Munshi, L

Abstract Background Acute hypoxemic and hypercapnic respiratory failure are among the most common reasons for ICU admission and need for invasive mechanical ventilation. Noninvasive respiratory support (NIRS) strategies—including high-flow nasal cannula (HFNC), noninvasive ventilation (NIV), and continuous positive airway pressure (CPAP)—may prevent intubation, improve outcomes, and reduce ICU utilization. However, there is uncertainty regarding optimal patient and modality selection, resulting in variable implementation. There are no clinical practice guidelines comprehensively addressing the use of the different noninvasive respiratory support strategies across the spectrum of acute respiratory failure. Objective To update and develop new evidence-based clinical practice recommendations informing noninvasive respiratory support use, including HFNC, NIV and CPAP, in adults with acute respiratory failure. Methods A multidisciplinary panel used the GRADE approach to address four PICO questions related to the use of NIRS for hypoxemic and hypercapnic respiratory failure, preoxygenation for intubation, and post-extubation respiratory support. Recommendations were informed by several systematic reviews and network meta-analyses. Results The panel made a strong recommendation for HFNC and a conditional recommendation for NIV or CPAP for adults with acute hypoxemic respiratory failure with close monitoring for the need for escalation of respiratory support, based primarily on effects on need for intubation. For acute hypercapnic respiratory failure, the panel made a strong recommendation for NIV to reduce mortality and need for invasive mechanical ventilation, and a conditional recommendation for HFNC only in patients with less severe hypercapnia and with mild acidemia (eg, pH > 7.25), provided that close monitoring and prompt escalation to NIV are available. The panel made a strong recommendation for HFNC or NIV for preoxygenation prior to endotracheal intubation to prevent peri-intubation hypoxemia. The panel also issued a risk-based recommendation, suggesting HFNC for low-risk patients and NIV for high-risk patients to reduce the need for re-intubation following extubation after critical illness. Conclusions Noninvasive respiratory support strategies are effective in improving outcomes in a range of clinical scenarios. We provide evidence-based recommendations, which can be further informed by patient risk, institutional capacity, and interface tolerance.

Abstract Rationale Major advances in precision medicine and immunotherapy have led to improved outcomes for patients with metastatic cancer, but have also introduced new, severe toxicities that can necessitate ICU admission. The overall mortality remains substantial; however, prior studies have often been single centered with small sample sizes. Therefore, we sought to provide an overview of the current literature with pooled estimates of short- and long-term mortality. Methods This study was registered a priori on Open Science Framework (10.17605/OSF.IO/438RY) and conducted per the Cochrane Handbook. A medical librarian searched MEDLINE, Embase, PubMed, Cochrane Central Register for Controlled Trials, supplemented by hand searching. Eligible studies included randomized and observational designs with >10 patients; studies were included if ≥ 70% of participants had metastatic disease when subgroup data were unavailable. The primary outcome was ICU mortality. Other outcomes included hospital, 30-day, 3-month, 6-month, and 12-month mortality. Pooled prevalence and 95% CIs were estimated using a random-effects model with Freeman-Tukey transformation. Risk of bias was assessed using the validated 10-item tool for prevalence studies. Results Of 4752 unique citations, 49 were included (N = 168,209). All studies were rated as having low risk of bias with the main threat to external validity being single-center design. Fourteen studies were multi-centered and 3 were multi-national, with the rest being single center. Across 23 studies (N = 2,918), the ICU mortality was 32% (95% CI: 25-39%; I2 = 89%; Figure 1). Across 5 studies (N = 400) with a focus on lung cancer, there was significantly higher ICU mortality (53% [43-64%]; I2 = 75% versus 26% [20-34%]; I2 = 86%; PQ = <0.01). Among 5 studies (N = 474) focusing on acute respiratory failure (ARF), ICU mortality was numerically higher but not statistically significant (49% [33-66%]; I2 = 91% versus 26% [20-34%]; I2 = 85%; PQ=0.07). Hospital mortality was 50% (44-56%; I2 = 99%; N = 157,682), 30-day mortality was 48% (39-57%; I2 = 84%; N = 1,067), 3-month mortality was 69% (62-74%; I2 = 80%; N = 1,296), 6-month mortality was 70% (59-79%; I2 = 91%; N = 1,310), and 12-month mortality was 84% (71-91%; I2 = 95%; N = 1,446; Figure 1). Conclusions Among critically ill adults with metastatic cancer, about one-third died in the ICU, half survived to hospital discharge, and 12-month mortality was 84%. There was considerable heterogeneity in mortality estimates, with worse outcomes in lung cancer. Given that ICU admission criteria and illness severity may differ, large multi-center studies are needed to clarify temporal trends and identify management strategies to improve outcomes. This abstract is funded by: None