Table 3 Characteristics of the included studies

Aliannejad et al. (2015) Iran [69]

Home

20 patients with severe mustard airway; 20 males, mean age 43 (standard deviation (SD) 5) years, mean FEV₁% predicted 25 (SD 9)

To assess the efficacy of non-invasive ventilation (NIV) in subjects with severe mustard airway disease

One group pre-post study

NIV improved health-related quality of life (HRQOL) in very severe patients. NIV reduced hospitalizations and exacerbations in severe patients. NIV did not improve exercise tolerance, pulmonary function, or dyspnoea

Beckert et al. (2020) New Zealand [62]

Hospital

15 patients with severe or very severe COPD; 9 males, mean age 69.2 (SD 8.2, range 55–89) years, mean FEV₁% predicted 26.4

To investigate the experiences and perceptions of participants using NIV

Qualitative approach using Grounded theory with inductive coding

NIV was uncomfortable and affected patients’ cognition. It was also considered as a life saver and a concern for others. Patients considered NIV as a viable option for future treatment and described a high level of trust in healthcare professionals and delegated decision-making to them regarding ongoing care

Borghi-Silva et al. (2010) Brazil [72]

Home; university physiotherapy department

14 patients with severe COPD used NIV; 9 males, mean age 68 (SD 9) years, mean FEV₁% predicted 34 (SD 10) 14 patients with severe COPD used supplemental oxygen; 9 males, age 67 (SD 7) years, mean FEV₁% predicted 33 (SD 7)

To investigate whether NIV alone could promote a true physiological training effect after training that is greater than that of oxygen supplementation

Randomized controlled trial (RCT)

There were significant differences between the NIV and supplemental-oxygen groups in lactate/speed ratio (33% vs -4%), maximum inspiratory pressure (80% vs 23%), 6-min walk distance (122 m vs 47 m), and leg fatigue (25% vs 11%). Changes in SpO₂/speed, VO₂, and dyspnoea were greater with NIV than with supplemental oxygen. HRQOL symptoms and disease impact were significant lowered in both groups. Activity and total St. George Respiratory Questionnaire scores were significantly reduced only in the NIV group

Budweiser et al. (2005) Germany [53]

Home

46 patients with stable COPD undergoing NIV treatment; 38 male, median age 65.2 (range 53.1–77.9) years, median FEV₁% predicted 29 (SD 8.2)

To evaluate the impact on lung deflation of patients receiving long-term home ventilation, by performing a retrospective analysis of different lung function parameters including inspiratory capacity and respiratory muscle function in a collective of severe symptomatic COPD patients in a stable status of their disease

Retrospective explorative study

One-year survival was 89.1%. There was a significant reduction in nocturnal and daytime PaCO2, a decrease in the ratio of residual volume to total lung capacity at 6 and 12 months. Significant improvements in inspiratory capacity, vital capacity and FEV₁ were found. For patients with the most severe hyperinflation a significant positive correlation between inspiratory positive airway pressure and reductions in PaCO2 and residual volume/total lung capacity were found

Budweiser et al. (2006) Germany [55]

Home

141 patients in stable state with severe COPD (at baseline); 106 males, median age was 65 (SD 8.4, range 41.7–80.0) years, FEV₁% predicted 29.7 (SD 9.1%)

To investigate whether initiation of NIV results in an alteration in body weight, particularly in malnourished patients, up to 12 months after initiation of treatment, and whether there is a link to changes in functional variables

Cohort study

Malnutrition (BMI of < 20 kg/m2) was found in 21% of the patients. BMI was significantly correlated with the severity of respiratory impairment, especially with hyperinflation. In malnourished patients there was a significant increase in body weight after 6 months and 12 months, while no significant changes in the overall study population. There was no correlation between changes in BMI and changes in blood-gas values, lung function, or inspiratory muscle function, either in the entire patient group or in the subgroup of malnourished patients

Budweiser et al. (2007) Germany [52]

Home

188 patients with COPD; 147 males, mean age 64.5 (SD 8.0) years, median FEV₁% predicted 30.0 (SD 9.6)

To focus on predictors of mortality in patients with chronic hypercapnic COPD receiving NIV

Cohort study

The mortality rate during follow-up was 44.7%, with 1-year, 2-year and 5-years survival rates of 84.0%, 65.3% and 26.4%. Death occurs mainly from respiratory causes (73.8%)

Carlucci et al. (2016) Italy [59]

Hospital; respiratory units; 2 rehabilitation centres and 1 respiratory critical care unit

43 patients with very severe COPD; 84% males, median age 72 (interquartile range [IQR] 65, 78) years, 44% use NIV, mean FEV₁% predicted NR, inclusion criteria: FEV₁% predicted < 30%

To assess the patients’ preferences regarding end-of-life sustaining interventions, the patient’s ‘comprehension and retention’ of their choices and if at the time of death the patients’ decision was respected

Prospective multicentre study

The choice of NIV a ‘ceiling’ treatment was associated with a current use of NIV and a recent family bereavement. A minority of subjects missed the meaning of ceiling NIV (19%). The wish of patients was respected in about half the patients: all these patients died under mechanical ventilation or NIV. Few relatives reported that patients’ preference changed

Casanova et al. (2000) Spain [70]

Hospital; pulmonary clinics

20 patients with severe COPD in NIV group; 20 males, mean age 64 (SD 65) years, mean FEV₁% predicted NR 24 patients with severe COPD in control group; 23 males, mean age 68 (SD 4) years, mean FEV₁% predicted NR

To determine the 1-year efficacy of NIV added to long-term oxygen therapy in patients with stable severe COPD

RCT

One-year survival and the number of acute exacerbations was similar in both groups. The only beneficial differences were observed in the Borg dyspnoea rating, which dropped from 6 to 5 in one of the neuropsychological tests (psychomotor coordination) for the NIV group at 6 months

Chakrabarti et al. (2009) United Kingdom [46]

Outpatient setting

50 patients with COPD patients; 34 males, age 69 (IRQ 14), median FEV₁% predicted 36

To understand the attitudes of patients with COPD toward acute ventilatory support and assess how aids to decision making regarding ventilation affect patients’ views of therapy

Standardized structured interview

86% found demonstration of NIV helpful in decision making compared to 24% with the photographic aid. 96% were willing to receive NIV after a verbal description of the technique. 76% consented when a photographic aid was shown. When NIV was demonstrated, willingness rose to 84%. Willingness to receive NIV was not significantly associated with gender, domiciliary oxygen use, prior participation in a pulmonary rehabilitation program, social status, whether currently smoking, MRC index, or WHO performance status

Christensen et al. (2017) Denmark [65]

Hospital

16 patients with severe COPD treated with NIV at least once during last 2 years; 6 males, mean age 69 (age range 47–86) years, mean FEV₁% 24 4 relatives7 HCPs; 3 nurses, 2 physicians, 1 healthcare worker, 1 NR (PhD student, principal researcher)

To investigate user perspectives on health care practice in the hospital concerning NIV treatment; to understand how patients with COPD and health professionals (HCPs) experience and evaluate treatment with NIV; to develop new management strategies for NIV treatment of patients with COPD based on patients’, relatives’ and HCPs’ perspectives on treatment

Qualitative approach using critical psychological practice research

15 patients evaluated treatment with NIV positively, 13 had experienced fear and 14 discomfort during treatment. The co-researcher group described HCPs’ perspectives and analyzed treatment practice based on data from patients’ perspectives developing new management strategies in clinical practice with NIV

Christensen et al. (2018) Denmark [66]

Hospital

16 patients with severe COPD treated with NIV at least once during last 2 years; 6 males, mean age 69 (age range 47–86) years, mean FEV₁% 24 4 relatives 7 HCPs; 3 nurses, 2 physicians, 1 healthcare worker, 1 NR (PhD student, principal researcher)

To clarify COPD patients’ perspectives on treatment with NIV ventilation and develop management strategies for the treatment based on these perspectives

Qualitative approach using critical psychological practice research

Patients regarded NIV treatment positively even though they experienced discomfort and anxiety. Patients conduct their everyday lives looking at COPD as a basic life condition rather than an illness. This approach had a major impact on patients’ attitudes to NIV treatment and hospitalization

Duenk et al. (2017) The Netherlands [67]

Hospital

33 patients with acute exacerbation of COPD; 19 males, mean age 72 (SD 10.4), mean FEV₁% predicted NR

To examine whether proactive indicators for palliative care are documented consistently in the medical records and explore the percentage of patients with a poor prognosis and prognostic value

Retrospective medical record review

NIV was always documented as 1 of 10 indicators for palliative care in the hospital setting

Elliott et al. (1992) England [51]

Home

12 patients with severe stable COPD and hypercapnic respiratory failure (HRF); 9 males, mean age 57.4 (SD 5.6) years, FEV₁% predicted NR

To evaluate the practicalities of nasal intermittent NIV at home in patients with COPD and the effect on sleep and quality of life

Prospective cohort

At 6 months 8 patients were continuing with NIV. At 6 months there was an increase in mean PaO2 of 11% and lower mean transcutaneous carbon dioxide tensions overnight compared with spontaneous breathing before the start of nasal NIV. Total sleep time and sleep efficiency changed during NIV by + 72, 5 min and + 5% respectively. Sleep architecture and the number of arousals were unchanged. QOL did not change but was no worse during NIV. At one year 7 patients were still using NIV and Paco2 and bicarbonate ion concentration during the day had improved further by comparison with the values at six months

Faes et al. (2018) Belgium [64]

Hospital

3872 patients died of COPD; 2597 males, mean age 78.8 (SD 10) years, FEV₁% predicted NR 19401 patients died with COPD (died of lung cancer or cardiovascular disease); mean age 76.6 (SD NR) years, FEV₁% predicted NR

To describe EOL resource use in people diagnosed with COPD in the last six months of life and compare this resource use between those dying of COPD, cardiovascular disease, and lung cancer

Retrospective, full- population analysis

Those who died of COPD (51%) were more likely to receive NIV than those who died of cardiovascular disease (22.5%) or lung cancer (37.9%) Those dying of COPD had significantly more days of NIV compared to the other two groups, with a mean of 60.2 days

Fahim & Kastelik (2014) United Kingdom [47]

Hospital

30 patients with COPD for at least 12 months duration; 18 males, mean age 70 (SD 8, age range 43–87) years, mean FEV₁% predicted 37 (SD 12.8)

To evaluate the COPD patients’ understanding of palliative care as a management option of COPD and to identify any barriers to resuscitation discussion in this group of patients

Prospective observational study

13 patients understood the term NIV. 11 of those would consider it again if needed

Fu et al. (2018) Taiwan [40]

Hospital; admitted for acute care

271 electronic health records from patients with terminal COPD; 249 males, median 83 (IQR 77–88), median FEV₁% predicted 60.5 (IQR 40–81.8)

To investigate factors associated with an early do-not-resuscitate (DNR) directive

Retrospective observational

Early DNR patients died less frequently in the intensive care unit, received less frequent invasive mechanical ventilation (IMV), more frequent non-invasive MV, and had a shorter length of hospital stay

Funk et al. (2011) Austria [58]

Home

13 patients with COPD in NIV group; 7 males, mean age 62 (SD 6) years, mean FEV₁% predicted 31 (SD 17) 13 patients with COPD in withdrawal group; 8 males, mean age 65 (SD 6) years, mean FEV₁% predicted 30 (SD 12)

To determine whether the withdrawal of longterm NIV causes clinical worsening in stable COPD patients who remained hypercapnic after an episode of acute respiratory failure requiring IMV

RCT

After randomisation the withdrawal group had a higher probability of clinical worsening compared to the ventilation group. After 12 months, ten patients in the withdrawal group, but only two patients in the NIV group, experienced a significant clinical worsening. 3 moths after randomization the 6-min walking distance increased in the NIV group and decreased in the withdrawal group

Gaber et al. (2004) United Kingdom [48]

Hospital

100 patients with COPD; 41 males, mean age 74.1 (age range 48–92) years, mean FEV₁% predicted NR, 44 patients had a FEV1 < 40%, 37 had a FEV1 between 40 and 59% predicted

To ascertain the views of patients with COPD in the community towards artificial ventilation and cardiopulmonary resuscitation (CPR) and whether this sensitive issue could be addressed by respiratory nurse specialists

Survey

48 patients wanted all additional treatments (NIV, IV, CPR) if needed and 12 wanted none. Nineteen patients said ‘no’ for CPR but ‘yes’ to NIV and IV. 10 patients said ‘no’ to CPR and IV but ‘yes’ to NIV. The remaining 11 patients gave other mixed answers. There were no significant differences between the “yes” and “no” group. 98% agreed that this sensitive issue should be discussed with all patients. 1 patient thought that it should be discussed only with seriously ill patients

Gale et al. (2015) UK [49]

Two hospitals

20 patients with COPD (either past or currently use of domiciliary NIV or minimum two episodes of acute NIV use); 8 were male, median age was 68 (age range 52–83) years 4 carers and 15 healthcare professionals (7 doctors, 4 specialist nurses, 2 physiotherapists and 2 physiologists): sex NR, age range 26–54 years

Explore experiences of domiciliary non-invasive ventilation in COPD, to understand decision-making processes and improve future palliative care

Qualitative interview study, based on the constructivist grounded theory

The study identified `adapting to NIV the central process enabling long-term use in palliative care, although the way in which this is approached by HCPs and patients do not always converge. Patients and HCPs actively negotiate the patient's adaption to NIV, although their experiences and views are not always convergent. While domiciliary NIV is valued by COPD patients, The process of adaption could be optimized by HCPs considering broader ways of explaining the process, other settings for initiation and generating more Patient-data on its benefits

Gäbler et al. (2019) Austria [41]

Hospital; ICU, pulmonology internal departments and geriatric/ palliative care

162 physicians (67 ICU, 51 pulmonology or internal departments, 44 geriatric or palliative care); 89 males, mean age 49 (SD 10, age range 27–65) years, 12 were physician in training, 18 general physician, 132 specialist, 110 had ⩾ 10 years of job experience

To investigate if the choice of treatment is influenced by the medical speciality

Cross-sectional survey

38 (23%) respondents chose NIV, 50 (31%) chose conservative treatment approach and 74 (46%) chose palliative approach. Intensivists had an almost 15-fold probability and pulmonologists/internists a nine-fold probability of inducing NIV in comparison with geriatricians/palliative physicians. Increasing age of the physician tended to correlate significantly against starting NIV. No effect was observed due to the following variables: amount (years) of professional experience, educational level and the importance of low patient stress due to the intervention

Gershon et al. (2018) Canada [42]

Population based (Ontario)

151 912 patients with advanced COPD between 2004 to 2014; 47.4% males, 80% were aged > 65 years, mean FEV₁% predicted NR

To describe trends in the use of EOL care strategies by people with advanced COPD in Ontario, Canada

Repeated cross-sectional study

The proportion admitted to the ICU who needed NIV slightly increased over time. In 2004 around 1% the patients used NIV, while in 2014 around 4% used NIV

Girault et al. (1997) France [71]

Hospital; medical intensive care unit (ICU)

15 patients with known COPD or a high probability of the disease; 12 males, mean age 64.5 (SD 6.75) years, mean FEV₁% predicted 29.20 (SD 10.43)

To investigate the effects of NIV assisted-control ventilation (ACV) by nasal mask on respiratory physiological parameters and comfort in acute on chronic respiratory failure (ACRF)

RCT

More COPD patients compared with other groups died in the ICU (1.54, p = 0.012). NIV ACV significantly decreased all the total inspiratory work of breathing parameters, pressure time product, and oesophageal pressure variation in comparison with spontaneous breathing (SP) mode. The ACV mode resulted in a significant reduction in surface diaphragmatic electromyographic activity to 36% of the control values and significantly improved the breathing pattern. The respiratory comfort was significantly lower with ACV than with SB

Gloeckl et al. (2019) Germany [43]

Pulmonary rehabilitation

20 patients; 12 males, mean age 60 (SD 6) years, mean FEV₁% predicted 19 (SD 4)

To investigate the acute effects of high-pressure NIV (along with oxygen supplementation) as an add-on tool during exercise in COPD patients with chronic hypercapnic respiratory failure

RCT, cross-over trial

On NIV COPD patients increased cycle endurance time by 39% compared to oxygen-use only. In NIV condition, TcPCO2 values were significantly lower at rest and at isotime compared to control condition. Oxygen saturation was significantly higher with NIV during exercise. All patients tolerated the use of NIV during exercise well and were able to perform cycle training with NIV. On NIV, TcPCO2 was significantly lower at rest and at isotime. Oxygen availability in the intercostal muscles remained relatively constant with NIV compared to oxygen-use only

Jerpseth et al. (2017) Norway [57]

Hospital; ICU, pulmonary ward

26 nurses (12 ICU and 14 pulmonary ward); 2 males; median age ICU nurses 38 (age range 31–55) years, median age pulmonary ward nurses 34 (age range 25–47), median years of experience ICU 8 (range 1–14) years, median years of experience pulmonary ward 6 years (range 9 months-15 years)

To investigate how nurses experienced their own role in decision-making processes regarding IMV in later stages of COPD and how they consider the patients’ role in these processes

Qualitative design

Nurses described the dilemma of being part of a medical treatment culture rather than being able to focus on the patients’ need for good care at the end of life. This medical culture focused on patients’ capacity to breathe, and the only solution offered to patients was either NIV or IMV. The patients’ situation was so complex that the nurses felt they needed care that extended beyond simply treatment with NIV or IMV. Nurses experienced lack of authority to act; they felt that they should have acted on the caring needs of the patients and felt like they acted against their caring values

Jerpseth et al. (2018) Norway [44]

Two university hospitals and three district hospitals

12 patients with severe COPD; 5 males, age range 63–87 years, FEV₁% predicted NR, 6 used NIV previous year

To explore the illness experiences of older patients with late-stage COPD and to develop knowledge about how patients perceive their preferences to be taken into account in decision-making processes concerning IMV and/or NIV

Qualitative design with hermeneutic–phenomenological approach

Patients clung to the hope that the NIV treatment would help them through what the experienced respiratory crisis. The mask was tiresome, unpleasant and a bother, but also seen as symbol of hope and survival even when there was no prospect of healing. Some described “waking up” on NIV several hours or days after hospitalisation which created a sense of vulnerability. Patients were not able to remember whether anyone had ever asked them if they wanted to use NIV, nor whether they had discussed the burden versus benefit of the treatment with either their physicians or their nurses

Jones, et al. (1998), UK [50]

Home based

11 patients (all ex-smokers in severe type II respiratory failure, were electively admitted between 1991 and 1995. All were diagnosed as having COPD and were receiving maximal drug therapy); 8 males, mean age 60 (range 45–73), FEV₁ predicted to 27 (SD 8.9) %

To test domiciliary nocturnal intermittent positive pressure ventilation (NIPPV) in patients with respiratory failure due to severe COPD

Follow up study for over two years (pre/ post)

Hospital admissions and GP consultations were halved after one year compared with the year before NIPPV and there was a sustained improvement in arterial blood gas tensions at 12 and 24 months when breathing air, despite progressive deterioration in ventilatory function. BMI did not change during the period of observation. The median survival was 920 days, with no patient dying within the first 500 days

Kuo et al. (2019) Taiwan [45]

Home, hospital

8640 patients with COPD; 69,4% males, mean age 79.97 (SD 9.87) year, FEV₁% predicted NR

To explore and compare EOL resource use during the last six months before death between COPD and LC patients (1) comparing EOL health care resource utilization and the use of intensive and supportive procedures during the last six months of life, (2) exploring changes in the trends of intensive procedures and palliative care between 2000 and 2012, and (3) examining predictive factors of the use of intensive procedures

Retrospective cohort study

Significantly more patients with COPD (16.54%) than patients with LC (13.53%) received non-invasive MV during the last six months of their life whole

Kvangarsnes et al. (2012) Norway

Hospital; ICU

10 patients with COPD; 5 males, age range 45–85 years, FEV₁% predicted NR

To explore patient perceptions of COPD exacerbation and the patients’ experiences of their relations with health personnel during care and treatment

Narrative inquiry

Patients’ perceptions of breathlessness were an essential theme, making them completely dependent on others regarding the mask treatment and breathing assistance. All patients had a positive experience with NIV treatment. Patients revealed stories of trust and distrust receiving NIV treatment

Landers et al. (2015) New Zealand [61]

Hospital; admitted respiratory specialist services

15 patients with severe COPD; 9 males, mean age 69.2 (SD 8.2, age range 55–89) years, mean FEV1% predicted 26.4 (SD 10)

To explore the experience of patients with advanced COPD after a life-threatening event, particularly focusing on end-of-life (EOL) issues

Grounded Theory

Some participants identified the need for acute hospital care to manage symptoms as a milestone (for example with a BiPAP in the ICU). These participants expressed confidence in the hospital to reduce their physical symptoms and related anxiety. Acute hospital care was often seen as a haven or place of security. Participants explained how such interventions (NIV) were required to keep them alive; however, the negative prognostic implication of these admissions were not explored by participants

McEvoy et al. (2009) Australia [63]

Hospital, sleep/ respiratory medicine departments and at home for the NIV long-term oxygen therapy (LTOT) group

144 patients with severe stable smoking-related COPD; 69% male, mean age 68, FEV1% predicted LTOT: 23.1 (21.4 to 24.8), NIV + LTOT: 25 (22.4 to 27.6)

To determine the effects of nocturnal non-invasive bi-level pressure support ventilation (NIV) on survival, lung function and quality of life in patients with severe hypercapnic COPD

A multicentre, open-label, RCT

NIV improved sleep quality and sleep-related hypercapnia acutely, and patients complied well with therapy (mean (SD) nightly use 4.5(3.2) h). Compared with LTOT alone, NIV (mean follow-up 2.21 years, range 0.01–5.59) showed an improvement in survival with the adjusted but not the unadjusted Cox model (adjusted hazard ratio (HR) 0.63, 95% CI 0.40 to 0.99, p = 0.045; unadjusted HR 0.82, 95% CI 0.53 to 1.25, p = NS). FEV 1.0 and PaCO2 measured at 6 and 12 months were not different between groups. Disease-specific QOL (SGRQ) at 12 months was not different between the two groups. Patients assigned to NIV + LTOT had reduced general and mental health and vigour on SF 36

Sinuff et al. (2008) Canada and US [68]

Academic or community centers

Intensivists, pulmonologists, and respiratory therapists (RTs). 104 of 183 (57%) physicians and 290 of 473 (61%) RTs participated

To determine clinicians’ attitudes to and stated use of NIV for patients with acute respiratory failure who have declined intubation and resuscitation or have chosen comfort measures only

Multi-center survey

2/3 of physicians include NIV during life support discussions with do-not-resuscitate patients at least sometimes, and 87% of RTs stated that NIV should be included in such discussions. For patients choosing comfort measures only, almost half of physicians reported including NIV as an option in their discussions at least sometimes, while fewer than half of RTs stated that these discussions should be conducted

Most (> 80%) physicians use NIV and most (> 80%) RTs are asked to initiate NIV for do-not-resuscitate patients with COPD

Volpato et al. (2022) Italy [60]

Inpatients and outpatients

90 patients with severe COPD; 46 males, mean age 76.2 (SD 8.03) years, mean FEV₁% predicted 50.7 (SD 27.0)

To analyze the impact of a brief psychological support intervention on adherence to NIV among patients with COPD

RCT

The psychological intervention was related to improvements in adherence to NIV and QOL after four to eight meetings with cognitive and behavioural therapy, with homework, during the NIV adaption compared with the control group (six sessions watching video related to COPD management). Results indicated a significant change in the QOL also over time

Windisch et al. (2005) Germany [54]

Hospital

34 patients with stable COPD and hypercapnic respiratory failure; 27 males, mean age 63.4 (SD 9.7, range 43–77),

To assess changes in blood gas levels and long-term outcome in a larger group of patients with COPD and chronic hypercapnic respiratory failure who were treated by controlled NIV aimed at achieving maximal improvement of PaCO₂

Retrospective study

Daytime Paco₂ during spontaneous breathing significantly decreased by 6.9 (SD 8.0) mm Hg, daytime Pao₂ significantly increased by 5.8 (SD 9.4) mm H, FEV₁ significantly increased by 0.14 (SD 0.16) L after 2 months of NIV. This was achieved with mean inspiratory pressures of 27.7 (SD 5.9) cm H₂O (range, 17 to 40 cm H₂O) at a mean respiratory rate of 20.8 (SD 2.5 breaths/min. The 2-year survival rate was 86%