Microcirculatory alterations in critically ill COVID-19 patients analyzed using artificial intelligence.

BACKGROUND: The sublingual microcirculation presumably exhibits disease-specific changes in function and morphology. Algorithm-based quantification of functional microcirculatory hemodynamic variables in handheld vital microscopy (HVM) has recently allowed identification of hemodynamic alterations in the microcirculation associated with COVID-19. In the present study we hypothesized that supervised deep machine learning could be used to identify previously unknown microcirculatory alterations, and combination with algorithmically quantified functional variables increases the model's performance to differentiate critically ill COVID-19 patients from healthy volunteers. METHODS: Four international, multi-central cohorts of critically ill COVID-19 patients and healthy volunteers (n = 59/n = 40) were used for neuronal network training and internal validation, alongside quantification of functional microcirculatory hemodynamic variables. Independent verification of the models was performed in a second cohort (n = 25/n = 33). RESULTS: Six thousand ninety-two image sequences in 157 individuals were included. Bootstrapped internal validation yielded AUROC(CI) for detection of COVID-19 status of 0.75 (0.69-0.79), 0.74 (0.69-0.79) and 0.84 (0.80-0.89) for the algorithm-based, deep learning-based and combined models. Individual model performance in external validation was 0.73 (0.71-0.76) and 0.61 (0.58-0.63). Combined neuronal network and algorithm-based identification yielded the highest externally validated AUROC of 0.75 (0.73-0.78) (P < 0.0001 versus internal validation and individual models). CONCLUSIONS: We successfully trained a deep learning-based model to differentiate critically ill COVID-19 patients from heathy volunteers in sublingual HVM image sequences. Internally validated, deep learning was superior to the algorithmic approach. However, combining the deep learning method with an algorithm-based approach to quantify the functional state of the microcirculation markedly increased the sensitivity and specificity as compared to either approach alone, and enabled successful external validation of the identification of the presence of microcirculatory alterations associated with COVID-19 status.

Renal Tubular Acidosis in Pregnant Critically Ill COVID-19 Patients: A Secondary Analysis of a Prospective Cohort.

BACKGROUND: Renal tubular acidosis (RTA) is an extremely rare cause of metabolic acidosis (10 in 100,000). RTA has been linked neither to pregnancy nor to severe coronavirus disease 2019 (COVID-19). The purpose of this study was to analyze the prevalence and clinical course of normal anion gap metabolic acidosis in critically ill pregnant COVID-19 patients and to compare them to an age-matched nonpregnant female patient cohort. METHODS: Secondary analysis was conducted on a prospective observational cohort of critically ill patients suffering from COVID-19 consecutively admitted to a tertiary intensive care unit (ICU) between February 2020 and April 2021. RESULTS: A total of 321 COVID-19 patients required admission to the ICU; 95 (30%) were female, and 18 (19%) were of childbearing age. Seven of eight (88%) pregnant women (all in the last trimester) required advanced respiratory support due to COVID-19. The estimated glomerular filtration rate was 135 (123-158) mL/min/m2 body surface area, and six pregnant women (86%) were diagnosed with a normal, respiratory compensated, anion gap metabolic acidosis (pHmin 7.3 (7.18-7.31), HCO3-min 14.8 (12.8-18.6) mmol/L, and paCO2 3.4 (3.3-4.5) kPa). Three (43%) acidotic pregnant women fulfilled diagnostic criteria for RTA. All women recovered spontaneously within less 7 days. CONCLUSIONS: Metabolic acidosis seems to be very common (85%) in pregnant critically ill COVID-19 patients, and the prevalence of RTA might be higher than normal. It remains to be demonstrated if this observation is an indirect epiphenomenon or due to a direct viral effect on the tubular epithelium.

Long-term ketamine infusion-induced cholestatic liver injury in COVID-19-associated acute respiratory distress syndrome.

BACKGROUND: A higher-than-usual resistance to standard sedation regimens in COVID-19 patients suffering from acute respiratory distress syndrome (ARDS) has led to the frequent use of the second-line anaesthetic agent ketamine. Simultaneously, an increased incidence of cholangiopathies in mechanically ventilated patients receiving prolonged infusion of high-dose ketamine has been noted. Therefore, the objective of this study was to investigate a potential dose-response relationship between ketamine and bilirubin levels. METHODS: Post hoc analysis of a prospective observational cohort of patients suffering from COVID-19-associated ARDS between March 2020 and August 2021. A time-varying, multivariable adjusted, cumulative weighted exposure mixed-effects model was employed to analyse the exposure-effect relationship between ketamine infusion and total bilirubin levels. RESULTS: Two-hundred forty-three critically ill patients were included into the analysis. Ketamine was infused to 170 (70%) patients at a rate of 1.4 [0.9-2.0] mg/kg/h for 9 [4-18] days. The mixed-effects model revealed a positively correlated infusion duration-effect as well as dose-effect relationship between ketamine infusion and rising bilirubin levels (p < 0.0001). In comparison, long-term infusion of propofol and sufentanil, even at high doses, was not associated with increasing bilirubin levels (p = 0.421, p = 0.258). Patients having received ketamine infusion had a multivariable adjusted competing risk hazard of developing a cholestatic liver injury during their ICU stay of 3.2 [95% confidence interval, 1.3-7.8] (p = 0.01). CONCLUSIONS: A causally plausible, dose-effect relationship between long-term infusion of ketamine and rising total bilirubin levels, as well as an augmented, ketamine-associated, hazard of cholestatic liver injury in critically ill COVID-19 patients could be shown. High-dose ketamine should be refrained from whenever possible for the long-term analgosedation of mechanically ventilated COVID-19 patients.

Closed-Loop Versus Conventional Mechanical Ventilation in COVID-19 ARDS.

BACKGROUND: Lung-protective ventilation is key in bridging patients suffering from COVID-19 acute respiratory distress syndrome (ARDS) to recovery. However, resource and personnel limitations during pandemics complicate the implementation of lung-protective protocols. Automated ventilation modes may prove decisive in these settings enabling higher degrees of lung-protective ventilation than conventional modes. METHOD: Prospective study at a Swiss university hospital. Critically ill, mechanically ventilated COVID-19 ARDS patients were allocated, by study-blinded coordinating staff, to either closed-loop or conventional mechanical ventilation, based on mechanical ventilator availability. Primary outcome was the overall achieved percentage of lung-protective ventilation in closed-loop versus conventional mechanical ventilation, assessed minute-by-minute, during the initial 7 days and overall mechanical ventilation time. Lung-protective ventilation was defined as the combined target of tidal volume <8 ml per kg of ideal body weight, dynamic driving pressure <15 cmH2O, peak pressure <30 cmH2O, peripheral oxygen saturation ≥88% and dynamic mechanical power <17 J/min. RESULTS: Forty COVID-19 ARDS patients, accounting for 1,048,630 minutes (728 days) of cumulative mechanical ventilation, allocated to either closed-loop (n = 23) or conventional ventilation (n = 17), presenting with a median paO2/ FiO2 ratio of 92 [72-147] mmHg and a static compliance of 18 [11-25] ml/cmH2O, were mechanically ventilated for 11 [4-25] days and had a 28-day mortality rate of 20%. During the initial 7 days of mechanical ventilation, patients in the closed-loop group were ventilated lung-protectively for 65% of the time versus 38% in the conventional group (Odds Ratio, 1.79; 95% CI, 1.76-1.82; P < 0.001) and for 45% versus 33% of overall mechanical ventilation time (Odds Ratio, 1.22; 95% CI, 1.21-1.23; P < 0.001). CONCLUSION: Among critically ill, mechanically ventilated COVID-19 ARDS patients during an early highpoint of the pandemic, mechanical ventilation using a closed-loop mode was associated with a higher degree of lung-protective ventilation than was conventional mechanical ventilation.

Prognostic factors associated with mortality risk and disease progression in 639 critically ill patients with COVID-19 in Europe: Initial report of the international RISC-19-ICU prospective observational cohort.

BACKGROUND: Coronavirus disease 2019 (COVID-19) is associated with a high disease burden with 10% of confirmed cases progressing towards critical illness. Nevertheless, the disease course and predictors of mortality in critically ill patients are poorly understood. METHODS: Following the critical developments in ICUs in regions experiencing early inception of the pandemic, the European-based, international RIsk Stratification in COVID-19 patients in the Intensive Care Unit (RISC-19-ICU) registry was created to provide near real-time assessment of patients developing critical illness due to COVID-19. FINDINGS: As of April 22, 2020, 639 critically ill patients with confirmed SARS-CoV-2 infection were included in the RISC-19-ICU registry. Of these, 398 had deceased or been discharged from the ICU. ICU-mortality was 24%, median length of stay 12 (IQR, 5-21) days. ARDS was diagnosed in 74%, with a minimum P/F-ratio of 110 (IQR, 80-148). Prone positioning, ECCO2R, or ECMO were applied in 57%. Off-label therapies were prescribed in 265 (67%) patients, and 89% of all bloodstream infections were observed in this subgroup (n = 66; RR=3·2, 95% CI [1·7-6·0]). While PCT and IL-6 levels remained similar in ICU survivors and non-survivors throughout the ICU stay (p = 0·35, 0·34), CRP, creatinine, troponin, d-dimer, lactate, neutrophil count, P/F-ratio diverged within the first seven days (p<0·01). On a multivariable Cox proportional-hazard regression model at admission, creatinine, d-dimer, lactate, potassium, P/F-ratio, alveolar-arterial gradient, and ischemic heart disease were independently associated with ICU-mortality. INTERPRETATION: The European RISC-19-ICU cohort demonstrates a moderate mortality of 24% in critically ill patients with COVID-19. Despite high ARDS severity, mechanical ventilation incidence was low and associated with more rescue therapies. In contrast to risk factors in hospitalized patients reported in other studies, the main mortality predictors in these critically ill patients were markers of oxygenation deficit, renal and microvascular dysfunction, and coagulatory activation. Elevated risk of bloodstream infections underscores the need to exercise caution with off-label therapies.