisations was significantly lower in the PREFER group (0.42 vs. 1.47). The total number of hospitalisations was lower in the PREFER group (15 vs. 53). The mean number of days spent in hospital was significantly lower in the PREFER group than in the control group (2.9 vs. 12.4). Person-centred care combined with active heart failure and palliative care at home has the potential to improve quality of life and morbidity substantially in patients with severe chronic heart failure. Intervention participants were significantly older than control group participants.
The study was small and at a single centre.
Chan et al.; 2014; Journal of Pain and Symptom Management [69] Prospective, longitudinal, observational study. 19 patients with end-stage renal disease who attended a renal palliative care clinic and had more than 1 emergency department visit in 3 months. Follow-up was more frequent and intensified. To assess whether intensified and more frequent follow-up visits affected the rate of renal palliative care clinic attendance, emergency department attendance and hospital admission. The rate of emergency department attendance (2.63 vs. 0.63) and acute hospital admission (1.59 vs. 0.58) was significantly reduced after intensified follow-up. Clinic attendance rates improved from 56 to 85%. Intensifying renal palliative care follow-up minimised the utilisation of acute medical services and improved outpatient attendance at the renal palliative care clinic. Small sample, selection bias.
Data originally collected for a different purpose.
Horton et al.; 2013; Journal of Palliative Medicine [70] Single-centre cohort, longitudinal, observational study 30 patients with advanced chronic obstructive pulmonary disease and 18 caregivers were followed in their home for 6 months whilst participating in an education programme followed by home-based palliative care. To determine the feasibility of 1) implementing a customised home-based palliative care service for patients and caregivers living with advanced chronic obstructive pulmonary disease; and 2) measuring outcomes of providing such services. 25 patients and 14 caregivers enrolled in the home-based palliative care after completing the education programme. 12/17 patients who chose home as their preferred place of death. None of the deaths occurred at home. A drop in hospital utilisation was observed in the first 100 patients before vs. after enrolment. Providing home-based palliative care for patients with advanced chronic obstructive pulmonary disease is feasible but requires further study to fulfil their place-of-death preference. Lack of standardisation of the intervention across participants and selection bias.
Hussain et al.; 2013; International Journal of Palliative Nursing [71] Retrospective study. Participation in a nurse-led palliative neurology service in the North of England was offered to 62 patients. Outcomes were compared with standard National End of Life Care Programme care To assess the key outcomes of a UK nurse-led palliative neurology service against standard National End of Life Care Programme standards. Mean hospital admissions in the intervention group were 0.9 vs. 3.5 nationally across all diagnoses. 26% of patients receiving the intervention died in hospital vs. 46% nationally. The service model fulfilled key standard National End of Life Care Programme recommendations and resulted in low hospital admissions and deaths. Small sample.
No control group.
Wu et al.; 2013; Journal of Palliative Medicine [72] Retrospective cohort study. 50 US patients who received a palliative care consultation in the emergency department before hospital admission were compared with 1385 patients who received a palliative care consultation after admission to hospital. To see whether inpatient admissions after palliative care initiated in the emergency department were associated with shorter length of stay than in patients whose palliative care was initiated after hospital admission. Length of stay was reduced by 3.6 in patients who received palliative care consultation prior to admission. Early initiation of palliative care in the emergency department resulted in shorter inpatient stays. Non-randomised and selection bias.
Desrosiers et al.; 2014; Journal of Pain and Symptom Management [73] Retrospective cohort study. 56 consecutive deaths under a new service at a hospital in Cape Town, South Africa, were compared with 48 consecutive deaths immediately before implementation of the new service. To assess whether a hospital-based palliative care service reduces admissions and increases home death rates. The mean number of admissions for the intervention group and control groups were 1.39 vs. 1.98. The mean total number of hospital days for the intervention and control groups were 4.52 vs. 9.3. 58.9% of the intervention group died at home, compared with 18.8% of the control group. An outpatient hospital-based service reduced admissions and improved the chance of home death, offering a feasible and cost-effective model for such settings. Selection bias.
  1. Increasing skills and knowledge of palliative care, increasing availability of palliative care services, coordinating palliative services and providing appropriate information to families and the