Risk prediction models for non-invasive ventilation failure in patients with chronic obstructive pulmonary disease: A systematic review.

Abstract

Background: Chronic obstructive pulmonary disease (COPD) is a common respiratory condition with high morbidity and mortality. Noninvasive mechanical ventilation (NIV) is often used to manage acute COPD exacerbations, but failure can lead to worse outcomes. This systematic review aimed to evaluate risk prediction models for NIV failure in patients with COPD.

Methods: PubMed, Embase, Web of Science, The Cochrane Library, CINAHL, CBM, CNKI, Wanfang, and VIP databases, from database inception to January 10, 2024, were searched for studies on risk prediction models for failure in NIV among COPD patients. Two reviewers independently screened the literature, extracted data, assessed the quality of included studies using the Prediction Model Risk of Bias Assessment Tool, and conducted a systematic evaluation of the prediction models.

Results: A total of 11 studies were included, encompassing 13 risk prediction models. The area under the receiver operating characteristic curve for the included models ranged from 0.810 to 0.978. Predictive factors in the models mainly included Acute Physiology And Chronic Health Evaluation II score, pH value, PaCO2, consciousness status, serum albumin level, and respiratory rate.

Conclusion: Existing risk prediction models for failure in NIV among patients with COPD demonstrated overall good predictive performance, but exhibited a risk of bias. Further validation is needed to assess the clinical applicability of these models.