Plasma glial fibrillary acidic protein and tau: predictors of neurological outcome after cardiac arrest.

BACKGROUND: The purpose was to evaluate glial fibrillary acidic protein (GFAP) and total-tau in plasma as predictors of poor neurological outcome after out-of-hospital (OHCA) and in-hospital cardiac arrest (IHCA), including comparisons with neurofilament light (NFL) and neuron-specific enolase (NSE). METHODS: Retrospective multicentre observational study of patients admitted to an intensive care unit (ICU) in three hospitals in Sweden 2014-2018. Blood samples were collected at ICU admission, 12 h, and 48 h post-cardiac arrest. Poor neurological outcome was defined as Cerebral Performance Category 3-5 at 2-6 months after cardiac arrest. Plasma samples were retrospectively analysed for GFAP, tau, and NFL. Serum NSE was analysed in clinical care. Prognostic performances were tested with the area under the receiver operating characteristics curve (AUC). RESULTS: Of the 428 included patients, 328 were OHCA, and 100 were IHCA. At ICU admission, 12 h and 48 h post-cardiac arrest, GFAP predicted neurological outcome after OHCA with AUC (95% CI) 0.76 (0.70-0.82), 0.86 (0.81-0.90) and 0.91 (0.87-0.96), and after IHCA with AUC (95% CI) 0.77 (0.66-0.87), 0.83 (0.74-0.92) and 0.83 (0.71-0.95). At the same time points, tau predicted outcome after OHCA with AUC (95% CI) 0.72 (0.66-0.79), 0.75 (0.69-0.81), and 0.93 (0.89-0.96) and after IHCA with AUC (95% CI) 0.61 (0.49-0.74), 0.68 (0.56-0.79), and 0.77 (0.65-0.90). Adding the change in biomarker levels between time points did not improve predictive accuracy compared to the last time point. In a subset of patients, GFAP at 12 h and 48 h, as well as tau at 48 h, offered similar predictive value as NSE at 48 h (the earliest time point NSE is recommended in guidelines) after both OHCA and IHCA. The predictive performance of NFL was similar or superior to GFAP and tau at all time points after OHCA and IHCA. CONCLUSION: GFAP and tau are promising biomarkers for neuroprognostication, with the highest predictive performance at 48 h after OHCA, but not superior to NFL. The predictive ability of GFAP may be sufficiently high for clinical use at 12 h after cardiac arrest.

Plasma neutrophil gelatinase-associated lipocalin independently predicts dialysis need and mortality in critical COVID-19.

Neutrophil gelatinase-associated lipocalin (NGAL) is a novel kidney injury and inflammation biomarker. We investigated whether NGAL could be used to predict continuous renal replacement therapy (CRRT) and mortality in critical coronavirus disease 2019 (COVID-19). This prospective multicenter cohort study included adult COVID-19 patients in six intensive care units (ICUs) in Sweden between May 11, 2020 and May 10, 2021. Blood was sampled at admission, days two and seven in the ICU. The samples were batch analyzed for NGAL, creatinine, and cystatin c after the end of the study period. Initiation of CRRT and 90-day survival were used as dependent variables in regression models. Of 498 included patients, 494 were analyzed regarding CRRT and 399 were analyzed regarding survival. Seventy patients received CRRT and 154 patients did not survive past 90 days. NGAL, in combination with creatinine and cystatin c, predicted the subsequent initiation of CRRT with an area under the curve (AUC) of 0.95. For mortality, NGAL, in combination with age and sex, had an AUC of 0.83. In conclusion, NGAL is a valuable biomarker for predicting subsequent initiation of CRRT and 90-day mortality in critical COVID-19. NGAL should be considered when developing future clinical scoring systems.

Sepsis mimics among presumed sepsis patients at intensive care admission: a retrospective observational study.

BACKGROUND: Diagnosing sepsis remains a challenge because of the lack of gold-standard diagnostics. Since there are no simple, broadly accepted criteria for infection, there is a risk of misclassifying sepsis patients (sepsis mimics) among patients with organ failure. The main objective of this study was to investigate the proportion of non-infected patients (sepsis mimics) in ICU patients with presumed sepsis at intensive care unit (ICU) admission. METHODS: Adult patients were screened retrospectively during 3.5 years in four ICUs in Sweden for fulfilment of the sepsis-3 criteria at ICU admission (presumed sepsis). Proxy criteria for suspected infection were sampled blood culture(s) and concomitant antibiotic administration. Culture-negative presumed sepsis patients were screened for infection according to the Linder-Mellhammar Criteria of Infection (LMCI). Sepsis mimics were defined as without probable infection according to the LMCI. Confirmed sepsis was defined as presumed sepsis after the exclusion of sepsis mimics. RESULTS: In the ICU presumed sepsis cohort (2664 patients), 25% were considered sepsis mimics. The most common reasons for ICU admission among sepsis mimics were acute heart failure and unspecific respiratory failure. Comparing sepsis mimics and confirmed sepsis showed that confirmed sepsis patients were slightly more severely ill but had similar mortality. C-reactive protein had modest discriminatory power (AUROC 0.71) with confirmed sepsis as the outcome. CONCLUSIONS: One-fourth of a presumed ICU sepsis population identified with the sepsis-3 criteria could be considered sepsis mimics. The high proportion of sepsis mimics has a potential dilutional effect on the presumed sepsis population, which threatens the validity of results from sepsis studies using recommended sepsis criteria.

[The Swecrit Biobank, associated clinical registries, and machine learning (artificial intelligence) improve critical care knowledge]. / Provsamlingen Swecrit och AI ger ny kunskap vid intensivvård.

The unique Swecrit Biobank and its associated clinical registries for sepsis, ARDS, cardiac arrest, trauma, and COVID-19 include more than 150,000 blood samples and descriptions of critically ill patients. These assets provide a unique opportunity to research and improve the care of the most seriously ill patients through biomarker analyses, proteomic studies, and genetic and epigenetic studies using modern machine learning techniques (artificial intelligence). Interested researchers are invited to submit their proposals and participate.

A simple mortality prediction model for sepsis patients in intensive care.

Background: Sepsis is common in the intensive care unit (ICU). Two of the ICU's most widely used mortality prediction models are the Simplified Acute Physiology Score 3 (SAPS-3) and the Sequential Organ Failure Assessment (SOFA) score. We aimed to assess the mortality prediction performance of SAPS-3 and SOFA upon ICU admission for sepsis and find a simpler mortality prediction model for these patients to be used in clinical practice and when conducting studies. Methods: A retrospective study of adult patients fulfilling the Sepsis-3 criteria admitted to four general ICUs was performed. A simple prognostic model was created using backward stepwise multivariate logistic regression. The area under the curve (AUC) of SAPS-3, SOFA and the simple model was assessed. Results: One thousand nine hundred eighty four admissions were included. A simple six-parameter model consisting of age, immunosuppression, Glasgow Coma Scale, body temperature, C-reactive protein and bilirubin had an AUC of 0.72 (95% confidence interval (CI) 0.69-0.75) for 30-day mortality, which was non-inferior to SAPS-3 (AUC 0.75, 95% CI 0.72-0.77) (p = 0.071). SOFA had an AUC of 0.67 (95% CI 0.64-0.70) and was inferior to SAPS-3 (p < 0.001) and our simple model (p = 0.0019). Conclusion: SAPS-3 has a lower prognostic value in sepsis than in the general ICU population. SOFA performs less well than SAPS-3. Our simple six-parameter model predicts mortality just as well as SAPS-3 upon ICU admission for sepsis, allowing the design of simple studies and performance monitoring.

Neurofilament light chain on intensive care admission is an independent predictor of mortality in COVID-19: a prospective multicenter study.

BACKGROUND: Neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), and total-tau protein (tau) are novel blood biomarkers of neurological injury, and may be used to predict outcomes in critical COVID-19. METHODS: A prospective multicentre cohort study of 117 consecutive and critically ill COVID-19 patients in six intensive care units (ICUs) in southern Sweden between May and November 2020. Serial NfL, GFAP and tau were analysed in relation to mortality, the Glasgow Outcome Scale Extended (GOSE) and the physical (PCS) and mental (MCS) components of health-related quality of life at one year. RESULTS: NfL, GFAP and tau on ICU admission predicted one-year mortality with an area under the curve (AUC) of 0.82 (95% confidence interval [CI] 0.74[Formula: see text]0.90), 0.72 (95% CI 0.62[Formula: see text]0.82) and 0.66 (95% CI 0.54[Formula: see text]0.77). NfL on admission was an independent predictor of one-year mortality (p = 0.039). Low NfL and GFAP values were associated with good PCS ([Formula: see text]45) at one year but not with good MCS ([Formula: see text]45) or GOSE ([Formula: see text]5). CONCLUSIONS: NfL on ICU admission was an independent predictor of mortality. High levels of NfL, GFAP and tau were associated with mortality but not with poor GOSE in survivors at one year. Low levels of NfL and GFAP were associated with improved physical health-related quality of life.

Plasma bioactive adrenomedullin on intensive care unit admission is associated with acute respiratory distress syndrome: an observational study.

BACKGROUND: Bioactive adrenomedullin (bio-ADM) is a vasoactive peptide with a key role in reducing vascular hyperpermeability and improving endothelial stability during infection, but it also has vasodilatory properties. Bioactive ADM has not been studied in conjunction with acute respiratory distress syndrome (ARDS), but it has recently been shown to correlate with outcomes after severe COVID-19. Therefore, this study investigated the association between circulating bio-ADM on intensive care unit (ICU) admission and ARDS. The secondary aim was the association between bio-ADM and ARDS mortality. METHODS: We analysed bio-ADM levels and assessed the presence of ARDS in adult patients admitted to two general intensive care units in southern Sweden. Medical records were manually screened for the ARDS Berlin criteria. The association between bio-ADM levels and ARDS and mortality in ARDS patients was analysed using logistic regression and receiver-operating characteristics analysis. The primary outcome was an ARDS diagnosis within 72 h of ICU admission, and the secondary outcome was 30-day mortality. RESULTS: Out of 1224 admissions, 11% (n = 132) developed ARDS within 72 h. We found that elevated admission bio-ADM level was associated with ARDS independently of sepsis status and of organ dysfunction as measured by the Sequential organ failure assessment (SOFA) score. Both lower levels (< 38 pg/L) and high (> 90 pg/L) levels of bio-ADM were independently (of the Simplified acute physiology score, SAPS-3) predictive of mortality. Patients with indirect mechanisms of lung injury had higher bio-ADM levels than those with a direct mechanism of injury, and bio-ADM increased with increasing ARDS severity. CONCLUSIONS: High levels of bio-ADM on admission are associated with ARDS, and bio-ADM levels significantly differ depending on the injury mechanism. In contrast, both high and low levels of bio-ADM are associated with mortality, possibly due to the dual action of bio-ADM in stabilising the endothelial barrier and causing vasodilation. These findings could lead to improved diagnostic accuracy of ARDS and potentially lead to new therapeutic strategies.

Predicting survival and neurological outcome in out-of-hospital cardiac arrest using machine learning: the SCARS model.

BACKGROUND: A prediction model that estimates survival and neurological outcome in out-of-hospital cardiac arrest patients has the potential to improve clinical management in emergency rooms. METHODS: We used the Swedish Registry for Cardiopulmonary Resuscitation to study all out-of-hospital cardiac arrest (OHCA) cases in Sweden from 2010 to 2020. We had 393 candidate predictors describing the circumstances at cardiac arrest, critical time intervals, patient demographics, initial presentation, spatiotemporal data, socioeconomic status, medications, and comorbidities before arrest. To develop, evaluate and test an array of prediction models, we created stratified (on the outcome measure) random samples of our study population. We created a training set (60% of data), evaluation set (20% of data), and test set (20% of data). We assessed the 30-day survival and cerebral performance category (CPC) score at discharge using several machine learning frameworks with hyperparameter tuning. Parsimonious models with the top 1 to 20 strongest predictors were tested. We calibrated the decision threshold to assess the cut-off yielding 95% sensitivity for survival. The final model was deployed as a web application. FINDINGS: We included 55,615 cases of OHCA. Initial presentation, prehospital interventions, and critical time intervals variables were the most important. At a sensitivity of 95%, specificity was 89%, positive predictive value 52%, and negative predictive value 99% in test data to predict 30-day survival. The area under the receiver characteristic curve was 0.97 in test data using all 393 predictors or only the ten most important predictors. The final model showed excellent calibration. The web application allowed for near-instantaneous survival calculations. INTERPRETATION: Thirty-day survival and neurological outcome in OHCA can rapidly and reliably be estimated during ongoing cardiopulmonary resuscitation in the emergency room using a machine learning model incorporating widely available variables. FUNDING: Swedish Research Council (2019-02019); Swedish state under the agreement between the Swedish government, and the county councils (ALFGBG-971482); The Wallenberg Centre for Molecular and Translational Medicine.

Plasma neurofilament light is a predictor of neurological outcome 12 h after cardiac arrest.

BACKGROUND: Previous studies have reported high prognostic accuracy of circulating neurofilament light (NfL) at 24-72 h after out-of-hospital cardiac arrest (OHCA), but performance at earlier time points and after in-hospital cardiac arrest (IHCA) is less investigated. We aimed to assess plasma NfL during the first 48 h after OHCA and IHCA to predict long-term outcomes. METHODS: Observational multicentre cohort study in adults admitted to intensive care after cardiac arrest. NfL was retrospectively analysed in plasma collected on admission to intensive care, 12 and 48 h after cardiac arrest. The outcome was assessed at two to six months using the Cerebral Performance Category (CPC) scale, where CPC 1-2 was considered a good outcome and CPC 3-5 a poor outcome. Predictive performance was measured with the area under the receiver operating characteristic curve (AUROC). RESULTS: Of 428 patients, 328 (77%) suffered OHCA and 100 (23%) IHCA. Poor outcome was found in 68% of OHCA and 55% of IHCA patients. The overall prognostic performance of NfL was excellent at 12 and 48 h after OHCA, with AUROCs of 0.93 and 0.97, respectively. The predictive ability was lower after IHCA than OHCA at 12 and 48 h, with AUROCs of 0.81 and 0.86 (p ≤ 0.03). AUROCs on admission were 0.77 and 0.67 after OHCA and IHCA, respectively. At 12 and 48 h after OHCA, high NfL levels predicted poor outcome at 95% specificity with 70 and 89% sensitivity, while low NfL levels predicted good outcome at 95% sensitivity with 71 and 74% specificity and negative predictive values of 86 and 88%. CONCLUSIONS: The prognostic accuracy of NfL for predicting good and poor outcomes is excellent as early as 12 h after OHCA. NfL is less reliable for the prediction of outcome after IHCA.

Intensive care unit burden is associated with increased mortality in critically ill COVID-19 patients.

BACKGROUND: Traditional models to predict intensive care outcomes do not perform well in COVID-19. We undertook a comprehensive study of factors affecting mortality and functional outcome after severe COVID-19. METHODS: In this prospective multicentre cohort study, we enrolled laboratory-confirmed, critically ill COVID-19 patients at six ICUs in the Skåne Region, Sweden, between May 11, 2020, and May 10, 2021. Demographics and clinical data were collected. ICU burden was defined as the total number of ICU-treated COVID-19 patients in the region on admission. Surviving patients had a follow-up at 90 days for assessment of functional outcome using the Glasgow Outcome Scale-Extended (GOSE), an ordinal scale (1-8) with GOSE ≥5 representing a favourable outcome. The primary outcome was 90-day mortality; the secondary outcome was functional outcome at 90 days. RESULTS: Among 498 included patients, 74% were male with a median age of 66 years and a median body mass index (BMI) of 30 kg/m2 . Invasive mechanical ventilation was employed in 72%. Mortality in the ICU, in-hospital and at 90 days was 30%, 38% and 39%, respectively. Mortality increased markedly at age 60 and older. Increasing ICU burden was independently associated with a two-fold increase in mortality. Higher BMI was not associated with increased mortality. Besides age and ICU burden, smoking status, cortisone use, Pa CO2 >7 kPa, and inflammatory markers on admission were independent factors of 90-day mortality. Lower GOSE at 90 days was associated with a longer stay in the ICU. CONCLUSION: In critically ill COVID-19 patients, the 90-day mortality was 39% and increased considerably at age 60 or older. The ICU burden was associated with mortality, whereas a high BMI was not. A longer stay in the ICU was associated with unfavourable functional outcomes at 90 days.

ABO and RhD blood group are not associated with mortality and morbidity in critically ill patients; a multicentre observational study of 29 512 patients.

BACKGROUND: The ABO and RhD blood group represent antigens on the surface of erythrocytes. The ABO blood group antigens are also present on multiple other cells. Interestingly, previous studies have demonstrated associations between the blood group and many types of disease. The present study aimed to identifying associations between the ABO blood group, the RhD blood group, and morbidity and mortality in a mixed cohort and in six pre-defined subgroups of critically ill patients. METHODS: Adult patients admitted to any of the five intensive care units (ICUs) in the Scania Region, Sweden, between February 2007 and April 2021 were eligible for inclusion. The outcomes were mortality analysed at 28- and 90-days as well as at the end of observation and morbidity measured using days alive and free of (DAF) invasive ventilation (DAF ventilation) and DAF circulatory support, including vasopressors or inotropes (DAF circulation), maximum Sequential Organ Failure Assessment score (SOFAmax) the first 28 days after admission and length of stay. All outcomes were analysed in separate multivariable regression models adjusted for age and sex. In addition, in a sensitivity analysis, five subgroups of patients with the main diagnoses sepsis, septic shock, acute respiratory distress syndrome, cardiac arrest and trauma were analysed using the same separate multivariable regression models. RESULTS: In total, 29,512 unique patients were included in the analyses. There were no significant differences for any of the outcomes between non-O blood groups and blood group O, or between RhD blood groups. In the sensitivity analysis of subgroups, there were no differences in mortality between non-O blood groups and blood group O or between the RhD blood groups. AB was the most common blood group in the COVID-19 cohort. CONCLUSIONS: The ABO and RhD blood group do not influence mortality or morbidity in a general critically ill patient population.

Neuron-specific enolase and long-term neurological outcome after OHCA - A validation study.

AIMS: To investigate what NSE levels predict long-term neurological prognosis at 24, 48 and 72 hours after ROSC in a cohort of out-of-hospital cardiac arrest and to validate previously suggested NSE cut-offs, including the latest ERC guidelines (2021). METHODS: Patients admitted to intensive care units in four hospitals in Southern Sweden between 2014-2018 were included. Blood samples were handled by a single local laboratory. The primary outcome was neurological outcome according to the Cerebral Performance Category (CPC) scale at 2-6 months after cardiac arrest. RESULTS: 368 patients were included for analysis. A ≤2% false positive rate for the prediction of poor neurological outcome was achieved with an NSE cut-off value of >101 µg/L at 48 hours and >80 µg/L at 72 hours. The cut-off suggested by the recent ERC guidelines of >60 µg/L at 48 and/or 72 hours generated a false positive rate of 4.3% (95 %CI 0.9-7.4%). CONCLUSION: A local validation study of the ability of serum levels of neuron-specific enolase to predict long-term poor neurological outcome after out-of-hospital cardiac arrest generated higher cut-offs than suggested by previous publications.

Circulating dipeptidyl peptidase 3 on intensive care unit admission is a predictor of organ dysfunction and mortality.

BACKGROUND: Our aim was to investigate the prognostic potential of circulating dipeptidyl peptidase 3 (cDPP3) to predict mortality and development of organ dysfunction in a mixed intensive care unit (ICU) population, and for this reason, we analysed prospectively collected admission blood samples from adult ICU patients at four Swedish hospitals. Blood samples were stored in a biobank for later batch analysis. The association of cDPP3 levels with 30-day mortality and Sequential Organ Failure Assessment (SOFA) scores on day two was investigated before and after adjustment for the simplified acute physiology score III (SAPS-3), using multivariable (ordinal) logistic regression. The predictive power of cDPP3 was assessed using the area under the receiver operating characteristic curve (AUROC). RESULTS: Of 1978 included consecutive patients in 1 year (2016), 632 fulfilled the sepsis 3-criteria, 190 were admitted after cardiac arrest, and 157 because of trauma. Admission cDPP3 was independently (of SAPS-3) associated with 30-day mortality with odds ratios of 1.45 (95% confidence interval (CI) 1.28-1.64) in the entire ICU population, 1.30 (95% CI 1.08-1.57) in the sepsis subgroup and 2.28 (95% CI 1.50-3.62) in cardiac arrest. For trauma, there was no clear association. Circulating DPP3 alone was a moderate predictor of 30-day mortality with AUROCs of 0.68, 0.62, and 0.72 in the entire group, the sepsis subgroup, and the cardiac arrest subgroup, respectively. By adding cDPP3 to SAPS-3, AUROC improved for the entire group, the sepsis subgroup, and the cardiac arrest subgroup (p = 0.023). CONCLUSION: Circulating DPP3 on admission is a SAPS-3 independent prognostic factor of day-two organ dysfunction and 30-day mortality in a mixed ICU population and needs further evaluation.

Plasma proenkephalin A 119-159 on intensive care unit admission is a predictor of organ failure and 30-day mortality.

BACKGROUND: Proenkephalin A 119-159 (penKid) has been suggested as a marker of renal failure and poor outcome. We aimed to investigate the association of penKid on ICU admission with organ dysfunction and mortality in a mixed ICU population. In this retrospective, observational study, admission penKid levels from prospectively collected blood samples of consecutive patients admitted to four Swedish ICUs were analysed. The association of penKid with day-two sequential organ failure assessment (SOFA) scores and 30-day mortality was investigated using (ordinal) logistic regression. The predictive power of penKid for 30-day mortality and dialysis was assessed using the area under the receiver operating characteristic curve (AUC). RESULTS: Of 1978 included patients, 632 fulfilled the sepsis 3-criteria, 190 had a cardiac arrest, and 157 had experienced trauma. Admission penKid was positively associated with 30-day mortality with an odds ratio of 1.95 (95% confidence interval 1.75-2.18, p < 0.001), and predicted 30-day mortality in the entire ICU population with an AUC of 0.71 (95% confidence interval 0.68-0.73) as well as in the sepsis, cardiac arrest and trauma subgroups (AUCs of 0.61-0.84). Correction for admission plasma creatinine revealed that penKid correlated with neurological dysfunction. CONCLUSION: Plasma penKid on ICU admission is associated with day-two organ dysfunction and predictive of 30-day mortality in a mixed ICU-population, as well as in sepsis, cardiac arrest and trauma subgroups. In addition to being a marker of renal dysfunction, plasma penKid is associated with neurologic dysfunction in the entire ICU population, and cardiovascular dysfunction in sepsis.

Proteome Profiling of Recombinant DNase Therapy in Reducing NETs and Aiding Recovery in COVID-19 Patients.

Severe coronavirus disease 2019 (COVID-19) can result in pneumonia and acute respiratory failure. Accumulation of mucus in the airways is a hallmark of the disease and can result in hypoxemia. Here, we show that quantitative proteome analysis of the sputum from severe patients with COVID-19 reveal high levels of neutrophil extracellular trap (NET) components, which was confirmed by microscopy. Extracellular DNA from excessive NET formation can increase sputum viscosity and lead to acute respiratory distress syndrome. Recombinant human DNase (Pulmozyme; Roche) has been shown to be beneficial in reducing sputum viscosity and improve lung function. We treated five patients pwith COVID-19 resenting acute symptoms with clinically approved aerosolized Pulmozyme. No adverse reactions to the drug were seen, and improved oxygen saturation and recovery in all severely ill patients with COVID-19 was observed after therapy. Immunofluorescence and proteome analysis of sputum and blood plasma samples after treatment revealed a marked reduction of NETs and a set of statistically significant proteome changes that indicate reduction of hemorrhage, plasma leakage and inflammation in the airways, and reduced systemic inflammatory state in the blood plasma of patients. Taken together, the results indicate that NETs contribute to acute respiratory failure in COVID-19 and that degrading NETs may reduce dependency on external high-flow oxygen therapy in patients. Targeting NETs using recombinant human DNase may have significant therapeutic implications in COVID-19 disease and warrants further studies.

Plasma proenkephalin A 119-159 and dipeptidyl peptidase 3 on admission after cardiac arrest help predict long-term neurological outcome.

BACKGROUND: A large proportion of adult survivors of cardiac arrest have a poor neurological outcome. Guidelines recommend multimodal neuro-prognostication no earlier than 72-96 h after cardiac arrest. There is great interest in earlier prognostic markers, including very early markers at admission. The novel blood biomarkers proenkephalin A 119-159 (penKid), bioactive adrenomedullin (bio-ADM) and circulating dipeptidyl peptidase 3 (cDPP3) have not been previously investigated for the early prognosis of cardiac arrest survivors. METHODS: This multicentre observational study included adult survivors of cardiac arrest admitted to intensive care at four Swedish intensive care units (ICUs) during 2016. Blood samples were collected at ICU admission and batch analysed. The association between admission plasma penKid, bio-ADM and cDPP3 and poor long-term neurological outcome, according to the Cerebral Performance Category (CPC) scale, was assessed by binary logistic regression. Their prognostic performance was assessed using the area under the receiver operating characteristic curve (AUC). RESULTS: A total of 190 patients were included, of which 136 patients had suffered out-of-hospital and 54 patients in-hospital cardiac arrest. Poor long-term neurological outcome was associated with elevated admission plasma concentrations of penKid and cDPP3, but not with bio-ADM. The association for penKid, but not for cDPP3, remained after adjusting for clinical cardiac arrest variables with prognostic value (time to return of spontaneous circulation (ROSC), initial rhythm, admission Glasgow coma scale (GCS) motor score and absence of pupillary reflexes). The prognostic performance of above mentioned clinical cardiac arrest variables alone was very good with an AUC of 0.90 (95% confidence interval, CI, 0.86-0.95), but improved further with the addition of penKid resulting in an AUC of 0.93 (95% CI 0.89-0.97, p < 0.026). Plasma penKid and cDPP3 alone provided moderate long-term prognostic information with AUCs of 0.70 and 0.71, respectively. CONCLUSION: After cardiac arrest, admission plasma levels of penKid and cDPP3, but not bio-ADM, predicted long-term neurological outcome. When added to clinical cardiac arrest variables, penKid further improved prognostic performance.

Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm.

BACKGROUND: Prognostication of neurological outcome in patients who remain comatose after cardiac arrest resuscitation is complex. Clinical variables, as well as biomarkers of brain injury, cardiac injury, and systemic inflammation, all yield some prognostic value. We hypothesised that cumulative information obtained during the first three days of intensive care could produce a reliable model for predicting neurological outcome following out-of-hospital cardiac arrest (OHCA) using artificial neural network (ANN) with and without biomarkers. METHODS: We performed a post hoc analysis of 932 patients from the Target Temperature Management trial. We focused on comatose patients at 24, 48, and 72 h post-cardiac arrest and excluded patients who were awake or deceased at these time points. 80% of the patients were allocated for model development (training set) and 20% for internal validation (test set). To investigate the prognostic potential of different levels of biomarkers (clinically available and research-grade), patients' background information, and intensive care observation and treatment, we created three models for each time point: (1) clinical variables, (2) adding clinically accessible biomarkers, e.g., neuron-specific enolase (NSE) and (3) adding research-grade biomarkers, e.g., neurofilament light (NFL). Patient outcome was the dichotomised Cerebral Performance Category (CPC) at six months; a good outcome was defined as CPC 1-2 whilst a poor outcome was defined as CPC 3-5. The area under the receiver operating characteristic curve (AUROC) was calculated for all test sets. RESULTS: AUROC remained below 90% when using only clinical variables throughout the first three days in the ICU. Adding clinically accessible biomarkers such as NSE, AUROC increased from 82 to 94% (p < 0.01). The prognostic accuracy remained excellent from day 1 to day 3 with an AUROC at approximately 95% when adding research-grade biomarkers. The models which included NSE after 72 h and NFL on any of the three days had a low risk of false-positive predictions while retaining a low number of false-negative predictions. CONCLUSIONS: In this exploratory study, ANNs provided good to excellent prognostic accuracy in predicting neurological outcome in comatose patients post OHCA. The models which included NSE after 72 h and NFL on all days showed promising prognostic performance.

Circulating bioactive adrenomedullin as a marker of sepsis, septic shock and critical illness.

BACKGROUND: Biomarkers can be of help to understand critical illness and to identify and stratify sepsis. Adrenomedullin is a vasoactive hormone, with reported prognostic and potentially therapeutic value in sepsis. The primary aim of this study was to investigate the association of circulating bioactive adrenomedullin (bio-ADM) levels at intensive care unit (ICU) admission with mortality in sepsis patients and in a general ICU population. Secondary aims included the association of bio-ADM with organ failure and the ability of bio-ADM to identify sepsis. METHODS: In this retrospective observational study, adult patients admitted to one of four ICUs during 2016 had admission bio-ADM levels analysed. Age-adjusted odds ratios (OR) with 95% CI for log-2 transformed bio-ADM, and Youden's index derived cut-offs were calculated. The primary outcome was 30-day mortality, and secondary outcomes included the need for organ support and the ability to identify sepsis. RESULTS: Bio-ADM in 1867 consecutive patients were analysed; 632 patients fulfilled the sepsis-3 criteria of whom 267 had septic shock. The median bio-ADM in the entire ICU population was 40 pg/mL, 74 pg/mL in sepsis patients, 107 pg/mL in septic shock and 29 pg/mL in non-septic patients. The association of elevated bio-ADM and mortality in sepsis patients and the ICU population resulted in ORs of 1.23 (95% CI 1.07-1.41) and 1.22 (95% CI 1.12-1.32), respectively. The association with mortality remained after additional adjustment for lactate in sepsis patients. Elevated bio-ADM was associated with an increased need for dialysis with ORs of 2.28 (95% CI 2.01-2.59) and 1.97 (95% CI 1.64-2.36) for the ICU population and sepsis patients, respectively, and with increased need of vasopressors, OR 1.33 (95% CI 1.23-1.42) (95% CI 1.17-1.50) for both populations. Sepsis was identified with an OR of 1.78 (95% CI 1.64-1.94) for bio-ADM, after additional adjustment for severity of disease. A bio-ADM cut-off of 70 pg/mL differentiated between survivors and non-survivors in sepsis, but a Youden's index derived threshold of 108 pg/mL performed better. CONCLUSIONS: Admission bio-ADM is associated with 30-day mortality and organ failure in sepsis patients as well as in a general ICU population. Bio-ADM may be a morbidity-independent sepsis biomarker.

Artificial neural networks improve early outcome prediction and risk classification in out-of-hospital cardiac arrest patients admitted to intensive care.

BACKGROUND: Pre-hospital circumstances, cardiac arrest characteristics, comorbidities and clinical status on admission are strongly associated with outcome after out-of-hospital cardiac arrest (OHCA). Early prediction of outcome may inform prognosis, tailor therapy and help in interpreting the intervention effect in heterogenous clinical trials. This study aimed to create a model for early prediction of outcome by artificial neural networks (ANN) and use this model to investigate intervention effects on classes of illness severity in cardiac arrest patients treated with targeted temperature management (TTM). METHODS: Using the cohort of the TTM trial, we performed a post hoc analysis of 932 unconscious patients from 36 centres with OHCA of a presumed cardiac cause. The patient outcome was the functional outcome, including survival at 180 days follow-up using a dichotomised Cerebral Performance Category (CPC) scale with good functional outcome defined as CPC 1-2 and poor functional outcome defined as CPC 3-5. Outcome prediction and severity class assignment were performed using a supervised machine learning model based on ANN. RESULTS: The outcome was predicted with an area under the receiver operating characteristic curve (AUC) of 0.891 using 54 clinical variables available on admission to hospital, categorised as background, pre-hospital and admission data. Corresponding models using background, pre-hospital or admission variables separately had inferior prediction performance. When comparing the ANN model with a logistic regression-based model on the same cohort, the ANN model performed significantly better (p = 0.029). A simplified ANN model showed promising performance with an AUC above 0.852 when using three variables only: age, time to ROSC and first monitored rhythm. The ANN-stratified analyses showed similar intervention effect of TTM to 33 °C or 36 °C in predefined classes with different risk of a poor outcome. CONCLUSION: A supervised machine learning model using ANN predicted neurological recovery, including survival excellently, and outperformed a conventional model based on logistic regression. Among the data available at the time of hospitalisation, factors related to the pre-hospital setting carried most information. ANN may be used to stratify a heterogenous trial population in risk classes and help determine intervention effects across subgroups.