We used CIS
] to synthesise a diverse and complex body of evidence consisting of both quantitative and qualitative studies. This methodology incorporates procedures of conventional systematic review methodology by applying explicit searching strategies to ensure replicable, specified inclusion criteria, and data extraction procedures to enhance clarity and comparability among studies. This was combined with recent methods for interpretive synthesis which drew on traditions of qualitative enquiry. This combined approach enabled both aggregation of the literature and the generation of theoretical categories that critically demonstrate the assumptions situated within the literature.
The search was conducted in four electronic databases: MEDLINE (1950 – January week 2 2010); EMBASE (1980 – January week 2 2010); PsycINFO (1806 – January week 2 2010); and CINAHL (1982 – January). Additionally the following databases were searched for relevant articles: Cochrane Database of Systematic Reviews, The Cochrane Central Register of Controlled Trials, Cochrane Pain, Palliative and Supportive Care Trials Register.
The combination of search terms was piloted with a number of trial searches to achieve a balance between recall and specificity. A pre-defined search strategy was decided upon, consisting of the following search terms, representing three groups which were combined with AND.
1. attitude$, OR view$, OR experience$, involv$, partic$, perspect$, concern$, challeng$
2. research (Focus)
3. palliative care, OR palliat$, OR terminal, OR terminally ill, OR supportive care, OR dying, OR death, OR end of life, OR end-of-life, OR hospice, OR bereavement.
We searched the reference lists of all relevant papers for further references. This was important as this topic concerns a body of evidence, which is not well delineated and is spread over the disparate research areas that address palliative care.
Three journals, in which the debate regarding whether or not to involve palliative care patients in research was conducted, were hand searched. These were:
· The Journal of Pain and Symptom Management (from 1999 to February 2010)
· Palliative Medicine (from 1989 to February 2010)
· Journal of Palliative Care (from 1985 to June 2010)
Inclusion and exclusion criteria
We included original research papers that enquired specifically about involvement in EoL care research or its impact on participants. The participants included a wide variety of stakeholders who come into contact with palliative care research, including: patients, informal caregivers, health professionals, managers, and researchers. We included all research designs and searched across the spectrum of palliative care from disease diagnosis to death.
We excluded studies on oncology patients participating in trials testing curative treatments. We excluded papers on children or adults with learning disabilities as these areas have their own specific issues; and studies on recruitment, informed consent, user involvement or bereavement. Non-English language papers were excluded. We did not include papers that were already included in previous reviews. However, we investigated the individual studies in these reviews and included them if they contributed to our review questions and responded to our quality criteria. Fatally flawed studies were excluded. To identify such low-quality studies we used the following criteria (as specified in
· Are the aims and objectives of the research clearly stated?
· Is the research design clearly specified and appropriate for the aims and objectives?
· Do researchers provide a clear account of the process by which their findings were produced?
· Do the researchers display enough data to support their interpretations and conclusions?
· Is the method of analysis appropriate and adequately explicated?
First, the identified studies were mapped according to country of origin, setting, participants, patient group, illness stage, and approach. The resultant studies were then counted. Three major categories were identified: studies that addressed issues relating to research in general; research with specific methods; and trials. Within these categories, studies were examined in more detail and the results compared. This revealed areas that were adequately covered and any remaining gaps.
In the second stage the research questions were applied to the literature across the categories. Results and emerging queries were listed. Recommendations from the studies were tabulated and organised inductively, thereby generating information on best practices.
A third stage focused on the normative assumptions underlying the studies and related these to the empirical evidence they generated. The characteristics attributed to ‘research’ and ‘care’ – and how these determined their conceptualization – were critically assessed. Whether the interpretation changed with each included study was continually tested. The findings are synthesised as a conceptual model for research participation for potentially vulnerable people.