Palliative care provision varies widely, and the effectiveness of palliative and hospice care teams (PCHCT) is unproven. To determine the effect of PCHCT, 10 electronic databases (to 2000), 4 relevant journals, associated reference lists, and the grey literature were searched. All PCHCT evaluations were included. Anecdotal and case reports were excluded. Forty-four studies evaluated PCHCT provision. Teams were home care (22), hospital-based (9), combined home/hospital care (4), inpatient units (3), and integrated teams (6). Studies were mostly Grade II or III quality. Funnel plots indicated slight publication bias. Meta-regression (26 studies) found slight positive effect, of approximately 0.1, of PCHCTs on patient outcomes, independent of team make-up, patient diagnosis, country, or study design. Meta-analysis (19 studies) demonstrated small benefit on patients' pain (odds ratio [OR]: 0.38, 95% confidence interval [CI]: 0.23-0.64), other symptoms (OR: 0.51, CI: 0.30-0.88), and a non-significant trend towards benefits for satisfaction, and therapeutic interventions. Data regarding home deaths were equivocal. Meta-synthesis (all studies) found wide variations in the type of service delivered by each team; there was no discernible difference in outcomes between city, urban, and rural areas. Evidence of benefit was strongest for home care. Only one study provided full economic cost-benefit evaluation. This is the first study to quantitatively demonstrate benefit from PCHCTs. Such comparisons were limited by the quality of the research.

To determine whether hospital-based palliative care teams improve the process or outcomes of care for patients and families at the end of life, a systematic literature review was performed employing a qualitative meta-synthesis and quantitative meta-analysis. Ten databases were searched. This was augmented by hand searching specific journals, contacting authors, and examining the reference lists of all papers retrieved. Studies were included if they evaluated palliative care teams working in hospitals. Data were extracted by two independent reviewers. Studies were graded using two independent hierarchies of evidence. A Signal score was used to assess the relevance of publications. Two analyses were conducted. In a qualitative meta-synthesis data were extracted into standardized tables to compare relevant features and findings. In quantitative meta-analysis we calculated the effect size of each outcome (dividing the estimated mean difference or difference in proportions by the sample's standard deviation). Nine studies specifically examined the intervention of a hospital-based palliative care team or studies. A further four studies considered interventions that included a component of a hospital or support team, although the total intervention was broader. The nature of the interventions varied. The studies were usually in large teaching hospitals, in cities, and mainly in the United Kingdom. Outcomes considered symptoms, quality of life, time in hospital, total length of time in palliative care, or professional changes, such as prescribing practices. Only one of the studies was a randomized controlled trial and this considered a hospital team as part of other services. Most method scores indicated limited research quality. Comparison groups were subject to bias and the analyses were not adjusted for confounding variables. In addition, there were problems of attrition and small sample sizes. Nevertheless, all studies indicated a small positive effect of the hospital team, except for one study in Italy, which documented deterioration in patient symptoms. The Signal scores indicated that the studies were relevant. No study compared different models of hospital team. This review suggests that hospital-based palliative care teams offer some benefits, although this finding should be interpreted with caution. The study designs need to be improved and different models of providing support at the end of life in hospital need comparison. Standardized outcome measures should be used in such research and in practice.

No abstract