The Predictive Potential of Elevated Serum Inflammatory Markers in Determining the Need for Intubation in COVID-19 Patients. Windham S, Hirsch K, Peterson R, Douin D, Chauhan L, Heery L, Fling C, Vukovic N, Holguin F, Zimmer S, Erlandson K. J Crit Care Med. 2021 Nov 13;8(1):14-22.
COVID-19 remains a critical public health crisis and ongoing pandemic, which has prodigiously demanded intensivists across the globe. The best ways to identify which patients with COVID-19 who may require intubation and have a rapid, unpredictable respiratory decompensation remains a challenge. The objective of this study was to determine the predictive potential of admission data and inflammatory markers in determining the need for intubation in patients with COVID-19. The authors performed a retrospective cohort study of the first 158 adult patients who were hospitalized with COVID-19 (without concomitant viral, bacterial, or fungal co-infections) between March and April 2020 at a single, large, academic hospital. The primary outcome was need for intubation. The institution’s criteria for intubation were based on patients who could not maintain >90% SpO2 on maximum settings of low-flow oxygen delivery devices. The minimax concave penalty (MCP) regularized logistic regression was employed to build a sparse predictive model using the covariates and confounders. After creating their predictive model, it was validated in the 102 subsequent patients admitted after the initial time period.
Amongst the 158 patients, 37% were discharged from the hospital without the need for invasive ventilation. 41% required intubation, 27% required vasopressor medications, 15% were prone positioned, and 2.5% died. Patients who required intubation were more likely to be diabetic (43 vs 22%, p=0.012), less likely to present with nausea or vomiting (16 vs 36%, p=0.014) or diarrhea (18 vs 32%, p=0.06), more likely to have bilateral lung infiltrates upon admission CXR (81 vs 54%, p<0.0001), and had higher inflammatory markers including CRP, LDH, ferritin, D-dimer, and neutrophil-lymphocyte ratio. After adjustment, the odds of intubation were significantly greater as inflammatory markers increased, per standard deviation increase in C-reactive protein (OR 2.81), lactate dehydrogenase (OR 2.10), and neutrophil-lymphocyte ratio (OR 2.21). When they stratified patients by low (<100) versus high (>100) CRP, the authors found that more than half of the patients in the high CRP group required intubation within the first 48 hours. Using a sparse predictive model, CRP, LDH, and diabetes mellitus were selected by the model as the most predictive factors for predicting the need for intubation. These three variables correctly classified intubated status 71% of the time, corresponding to a sensitivity of 86%, specificity of 63%, positive predictive value (PPV) of 61%, and negative predictive value (NPV) of 87% using a 30% threshold. The area under the ROC curve was 78%. In its subsequent validation, it showed a sensitivity of 91%, specificity of 41%, PPV of 44%, and NPV of 90%. The study is limited by single center perspective, potential variation in intubation practices by provider, and the fact that during the study period, positive pressure ventilation was not utilized (due to initial concerns regarding aerosolization of SARS-CoV-2). However, limitations aside, the present findings that high CRP, LDH, and diabetes mellitus can quickly identify those at highest risk for intubation may inform resource allocation and dispo from ED when ICU beds are not be as readily available.
Prophylactic melatonin for delirium in intensive care (Pro-MEDIC): a randomized controlled trial. Wibrow B, Martinez FE, Myers E, Chapman A, Litton E, Ho KM, Regli A, Hawkins D, Ford A, van Haren FMP, Wyer S, McCaffrey J, Rashid A, Kelty E, Murray K, Anstey M. Intensive Care Med. 2022 Feb 27.
Delirium is a common problem among critically ill patients and its presence has significant deleterious effects on hospital and ICU length of stay, duration of mechanical ventilation, functional status, morbidity, and mortality. A major risk factor for delirium is poor sleep and dysregulated circadian rhythm which is a known issue in critically ill patients. Melatonin is thought to enhance sleep quality thus the authors of this trial sought to better define its effects on delirium in critically ill patients. This is a well done multicenter, randomized, placebo-controlled, double-blind trial involving 12 ICU’s in Australia, analyzing the outcomes from 841 critically ill patients and a total of 5821 CAM-ICU assessments. Patients were assigned via 1:1 randomization to either receive 4mg of liquid melatonin vs. placebo every evening for 14 consecutive nights or until ICU discharge. Delirium was assessed via twice daily CAM-ICU assessments and their primary outcome was the proportion of delirium-free assessments within 14 days or before ICU discharge.
The authors found that melatonin was not associated with increased delirium-free assessments across all subgroups of medical and surgical critically ill patients. There was no significant difference in the average proportion of delirium-free assessments per patient between the two groups (79.2% for melatonin group and 80% for the placebo group). There was no benefit to melatonin administration in any of the secondary outcomes as well (such as ICU LOS, hospital LOC, duration of mechanical ventilation, 28 day or 90 day mortality). They also assessed absorption of melatonin via a subset of patients via blinded, independent biochemist collecting and assessing blood samples for serum melatonin concentrations 2-3 hours post-administration. Oral melatonin administration led to increased serum melatonin concentrations indicating intact absorption. Interestingly, the authors also took into account missing CAM-ICU scores in patients which occurred in 17% of the total assessments and because of this, they increased their sample size by 15%. This was the largest randomized control trial of melatonin for delirium prevention and overall, the study is complete and comprehensive.