Table 2 Characteristics of the included studies according to different disease categories

 Jordhøy et al. 2001, (Norway) [20]

Cluster Randomised Trial

33

Collaboration between palliative medicine unit and community service

HRQL*: physical, emotional functioning, pain, psychological distress. Place of death, hospital utilisation.

There was no evidence of any impact on the patients’ HRQL*. There was no tendency in favour of any treatment group on the main outcomes in assessments that were made within 3 months before death.

 Smeenk et al. 1998, (The Netherlands) [21]

Quasi-experimental study

31

Transmural home care programme: collaboration primary care team and supporting hospital care team.

Re-hospitalization, QoL*, home death

Patients in the intervention group underwent significantly less re-hospitalization during the terminal phase of their illness; the intervention contributed significantly positive to the patients physical QoL; A higher, not significant percentage died at home.

 Colombet et al. 2012, (France) [22]

Case series study nested in a cohort

30

Impact of oncologist’s awareness of PC, clinical intervention of PCT* and timing, multidisciplinary decision-making.

Indicators: location of death, number of ER* visits in last month of life, chemotherapy administration in last 14 days of life.

58 patients died at home, 45 in an ICU or ER, and 253 in an acute care hospital; 185 patients visited the ER* in last month of life and 75 received chemotherapy in last 14 days of life. OPM* independently decreases the odds of receiving chemotherapy in last 14 days of life. Early PCT* intervention had no impact on indicators, whereas OPM* reduced the odds of persistent chemotherapy in the last 14 days of life. Decision-making with oncologists and the PC* team is the most critical parameter for improving EoLC*.

 Schreml et al. 2000, (Germany) [23]

Observational study

25

Integration of PC into regular internal ward services in general hospital.

Release of pain, respiratory distress, dying in hospital, length of hospital stay.

It is possible to integrate PC into a regular internal medical ward, with positive impact on recorded outcome measures.

 Constantini et al. 2013, (Italy) [24]

Cluster Randomised Trial

34

LCP* for cancer patients dying in hospital medical wards.

Quality of EoLC* from perspective of bereaved family member; communication between ward staff and GPs* (VOICES)*.

Aspects of quality of EoLC* improved (emotional, spiritual needs, self-efficacy); slight improvement in communication, no significant improvement in symptom control.

 Veerbeek et al. 2008, (The Netherlands) [25]

Uncontrolled before and after study

27

Effect of the LCP* on 3 health care settings (hospital, nursing home, home).

Comparison of level of documentation, symptom burden (EORTC QLQ-C30* questionnaire) and aspects of communication (VOICES)* before and after introduction of LCP*.

Introduction of LCP* increased documentation, decreased symptom burden.

 Grande et al. 2000, (UK) [26]

Randomised controlled trial

33

Cambridge Hospital at Home for palliative care (CHAH).

Symptom control, adequacy of care, likelihood of remaining at home in their final 2 weeks, GP visits.

CHAH appeared to be associated with better quality home care.

 Vicente et al. 2010, (Spain) [27]

Retrospective and prospective cohort study

30

Influence of the Integrated Plan of PC* of the Autonomous Community of Madrid in the medical activity of a hospital based PC* unit.

Improvement in continuity of care, coordination amongst assistant bodies, increase in mean stay at the PCU*, increase in number of home deaths, etc.

PC home care improves continuity in care of patients. Transfers to intermediate stay care centers and deaths at home increased. Median stay at the PCU* decreased.

 Sampson et al. 2011, (UK) [28]

Randomised controlled trial

28

Pilot implementation of the assessment of PC needs of patients with severe dementia and discussion with principal carers to improve EoLC*.

Kessler Distress Scale, EQ-5D*, Decision Conflicts Scale, Decision Satisfaction Inventory, State Anger Scale, Life Satisfaction Scale, Satisfaction with EoLC*, Advanced Dementia Scale (FAST Scale); Pain and distress (the Abbey pain scale, the PACSLAC and the Doloplus), delirium (Confussion Assessment Method),

General unwillingness to address EoL* issues. All carers were keen to receive more information about EoL* issues in dementia, found discussions very helpful. Participation of clinical MD* team facilitated integration of intervention with the clinical service.

 Higginson et al. 2009, (UK) [29]

Randomised controlled trial

36

Evaluation of cost-effectiveness of a new PC service for people with MS*.

Use of services, patient symptoms (UNDS*, EDSS* and POS-8*), other outcomes, caregiver burden (ZBI*).

Short-term PCT* was found to be cost-effective, reducing inpatient and community costs, caregiver burden and possibly patient pain.

 Edmonds et al. 2010, (UK) [30]

Randomised controlled trial

36

Evaluation of a novel PC service.

MS Impact Scale, POS-8*, ZBI*, Modified Lawton positivity questionnaire.

MS patients who received PC* service had improvements in 5 key symptoms (pain, nausea, vomiting, mouth problems and sleeping difficulties) on the POS and improved informal caregiver wellbeing.

 Koffman et al. 1996, (UK) [31]

A descriptive pilot evaluation

23

Pilot evaluation of hospice at home service for patients with advanced HIV/AIDS; 24-h terminal care.

STAS*, evaluating pain control, other symptom control, patient/family anxiety, patient/family insight and communication between patient and family, between professionals, between professionals and patient and family.

80 % died at home; STAS* showed improvements in items ‘other symptoms control’ and family insight.

 Pattenden et al. 2012, (UK) [32]

Non-randomised pilot evaluation

30

Collaborative PC* for advanced heart failure.

Death in preferred place of care; hospital admissions averted; costs of medical procedures, inpatient care and directs costs of intervention.

This pilot study provides tentative evidence that a collaborative home-based PC* service for patients with advanced CHF may increase the likelihood of death in place of choice and reduce inpatient admissions.

 Navarro et al., 2011, (Spain) [33]

Observational, retrospective and descriptive study

26

EoLC* of advanced chronic non-cancer patients identified by multidimensional evaluation and interdisciplinary teamwork in a medium and long term hospital.

General data, terminal criteria, diagnostic and prognostic information, development of advance directives, limiting levels of effort care, times from admission, risk of complicated bereavement.

Identification of advanced chronic non-cancer patients and their needs by interdisciplinary teamwork enabled indication for PC soon after admission and ensured appropriate care during their stay.