Patient reported outcome measures for measuring dignity in palliative and end of life care: a scoping review

Patient reported outcome measures (PROMs) are standard questionnaires or tools used to acquire information from patients regarding perceptions of their wellbeing and functional status [1, 2]. They operate by recording changes over time, following an initial baseline measurement being taken [3]. These tools can be used in areas such as palliative care to provide a means of assessing and monitoring care [3]. Measuring outcomes in healthcare require patients to be the primary givers of information, in order to gain accurate and relevant insight into changes in health, quality of life, and symptom severity [3]. PROMs are increasingly encouraged for use in clinical practice, audit and research [3, 4]. In research, PROMs are pivotal for testing the effect of an intervention [4‐6]. PROMs are advocated for their ability to elicit information directly from the patient, prioritise the perspective of the patient, and when sufficiently validated, they are able to deliver an accurate assessment of the clinical population [1, 3]. PROMs are the ideal strategy to measure patient-centredness by giving patients the opportunity to assess and convey the extent to which they feel that care received meets their values and needs [7]. Furthermore, PROMs can help highlight aspects of care that need to be improved [7]. In addition, assessments allowed through PROMs are crucial to care provision [8].

In palliative and end of life care, the use of PROMs is encouraged to allow palliative care interventions to be assessed, which ultimately provides guidance for teams working in this area. Furthermore, UK supportive and palliative care guidance recommends that systematic assessments are crucial to the provision of supportive care [8, 9]. It has been suggested that the growing landscape of research into experiences of dying have highlighted the importance of focusing on quality improvement in the area of end of life care [10].

The primary goal of palliative care should be to ensure high quality of life for patients who have advanced incurable disease. This involves attending to the person’s psychological, social, and spiritual needs, so ultimately the endeavour of palliative care is to generate the best quality of life for patients and their families [11]. Outcome measures should then strive to assess whether this has been achieved [11]. Hearn and Higginson [12] suggest that PROMs in palliative care should be developed to capture key goals of palliative care philosophies including improving quality of life before death, attending to symptoms, and providing support for loved ones. The information gained from outcome measures is useful since it produces a clinical picture of patients, helps improve symptom assessment, enhances communication between service users and staff, achieves better patient satisfaction, and ultimately supports the delivery of person-centred care [6, 13‐15]. In turn, this can help improve the quality of the service [6], and supports shared decision making between patients and staff [15]. In addition, data elicited from PROM reporting can also generate key information that may not be otherwise routinely recorded or available in medical and nursing records [14].

The palliative care environment can pose challenges to the effective use of PROMs [12]. For example, the patient with palliative care needs may be unable to complete PROMs due to advanced illness or cognitive impairment. This potentially leads to situations where PROMs are only used for patients with less problems and therefore excludes those with more severe issues. This also compromises the validity of the PROM, since the use of proxy individuals may not reflect the true perceptions of the patient [16]. Collins et al. [6] propose the use of the term “person-centred outcome measures”, to reflect situations where someone other than the patient (such as a family member of healthcare professional) fills in the measure, but still strives to capture the patient’s priorities. In their systematic review on the use of outcome measures in palliative care, Hearn and Higginson [12] found that no outcome measure in their final included selection was designed to cover more than one relevant domain in palliative care. This is echoed by a more recent review [10] focusing on end of life care, in which they identify a need to take into account “the full spectrum of patients’ and caregivers’ end-of-life experience” (2007:1849). In addition, measuring the diverse outcomes of palliative care requires a holistic approach that includes consideration of what may be seen as obscure domains such as psychosocial and spiritual dimensions [3]. These authors also propose that the realm of palliative care is increasingly complex because of the spectrum of needs and conditions that are present, which then makes measurement difficult. Moreover, certain outcomes, such as quality of life, may be difficult to measure, since this is a multidimensional and subjective concept [7].

There may also be challenges to both the selection of PROMs and implementing them in clinical practice, due to lack of time, resources and training [1]. The unique needs of patients requiring palliative care means that outcome measurement can be challenging [3]. In addition, ethical concerns arise around issues such as whether measuring symptoms might intrude on the patient’s preferred use of time at the end of life; whether it is appropriate to measure symptoms that may be complex and inter-related in presentation; [3]. It is also highlighted that [13] the possible drawbacks of using paper-based outcome measures, suggesting that they may not be flexible; are limited in terms of language and literacy requirement; may not be appropriate for individuals from minority groups; and the inability to generate instant clinical information. Similarly others question the usefulness of questionnaire content, by suggesting that they do not offer clear distinction between statistical and clinical significances [16]. However, in cancer settings, [12] PROMs are useful in helping to distinguish between physical, emotional, and social problems and also providing a means to monitor effects.

A substantial body of work has been undertaken which provides guidance and recommendations for the development, selection, implementation and use of PROMs in palliative and end of life care, in addition to, addressing some of the identified challenges associated with the use of PROMS. Ideally, PROMs in end of life care should aim to cover the key areas of hope, spirituality, symptom control, self-concept, the therapeutic consultation, and dignity [12]. In order to enhance the evidence base for end of life care, PROMs should capture patient and caregiver experiences and undergo reliability and validity testing [10]. As well as, being selected and implemented from a reliable evidence base [4] Specific projects have prioritised developing and providing guidance for PROMs in palliative and end of life care. These are summarised in Table 1.

As part of the Methods of Researching End of life Care (MORECare project) [17] a checklist of components was generated to help researchers in designing and conducting intervention studies in end of life care. This work provides specific recommendations for the design of outcome measures and proposes guidelines for format aspects including short lengths; ease in administering and interpreting; and ensuring that the tool is adaptable to different settings. In addition, this body of work emphasises the importance of the tool being validated and tested for reliability. Similarly, the PRISMA project [18] highlights the importance of validity and reliability and proposes what types of validity and reliability testing should be done. In addition, this work recommends that outcome measures have the ability to be translated into other languages. Both MORECare and PRISMA state that responsiveness to change over time is also a priority for outcome measures to demonstrate [3, 12]. A further key body of work is Witt et al.’s Outcome Assessment and Complexity Collaborative (OACC) Suite of Measures [19], which emphasises the need for outcome measures to capture the domains relevant to palliative care. These recommendations were taken into account when formulating the critical appraisal tool for this current scoping review (Please see section Critical appraisal of PROMs under “Methods”, for further detail).

However, as well as meeting particular quality standards, PROMs also have a responsibility to measure phenomena that are significant for, and prioritised by patients [19]. Dignity is recognised as being integral to human rights and a priority for recipients of palliative and end of life care [20, 21]. Dignity is defined as having a quality that is deemed worthy of receiving respect and, in turn, also promotes self-respect [22]. The tenets of dignity, which revolve around kindness, humanity, and respect, are arguably neglected by professionals working under challenging conditions, including time pressure [23]. Attention to conserving dignity in palliative and end of life care settings is welcomed by patients and the research landscape is increasingly dedicated to developing clinically-appropriate interventions which fulfil this aim [21, 24‐26]. Nevertheless, dignity is not always made a priority when developing PROMs [27]. Therefore, PROMs which extract dignity-related outcomes are important in order to assess the extent to which patients are satisfied that their dignity is being conserved by care-giving activities and approaches.

This current review intended to identify key PROMs used in palliative and/or end of life care that related to dignity and were related to and relevant for the nursing context, i.e. care delivered by nurses as part of a multi-professional team caring for people with palliative care needs in the last months of life. In addition, this review, uniquely, seeks to find PROMs which specifically help to measure patient perceptions of dignity, a focus which has not been exclusively attended to by previous reviews. Moreover, as evidence-based practice is central to both medical and nursing care environments [28‐31], it was vital that a review of available evidence was undertaken to ensure that appropriate outcome measures are being used in research and clinical practice.

An early review [3] sought to identify outcome measures which enable palliative care interventions used for patients with advanced cancer to be evaluated. Forty-one measures made it into the final review, including the Edmonton Symptom Assessment Schedule (ESAS), and the McGill Quality of Life Questionnaire (MQQL). These authors proposed key criteria to use when assessing measures, which revolve around validity and reliability considerations, as well as responsiveness to change. The final seven included measures were found to target physical, psychological and spiritual domains to some extent, but no one measure fulfilled all the requirements of an ideal tool. Hearn and Higginson [12] questioned whether such an ideal tool can be achieved, but also stand firm that this endeavour should not be neglected. Mularski et al.’s [10] review had similar aims in seeking to identify appropriate end of life measures, which had undergone good psychometric testing. A larger number of measures were found with 95 being identified from the review; thirty-five from a previous systematic review; and finally 64, which measured end of life experience. The measures varied in the focus, and revolved around quality of life; physical, emotional, spiritual areas; advanced care planning; and caregiver wellbeing. The review highlighted the lack of robust testing for most measures, and a significant gap in appropriate measures to address continuity of care, advanced care planning and spiritual issues. However, the authors acknowledge that amongst the limitations of their review, their broad definition of “end of life” risks being both over-inclusive and possibly not inclusive enough.

Other reviews chose specific measures to focus on; Collins et al. [6], aimed to appraise evidence for use of the Support Team Assessment Schedule (STAS) and the Palliative care/ Patient Outcome Scale (POS). The review highlighted a clear increase in use of both measures in diverse settings and countries, with both being accepted as validated and reliable tools used to assess symptoms and needs. In addition, as well as being part of clinical assessments, the tools were also used in the evaluation of interventions. The POS remained more popular than the STAS, and the authors propose that this may be because the POS is inherently patient-focused by being a PROM. However, the review risks having an element of bias, since their focus on these two measures could arguably have been influenced by the fact that most of the authors had involvement in the development of them.

Finally, Parker and Hodgkinson [3] were interested in determining the reliability, validity, and feasibility of outcome measures used specifically in long term care facilities. Motivated by the lack of work previously addressing the area of long term facilities, the review identified ten outcome measures appropriate for this setting, with the Family Perceptions of end-of-life Care Scale (FPCS) being deemed most appropriate, based on the rigorous development and testing it has undergone. However, the authors found that some validity aspects were difficult to determine, thus compromising this aspect of appraisal.

Development of PROMs research is in line with recent health policy recommendations [9, 32, 33], and acknowledges the importance of prioritising the patient’s perspective, thus supporting patient-centred care. This current scoping review is both relevant and timely, in particular, because no recent reviews focus exclusively on PROMs within the nursing context nor with a focus on dignity. Undertaking a scoping review, as opposed to any other form of knowledge appraisal, allowed us to specifically address an exploratory research question which aimed to map key types of evidence as well as identifying any gaps in that evidence [34]. Moreover, this review intends to identify high quality PROMs that can potentially be used to confidently evaluate palliative and/or end of life interventions, aligning with the MRC’s recommendations [35] for evaluating complex interventions. This will strengthen the evidence-base of these interventions and support their implementation to practice.

To map the evidence and quality of PROMs which assess dignity, and have been used for patients in palliative and/or end of life settings. The key question guiding this scoping review was as follows:

What are the key PROMs available that are used to measure dignity in palliative and end of life care for clinical practice and/or research purposes?

As well as measuring dignity, PROMs were expected to show high standards of development, evidenced by their ability to meet key critical appraisal requirements (see Table 5).

Inclusion criteria were developed early on to guide the narrowing down process of retrieved articles. This conveyed the key criteria that papers were expected to achieve in order to be considered for inclusion. The inclusion criteria are displayed in Table 2 and were used in Stages 1 and 2 of the narrowing down process (please see section Study selection ).

The methods for scoping reviews have come under recent scrutiny [34] and have been developed from the early work of Arksey & O′ Malley [36] with an aim of producing consistent use of methodological guidance and reporting. We followed five of the six methodological steps outlined in the Arksey & O’Malley [36] and Levac et al. [37] framework recommended by Colquhoun et al. [34]. The final stage ‘Consultation’ is optional and involves opportunities for consumer and stakeholder involvement which were beyond the scope of this review:

Figure 1 provides an overview of the key stages of the review process that we undertook.

The search strategy was developed through collaborations between the research team (including a senior clinical academic with expertise in palliative and supportive care; a senior academic with expertise in evidence synthesis; a lecturer in adult nursing with clinical and research experience in community nursing), and an information scientist. In addition to consultations with the expert in evidence synthesis and information scientist, keywords and free text terms were also informed by an early scoping exercise, from which literature was scrutinised for relevant terminology and synonyms. The Medical Subject Headings (MeSH) term browser, provided by the United States’ National Library of Medicine, was also used to identify appropriate index terms. The final search strategy consisted of dignity related terms such as personhood; dignity or dignified; patient-centred care; which were linked (via the Boolean operator “AND”) to care-related terms such as terminal care; hospice care; palliative care; end of life. A sample search strategy from the MEDLINE database is displayed in Table 3.

Searches were conducted within key nursing, medical, psychological, and social sciences databases to identify papers from the period of 2005–2015. This period was deemed appropriate, since a preliminary exercise and consultation with the research team indicated that relevant papers, featuring the most advanced dignity-related PROMs, were most likely to emerge within this ten-year timescale. The databases that were consulted were ASSIA; CINAHL; Cochrane Central Register of Controlled Trials (CENTRAL); Cochrane Database of Systematic Reviews (CDSR); Database of Abstracts of Reviews of Effects (DARE); EMBASE; Health Technology Assessment Database (HTA); MEDLINE; Oxford PROM Bibliography; PROQOLD; PsycINFO; and Social Science Citation Index.

In addition, grey literature searching was carried out using Google search engine, grey literature databases, and relevant charity and organisation websites. Databases consulted were the Bielefeld Academic Search Engine (BASE) and OpenGrey. Organisation and charity websites, which included Royal College of Nursing, Age UK and Department of Health were consulted. Grey literature searching was supported by consultations with experts in the field. Key terms were adapted from the main search strategy. The grey literature failed to provide any appropriate PROMs which stood up against the criteria outlined in the critical appraisal tool.

Database searches retrieved 7845 results which were exported to a referencing management program (Endnote ×6). The numbers of retrieved articles from each database are displayed in Table 4. Thirty-nine duplicate papers were identified and removed, leaving a total of 7806. A narrowing down process was established which consisted of systematic stages. Stage 1 involved members of the research team (MN, BJ, CC) assessing title and abstract against the inclusion criteria. If the titles and abstracts failed to meet the inclusion criteria, the paper was excluded. In Stage 2, the full texts were assessed against the inclusion criteria (BJ; CC; KF; BH; MN). Papers which failed to meet the inclusion criteria were excluded. There were three rounds of eligibility assessments for full texts papers, before the final seven papers were agreed. The narrowing down process is captured in a PRISMA flow diagram as conveyed in Fig. 2.

Critical appraisal of the PROMS was undertaken as part of the scoping review as recommended by Daudt et al. [38]. The critical appraisal guidance was developed based on previous work in the field [3, 12, 13, 39] and consisted of key qualities that the patient reported outcome measures being discussed in the papers were expected to have. In line with the recommendations and guidance offered by MORECare, PRISMA, and OACC projects (please see Table 1), the critical appraisal tool of this current review takes into account aspects of formatting; validity and reliability testing; clinical responsiveness; translation expectations; and other issues deemed relevant when appraising PROMs in palliative and end of life care. This tool is displayed in Table 5, and was used in stage 3 of the narrowing down process. This consisted of members of the research team (MN; BJ) assessing the PROMs identified in the papers from stages one and two, against the critical appraisal tool.

The specially developed critical appraisal tool comprises of seven key screening questions, with 17 specific areas to be scored. Where necessary, bullet pointed guidance is given under areas where it was anticipated that further clarification would be needed. The PROMs were scored based on the ability to answer “yes” to the relevant areas. For PROMs where translation was not relevant, there were 15 potential areas to score “yes” on (three under FORMAT; one under DATA COLLECTION TIME POINTS; four under VALIDITY TESTING; four under RELIABILITY TESTING; one under CLINICAL RESPONSIVENESS; and two under ACCEPTABILITY AND APPLICABILITY. Where considerations for translation were relevant, there were 17 areas to score “yes” on (15 as before with the additional two areas that come under TRANSLATION). In addition to using the tool, descriptive summaries of each PROM was produced to ensure that strengths and limitations not captured by the tool were identified.

Table 6 provides a means of charting the data and is an overview of the final included articles. The table headings which assisted the charting process are informed by data extraction techniques employed by previous researchers undertaken similar work [3, 10]. Charting the data around PROMs has allowed useful summaries to be produced, which served to compliment the critical appraisal performed using the tool.

Five PROMs had explicit dignity foci, as evident in the title of the tool, dignity being directly mentioned in items, and/or dignity being addressed and described in the developmental process. These were the Palliative Nursing Quality Measure [40]; the Patient Dignity Inventory [41] the Jacelon Attributed Dignity Scale [42]; the MIDAM-LTC [43]; and the QCQ-EOL [27]. These PROMs were specifically developed to take into account the need to preserve dignity-conserving care [27]; acknowledge the broad landscape of distress-related concerns of people nearing end of life, and address the physical, psychosocial, existential, and spiritual aspects of the patient’s experience accepted that dignity was an integral theme within palliative care, and that commitment to this is a vital quality of palliative care nurses that enables them to respect humanity [40]; and appreciated the complex and unique nature of dignity [42, 43]. Two PROMs honed in on two particular types of dignity, “attributed dignity” (behaviour with respect to self and to others) [42] and “personal dignity” (individualistic, related to personal circumstances) [43, 44], thus offering a nuanced approach to measuring dignity.

The three other PROMs targeted dignity-related themes, such as kindness, humanity, respect [26]) compassion and person centred care [45]. The short version Problems and Needs in Palliative Care questionnaire [46] contains items that explore the impact of fear and difficulties upon the person’s wellbeing, the extent to which these are problematic, as well as allowing the person to indicate whether they wish for the healthcare professional to attend to this. The term dignity is not mentioned in the development of the questionnaire and does not appear in any items, but it was designed to be patient-centred [45]. Likewise, although The Quality of Communication Questionnaire [47] does not directly reference dignity, communication has been cited as main factor of dignity [48], so the tool can be regarded as relevant for assessing dignity. The Missoula-VITAS Quality of Life Index [49] includes a number of items under five domains, to assess the impact of aspects such as interpersonal and wellbeing factors on quality of life. However, since dignity has been recognised as a potential indicator of quality of life [50], it might have been helpful for the tool to more explicitly include this phenomenon.

The scoping review has successfully identified eight PROMs currently being used in palliative and/or end of life settings to measure dignity amongst people receiving care. The critical appraisal process highlighted some shortcomings in our critical appraisal tool, namely that not all psychometric criteria were covered, for example, ecological validity and factor analysis. This meant that if a study had subjected measures to those types of testing, this could not be acknowledged by our scoring. In addition, the Missoula-VITAS Quality of Life Index- Revised achieved a high score despite the authors suggesting that it was not suitable for use as a research outcome measure [49]. Moreover, the critical appraisal tool did not differentiate between the types of participants used during face validity testing. Therefore claims by researchers to have addressed face validity should be looked at more critically, since this validation is not always attained from the perceptions and responses of patient groups i.e. the end-users; rather it may be that the tool was tested amongst healthcare providers. This may result in a measure consisting of items that do not reflect areas that matter to patients, despite having undergone some form of face validity. However, the critical appraisal tool only partly informed the overall critical appraisal, and the description of measures allowed strengths and limitations to be considered beyond the tool’s criteria.

Based on the psychometric qualities and critical appraisal the outcome measure of choice from our review is the Patient Dignity Inventory developed by Chochinov et al. [41] which scored 14/15 and the Missoula-VITAS Quality of Life Index Schwartz et al. [49] which scored 12/15. We recommend researchers and clinicians use the measures with the highest scores. We recommend further validity and reliability testing and clinical research for the potentially useful other measures which scored lower regarding psychometric properties and in our critical appraisal. Findings of this review have useful implications for future research and practice. We have identified measures that can be used to measure dignity issues at the end of life. We recommend further PROM developed to improve person centred care at the end of life.