Impact of end-of-life respiratory modalities on quality of dying and death and symptom relief in patients with interstitial lung disease: a multicenter descriptive cross-sectional study

BACKGROUND: Respiratory modalities applied at the end of life may affect the burden of distressing symptoms and quality of dying and death (QODD) among patients with end-stage interstitial lung disease (ILD); however, there have been few studies into respiratory modalities applied to these patients near death. We hypothesized that high-flow nasal cannula (HFNC) might contribute to improved QODD and symptom relief in patients with end-stage ILD. OBJECTIVES: This multicenter study examined the proportion of end-of-life respiratory modalities in a hospital setting and explored its impact on QODD and symptom relief among patients dying with ILD. METHODS: Consecutive patients with ILD who died in four participating hospitals in Japan from 2015 to 2019 were identified and divided into four groups according to end-of-life respiratory modality: conventional oxygen therapy (COT), HFNC, non-invasive ventilation (NIV), and invasive mechanical ventilation (IMV). In addition, a mail survey was performed to quantify the QODD and symptom relief at their end of life from a bereaved family’s perspective. QODD and symptom relief were quantified using the Good Death Inventory (GDI) for patients with a completed bereavement survey. The impact of end-of-life respiratory modalities on QODD and symptom relief was measured by multivariable linear regression using COT as a reference. RESULTS: Among 177 patients analyzed for end-of-life respiratory modalities, 80 had a completed bereavement survey. The most common end-of-life respiratory modality was HFNC (n = 76, 42.9%), followed by COT (n = 62, 35.0%), NIV (n = 27, 15.3%), and IMV (n = 12, 6.8%). Regarding the place of death, 98.7% of patients treated with HFNC died outside the intensive care unit. Multivariable regression analyses revealed patients treated with HFNC had a higher GDI score for QODD [partial regression coefficient (B) = 0.46, 95% CI 0.07–0.86] and domain score related to symptom relief (B = 1.37, 95% CI 0.54–2.20) than those treated with COT. CONCLUSION: HFNC was commonly used in patients with end-stage ILD who died in the hospital and was associated with higher bereaved family ratings of QODD and symptom relief. HFNC might contribute to improved QODD and symptom relief in these patients who die in a hospital setting. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12931-022-02004-x.