Utility of illness severity scores to predict mortality in patients hospitalized with respiratory deterioration of idiopathic pulmonary fibrosis

• Published in: QJM : An International Journal of Medicine Vol. 114; no. 8; pp. 559 - 567
• Main Authors: Hyams, C, Hettle, D, Bibby, A, Adamali, H A, Barratt, S L
• Format: Journal Article
• Language: English
• Published: Oxford University Press 05.11.2021
• ISSN: 1460-2725; 1460-2393
• DOI: 10.1093/qjmed/hcaa214

Summary Introduction In the context of idiopathic pulmonary fibrosis (IPF), respiratory-related admissions to hospital are associated with a high morbidity and short-term mortality with significant burden on secondary care services. It has yet to be determined how to accurately identify patients at risk of acute respiratory deterioration (ARD) or the prognosticating factors. Aim We sought to define the characteristics of hospitalized ARD-IPF patients in a real-world cohort and investigate factors associated with worse outcomes. Specifically, we wished to determine the association between baseline CURB-65 and NEWS-2 and mortality in IPF, given illness severity scores have not previously been validated in this cohort. Methods Single-centre retrospective observational cohort study. Results Of 172 first hospitalizations for ARD, 27 admissions (15.7%) were due to an acute exacerbation of IPF (AE-IPF), 28 (16.3%) secondary to cardiac failure/fluid overload and 17 due to pneumonia (9.9%). Other admissions related to lower respiratory tract infection, extra-parenchymal causes and those without a specific trigger. Baseline patient characteristics were comparable for all underlying aetiologies of ARD-IPF. Treatment pathways did not differ significantly between AE-IPF and other causes of ARD-IPF. Short-term mortality was high, with ∼22% patients dying within 30 days. Illness severity scores (NEWS-2 and CURB-65) were independent predictors of mortality in multivariable logistic regression modelling. Conclusions Our findings suggest significant mortality related to hospitalization with ARD-IPF of any underlying cause. Our data support the use of CURB-65 and NEWS-2 scores as illness severity scores that can provide a simple tool to help future prognostication in IPF. Research should be aimed at refining the management of these episodes, to try to reduce mortality, where possible, or to facilitate palliative care for those with adverse prognostic characteristics.