Abstract
Interstitial lung disease (ILD) confers a high mortality and symptom burden, substantially impacting quality of life. Studies evaluating palliative care in ILD are rapidly expanding. Uniform outcome measures are crucial to assessing the impact of palliative care in ILD. This scoping review evaluates existing outcome measures in general health-related quality of life (HRQoL), physical health, mental health, social health and advance care planning (ACP) domains in patients with ILD. Articles in English with quantitative assessment of at least one measure of general HRQoL, physical health, mental health, social health or ACP in patients with ILD were included. Searches across three databases yielded 3488 non-duplicate articles. 23 met eligibility criteria and included three randomised controlled trials (RCTs) or secondary analysis of an RCT (13%), three cross-sectional studies or secondary analysis of cross-sectional study (13%), one prospective study (4%) and 16 retrospective studies (70%). Among eligible articles, 25 distinct instruments were identified. Six studies assessed general HRQoL (26%), 16 assessed physical health (70%), 11 assessed mental health (48%), six assessed social health (26%) and 16 assessed ACP (70%). The ability to compare results across studies remains challenging given the heterogeneity in outcome measures. Future work is needed to develop core palliative care outcome measures in ILD.
Abstract
Palliative care is a key component of quality care for patients with interstitial lung disease (ILD). This study delineates outcome measures used to assess palliative care in ILD and highlights the current heterogeneity, which hinders comparison between studies. https://bit.ly/3p8CbXQ
Introduction
Rationale
Interstitial lung disease (ILD) is a group of lung diseases with distinct aetiologies but common pathways of inflammation and fibrosis. The prevalence of ILD is 97.9 per 100 000 persons with an incidence of 19.4 per 100 000 person-years based on a recent study from France [1]. ILD affects individuals across a range of ages, races and sexes. Treatment is limited and, even with advanced therapy, the median survival ranges from 2–3 years in idiopathic pulmonary fibrosis [2] to 7 years in both chronic hypersensitivity pneumonitis [3] and rheumatoid arthritis-associated ILD [4]. For patients who require supplemental oxygen, the median survival decreases to 8.4 months [5]. In addition to the high mortality rate, ILD is associated with a profound symptom burden, including intractable cough, fatigue and immobilising dyspnoea [6]. In fact, ILD patients have more severe dyspnoea at the end of life than terminal lung cancer patients [7]. In addition to physical suffering, ILD patients may experience psychological distress, social impairment and difficult decisions regarding advance care planning (ACP) [8].
Palliative care aims to improve quality of life (QoL) through identification, assessment and treatment of physical, psychological, emotional and spiritual distress, while determining patients’ values, life goals and preferences around dying [9]. The American Thoracic Society recommends that palliative care be available and tailored to patients at all stages of illness and has prioritised research funding that advances high-quality palliative care and improves clinically relevant outcomes [9]. Clinically relevant palliative care outcomes assess patients’ symptoms and needs, monitor changes in health status or QoL, facilitate communication, aid in clinical decision-making and evaluate the effect of interventions or care [10]. Heterogeneity of outcome measures limits the ability of comparing the efficacy of interventions across studies. Thus, standardisation of a core outcome set is needed [10]. An essential first step in this process is to identify key domains and summarise existing palliative care outcome measures in patients with ILD via a scoping review.
Objectives
This scoping review seeks to delineate the palliative care outcome measures used in palliative care-related observational studies and clinical trials of patients with ILD and to classify them into five distinct domains. We use the validated patient-reported outcomes measurement information system (PROMIS) domains: physical health, mental health and social health. The PROMIS domains were used to provide a consistent, recognised methodology [11, 12]. In this review, it was recognised that there are data that fall into two important domains not easily categorised under the PROMIS structure. Thus, we propose two additional domains: general health-related quality of life (HRQoL) and ACP. This scoping review may serve as a reference for future researchers selecting palliative care outcome measures in the ILD population.
Methods
Identification of eligible studies
We conducted a search of palliative care interventions in patients with ILD. Our initial search was conducted on 19 August 2019 and included PubMed, Embase and the Cochrane Library. Major search terms included: “palliative care”, “hospice and palliative care” and “palliation”, in addition to general and more specific search terms for the major types of advanced lung disease, including ILD. We limited searches to citations in English (supplementary material). An updated search was conducted on 31 October 2020 and we identified two additional studies that met inclusion criteria.
Study selection
Exclusion criteria included articles that addressed a population aged <18 years, did not address palliative care, only included qualitative data, described surgical procedures, case studies of less than 10 cases and editorials or other descriptive nonclinical articles. We screened article titles and abstracts, reviewed full texts of articles, and finally performed data abstraction. Reviewers were all educated on inclusion and exclusion criteria, and we conducted regular team meetings to identify any concerns. Two independent reviewers screened all citations. Discrepancies were resolved through a third reviewer or consensus agreement. Abstract and full-text screening was managed using Covidence Software (Veritas Health Innovation, Melbourne, Australia). Reviewers (A.C. Moale, B. Seth, H. Brown, J.B. Vick and R.A. Gersten) initially screened titles and abstracts of each retrieved citation and subsequently reviewed eligible full text articles for inclusion. Only studies that addressed at least one palliative care outcome in patients with ILD with quantitative data were included. Studies that combined both qualitative and quantitative data were included, but only quantitative data were abstracted.
Data abstraction
Data abstraction was performed by A.C. Moale and R.A. Gersten between February and October 2020. The following data were collected from each of the eligible articles: author, year, study location, study design, aims of study, outcomes measured and instruments used to assess each outcome (table 1). Validated tools, custom-made instruments/questionnaires and numerical scales were all included under instruments. The term “retrospective data” includes outcomes assessed via retrospective chart review, the medical record and/or care provider discussions.
Description of all 23 included studies
Outcome domains were categorised according to the physical, mental and social health PROMIS Adult Self-Reported Health domains with the addition of general HRQoL and ACP, as fourth and fifth domains, respectively (table 2). HRQoL was included as a separate domain for the measurement of general HRQoL, rather than a specific physical, mental or social health dimension of QoL (figure 1). Social health includes social function, defined by the ability to participate in and satisfaction with social roles and social relationships, such as social support rather than isolation [34]. Table 2 lists the outcomes measured under each domain. The instrument(s) used to measure each outcome was identified (tables 1 and 3) according to which outcome and domain each article used the instrument to assess. Instruments were included under multiple domains if the authors used the instrument to assess outcomes in multiple domains.
Categorisation of domains and outcome measures
Physical, mental and social domains as categorised by patient-reported outcomes measurement information system (PROMIS) adult self-reported health. General health-related quality of life (HRQoL) was added as a fourth domain, realising that although HRQoL includes physical, mental and social functioning, it can also be assessed generally. Advance care planning was added as a fifth domain.
Palliative care outcome measures in ILD
Results
Literature review process
We examined 4782 total articles (1667 from PubMed, 2874 from Embase and 241 from the Cochrane Library). Of these, 1294 duplicates were removed and there were 3488 remaining articles. 760 studies were deemed eligible for full-text review. Reviewers screened these full-text articles and selected 264 articles under the initial search strategy for palliative care outcomes in advanced lung disease. Reviewers (A.C. Moale, B. Seth and R.A. Gersten) identified those articles specific to ILD, which yielded 68 total articles. Of these, reviewers (A.C. Moale, B. Seth and R.A. Gersten) selected studies with quantitative results, for which there were initially a total of 21. An updated search prior to paper submission yielded two additional articles, resulting in a total of 23 included articles (figure 2).
Literature flow of articles in the scoping review. PC: palliative care: ILD: interstitial lung disease.
Study designs of eligible studies
Of the 23 eligible articles, the majority (16) were retrospective (70%), followed by three randomised controlled trials (RCTs) or secondary analysis of an RCT (13%), three cross-sectional studies or secondary analysis of a cross-sectional study (13%) and one prospective study (4%) (table 1).
Outcomes assessed by domain
Among the 23 eligible articles, six (26%) assessed general QoL, 16 (70%) assessed physical health, 11 (48%) assessed mental health, six (26%) assessed social health and 16 (70%) evaluated ACP. The majority of articles (61%) measured outcomes in more than one domain and 43% measured more than two domains. Among the 23 papers, 76 total analyses were performed (table 1). There were 25 distinct instruments used in 53 analyses (70%). Retrospective data, as defined above, was used in 23 analyses (30%). Instruments may be used to assess multiple domains. Table 3 describes each instrument and identifies the domains authors used the instrument to assess.
HRQoL
General HRQoL was assessed in six articles (26%) with eight distinct instruments (table 1). Of these eight instruments, only King's Brief Interstitial Lung Disease (KB-ILD) and St George's Respiratory Disease Questionnaire (SGRQ) are validated in the ILD population (table 3) [16, 38, 49]. The Chronic Respiratory Disease Questionnaire and the RAND 36-Item Health Survey (RAND-36) were used in multiple studies. The remaining six instruments (table 2) were each used only once. Retrospective data was never used to assess HRQoL.
Physical health
Physical health was assessed in 16 articles (70%) with 16 distinct instruments. The modified Medical Research Council Dyspnoea Scale was used four times and the Numerical Rating Scale was used three times to measure physical health. The London Chest Activity of Daily Living, the modified Edmonton Symptom Assessment System (mESAS), RAND-36 and the Palliative Care Outcome Scale were each used twice. The remaining 10 instruments (table 2) were all used once. Retrospective data was used four times.
Among the physical health outcomes, dyspnoea was the most commonly measured (15 times). Other symptoms (including restlessness, agitation, cough and dry mouth) were assessed nine times. Fatigue and pain were assessed eight times. Gastrointestinal symptoms (including anorexia, nausea, diarrhoea, constipation, dysphagia and loss of appetite) were assessed seven times. Sleep disturbances, including insomnia, were assessed six times. Physical function and activities of daily living were assessed a total of four times. Energy level was only assessed once.
Mental health
Mental health was assessed in 11 articles (48%) with eight distinct instruments. The Hospital Anxiety and Depression Scale was used three times and the mESAS and RAND-36 were each used twice to measure mental health. The remaining five instruments (table 2) were each used once. While some of these instruments have been validated in the general or cancer population, none of these instruments are validated in the ILD population. Retrospective data was used three times. Anxiety and depression were each assessed nine times. Cognitive function (including confusion, loss of consciousness and delirium) was assessed three times.
Social health
Social health was assessed in six articles (26%) with four distinct instruments. RAND-36 was used twice. The remaining three instruments (table 2) were each used once. Retrospective data was used once. Ability to participate in social roles and activities was assessed three times. Social isolation, social function and other (including receipt of health, voluntary or social care services) were each assessed once.
ACP
ACP was assessed in 16 articles (70%). The Swedish Register of Palliative Care end-of-life questionnaire was the only instrument used. Retrospective data was used 15 times. Specialist palliative care was assessed 11 times, location of death was assessed nine times and ACP activities (including code status and advance directive discussions) were assessed eight times. Mechanical support (including intubation and removal of endotracheal tube) was assessed four times. Less commonly, goals of care discussions (including prognosis and end-of-life care) and location of care were assessed three times. Expected versus unexpected death was only assessed once.
Discussion
This scoping review delineates multi-dimensional outcome measures of palliative care in patients with ILD. Palliative care is a comprehensive approach to improving the QoL of patients with serious illness through the identification and assessment of physical, psychosocial and spiritual issues, which may evolve throughout the course of an illness [9]. Palliative care has shown to benefit a number of chronic medical conditions and has been integrated into standard care. In general, for adults with advanced illness, palliative care may improve a patient's QoL, symptom burden, patient satisfaction with care and chances of dying in a preferred location [50]. For patients with congestive heart failure (CHF), palliative care has demonstrated statistically and clinically significant improvements in QoL and symptom burden leading to the incorporation of palliative care into the treatment guidelines for CHF [51, 52]. Palliative care in lung cancer improves QoL, lessens depressive symptoms and actually increases median survival despite less aggressive care at the end of life [53]. When palliative care is embedded in an ILD clinic, there are significantly higher rates of ACP. Higher rates of ACP increase the likelihood of patients receiving end-of-life care in concordance with their preferences and may increase satisfaction with healthcare and decrease healthcare cost [54]. The importance of palliative care for patients with ILD is of increasing interest [9]. While we speculate that palliative care improves QoL in patients with ILD, we currently lack conclusive data. Evaluation of outcome measurement plays a significant role in assessing quality palliative care interventions. Currently, the diversity of outcome measures is vast. This lack of uniformity makes the selection of appropriate outcome measures and interpretation of results across studies challenging.
The majority of articles included were retrospective analyses. Only three RCTs have been conducted on this important subject. RCTs are the “gold standard” for study design [55], and we hope to see the emergence of high-quality trials to further evaluate the impact of palliative care interventions in ILD. Standardisation of domains and outcome measures is a key first step to expanding this evidence base.
In this study, we organised data into the PROMIS domains of physical health, mental health and social health. Not all critically important data to the field of palliative care fit into the PROMIS domains, thus we proposed two new domains of general HRQoL and ACP (figure 1 and table 2). Physical health was the most frequently assessed domain, with dyspnoea being most often assessed (table 3). This may be because physical symptoms are more easily and commonly described by patients, and clinicians often focus more on patients’ symptom burden and physical limitations than psychologic well-being [56]. Physical health was primarily evaluated via instruments, which is likely due to the greater number of standardised instruments designed to assess physical health than any other QoL domain.
ACP was also frequently assessed. However, the majority of assessments were conducted by retrospective chart review rather than by use of a specific instrument. In order to consistently assess patients’ goals, which govern care and treatment options, it is critical to develop standardised, validated instruments to assess adequate ACP rather than relying on inconsistent provider-specific documentation. No such tool currently exists.
After physical health and ACP, mental health was most frequently evaluated. Many patients with chronic diseases, such as ILD, experience mental distress; however, symptom assessment can be challenging [56]. Anxiety and depression were commonly assessed. Social health was the least frequently evaluated domain. Patients’ psychosocial issues may emotionally burden clinicians [56] and are often overlooked, despite a recent systematic review demonstrating the critical role of social health from patients’ perspectives [57].
Despite the variety of existing outcomes measures, most studies used generic outcome measures not specifically designed for or validated in ILD to evaluate multiple domains. On the contrary, there are widely used and validated outcome measures that are not being consistently used in palliative care research in patients with ILD. Of the two instruments validated in ILD (KB-ILD and SGRQ), both assess the general HRQoL domain. Some articles introduced new unvalidated instruments. Ideally, validated instruments specific to the domain(s) under investigation should be selected. For instruments that measure several domains, clear delineation of section scores by domain would allow for independent assessment of more than one domain per instrument. Likewise, established instruments are preferred to the development of new instruments [10] in order to minimise collection burden [58]. It may be adventitious to validate established outcome measures in the ILD population.
Currently, there are no standardised outcome measures to assess palliative care interventions. In order to conduct urgently needed research on the impact of palliative care interventions in ILD, we must establish a core set of outcome measures. This would include consensus agreement upon our five proposed domains, validation of more instruments in the ILD population and concordance on which domains may be assessed with each instrument. We hope this scoping review serves as a first step in accomplishing these goals.
There are strengths and limitations to this scoping review. To our knowledge, this is the only review of palliative care outcome measures in ILD patients. Our initial search was broad, including all studies evaluating palliative care interventions in advanced lung disease. We subsequently selected studies specific to the ILD population that included quantitative data. Although every effort was made to include all eligible studies, it remains possible that eligible studies may have been unintentionally omitted. For example, although palliative care encompasses end-of-life care and symptom management, we did not specifically search for “end-of-life care” or particular interventions focused on symptom control. Additionally, since this study only evaluated quantitative data, valuable qualitative data that may be helpful in understanding patient-centred outcome measures were not included. Furthermore, although the caregiver burden is high and the development of caregiver outcome measures is important, this review focused on patient outcome measures. It should also be acknowledged that only articles available in English were included, which primarily biases towards studies conducted in the UK and the USA. Palliative care is dedicated to the provision of culturally sensitive care and we may have omitted studies from other countries that might explore interventions or outcomes of particular importance in other cultures.
Conclusion
This scoping review of 23 articles demonstrates the current diversity of palliative care outcome measures in ILD. We identified 25 distinct instruments and categorised them under one or more of five domains. We introduce a new domain of ACP given the critical need for standardised assessment of patients’ goals and preferences. The majority of these instruments are generic outcome measures, assessing more than one domain. This review can serve as a reference for researchers seeking guidance in choosing outcome measures of palliative care interventions in the ILD population. It is important to recognise that the heterogeneity of outcome measures creates challenges in selecting appropriate outcome measures and comparing results across studies. Development and utilisation of a core set of outcome measures is urgently needed to advance research aimed at improving patient-centred outcomes and quality of care for patients with ILD.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Online supplement ERR-0080-2021_Online_supplement
Acknowledgements
We thank clinical informationist Carrie Price, MLS (Johns Hopkins University, Baltimore, MD, USA) for her assistance with our literature search.
Footnotes
Provenance: Submitted article, peer reviewed
This article has supplementary material available from err.ersjournals.com
Conflict of interest: R.A. Gersten has nothing to disclose.
Conflict of interest: A.C. Moale has nothing to disclose.
Conflict of interest: B. Seth has nothing to disclose.
Conflict of interest: J.B. Vick has nothing to disclose.
Conflict of interest: H. Brown has nothing to disclose.
Conflict of interest: M.N. Eakin has nothing to disclose.
Conflict of interest: S.C. Mathai has nothing to disclose.
Conflict of interest: S.K. Danoff has nothing to disclose.
- Received March 19, 2021.
- Accepted May 26, 2021.
- Copyright ©The authors 2021
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org
References
