How should we manage information needs, family anxiety, depression, and breathlessness for those affected by advanced disease: development of a Clinical Decision Support Tool using a Delphi design

Providing optimal care to patients facing life-threatening and progressing illnesses requires a focus on patients’ and their family caregivers’ physical, psychological, emotional, and spiritual needs [1]. It is not solely a task of experts in palliative care. The number of people affected by chronic and progressive illness is escalating, fueled by population ageing and improved life expectancy [2]. While specialist palliative care services, such as hospices and home care teams, provide an extra layer of support for those with the most complex needs, patients and their families with progressive and life threatening illness come into contact with health and social care workers in all settings – primary, secondary, and tertiary. Providing holistic care for seriously ill patients and their families is now a central component of health care [3]. This can, however, be daunting for many doctors, nurses and other staff, especially with limited formal (undergraduate and postgraduate) training [4, 5].

An important starting point in effective clinical management is the assessment and identification of problems [6]. Unfortunately, practitioners often miss many of patients’ needs and symptoms, especially when these are non-physical [7‐11]. This problem is compounded because patients have multiple symptoms and concerns [12‐14]. Asking patients to report their concerns using Patient Reported Outcome Measures (PROMs) has been proposed as a way to overcome this [15]. When repeated over time, such assessments can become measurable outcomes, revealing how patients’ health status changed following the care provided [16].

Although practitioners express positive attitudes towards measuring PROMs [17, 18], routine measurement of concerns is hampered by a lack of training and guidance on how to use and respond to PROMs in clinical care [17‐19]. Published guidance on using PROMs [20‐22] is fragmented; it lacks specified steps and recommendations to follow when implementing PROMs [20, 21] or does not focus on advance disease [22]. Even more pressing, there is a lack of guidance on how to respond to specific PROM scores. There is often an unfamiliarity with the score interpretation [23], while PROM scores are potentially not fed back to the right person, or reported as often as needed [24]. Clinical significance of health-related quality of life scores are not always reported in studies [25, 26]. Unsurprisingly, therefore, clinical decision-making is not often based on outcome scores [27] and measuring PROMS seems to have a stronger impact on process-aspects of care, i.e. detection of symptoms, than on outcome-aspects of care, i.e. patients’ health status (e.g. [15, 23, 28, 29]). Ideally, PROMs should routinely be used and assist in detecting problems, planning treatment, and monitoring how well concerns are alleviated.

To fill this gap we need specific clinical decision support aids for specific PROM scores. This should be linked with better focused guidance on how to implement PROMs in routine clinical practice. It is important that both sets of guidance are developed together. This will ensure specific PROMs are successfully implemented in clinical care, while at the same time clinicians are supported in using the PROMs when responding to concerns.

One of the most widely used outcome measures in clinical practice in advanced illness is the Palliative (or Patient) care Outcome Scale (POS) family of measures (consisting of the POS, Integrated POS (IPOS), African Palliative Care Association African POS (APCA African POS), and POS-Symptoms (POS-S)) [30]. It assesses physical symptoms, emotional, psychological, and spiritual concerns, and needs for information and support [31]. It is brief (<10 minutes to complete), widely validated, able to transfer across settings, has good responsiveness to change, and has been translated and/or culturally adapted and revalidated in many different languages and cultures (e.g. [32‐34]). A full suite of free user support resources is available at www.​pos-pal.​org [35]. During a training day on POS in 2013, we identified clinicians’ need for guidance in the interpretation of and responding to POS scores. Most difficulties were encountered with interpreting the questions regarding psychological functioning (i.e. depression/feeling worthwhile, n = 18/36 comments), information provision (n = 7/36 comments), and family anxiety (n = 4/36 comments). Breathlessness is a very disturbing symptom for patients [36, 37] and their families [38]. Creation of evidence-based clinical guidance on how to respond to these concerns is needed.

The aim of this study was therefore to develop a Clinical Decision Support Tool (CDST) for the specific POS items of most concern to clinicians: information needs, family anxiety, depression, and breathlessness. The guidance is aimed for all practitioners and settings, and is applicable to all patients/family caregivers with complex needs and progressive, life-threatening and serious disease. Because successful implementation of PROMs in clinical practice is a precondition to using a CDST, but current guidance is fragmented, in parallel we also developed accompanying guidance on PROM implementation. This article reports the development and final formats of the CDST and implementation guidance. It also offers a novel methodological approach for CDSTs in other areas.

This is the first study to develop a CDST for some of the complex problems faced in advanced disease, in particular information needs, family anxiety, depression, and breathlessness, linking these to specific actions. The CDST has an evidence-based approach, responding to different levels of patient-reported symptoms. We found it was necessary to have both core recommendations, applicable for all symptom levels, as well as responses that could be titrated up with increasingly severe symptom scores. Accompanying guidance on PROM implementation was also developed, using an 8-step approach. By producing both types of evidence, PROMs (and specifically POS, or similar tools) might be more widely integrated in clinical care and their effects on patients’ outcomes can be strengthened.

The potential of our guidance and CDST to improve provided care and patient outcomes is supported by accumulating evidence. The Australian Palliative Care Outcome Collaboration showed that implementing PROMs nationally in palliative care improved patients’ reported symptoms year after year [89, 90]. Expert consensus underlines the importance of coming to a routinely collected set of outcome measures, which can be used to make comparisons across services and countries [21]. Our provided guidance might assist in this development while overcoming some of the perceived barriers of implementing PROMs, most notably a lack of knowledge and education [6, 91].

The CDST can assist clinicians in responding to the psychosocial aspects of clinical care which they may feel less confident in delivering. Across the several symptoms included in our Delphi, for higher scores, more intensive clinical responses seem warranted. Ensuring the core recommendations (good patient care; providing psychosocial support and empathy; the use of open communication) are in place remains vital in these situations. These core recommendations reflect previous study findings, showing that patients expect their clinicians to show both technical competence [92] as well as seeing them as an individual person [93]. Empathy becomes increasingly important in progressing disease [94], while most medical complaints are related to communication deficits [95, 96], adding not only to patient but also caregiver distress [97]. Our CDST postulates that proper assessment should be followed by non-pharmacological interventions which can be progressed to pharmacological interventions in the more medical oriented symptoms of breathlessness and depression. However, also in these domains, there seems a preference for psychosocial interventions. Indeed, a recent RCT showed that a short-term service provided by palliative care, respiratory, physiotherapy, and occupational therapy improved breathless patients’ outcomes [98]. This service combined pharmacological review with a focus on non-pharmacological interventions such as pacing and relaxation.

Although originally developed for the POS family of measures, the core recommendations point towards the potential usefulness of our CDST for other PROMs covering similar domains (e.g. the Edmonton Symptom Assessment System [99], the McGill Quality of Life Questionnaire [100], or quality of life measures such as the EORTC-QLQC30 [101]). It is a wider problem that in advanced illness patients’ and caregivers’ psychosocial needs often remain unmet [102, 103]. A recent systematic review, including but not limited to POS, concluded that feeding back PROM data in palliative care influenced patients’ psychosocial outcomes the most [104]. Our CDST might strengthen this finding by generalizing its applicability beyond POS.

Although the CDST proposals are based on scientific evidence combined with expert opinion, this tool is not intended to be prescriptive. Instead, it aims to help practitioners think through the best decision towards difficult and complex encountered problems. Being too prescriptive can be counterproductive, especially because in advanced illness patients and families have complex and often quite individual needs, circumstances, and trajectories, which interact [1, 105]. Instead the CDST should be used as the name implies, as a support, to aid a wider range of aspects to be considered when making treatment choices, alongside skilled individual clinical assessment and knowledge, taking into account patients’ and families’ individual preferences, circumstances and available resources. In areas of uncertainty or conflict, specialist support or a second opinion should be obtained. The CDST should not be used as an endpoint (‘tick-box’ exercise) in itself but as a starting point to achieve high quality person-centered care and good clinical practice. As the implementation guide makes clear, training and ongoing support will be an essential component.

This study has limitations. Our sample size was relatively small, although it should be noted that the responses rates in the online Delphi (66 % vs 62 %) are in line with a previous similar study in this field [106]. Next, we asked participants in the Delphi study to review a wide area of topics, which might have been challenging, especially surrounding the specific topic of breathlessness. However, as palliative care focuses on physical, psychological, emotional, and spiritual issues, we anticipated most participants to have (some) knowledge in all domains and added a ‘do not know’ option in the Delphi. The quality rating should also be interpreted with caution as we used an adapted GRADE approach. Many sources used different rating systems and we followed sources’ own evidence-levels if already provided. No strength of recommendation (representing whether desirable effects of recommendations clearly outweigh (or not) undesirable effects [44]) was provided due to the heterogeneity of used quality ratings and the envisaged applicability of our recommendations in different settings and countries. Lower ratings do not necessarily imply that these recommendations are not important, as they were among the highest rated recommendations by our expert panel (e.g. the provision of emotional support and respect for cultural/religious traditions when handling family anxiety). These ratings also highlight the pressing need for more high-quality research studies to build the evidence-base of these and other key components of palliative care. For example, the use of a fan was labelled as ‘no evidence’ based on a Cochrane review [59]. However, this is still in the early stages of evaluation, with recent studies showing conflicting effects (e.g. [107, 108]). As fans are inexpensive, unlike to cause any harm and have low side effects, they can be worth trying. Another limitation entailed that we only included sources that focused on palliative care in general and not on specific diseases as we sought recommendations that were applicable across diseases. For some diseases, the evidence might propose alternative strategies; however, we would expect experts in those fields to be aware of this evidence to take it into consideration. Although we used a broad search strategy, relevant sources might have been missed. Furthermore, we refer in footnotes to the Cicely Saunders Institute’s Breathlessness Intervention Service and the Vitaltalk organization for more resources on breathlessness and information needs. These were not found via our search strategies (as they are very recent), but we believed them to be of importance to clinical care. Finally, due to time and resource constraints, only one author identified resources and extracted data of included sources. That being said, a major strength of this work includes the involvement of patients and families in the development of the CDST and the combination of evidence-based recommendations with clinician expertise. Future studies can use this study as a starting point in developing evidence-based CDST for advanced disease, and should now refine our proposed CDST, test its effect on patient and family outcomes, and develop CDST for other POS items.

Our findings underline the importance of providing good patient care, psychosocial support and empathy, and communication in advanced disease for all patients and families, irrespective of POS scores, in the domains of information needs, family anxiety, depression, and breathlessness. For increasing scores, patients’ symptoms should be assessed and responded to with non-pharmacological interventions, followed by pharmacological interventions (for breathlessness and depression). We were able to develop these recommendations into a CDST. Our novel evidenced-based approach to develop a CDST offers a replicable method for other areas. By using the presented recommendations alongside skilled clinical knowledge and patient preferences and ongoing training, it aims to help support clinicians provide the best possible patient-centered care and patients and their families achieve the best possible outcomes in highly threatening times. Systematically following the 8-step framework can support successful implementation of PROMS, and POS, in advanced disease.