Implementation of improvement strategies in palliative care: an integrative review

A comprehensive literature review was conducted, including Medline, CINAHL, British Nursing Index, PsycINFO, and by searching for grey literature [10] (e.g. literature that has not been published in peer-reviewed literature). The search strategy was limited to English literature only, to publications that concerned palliative care for adults (aged 18 or above) and to publications that were published between 2000 and August 2011. Various search terms were used that referred to subject-specific keywords describing palliative care, as well as the type of implementation strategy and outcomes of the implementation, including synonyms and Medical SubHeadings (MeSH) to include all relevant literature. Table 1 provides an overview of the search terms (the search string is available in Additional file 1). Disease-specific search terms, such as cancer or dementia, were not included because the organisation of palliative care goes beyond a specific disease [11].

Publications were included when they described (1) improvements to the organisation of palliative care, (2) which implementation strategies were used, (3) how these strategies were implemented and (4) the effectiveness of these strategies. Publications were excluded when (1) no abstract was available, (2) the implementation strategies were not directed at health care professionals or volunteers, (3) they were not directed at adult health care services or (4) educational curricula were developed.

Two of the authors (JvRP and RS, one with a background in nursing and health sciences and the other in sociology) independently screened title and abstract and reviewed the full-text articles of the included studies to identify implementation strategies to improve the organisation of palliative care. A data collection form was used to extract information about the country and year in which the study was published, study design, setting, type of disease, health care professionals involved, and type, description and impact of the implementation strategies used. Subsequently, implementation strategies were categorised into subgroups, according to the type of implementation strategy, similar to the approach of Grol and Grimshaw [5]. Data from the subgroup classification was coded and compiled into a matrix, whereby the effect of each implementation strategy was summarized as a significant improvement (++), improvement (+), mixed or limited effect (+/−) or no effect (−). To ensure the trustworthiness and rigor of the analysis, peer debriefing took place with the other authors throughout the entire process of data analysis.

Of the 2379 initially identified publications, 241 were selected for full-text assessment (Fig. 1). A first assessment of the full-text of these publications revealed that 156 publications could be excluded, as they did not meet the inclusion criteria (for example, because there were no details given about the improvement strategy used). Reference lists of all publications eligible for inclusion as well as a hand search in grey literature databases revealed an additional 27 publications eligible for inclusion. The remaining publications represented a wide variety of research methods and designs: interviews, focus groups, strategy development designs, case descriptions, surveys, process evaluations, RCTs, pre-post-test interventions, review papers as well as theoretical papers. Because of the large number of identified studies, of which many were of low scientific quality and with incomparable outcome measures, only studies with an experimental (n = 12) or quasi-experimental (n = 56) study design were selected for further analysis.

A total of 17 single intervention studies and 51 multi-faceted intervention studies were identified. Most studies were conducted in the USA (n = 29), UK (n = 19) and Australia (n = 8), but studies were also included from countries such as Japan, Taiwan, Italy and the Netherlands. Studies were conducted within the entire range of palliative care services, from home care services to advanced palliative care units in hospitals. Fifty-one studies were conducted in one setting (primary care: n = 2, hospital: n = 38, nursing home: n = 9, hospice: n = 1 and other: n = 1) and 11 were conducted in multiple settings. For six studies, the type of setting could not be identified. Within the included studies, a large variety of professionals participated. Thirty-five studies were directed at a single type of professionals (e.g. nurses only), 29 at two or more different groups of professionals (e.g. nurses and physicians), and four studies did not report the target group of professionals.

Additional file 2 provides a summarized description of the methodology, setting and country, number and type of participants, the implementation strategy and the impact of the strategy of each individual study. Table 2 provides an overview of the results of studies with an experimental or quasi-experimental study design.

Two experimental and 12 quasi-experimental studies used different types of educational strategies to improve palliative care, including lectures [12‐14], study days [15, 16], role play sessions [17, 18], interactive education [19, 20], educational outreach visits [21] and computer-facilitated education [22‐25]. Eight studies were targeted at a single profession, while six were targeted at multiple professionals. Four studies reported significant improvements, eight reported improvement, and two studies had limited or no effect.

One study used process mapping to improve the organisation of palliative care in a nursing home [26]. Before implementing the Liverpool Care Pathway, nursing home staff organized interdisciplinary team discussions where they answered the question, ‘If your patient is diagnosed as dying at 10 am on Monday morning and they are in pain, what happens?’ or in other words, ‘What is the process?’ in order to identify bottlenecks to be expected [26]. A repeated process measure post-implementation reduced the numbers of expected bottlenecks.

One study addressed feedback to improve the organisation of palliative care [27]. In an RCT, patients completed a health-related quality of life questionnaire. In the intervention group, hospital physicians received automated feedback upon completion of this questionnaire. The RCT showed improved health-related quality of life in the intervention group compared to the control group.

One quasi-experimental study performed by Lilly et al. described family and multidisciplinary meetings to improve the communication and shared-decision making at the intensive care unit (ICU) in a hospital [28, 29]. Pre- and post-intervention measurements showed that the use of such meetings reduced length of ICU stay.

Fifty-one studies used a combination of strategies. Half of these studies used a combination of solely educational strategies (for example, lecture combined with role play sessions) [30‐58], while the other half combined a variety of strategies (for example, education combined with feedback and reminders) [59‐86]. Nine studies had an experimental design. Based on the conclusions of the authors, there were 13 studies that had no or a limited effect on the specified outcomes, 21 studies that had a positive effect, and the remaining 17 studies reported a significant improvement.

This is one of the first reviews that provides an overview of implementation strategies used to improve the organisation of palliative care. The results of this review can be used as a starting point for further research. However, some limitations should be taken into account. Firstly, this review used the integrative review methodology. Although this approach allows for the combination of diverse methodologies (including non-experimental research), only studies with an experimental and quasi-experimental design were included because of the unexpected high number of publications on the highest evidence level. Since a variety of methods was used in these studies, a quantitative comparison of effect size was considered impossible. Secondly, because it was the aim of this review to provide an overview of strategies used to improve the organisation of palliative care rather than the effectiveness, we did not assess each individual study for risk of bias or effect estimates. The effects of the strategies presented in this paper should therefore be interpreted with caution. Thirdly, we have limited the search strategy to English literature only. Although publications have been included from non-English speaking countries such as Japan and Italy, it is likely that we have missed potentially interesting publications from countries that often publish in their own language. Fourthly, there is no generic set of search terms to identify literature about improvement strategies, despite the fact that improvements are now being recognized as a science [102, 103]. Although we have captured a broad selection of literature with our search strategy, it is possible that it did not identify all available publications on this topic.