Background

The scope of palliative care (PC) has broadened in the last decade to include care of patients with a life-limiting disease, such as chronic obstructive pulmonary disease (COPD) [1,2,3]. PC is recommended to be initiated early in the disease trajectory [4]; however, as there is no accepted method for predicting prognosis or defining end-stage COPD [5], it is difficult to initiate PC [6]. Consequently, PC should instead be initiated on the basis of refractory symptoms, patients’ preferences, physical, psychological, social or spiritual/existential unmet needs [7, 8]. According to a recent task force report from the European Respiratory Society, PC in patients with COPD should start when an unmet need arises [8]. PC can be provided to these patients alongside disease-modifying therapies [9].

Dyspnoea is the most frequent and burdensome symptom in patients with severe COPD and may impact patients’ ability to perform daily activities, such as dressing, walking, and eating [10,11,12]. Experiencing dyspnoea often leads to anxiety and panic as well as concern about death and dying [10, 13].

Non-invasive ventilation (NIV) is the standard treatment for patients admitted to hospitals with COPD exacerbation and acute respiratory failure [14]. NIV may be appropriate for managing severe dyspnoea by improving ventilation, oxygenation, and the resistive load on the ventilatory muscles, all of which reduce the work of breathing [14,15,16]. In recent decades, a substantial increase in the use of NIV has been found among older hospitalised patients with terminal respiratory illness, suggesting a major shift in the way healthcare professionals (HCPs) provide ventilatory support in the palliative phase [17]. NIV seems to be used in around 33% of patients with COPD with poor life expectancy [18]. Home-based NIV is frequently used in patients with COPD and chronic hypercapnic respiratory failure [16]. Physicians involved in home-based NIV have reported that the main expected benefits of NIV treatment for patients with COPD are alleviation of dyspnoea, improvement in quality of life (QOL), and a reduction in the number of hospitalisations [19]. Alleviation of dyspnoea is considered important when NIV is used as life support for patients who have decided to forgo invasive mechanical ventilation (IMV) or as a palliative measure [20]. However, the use of NIV as a palliative measure is controversial. NIV could improve QOL and provide comfort, however, NIV could also be a futile treatment that may prolong the dying process without improving QOL [21, 22]. Concerns have been raised by HCPs as to whether patients and relatives fully understand the goals of care when NIV is used as a PC measure [23]. In this paper, patients with severe or very severe COPD undergoing NIV treatment are understood as patients with PC needs.

Previous reviews and meta-analyses have explored the impact of home-based NIV on clinical outcomes in patients with COPD [24, 25], including its effect on mortality in acute settings [26], and addressing dyspnoea during acute exacerbations, and respiratory failure [27].

One systematic review and meta-analysis investigated NIV use across various diagnosis, encompassing acute respiratory failure, with do-not-intubate or comfort-measures-only orders [28]. An integrated review found that combining NIV with long-term oxygen treatment (LTOT) may alleviate hypercapnia and dyspnea in patients with COPD [29]. However, most studies on NIV have focused on avoiding intubation in end-stage COPD rather than providing symptom relief [29]. Another systematic review on NIV in PC suggested potential improvements in QOL and dyspnoea for end-stage lung disease patients [30].

Previous reviews of the use of NIV in patients with PC needs have not solely focused on patients with COPD and have included only quantitative studies. In our initial search of the literature, we did not find any scoping review that had mapped studies on the use of NIV in patients with COPD with PC needs. Conducting a scoping review that includes various research methods and study designs, as well as the experiences and perspectives of patients, relatives, and HCPs, is important to achieve a more comprehensive understanding [31]. Such a review could inform clinical practice by summarizing exiting evidence regarding the use of NIV to these patients. Furthermore, our review may identify research gaps that are important to develop interventions to improve patients’ compliance to NIV, improve the effectiveness of NIV, and to prioritize future research in this field. This scoping review aimed to systematically map published studies on the use of NIV to patients with COPD with PC needs. Our research question was: What is known about the use of NIV from the perspectives and experiences of patients with COPD with PC needs, their relatives, and HCPs?

Methods

This scoping review employed the methodological framework described by Arksey and O’Malley [31]. We performed the optional stage consultation exercise to make the findings more useful and relevant for clinical practice [31]. The reporting of the review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension for Scoping Reviews checklist [32]. Deviations from the published protocol [33] are described in Additional file 1.

Eligibility criteria

We used the Sample, Phenomenon of Interest, Design, Evaluation, and Research Type (SPIDER) framework [34] to describe the eligibility criteria (Table 1).

Table 1 Eligibility criteria using the SPIDER framework
Full size table

Information sources

We searched AMED, CINAHL, Embase, MEDLINE, PEDro, and PsycInfo from inception to February 1, 2020 to identify relevant studies published in peer-reviewed journals. The database search was updated on November 14, 2022. We set no limit on the year of publication, as we wanted to describe the entire range of published studies relevant for our research question [35].

Search

The fourth author, an experienced librarian, and the first and last author built a comprehensive and systematic search strategy in MEDLINE using medical subject headings and text words and the search strategy was discussed with the other authors and tested. The final search strategy in MEDLINE was adapted for the other databases, which were peer-reviewed by a second librarian using the Peer Review of Electronic Search Strategies checklist [36]. The search strategy for all the databases is described in Additional file 2. The database search was limited to papers published in Danish, English, German, Norwegian, Spanish, and Swedish since we speak these languages fluently. We applied publication type filters to exclude editorials, letters, comments, and conference abstracts according to the functionality of each database to ensure that only studies published in peer-reviewed journals were included.

We performed manual searches to screen the reference lists of the included papers.

Selection of the sources of evidence

The fourth author transferred the search results to EndNote for duplicate removal and then transferred the search results to Covidence (covidence.org), a web-based systematic review application, to facilitate storage and blinding of the study selection process. Covidence ensured that two authors independently assessed whether titles and abstracts and then full-text papers met the inclusion criteria (Table 1). In cases of disagreement, two authors (SAS, MHL) performed an independent assessment, and a final decision was based on consensus between these two authors.

Data charting

We developed a standardised data-charting form in Covidence using the SPIDER framework. The data-charting form was piloted by two pairs of authors who independently extracted data from five of the included papers. After the pilot we removed research type (R) from the data-charting form as it overlapped with design (D). The final data-charting form included the following information: author, year of publication, country, setting, sample; aim; design; findings. One author extracted data, while another checked the accuracy of the extracted data against the papers. In cases of disagreement, a third author (SAS, MHL) independently extracted data and made the final decision.

Synthesis of results

We inductively summarised and organised the data thematically [31]. First, we extracted NIV-related findings from the results section of the included studies, and these data were read several times to identify patterns of similarities and differences in perspectives and experiences regarding the use of NIV to patients with COPD with PC needs across the included studies. We sorted text related to the identified patterns into thematic groups. The first and last author analysed the data, while all the authors discussed the emerging patterns and agreed upon the final thematic grouping [37,38,39].

Consultation exercise

To enhance the review findings’ relevance for clinical practice, we conducted a consultation exercise [31]. We recruited respiratory nurses and critical care nurses to an advisory board through information posted on the university college’s web page and on social media (i.e., Facebook and Instagram) as well as asking key persons in the respiratory nursing and critical care nursing community to share information about the consultation exercise in their networks. Five nurses contacted the first author and wanted to participate. The characteristics of the participants are described in Table 2.

Table 2 Description of the participants
Full size table

We facilitated four one-hour workshops using videoconferencing (Zoom); three workshops consisted of one participant in each workshop, while two participants participated in one workshop. The first and second author attended and presented the findings related to the thematic groupings (one acted as moderator and the other as secretary). We used an interview guide to facilitate reflection and dialogue with the participants regarding which of the review findings were most important, relevance for clinical practice, whether the findings were surprising or expected, findings that were missing, and what future research regarding NIV in PC by patients with COPD should address. The audio recording from the videoconference was transcribed verbatim by an external transcriber.

Ethical considerations

We attained approval from the Norwegian Centre for Research Data (reference number 480222) before we conducted the workshop with the advisory board. The information highlighted that participation was voluntary and that anonymity and confidentiality would be safeguarded. The participants signed informed written consent.

Results

Study selection

The database searches yielded 1193 publications. After 346 duplicates were removed, titles and abstracts for 847 publications were screened. Based on the inclusion and exclusion criteria, the full text of 120 publications was read, and 23 studies were included. The manual search identified nine studies from 10 papers. The reasons for the exclusion of full-text papers are presented in Fig. 1.

Fig. 1
figure 1

PRISMA flow diagram of the search and screening process

Full size image

Study characteristics

In total, 33 papers from 32 studies were included. The studies were published between 1992 and 2022, and nine (32%) were published in 2018 or later [40,41,42,43,44,45]. In total, 17 countries were represented. Six studies were from the United Kingdom [46,47,48,49,50,51], five from Germany [43, 52,53,54,55], three from Norway [44, 56, 57], and two from Austria [41, 58], Italy [59, 60], Taiwan [40, 45], and New Zealand [61, 62], respectively. Furthermore, the following countries were represented with one study each: Australia [63], Belgium [64], Denmark [65, 66] (two papers from the same qualitative study), The Netherlands [67], both the US and Canada [68], Canada [42], Iran [69], Spain [70], France [71], and Brazil [72].

Eight studies had a qualitative design [44, 49, 56, 57, 61, 62, 65, 66], six were randomised controlled trials (RCTs) [43, 58, 63, 70,71,72], four had a cross-sectional design [41, 42, 48, 68], five were retrospective cohort studies [40, 45, 53, 54, 67], and four cohort studies had a prospective design [47, 51, 52, 59]. In addition, there was one pre-post study [69], one with standardised structured interviews [46], and one follow-up study over two years (pre/post) [50].

A total of 166230 patients with COPD participated in the studies, ranging from 10 to 151912 participants, of whom 81642 (49.1%) were men. Thirty-one studies reported mean age, which was 65.2 years across the studies, ranging from 43 to 83 years.

Five of the studies [48, 49, 57, 65, 66, 68] included 604 HCPs, and two studies included four relatives [49, 65, 66]. The majority of HCPs were different types of physicians (n = 131), physicians in training (n = 12), general practitioners (n = 18), and specialists (n = 132). Other HCPs represented in the studies were respiratory therapists (RT) (n = 290), nurses (n = 33), physiotherapists (n = 2), psychologists (n = 2), and health care workers (n = 1), and one study did not report occupation.

The majority of the studies were conducted in a hospital setting (n = 20), including one outpatient setting [46]; eight studies were home-based [50,51,52,53, 55, 58, 60, 69], and two were a combination of home-based and hospital [45, 72]. The other three were from various settings (i.e., pulmonary rehabilitation [43], community centre [68], and population-based [42]).

NIV was used in different situations applicable for patients with COPD. Three studies tested nocturnal NIV plus LTOT [63, 70], one study tested nocturnal NIV [50], and two studies tested NIV during exercise [43, 72]. Other studies investigated NIV as a factor related to cardiopulmonary resuscitation or a do-not-resuscitate order [40, 47, 48]. Additional studies focused on special diseases, such as mustard airway disease [69], or on acute ventilatory support [46]. The characteristics of the included studies are shown in Table 3.

Table 3 Characteristics of the included studies
Full size table

Outcome measures used in the studies varied according to study aim and design (Table 4). All RCT and pre-post studies (n = 9) used lung and respiratory muscle function tests to measure the effect of NIV. All nine studies included spirometry; other tests varied (i.e., maximum inspiratory pressure, diaphragmatic activity). Six of the nine studies measured dyspnoea and exercise tolerance (for example, by using the 6-min walking test or Borg Dyspnoea Scales) [43, 58, 60, 69, 70, 72]. Arterial blood gases, lactate, or haematocrit were measured in eight studies [50, 58, 60, 63, 69,70,71,72], while sleep quality was measured in four studies, mostly using polysomnography [58, 63, 69, 70]. Hospital admittance was measured in five studies [50, 58, 63, 69, 70]. Seven of these nine studies used patient-reported outcome measures (PROMS), such as health-related quality of life (HRQOL) (n = 3) [63, 69, 72] and different measures of satisfaction, mood, or well-being (n = 4) [50, 60, 63, 71]. In addition, these studies applied a variety of anthropometric measurements, such as BMI, blood pressure, and SpO2. Four of the ten studies with a cross-sectional or survey design used only a few outcome measures: three included HRQOL PROMS [47, 48, 59], and one measured depression [59]. Four studies included lung and respiratory muscle function tests [40, 53, 54, 67]. Four cross-sectional studies measured different types of medical utilization, such as readmission and the number of emergency room visits [40, 42, 54, 67].. In addition, studies mapped treatment preferences, opioid use, and PC use. In addition, studies mapped treatment preferences, opioid-use, and PC use. The two cross-sectional studies, including HCPs [41, 68], both mapped attitudes, considerations, and decision-making regarding NIV treatment.

Table 4 Overview of the outcome measures in the included studies
Full size table

In the five cohort studies, two studies measured lung and respiratory muscle function using body plethysmography [52, 55], another used spirometry [51], and the same three studies also measured blood gas. Three of the cohort studies measured different types of resource use, such as the numbers of out-patient visits, palliative care and number of hospitalizations [45, 52, 64]. One study measured SaO2 and BMI [51].

The findings are presented in four thematic groupings: 1) preferences and attitudes towards the use of NIV, 2) patient participation in the decision-making process of NIV treatment, 3) conflicting results on perceived the benefits and burdens of NIV treatment, and 4) heterogenous clinical outcomes in experimental studies.

Preferences and attitudes towards the use of NIV

Seventeen studies reported patients’ and HCPs’ preferences and attitudes towards the use of NIV in hospital or at home [40,41,42, 44, 45, 47,48,49,50, 59,60,61,62, 64,65,66,67,68].

When a scenario-based approach was used to elicit patients’ end-of-life preferences and preferences regarding artificial ventilation and resuscitation, 40–50% of the patients chose NIV [48, 59]. Patients perceived the NIV mask as a ‘life buoy’, a symbol of survival and hope that could help them through a respiratory crisis [44, 62, 65] and as helpful even when there was no chance of cure [44]. Patients interviewed in a hospital setting believed that NIV treatment in the intensive care unit (ICU) or the use of home-based NIV was necessary to manage their symptoms, keep them alive, and prolong their life [44, 49, 61]. In one study, 75% of the home-based patients that used NIV over two years reported that they were very satisfied with the treatment [50]. An RCT found that a brief psychological intervention significantly improved patients’ acceptance of and adherence to NIV [60]. Hospital-based patients who received NIV treatment more than once felt that familiarity with the treatment made it easier to cooperate in acute situations [65]. Patients who experienced the NIV treatment as awful changed their attitudes towards it after hospital discharge and wanted NIV treatment again in the future [66]. A study that compared end-of-life care utilisation between patients with COPD and patients dying of other diseases, such as lung cancer and cardiovascular disease, found that patients with COPD were more likely to receive NIV in the last six months of life than the patient groups with lung cancer and cardiovascular disease [45, 64]. Another study found that in 2004, 1% of the patients with COPD admitted to the ICU needed NIV, while approximately 4% of the patients needed NIV in 2014 [42].

A study [41] investigating physicians’ treatment choices by using a case vignette featuring a geriatric patient with end-stage COPD and acute respiratory failure showed large differences between the choices made by the doctors. Twenty-three percent of the physicians chose NIV, while 46% chose a palliative approach without respiratory assistance. Pulmonologists and internists were more likely to choose NIV than palliative physicians or geriatricians. Increasing age of physicians was associated with not starting with NIV; no effect was found regarding educational level and years of professional experience [41]. A survey showed that the majority of RT and physicians were more likely to choose NIV for patients with COPD or cardiogenic pulmonary oedema with a do-not-resuscitate order near the end of life than for patients with underlying malignancy with a do-not-resuscitate order [68]. HCPs felt limited by the lack of clear evidence of measurable benefits of home-based NIV, such as hospital admissions and mortality, and patient selection for NIV treatment was carried out on a case-by-case basis [49]. HCPs discussed ethical challenges, such as patients’ resistance to using NIV or terminating the NIV treatment when patients were in the end-stage of COPD [65].

Patient participation in the decision-making process of NIV treatment

Ten studies reported patients’ and HCPs’ experiences of challenges regarding patient participation in the decision-making process of NIV treatment in hospital or at home [40, 41, 44, 46, 48, 49, 56, 57, 62, 65, 66].

Hospital-based patients who had used NIV described circumstances in which they experienced lack of control, felt vulnerable, recalled feelings of impending death, and were totally dependent on HCPs due to breathlessness, anxiety, and panic [44, 56, 62]. Moreover, patients felt that they were dependent on HCPs for survival [44]. Many wanted to take part in the decision-making process regarding NIV treatment [44, 48, 49, 56], and in one study most patients expressed that treatment such as NIV should be discussed with all patients [48]. Patients deemed that demonstration of NIV was more helpful than the use of a photograph of an NIV machine in use on a patient and increased patients’ willingness to use NIV [46].

Patients expressed that they had been included by HCPs in such decision-making to varying degrees [44, 48, 49, 56], and they often felt that they had no choice other than to use NIV because physicians recommended it or because their deteriorating health condition demanded it [44, 49]. Other patients expressed that the HCPs should make the treatment decisions as they believed that physicians were doing their best and acted in their best interest [44, 49, 56, 62, 65].

Although hospitalised patients handed over the decision-making to HCPs, they still interacted with nurses and participated actively in what was happening to them [56]. The medical records of patients who died in hospital showed that patients with an early do-not-resuscitate order used less IMV, used NIV more often, died less often in the ICU, and had a shorter length of stay in hospital than those with a late do-not-resuscitate order [40].

HCPs pointed out the importance of using time to communicate the aim of home-based NIV to the patients to facilitate the patients’ choices and enhance adherence. In this way, the HCPs underpinned that the decision to use home-based NIV ultimately lay with the patients [49]. Responding to a case vignette, around 40% of the physicians commented that they would have wanted to know if the patient had any sort of precautionary directive [41]. Nurses perceived that they were part of a treatment culture that focused more on medical treatment than creating optimal end-of-life care and that the only measures offered to support patients’ capacity to breath were NIV or IMV. According to these nurses, the interdisciplinary meetings were short and mostly involved information about patients’ objective symptoms regarding their lungs and laboratory tests. In this context, there was no room to discuss patients’ participation in treatment decisions, undertake care planning, or pay attention to the patients’ QOL, suffering, or functional status [57].

Conflicting and divergent results on the perceived benefits and burdens of NIV treatment

Fifteen studies reported on conflicting and divergent results on the perceived benefits and burdens in NIV treatment in hospital and at home from the perspective of patients, relatives, and HCPs [43, 44, 49,50,51, 56, 58, 62, 63, 65, 66, 68,69,70,71,72] (the PROMS are described in Table 4).

Patients spoke of the benefits of NIV treatment, such as easing chest pressure [56], and experienced the spontaneous breathing mode as more comfortable than the assist-control ventilation mode [71]. Patients reported that they complied well with nocturnal NIV plus LTOT [63]. Several burdens were described for NIV treatment, such as the NIV mask being a bother, unpleasant, and tiresome as it was often too tight, caused facial sores, dryness in the mouth, and claustrophobia with feelings of suffocation and made it difficult to communicate [44, 49, 50, 62, 65, 66]. Patients were afraid of being alone with the ventilator and underlined the importance of knowing that they were not alone [65, 66]. A few patients did not tolerate the NIV treatment within the first weeks and expressed that the experienced pressure of the NIV was too high [70]. Some patients withdrew from a study as they were not able to sleep using NIV [51].

Studies reported conflicting results on whether the use of NIV improved HRQOL and dyspnoea. Patients using home-based NIV reported improved QOL and well-being, reduction of dyspnoea, and having more energy to carry out activities in everyday life. Some expressed that they got their old life back while using home-based NIV [49]. The relatives had similar thoughts [49]. In a pre-post study, home-based NIV improved HRQOL and the symptom and total St George Respiratory Questionnaire (SGRQ) scores were improved [69]. An RCT that consisted of a 6-week physical training program (treadmill walking) for patients using NIV or supplemental oxygen found that the HRQOL measured using the SGRQ improved in patients in both the NIV and oxygen group, while the total and activity SGRQ score significantly improved in only the NIV group [72]. An RCT that determined the effect of withdrawal of NIV in patients who remained hypercapnic after acute respiratory failure found no significant differences in HRQOL using the SGRQ three months after randomisation between the NIV group and those who stopped using NIV. It was not possible to analyse long-term effects on HRQOL due to a small remaining sample [58]. Furthermore, an RCT found that patients who used nocturnal NIV plus LTOT reported less vigour and more confusion and bewilderment (measured with SF 36 and the Profiles of Mood States) than the controls (LTOT) [63]. Another study found no change in QOL in patients using nocturnal home-based NIV [51]. Patients using nocturnal NIV plus LTOT and NIV or NIV only reported significantly improved sleep quality [50, 63, 69].

Additionally, in another RCT patients in the nocturnal NIV plus LTOT group reported significant improvement of dyspnoea on both the Medical Research Council scale and the Borg Scale by the third month compared with the control group (LTOT). By the sixth month, improvement of dyspnoea in favour of the nocturnal NIV was evident on the Borg Scale [70]. In a follow-up study stretching over two years, the patients using NIV reported no improvement in dyspnoea [50, 69]. Patients using NIV while participating in physical training reported greater improvement than the supplemental oxygen group regarding dyspnoea [72] and showed a lower increase in exertional dyspnoea in a cross-over trial [43].

In a survey involving physicians, more than 50% reported that for patients with a do-not-resuscitate order, NIV added to the alleviation of dyspnoea provided by anxiolytics and analgesics and that NIV facilitated communication with HCPs and relatives. Regarding patients with a comfort-measures-only order, a minority of RT and physicians agreed that NIV provided either of these two benefits [68]. In another study, HCPs reported that communicating with patients using NIV was difficult due to the tight mask and the patients’ condition [65].

Results from a qualitative study reported that HCPs struggled to balance the potential of NIV to reduce QOL with its uncertain clinical effectiveness [49].

Heterogenous clinical outcomes in experimental studies

Eleven studies reported on heterogenous clinical outcomes in experimental studies regarding the use of NIV [43, 50, 52,53,54,55, 58, 63, 69, 70, 72] (the clinical outcomes are shown in Table 4).

A randomised cross-over trial found that patients using NIV plus oxygen showed a 39% increase in cycle endurance time and reduced exercise-induced hypercapnia compared with the control group (oxygen only) [43]. In another RCT, after a physical training programme (treadmill walking) the patients in the NIV group significantly increased their respiratory muscle strength, peripheral muscle strength, and 6-min walking distance and showed decreased leg fatigue and reduced lactate/speed ratio compared with controls (oxygen) [72].

An RCT found that patients who were hypercapnic after acute respiratory failure who needed NIV and stopped using it had a higher chance of clinical worsening (escalation of IMV) than those who continued to use NIV. The 6-min walking distance increased in the NIV group, while it decreased in those who stopped using NIV [58]. A pre–post study found that the number of hospital admissions and exacerbations was significantly reduced in patients with severe mustard airway disease using NIV [69].

In a follow-up study of patients who used nocturnal home-based NIV, the use of general practitioners and hospitalisations halved after one year compared with the year before NIV. The median survival was 920 days, and no patients died during the first 500 days [50]. Studies also showed that nocturnal NIV significantly reduced paCO2 [50,51,52,53,54]. A retrospective study found that hypercapnia improved significantly during nocturnal NIV in hospitalised stabile patents with hypercapnic COPD using respiratory pressures with a mean of 28 cm H2O [54]. An RCT showed improved survival for patients using nocturnal NIV plus LTOT compared with controls and found improved sleep-related hypercapnia in favour of the NIV groups compared with controls [63].

Another RCT that evaluated the one-year efficacy of NIV plus LTOT found that the one-year survival and number of acute COPD exacerbations were similar in the NIV plus oxygen group and the control group (97% used LTOT). The number of hospitalisations decreased at three months in the NIV plus LTOT group compared to the control group; however, this effect was not seen at six months or at one year [70]. A study found that the mortality rate in patients with hypercapnic respiratory failure using home-based NIV was 44.7% with one-year, two-year, and five-year survival rates of 84.0%, 65.3%, and 26.4%, respectively [52].

A study evaluated the prevalence of malnutrition (body mass index < 20 kg/m2) and the longitudinal changes in nutritional status in patients undergoing NIV. There was a significant weight gain among the malnourished patients after the implementation of NIV. The weight gain was not correlated with improvements in lung function or blood-gas values [55].

Consultation exercise

The nurses expressed great recognition of the presented findings and their relevance for clinical practice and supplemented the review findings with reflections regarding their own experiences of administering NIV treatment to these patients. These experiences concerned lack of information on how to use home-based NIV to enhance QOL; tailoring the use of NIV to prevent complications; professional competency regarding NIV as palliative treatment; and ethical challenges using NIV as a palliative measure. Sharing these preliminary findings with peers increased the relevance of our findings for future research and COPD practice. Additionally, we believe that the consultation exercise served as a mechanism to keep us informed about evolving trends in NIV treatment. It would have alerted us to any changing priorities within the field that might not have been covered by the included studies.

Discussion

Our findings suggest that patients perceived NIV as a ‘life buoy’ to keep them alive. Even though HCPs would choose and use NIV in patients with severe COPD, they may not perceive NIV as a palliative measure. Many patients wanted to take part in the decision-making process regarding NIV treatment but expressed being included to varying degrees by HCPs in such decision-making. Furthermore, there were conflicting findings regarding the perceived benefits and burdens of NIV treatment and diverse heterogeneous clinical outcomes in experimental studies.

Our review suggest that many patients seemed to be satisfied with NIV treatment and perceived it as a way to keep them alive, prolong their lives, and alleviate symptoms. This preference for and attitude towards NIV may be due to patients’ lack of knowledge regarding diagnosis, prognosis, and treatment as well as the absence of discussions with HCPs about PC [2, 73]. Patients often want to focus on staying alive rather than talk about end of life [74] and may be willing to undergo life-saving treatment despite discomforts with the expectation that they will recover [75, 76]. Many patients may not view NIV as a burdensome intervention [76]. Furthermore, dyspnoea is a frequent and tiring symptom that is challenging to alleviate [10, 11], and patients may have experienced that NIV treatment provided adequate alleviation of this demanding symptom.

Our findings were inconclusive as to whether HCPs perceived NIV as a palliative measure. Traditionally, PC for patients with COPD has been associated with care of the dying, and the prognostic uncertainty makes it challenging for HCPs to identify when PC should be introduced [9]. The literature suggests that NIV can also be used when intubation is not an option to prolong life or to alleviate dyspnoea and provide comfort measures only [20]. However, NIV at the end of life is controversial and has been portrayed both as a measure to improve QOL and dyspnoea and provide comfort and as a futile treatment that may prolong the dying process and increase suffering [21, 22]. A guideline suggests that NIV could be used in a palliative setting if it does not have negative consequences, such as mask discomfort or prolonging the dying process. HCP training is a prerequisite for NIV to be used as a palliative measure [77]. Furthermore, careful patient selection and setting goals in advance for such treatment, a timeframe for revaluations, and criteria for success or failure are important [77, 78].

Our findings show that patient participation in the decision-making process regarding NIV may be challenging: some patients want to participate, while others would prefer physicians to act in their best interest. Autonomy and patient participation are significant to enable patients to live a good life based on their needs and preferences [79]. Our findings describe that patients felt they were included to varying degrees in the decision-making process or that they had no choice other than to use NIV since HCPs recommended it or due to their deteriorating health [80]. Patients may not consider future treatment, be reluctant to make binding decisions regarding future care [81], or feel that such discussions are pointless as the future seems beyond their control due to their disease [75]. The timing of such discussions needs to be tailored, but they should be carried out early in the disease trajectory when patients are in a stable phase of the disease by a physician who has a trusting relationship with them [82]. However, HCPs may be reluctant to discuss PC and future treatment since they mainly focus on curative treatments or may be concerned about destroying patients’ hope [2, 9].

Our findings show that patients experience burdens regarding NIV treatment, such as the negative consequences of wearing a mask, and that HCPs were concerned that NIV could reduce patient QOL. In a hospital-based study, patients at all stages of COPD and acute respiratory failure undergoing NIV treatment described ambivalent feelings regarding NIV due to concerns over whether the struggle and discomfort experienced when using the mask and ventilation were worthwhile. For some patients, the discomfort caused by the mask and ventilation was perceived as minor compared with the feeling of dyspnoea [83]. Dyspnoea and HRQOL are recognised as among the most important patient-centred outcomes in patients with COPD [84]. However, our findings showed conflicting results as to whether NIV improved these outcomes. This could be due to the heterogeneity of these studies, such as different study designs, outcome measures, data collection points, and duration of NIV. In contrast to a randomised feasibility study that suggested that NIV was more effective than oxygen in reducing dyspnoea in patients with cancer (life expectancy less than six months) [85], none of the studies in our review were designed to assess the acceptability of NIV as the only palliative measure.

Some of the studies in our review suggested that NIV may reduce the number of hospitalisations and COPD exacerbation as well as increase exercise capacity. Many patients want to spend as much time as possible at home and even want to die there [86] as patients may feel more comfortable and in control at home than in the hospital [87]. However, patients with COPD are more likely to die in hospital or nursing homes than patients with lung cancer [2, 86]. Relatives reported that during the last year of life patients with COPD were afraid of being left alone and of dying due to dyspnoea, anxiety, and panic [88]. Such fear may contribute to hospitalisation. Furthermore, exercise capacity is recognised as another important outcome in patients with COPD [84]. By alleviating dyspnoea and increasing exercise capacity, home-based NIV may enable patients to spend more time at home and participate in a few meaningful activities. Moreover, the use of telehealth may support and enable these patients to spend more time at home. A systematic mixed-studies review suggests that home-based patients with PC needs experience telehealth as a potential support system that could enable them to remain at home and that self-reporting could provide HCPs with information about patients’ symptoms and circumstances that can be used to tailor care [89].

Based on the findings from this review, we recommend that future studies should be designed to examine the acceptability, experiences with, and effectiveness of the use of NIV as a palliative measure to alleviate dyspnoea and improve QOL and other PROMS. As diverse patient-centred outcomes and clinical outcomes are reported, future studies should determine which are most important. Studies should also explore how palliative home-based NIV treatment and dependence on masks impact patients’ and relatives’ daily life and QOL. Due to challenges to including patients in the decision-making process of NIV treatment, patients, relatives, and HCPs should be included in the co-design of interventions to enhance patient health literacy and participation. The relatives’ voice was only included in two studies, and their experiences of providing informal care at home and experience with NIV at end of life should be investigated. Studies that examined HCPs’ perspectives and experiences with the use of NIV were mostly limited to physicians. However, nurses administer NIV treatment and observe the effects and adverse events of the treatment. Consequently, studies on nurse training and competence in NIV and studies with a multidisciplinary approach are warranted. We emphasize that by doing the consultation exercise including respiratory and critical care nurses contributed to nuance and validate the review findings, and also highlighted that this topic is highly relevant within nursing care.

Limitations

Synonyms of both PC and NIV may exist that we were unable to identify and include in our database search strategy. In addition, as our review had some language limitations, there may be studies that we were unable to identify. Another limitation may be that in the included studies the patients may not have been regarded as having PC needs and NIV treatment may have had a curative or restorative intention rather than a palliative intention.

We chose to include nurses from the hospital setting in the consultation exercise as they care for patients using NIV around the clock, and their voice was included in very few studies. Nurses have extensive experience and competence regarding administering NIV to patients with COPD. However, a limitation may be that we did not include nurses from homecare setting, physicians as they have the ultimate responsibility for medical decision-making or physiotherapist or respiratory therapist as they also have a crucial role in NIV to these patients. Including multidisciplinary HCPs, as well as patients and relatives could have provided other type of feedback regarding the clinical relevance of our findings. The sample size of the consultation exercise may be a limitation. However, the participants had extensive experience regarding the use of NIV to these patients and provided rich input regarding the review findings.

We conducted the scoping review in line with an acknowledged methodological framework and the PRISMA extension for Scoping Reviews checklist and did not appraise the methodological quality of the included studies and synthesise the data. A potential source for heterogeneity in our scoping review is studies included from different countries as there are differences in cultural backgrounds, healthcare services, resources available, healthcare guidelines, policies and approaches. Furthermore, we also included older studies as we wanted to describe the entire range of studies relevant for our research question. Thus, our findings and how the findings can inform clinical practice (e.g. the effectiveness of NIV) need to be interpreted with caution.

Conclusion

Patients with COPD with PC needs seem to view NIV as a treatment that can keep them alive rather than as a palliative measure. HCPs use NIV in these patients but may not consider it a palliative measure, suggesting that NIV may be introduced with a curative or restorative intention rather than a palliative one. The included studies mostly mirror the physicians’ views on NIV as a treatment option and its physiological outcomes. There is a knowledge gap related to competence development in the palliative use of NIV and to the experience of other stakeholders, especially homecare nurses. Future studies should be designed to explore and evaluate the use of palliative NIV treatment more clearly. As some patients may be reluctant to participate, or may not be included, in the decision-making process regarding NIV, measures for enhanced communication and information about NIV and the inclusion of patients in such processes early in the disease trajectory are warranted. The included studies reported the conflicting benefits and burdens of NIV treatment and heterogeneous clinical outcomes. Consequently, future studies should investigate which outcomes are the most important, effective, and relevant for optimal treatment when NIV is used in patients with COPD with PC needs.