Analysis of physiologic alterations in intensive care unit patients and their relationship with mortality

Machine learning-based prediction of in-hospital mortality for critically ill patients with sepsis-associated acute kidney injury

OBJECTIVES

This study aims to develop and validate a prediction model in-hospital mortality in critically ill patients with sepsis-associated acute kidney injury (SA-AKI) based on machine learning algorithms.

METHODS

Patients who met the criteria for inclusion were identified in the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database and divided according to the validation (n = 2440) and development (n = 9756, 80%) queues. Ensemble stepwise feature selection method was used to screen for effective features. The prediction models of short-term mortality were developed by seven machine learning algorithms. Ten-fold cross-validation was used to verify the performance of the algorithm in the development queue. The area under the receiver operating characteristic curve (ROC-AUC) was used to evaluate the differentiation accuracy and performance of the prediction model in the validation queue. The best-performing model was interpreted by Shapley additive explanations (SHAP).

RESULTS

A total of 12,196 patients were enrolled in this study. Eleven variables were finally chosen to develop the prediction model. The AUC of the random forest (RF) model was the highest value both in the Ten-fold cross-validation and evaluation (AUC: 0.798, 95% CI: 0.774-0.821). According to the SHAP plots, old age, low Glasgow Coma Scale (GCS) score, high AKI stage, reduced urine output, high Simplified Acute Physiology Score (SAPS II), high respiratory rate, low temperature, low absolute lymphocyte count, high creatinine level, dysnatremia, and low body mass index (BMI) increased the risk of poor prognosis.

CONCLUSIONS

The RF model developed in this study is a good predictor of in-hospital mortality for patients with SA-AKI in the intensive care unit (ICU), which may have potential applications in mortality prediction.

Shared decision-making in the ICU from the perspective of physicians, nurses and patients: a qualitative interview study

OBJECTIVE

To identify views, experiences and needs for shared decision-making (SDM) in the intensive care unit (ICU) according to ICU physicians, ICU nurses and former ICU patients and their close family members.

DESIGN

Qualitative study.

SETTING

Two Dutch tertiary centres.

PARTICIPANTS

19 interviews were held with 29 participants: seven with ICU physicians from two tertiary centres, five with ICU nurses from one tertiary centre and nine with former ICU patients, of whom seven brought one or two of their close family members who had been involved in the ICU stay.

RESULTS

Three themes, encompassing a total of 16 categories, were identified pertaining to struggles of ICU physicians, needs of former ICU patients and their family members and the preferred role of ICU nurses. The main struggles ICU physicians encountered with SDM include uncertainty about long-term health outcomes, time constraints, feeling pressure because of having final responsibility and a fear of losing control. Former patients and family members mainly expressed aspects they missed, such as not feeling included in ICU treatment decisions and a lack of information about long-term outcomes and recovery. ICU nurses reported mainly opportunities to strengthen their role in incorporating non-medical information in the ICU decision-making process and as liaison between physicians and patients and family.

CONCLUSIONS

Interviewed stakeholders reported struggles, needs and an elucidation of their current and preferred role in the SDM process in the ICU. This study signals an essential need for more long-term outcome information, a more informal inclusion of patients and their family members in decision-making processes and a more substantial role for ICU nurses to integrate patients' values and needs in the decision-making process.

Nursing workload assessment in an intensive care unit: A 5-year retrospective analysis

AIMS

To study the correlation between the workload of intensive care nursing teams and the sociodemographic, anthropometric and clinical characteristics of patients in critical condition in a Portuguese Intensive Care Unit (ICU) during a 5-year period.

BACKGROUND

Currently, indices of nursing workload quantification are one of the resources used for planning and evaluating ICUs. Evidence shows that there are several factors related to critical patients and their hospitalisation which potentially influence the nursing workload.

DESIGN

Retrospective cohort analysis of a health record database from adult patients admitted to a Portuguese ICU between 1 January 2015-31 December 2019.

METHODS

Simplified Therapeutic Intervention Scoring System (TISS-28) scores of 730 adult patients. Three TISS-28 assessments were considered: first assessment, last assessment and average. The STROBE guidelines were used in reporting this study.

RESULTS

The TISS-28 has an average of 34.2 ± 6.9 points at admission, which is considered a high nursing workload. A somewhat lower result was found for the discharge and average assessments. It shows that basic activities accounted for the highest percentage of time spent (38.0%), followed by the cardiovascular support category (26.5%). The TISS-28 shows consistent results throughout the study period, despite a small trend reduction in the last 2 years.

CONCLUSIONS

Lower workloads were found for age ≤44 years and with a shorter length of stay. Higher workload was more probable in patients classified in Cullen Class IV (OR = 2.5) and with a normal to higher weight percentile (OR = 1.9 and 1.5, respectively).

RELEVANCE TO CLINICAL PRACTICE

Knowledge of the factors influencing the nursing workload facilitates the implementation of rules to improve performance in nursing interventions, based on the redefinition of care priorities, increased productivity, human resources management and reduction of additional costs to the organisation, related to possible adverse events, among others.

Validity of Sequential Organ Failure Assessment and Quick Sequential Organ Failure Assessment in Assessing Mortality Rate in the Intensive Care Unit With or Without Sepsis

Introduction

Sepsis and septic shock (sepsis-induced hypotension not improved by adequate fluid resuscitation) are among the most common reasons for admission to an intensive care unit (ICU) and display high mortality rates. Different scoring systems are used to diagnose and predict the mortality of patients having sepsis. This study aims to validate the prognostic accuracy of Sequential Organ Failure Assessment (SOFA) and Quick Sequential Organ Failure Assessment (qSOFA) in determining the mortality of both septic and non-septic patients.

Materials and methods

This retrospective cohort study was conducted in May 2018 in the Surgical Intensive Care Unit (SICU) of a tertiary care hospital in Karachi, Pakistan. Past 200 patient records, from January 2018 to April 2018, were examined, and 20 records were discarded due to insufficient data. Sufficient observational data were collected, which was used to assess the validity of the SOFA and qSOFA in determining the mortality rate of sepsis. A comparison of the two modalities was made.

Results

Out of the 200 patients, 180 were enrolled. Data from their entire ICU stay were used to calculate their initial, highest, and mean SOFA and qSOFA. Mean SOFA score up to nine correlated with a mortality rate of up to <79%, while scores 10 and above predicted a 100% mortality rate. A mean qSOFA score of three predicted a 67% mortality rate. Univariate logistic analysis performed with odds ratio showed that the mean qSOFA score was in comparison more closely able to predict mortality, followed by mean SOFA score (p values < 0.01).

Conclusions

This study concluded that both SOFA and qSOFA scores are good predictors of mortality. However, qSOFA is more closely accurate in predicting mortality than SOFA. But further analysis with larger sample size for a longer duration as well as the application of these scores in the emergency departments and general wards can prove the precision of this study.

Testing the Reliability and Validity of the Nursing Activities Score in Critical Care Nursing

BACKGROUND AND PURPOSE

The purpose of this study was to test the reliability and validity of the conceptual model of critical care nursing using the Nursing Activities Score (NAS) for intensive care unit (ICU) patients.

METHODS

An observational study conducted in 2011 using the NAS scores of 219 patients in Norway. The inter-rater reliability (IRR) was tested by parallel classifications. The validity was explored by an exploratory factor analysis (EFA) and a confirmatory factor analysis (CFA) as the measurement models in the structural equation model.

RESULTS

Within the paired ratings the Spearman's correlation coefficient was 0.39. The EFA results explained 77% of the variance with six factors. The reduced CFA model resulted in a three-factor model: relationship, prevention and treatment.

CONCLUSION

The findings supported the IRR and construct validity of the conceptual model of the NAS.

Mortality Risk Profiles for Sepsis: A Novel Longitudinal and Multivariable Approach

Supplemental Digital Content is available in the text.

To determine if a set of time-varying biological indicators can be used to: 1) predict the sepsis mortality risk over time and 2) generate mortality risk profiles.

SOFA Score prognostic performance among patients admitted to High-Dependency Units

BACKGROUND

The aim of this study was to assess prognostic stratification in patients admitted in two Italian Emergency-Department High-Dependency Units (ED-HDU).

METHODS

From June 2014, to July 2016, we recorded all patients admitted in the ED-HDU of the Careggi University Hospital and the Vittorio Emanuele University Hospital in a standardized database. Charlson Index and SOFA Score were calculated to evaluate comorbidity burden and severity of organ dysfunction. End-points were HDU and in-hospital mortality rate and need of Intensive Care Unit (ICU) transfer.

RESULTS

The overall number of patients admitted in the two Units was 3311, 1822 in Florence and 1489 in Catania. HDU mortality rate was 5% (N.=171); compared with survivors, non-survivors showed a higher SOFA Score (10.0±4.2 vs. 3.5±2.9, P<0.001) and a higher number of organ dysfunctions (1.6±0.9 vs. 0.6±0.8, P<0.001). All patients with a SOFA Score in the first and second quartile survived HDU admission (only two non-survivors among patients in the second quartile), while mortality was disproportionally high in the group with a score value in the fourth quartile (0%, 0.2%, 3% and 14%, P<0.001). Presence and number of organ failure, as well as SOFA Score (5.6±4.0 vs. 3.4±2.8, P<0.001), were significantly higher in patients transferred to ICU than in those admitted in an ordinary ward or discharged. A higher SOFA Score (RR 1.55, 95% CI: 1.47-1.63, P<0.001) was associated with an increased HDU mortality, independent of age and Charlson Index.

CONCLUSIONS

SOFA Score showed a good discrimination ability for both HDU - mortality and indication to increase the level of care.

Prognostic accuracy of the sequential organ failure assessment (SOFA) and quick SOFA for mortality in cancer patients with sepsis defined by systemic inflammatory response syndrome (SIRS)

PURPOSE

We aimed to assess the prognostic accuracy of SOFA and qSOFA scores in cancer patients with sepsis, and also to determine if the addition of hyperlactatemia to qSOFA increases the accuracy in predicting the 30-day mortality.

MATERIAL AND METHOD

We consecutively included adult active cancer patients (age ≥ 18 years) with sepsis defined by SIRS who visited the emergency department (ED) from May 1st to July 30th, 2017. Data were collected retrospectively through reviewing medical records. The SOFA and qSOFA scores were calculated with the initial variables at the time of ED admission. The primary endpoint was 30-day mortality.

RESULT

Of 1137 screened, 301 were included. The 30-day mortality was 14.3% (43 patients). Among the total 301, the SOFA score was ≥ 2 in 168 and qSOFA ≥ 2 in 23. For those with SOFA ≥ 2 and < 2, the mortality was 23.2% and 3%, respectively (P < 0.001). For those with qSOFA ≥ 2 and < 2, the mortality was 47.8% and 11.5%, respectively (P < 0.001). The AUROC of 30-day mortality for qSOFA was lower than that for SOFA (0.66 (95% CI, 0.56-0.75) vs. 0.79 (95% CI, 0.72-0.87), P = 0.004)). However, the combination of qSOFA with lactate ≥ 2 threshold considerably enhanced a discrimination capacity for mortality with an AUROC 0.77 (95% CI, 0.69-0.85), which was similar to SOFA (P = 0.11).

CONCLUSION

In cancer patients with sepsis, qSOFA was inferior to SOFA in predicting mortality. However, adding lactate to qSOFA resulted in greater prognostic accuracy for short-term mortality, comparable with SOFA.

Outcomes of Patients Presenting with Mild Acute Respiratory Distress Syndrome

Editor’s Perspective

What We Already Know about This Topic

What This Article Tells Us That Is New

Background

Patients with initial mild acute respiratory distress syndrome are often underrecognized and mistakenly considered to have low disease severity and favorable outcomes. They represent a relatively poorly characterized population that was only classified as having acute respiratory distress syndrome in the most recent definition. Our primary objective was to describe the natural course and the factors associated with worsening and mortality in this population.

Methods

This study analyzed patients from the international prospective Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure (LUNG SAFE) who had initial mild acute respiratory distress syndrome in the first day of inclusion. This study defined three groups based on the evolution of severity in the first week: “worsening” if moderate or severe acute respiratory distress syndrome criteria were met, “persisting” if mild acute respiratory distress syndrome criteria were the most severe category, and “improving” if patients did not fulfill acute respiratory distress syndrome criteria any more from day 2.

Results

Among 580 patients with initial mild acute respiratory distress syndrome, 18% (103 of 580) continuously improved, 36% (210 of 580) had persisting mild acute respiratory distress syndrome, and 46% (267 of 580) worsened in the first week after acute respiratory distress syndrome onset. Global in-hospital mortality was 30% (172 of 576; specifically 10% [10 of 101], 30% [63 of 210], and 37% [99 of 265] for patients with improving, persisting, and worsening acute respiratory distress syndrome, respectively), and the median (interquartile range) duration of mechanical ventilation was 7 (4, 14) days (specifically 3 [2, 5], 7 [4, 14], and 11 [6, 18] days for patients with improving, persisting, and worsening acute respiratory distress syndrome, respectively). Admissions for trauma or pneumonia, higher nonpulmonary sequential organ failure assessment score, lower partial pressure of alveolar oxygen/fraction of inspired oxygen, and higher peak inspiratory pressure were independently associated with worsening.

Conclusions

Most patients with initial mild acute respiratory distress syndrome continue to fulfill acute respiratory distress syndrome criteria in the first week, and nearly half worsen in severity. Their mortality is high, particularly in patients with worsening acute respiratory distress syndrome, emphasizing the need for close attention to this patient population.

Feature selection for the accurate prediction of septic and cardiogenic shock ICU mortality in the acute phase

Circulatory shock is a life-threatening disease that accounts for around one-third of all admissions to intensive care units (ICU). It requires immediate treatment, which is why the development of tools for planning therapeutic interventions is required to deal with shock in the critical care environment. In this study, the ShockOmics European project original database is used to extract attributes capable of predicting mortality due to shock in the ICU. Missing data imputation techniques and machine learning models were used, followed by feature selection from different data subsets. Selected features were later used to build Bayesian Networks, revealing causal relationships between features and ICU outcome. The main result is a subset of predictive features that includes well-known indicators such as the SOFA and APACHE II scores, but also less commonly considered ones related to cardiovascular function assessed through echocardiograpy or shock treatment with pressors. Importantly, certain selected features are shown to be most predictive at certain time-steps. This means that, as shock progresses, different attributes could be prioritized. Clinical traits obtained at 24h. from ICU admission are shown to accurately predict cardiogenic and septic shock mortality, suggesting that relevant life-saving decisions could be made shortly after ICU admission.

Addition of lactic acid levels improves the accuracy of quick sequential organ failure assessment in predicting mortality in surgical patients with complicated intra-abdominal infections: a retrospective study

Background

The quick sequential organ failure assessment (qSOFA) alone has a poor sensitivity for predicting mortality in patients with complicated intra-abdominal infections, and plasma lactate levels have been shown to have a strong association with mortality in critically ill patients. Therefore, this study aimed to compare the performance of qSOFA with a score derived from a combination of qSOFA and serum lactate levels for predicting mortality in surgical patients with complicated intra-abdominal infections.

Methods

This retrospective study was performed at a university hospital. The medical records of 457 patients who presented to the emergency department (ED) between January 2008 and December 2016 and required emergency gastrointestinal surgery for a complicated intra-abdominal infection were reviewed retrospectively. qSOFA criteria, sequential organ failure assessment (SOFA) scores, and plasma lactate levels during their ED stay were collected. We performed area under receiver operating characteristic (AUROC) curve and sensitivity analysis to compare the performance of qSOFA alone with that of a score derived from the use of a combination of the qSOFA and lactate levels for predicting patient mortality.

Results

Fifty patients (10.9%) died during hospitalization. The combined qSOFA and lactate level score was superior to qSOFA alone (AUROC = 0.754 vs. 0.717, p = 0.039, respectively) and comparable to the full SOFA score (AUROC = 0.754 vs. 0.795, p = 0.127, respectively) in predicting mortality. Sensitivity and specificity of qSOFA alone were 46 and 86%, respectively, and those of the combined score were 72 and 73%, respectively (p < 0.001).

Conclusion

A score derived from the qSOFA and serum lactate levels had better predictive performance with higher sensitivity than the qSOFA alone in predicting mortality in patients with complicated intra-abdominal infections and had a comparable predictive performance to that of the full SOFA score.

Patients Admitted to Three Spanish Intensive Care Units for Poisoning: Type of Poisoning, Mortality, and Functioning of Prognostic Scores Commonly Used

Objectives. To evaluate the gravity and mortality of those patients admitted to the intensive care unit for poisoning. Also, the applicability and predicted capacity of prognostic scales most frequently used in ICU must be evaluated. Methods. Multicentre study between 2008 and 2013 on all patients admitted for poisoning. Results. The results are from 119 patients. The causes of poisoning were medication, 92 patients (77.3%), caustics, 11 (9.2%), and alcohol, 20 (16,8%). 78.3% attempted suicides. Mean age was 44.42 ± 13.85 years. 72.5% had a Glasgow Coma Scale (GCS) ≤8 points. The ICU mortality was 5.9% and the hospital mortality was 6.7%. The mortality from caustic poisoning was 54.5%, and it was 1.9% for noncaustic poisoning (p < 0.001). After adjusting for SAPS-3 (OR: 1.19 (1.02–1.39)) the mortality of patients who had ingested caustics was far higher than the rest (OR: 560.34 (11.64–26973.83)). There was considerable discrepancy between mortality predicted by SAPS-3 (26.8%) and observed (6.7%) (Hosmer-Lemeshow test: H = 35.10; p < 0.001). The APACHE-II (7,57%) and APACHE-III (8,15%) were no discrepancies. Conclusions. Admission to ICU for poisoning is rare in our country. Medication is the most frequent cause, but mortality of caustic poisoning is higher. APACHE-II and APACHE-III provide adequate predictions about mortality, while SAPS-3 tends to overestimate.

Daily estimation of the severity of organ dysfunctions in critically ill children by using the PELOD-2 score

INTRODUCTION

Daily or serial evaluation of multiple organ dysfunction syndrome (MODS) scores may provide useful information. We aimed to validate the daily (d) PELOD-2 score using the set of seven days proposed with the previous version of the score.

METHODS

In all consecutive patients admitted to nine pediatric intensive care units (PICUs) we prospectively measured the dPELOD-2 score at day 1, 2, 5, 8, 12, 16, and 18. PICU mortality was used as the outcome dependent variable. The discriminant power of the dPELOD-2 scores was estimated using the area under the ROC curve and the calibration using the Hosmer-Lemeshow chi-square test. We used a logistic regression to investigate the relationship between the dPELOD-2 scores and outcome, and between the change in PELOD-2 score from day 1 and outcome.

RESULTS

We included 3669 patients (median age 15.5 months, mortality rate 6.1%, median length of PICU stay 3 days). Median dPELOD-2 scores were significantly higher in nonsurvivors than in survivors (p < 0.0001). The dPELOD-2 score was available at least at day 2 in 2057 patients: among the 796 patients without MODS on day 1, 186 (23.3%) acquired the syndrome during their PICU stay (mortality 4.9% vs. 0.3% among the 610 who did not; p < 0.0001). Among the 1261 patients with MODS on day 1, the syndrome worsened in 157 (12.4%) and remained unchanged or improved in 1104 (87.6%) (mortality 22.9% vs. 6.6%; p < 0.0001). The AUC of the dPELOD-2 scores ranged from 0.75 (95% CI: 0.67-0.83) to 0.89 (95% CI: 0.86-0.91). The calibration was good with a chi-square test between 13.5 (p = 0.06) and 0.9 (p = 0.99). The PELOD-2 score on day 1 was a significant prognostic factor; the serial evaluation of the change in the dPELOD-2 score from day1, adjusted for baseline value, demonstrated a significant odds ratio of death for each of the 7 days.

CONCLUSION

This study suggests that the progression of the severity of organ dysfunctions can be evaluated by measuring the dPELOD-2 score during a set of 7 days in PICU, providing useful information on outcome in critically ill children. Its external validation would be useful.

Which risk-adjustment index performs better in predicting 30-day mortality? A systematic review and meta-analysis

RATIONALE, AIMS AND OBJECTIVES

Individual comparisons of the performance of risk-adjustment indices have been widely conducted. Few reviews have been conducted to summarize the performance of different risk-adjustment indices. A 30-day mortality rate is widely used to evaluate the quality of care in hospitals by federal agencies like the Centers for Medicare and Medicaid Services. This study examined relative performance of risk-adjustment indices that predict 30-day mortality.

METHODS

Databases including Medline, PubMed and PsycINFO were searched for studies that compared risk-adjustment indices. The search protocol included comparative studies in which the performance of risk-adjustment indices were compared across any defined cohort to compare 30-day mortality, including mortality within 30 days and intensive care unit mortality, which lasts less than 30 days. Data were extracted using a structured form and abstract data included author and publication year, population studied (including location, sample size, study time period), comparison indices, outcome studied, results and conclusions from the results. A meta-analytical approach was used to summarize all the studies. Scaled ranking score was used to estimate the relative superiority of any given risk-adjustment indices. A hypergeometric test was carried out to evaluate the performance of risk-adjustment measures.

RESULTS

Out of 2805 studies identified, 23 studies met the eligibility criteria. Main risk-adjustment indices used for comparison included Acute Physiology and Chronic Health Evaluation, Sequential Organ Failure Assessment score, Charlson co-morbidity index, Model for End-Stage Liver Disease score and Simplified Acute Physiology Score (SAPS). Based on scaled ranking score, SAPS performed best (score 0.510) among all the risk-adjustment indices. However, based on hypergeometric test, the five measures performed equally well.

CONCLUSIONS

Although all the selected risk-adjustment indices perform equally well, SAPS seems better than other indices for short-term mortality based on scaled ranking score.

Evaluation of the PaO2/FiO2 ratio after cardiac surgery as a predictor of outcome during hospital stay

BACKGROUND

The arterial partial pressure of O2 and the fraction of inspired oxygen (PaO2/FiO2) ratio is widely used in ICUs as an indicator of oxygenation status. Although cardiac surgery and ICU scores can predict mortality, during the first hours after cardiac surgery few instruments are available to assess outcome. The aim of this study was to evaluate the usefulness of PaO2/FIO2 ratio to predict mortality in patients immediately after cardiac surgery.

METHODS

We prospectively studied 2725 consecutive cardiac surgery patients between 2004 and 2009. PaO2/FiO2 ratio was measured on admission and at 3 h, 6 h, 12 h and 24 h after ICU admission, together with clinical data and outcomes.

RESULTS

All PaO2/FIO2 ratio measurements differed between survivors and non-survivors (p < 0.001). The PaO2/FIO2 at 3 h after ICU admission was the best predictor of mortality based on area under the curve (p < 0.001) and the optimum threshold estimation gave an optimal cut-off of 222 (95% Confidence interval (CI): 202-242), yielding three groups of patients: Group 1, with PaO2/FIO2 > 242; Group 2, with PaO2/FIO2 from 202 to 242; and Group 3, with PaO2/FIO2 < 202. Group 3 showed higher in-ICU mortality and ICU length of stay and Groups 2 and 3 also showed higher respiratory complication rates. The presence of a PaO2/FIO2 ratio < 202 at 3 h after admission was shown to be a predictor of in-ICU mortality (OR:1.364; 95% CI:1.212-1.625, p < 0.001) and of worse long-term survival (88.8% vs. 95.8%; Log rank p = 0.002. Adjusted Hazard ratio: 1.48; 95% CI:1.293-1.786; p = 0.004).

CONCLUSIONS

A simple determination of PaO2/FIO2 at 3 h after ICU admission may be useful to identify patients at risk immediately after cardiac surgery.

Prediction of survival of ICU patients using computational intelligence

This paper presents a computational-intelligence-based model to predict the survival rate of critically ill patients who were admitted to an intensive care unit (ICU). The prediction input variables were based on the first 24 h admission physiological data of ICU patients to forecast whether the final outcome was survival or not. The prediction model was based on a particle swarm optimization (PSO)-based Fuzzy Hyper-Rectangular Composite Neural Network (PFHRCNN) that integrates three computational intelligence tools including hyper-rectangular composite neural networks, fuzzy systems and PSO. It could help doctors to make appropriate treatment decisions without excessive laboratory tests. The performance of the proposed prediction model was evaluated on the data set collected from 300 ICU patients in the Cathy General Hospital in 2012. There were 10 input variables in total for the prediction model. Nine of these variables (e.g. systolic arterial blood pressures, systolic non-invasive blood pressures, respiratory rate, heart rate, and body temperature) were routinely available for 24 h in ICU and the last variable is patient's age. The proposed model could achieve a 96% and 86% accuracy rate for the training data and testing data, respectively.

Consideration of additional factors in Sequential Organ Failure Assessment score

PURPOSE

The Sequential Organ Failure Assessment (SOFA) score, originally developed to assess organ failure status, is widely used as a prognostic indicator in intensive care unit patients. Additional prognostic factors, such as age and comorbidities, may complement the predictive performance of the SOFA.

METHODS

In total, 1049 consecutive patients were enrolled prospectively. SOFA and other admission-based intensive care unit scores were recorded during the first 24 hours. A complemented SOFA (cSOFA) score model was constructed by adding age and comorbidity scores to the original SOFA score, based on logistic regression analysis. The predictive performance was evaluated with regard to hospital mortality by receiver operating characteristics analysis. The Hosmer-Lemeshow goodness-of-fit test was used to assess calibration of the model, and leave-one-out cross-validation was performed.

RESULTS

The cSOFA score (maximum 30 points) was calculated as the SOFA score (24 points) + age score (2 points) + comorbidity score (4 points). The cSOFA score model showed satisfactory calibration and cross-validation performance. The AUC (95% CI) of the cSOFA score (0.812 [0.787-0.835]) was higher than the SOFA score (0.743 [0.715-0.769], P < .0001).

CONCLUSION

The performance of the SOFA score to predict hospital mortality can be improved by considering age and comorbidity factors.

Measurement of multiple biomarkers in advanced stage heart failure patients treated with pulmonary artery catheter guided therapy

Introduction

This study was carried out to investigate the prognostic utility of biomarkers in advanced stage heart failure (HF) patients requiring ICU admission for pulmonary artery catheter (PAC) guided therapy.

Methods

Thirty patients admitted to an ICU for PAC guided HF therapy were enrolled; concentrations of soluble ST2 (sST2), highly sensitive troponin I, an experimental ultrasensitive troponin I, amino-terminal pro-B type natriuretic peptide, cystatin C, and myeloperoxidase were measured over the first 48 hours. Outcomes included response of filling pressures and hemodynamics to tailored therapy and 90-day event-free survival (death, left ventricular assist device implantation, transplant).

Results

Of the biomarkers evaluated, only sST2 concentrations were higher in those who failed to achieve goals for central venous pressure ((CVP), 225.3 versus 104.6 ng/mL; P = 0.003) and pulmonary capillary wedge pressure ((PCWP), 181.7 versus 88.2 ng/mL; P = 0.05). Only sST2 concentrations were associated with adverse events (186.7 versus 92.2 ng/mL; P = 0.01). In age-adjusted Cox proportional hazards analysis, an elevated sST2 during the first 48 hours following ICU admission independently predicted 90-day outcomes (Hazard Ratio = 5.53; P = 0.03) superior to the Simplified Acute Physiology Score for this application; in Kaplan-Meier analysis the risk associated with elevated sST2 concentrations was present early and sustained through the duration of follow-up (log rank P = 0.01).

Conclusions

In patients undergoing HF therapy guided by invasive monitoring, sST2 concentrations were associated with impending failure to reduce filling pressures and predicted impending events. Elevated sST2 values early in the ICU course theoretically could assist therapeutic decision-making in advanced stage HF patients.

Trial registration

ClinicalTrials.gov Identifier: NCT00595738

Cross-validation of a Sequential Organ Failure Assessment score-based model to predict mortality in patients with cancer admitted to the intensive care unit

PURPOSE

This study aims to validate the performance of the Sequential Organ Failure Assessment (SOFA) score to predict death of critically ill patients with cancer.

MATERIAL AND METHODS

We conducted a retrospective observational study including adults admitted to the intensive care unit (ICU) between January 1, 2006, and December 31, 2008. We randomly selected training and validation samples in medical and surgical admissions to predict ICU and in-hospital mortality. By using logistic regression, we calculated the probabilities of death in the training samples and applied them to the validation samples to test the goodness-of-fit of the models, construct receiver operator characteristics curves, and calculate the areas under the curve (AUCs).

RESULTS

In predicting mortality at discharge from the unit, the AUC from the validation group of medical admissions was 0.7851 (95% confidence interval [CI], 0.7437-0.8264), and the AUC from the surgical admissions was 0.7847 (95% CI, 0.6319-0.937). The AUCs of the SOFA score to predict mortality in the hospital after ICU admission were 0.7789 (95% CI, 0.74-0.8177) and 0.7572 (95% CI, 0.6719-0.8424) for the medical and surgical validations groups, respectively.

CONCLUSIONS

The SOFA score had good discrimination to predict ICU and hospital mortality. However, the observed underestimation of ICU deaths and unsatisfactory goodness-of-fit test of the model in surgical patients to indicate calibration of the score to predict ICU mortality is advised in this group.

Relationship between mortality and first-day events index from routinely gathered physiological variables in ICU patients

OBJECTIVE

To test the hypothesis that the degree and duration of alterations in physiological variables routinely gathered by intensive care unit (ICU) monitoring systems during the first day of admission to the ICU, together with a few additional routinely recorded data, yield information similar to that obtained by traditional mortality prediction systems.

DESIGN

A prospective observational multicenter study (EURICUS II) was carried out.

SETTING

Fifty-five European ICUs.

PATIENTS

A total of 17,598 consecutive patients admitted to the ICU over a 10-month period.

INTERVENTIONS

None.

MAIN VARIABLES OF INTEREST

Hourly data were manually gathered on alterations or "events" in systolic blood pressure, heart rate and oxygen saturation throughout ICU stay to construct an events index and mortality prediction models.

RESULTS

The mean first-day events index was 6.37±10.47 points, and was significantly associated to mortality (p<0.001), with a discrimination capacity for hospital mortality of 0.666 (area under the ROC curve). A second index included this first-day events index, age, pre-admission location, and the Glasgow coma score. A model constructed with this second index plus diagnosis upon admission was validated by using the Jackknife method (Hosmer-Lemeshow, H=13.8554, insignificant); the area under ROC curve was 0.818.

CONCLUSIONS

A prognostic index with performance very similar to that of habitual systems can be constructed from routine ICU data with only a few patient characteristics. These results may serve as a guide for the possible automated construction of ICU prognostic indexes.

Is there a relationship between bone and critical illness?

Critical illness is a result of a series of severe diseases of complicated causes. Although accumulating therapeutic strategies have been carried out for the prevention and treatment of critical illness, mortality remains high. Inflammations, electrolyte disturbances, hyperglycemia, and depressed immune function are factors that may be responsible. Recent studies have suggested that the bone influences hematopoiesis, immunity, and metabolism. Published data and our own work have led us to hypothesize that the bone plays an important role in the pathogenesis and outcome of critical illness.

Prediction of clinical conditions after coronary bypass surgery using dynamic data analysis

This work studies the impact of using dynamic information as features in a machine learning algorithm for the prediction task of classifying critically ill patients in two classes according to the time they need to reach a stable state after coronary bypass surgery: less or more than 9 h. On the basis of five physiological variables (heart rate, systolic arterial blood pressure, systolic pulmonary pressure, blood temperature and oxygen saturation), different dynamic features were extracted, namely the means and standard deviations at different moments in time, coefficients of multivariate autoregressive models and cepstral coefficients. These sets of features served subsequently as inputs for a Gaussian process and the prediction results were compared with the case where only admission data was used for the classification. The dynamic features, especially the cepstral coefficients (aROC: 0.749, Brier score: 0.206), resulted in higher performances when compared to static admission data (aROC: 0.547, Brier score: 0.247). The differences in performance are shown to be significant. In all cases, the Gaussian process classifier outperformed to logistic regression.

Daily estimation of the severity of multiple organ dysfunction syndrome in critically ill children

BACKGROUND

Daily evaluation of multiple organ dysfunction syndrome has been performed in critically ill adults. We evaluated the clinical course of multiple organ dysfunction over time in critically ill children using the Pediatric Logistic Organ Dysfunction (PELOD) score and determined the optimal days for measuring scores.

METHODS

We prospectively measured daily PELOD scores and calculated the change in scores over time for 1806 consecutive patients admitted to seven pediatric intensive care units (PICUs) between September 1998 and February 2000. To study the relationship between daily scores and mortality in the PICU, we evaluated changes in daily scores during the first four days; the mean rate of change in scores during the entire PICU stay between survivors and nonsurvivors; and Cox survival analyses using a change in PELOD score as a time-dependent covariate to determine the optimal days for measuring daily scores.

RESULTS

The overall mortality among the 1806 patients was 6.4%. A high PELOD score (>or=20 points) on day 1 was associated with an odds ratio (OR) for death of 40.7 (95% confidence interval [CI] 20.3-81.4); a medium score (10-19 points) on day 1 was associated with an OR for death of 4.2 (95% CI 2.0-8.7). Mortality was 50% when a high score on day 1 increased on day 2. The course of daily PELOD scores differed between survivors and nonsurvivors. A set of seven days (days 1, 2, 5, 8, 12, 16 and 18) was identified as the optimal period for measurement of daily PELOD scores.

INTERPRETATION

PELOD scores indicating a worsening condition or no improvement over time were indicators of a poor prognosis in the PICU. A set of seven days for measurement of the PELOD score during the PICU stay provided optimal information on the progression of multiple-organ dysfunction syndrome in critically ill children.

Day 1 multiple organ dysfunction syndrome is associated with poor functional outcome and mortality in the pediatric intensive care unit

OBJECTIVE

The epidemiology and outcomes of multiple organ dysfunction syndrome (MODS) are incompletely characterized in the pediatric population due to small sample size and conflicting diagnoses of organ failure. We sought to describe the epidemiology and outcomes of early MODS in a large clinical database of pediatric intensive care unit (PICU) patients based on consensus definitions of organ failure.

DESIGN

Retrospective analysis of a contemporaneously collected clinical PICU database.

SETTING

Virtual Pediatric Intensive Care Unit Performance System database patient admissions from January 2004 to December 2005 for 35 U.S. children's hospitals.

PATIENTS

: We evaluated 63,285 consecutive PICU admissions from January 2004 to December 2005 in the Virtual Pediatric Intensive Care Unit Performance System database. We excluded patients younger than 1 month or older than 18 years of age, and hospitals with >10% missing values for MODS variables. We identified day 1 MODS by International Pediatric Sepsis Consensus Conference criteria with day 1 laboratory and vital sign values. We evaluated functional status using Pediatric Overall Performance Category and Pediatric Cerebral Performance Category scores from PICU admission and discharge.

ANALYSIS

Student's t test, chi-square test, Mann-Whitney rank sum, Kruskal-Wallis, and linear and logistic regression.

MEASUREMENTS AND MAIN RESULTS

We analyzed 44,693 admissions from 28 hospitals meeting inclusion criteria. Overall PICU mortality was 2.8%. We identified day 1 MODS in 18.6% of admissions. Patients with day 1 MODS had higher mortality (10.0% vs. 1.2%, p < .001), longer PICU length of stay (3.6 vs. 1.3 days, p < .001), and larger change from baseline Pediatric Overall Performance Category and Pediatric Cerebral Performance Category scores at time of PICU discharge (p < .001). Infants had the highest incidence of day 1 MODS (25.2% vs. 16.5%, p < .001) compared with other age groups.

CONCLUSIONS

Using the largest clinical dataset to date and consensus definitions for organ failure, we found that children with MODS present on day 1 of intensive care unit admission have worse functional outcomes, higher mortality, and longer PICU length of stay than children who do not have MODS on day 1. Infants are disproportionally affected by MODS.

Plasma fibrinolysis is related to the degree of organ dysfunction but not to the concentration of von Willebrand Factor in critically ill patients

Background

Endothelial cell dysfunction, by promoting fibrin deposition, has been implicated in the development of multiple organ failure. Altered fibrinolysis during inflammation may participate in microvascular alterations. We sought to determine whether plasma fibrinolysis was related to the severity of organ dysfunction and/or to the levels of von Willebrand factor (vWF antigen), as a marker of endothelium dysfunction, in critically ill patients.

Methods

Forty-nine consecutive patients admitted to an adult medico-surgical intensive care unit (ICU) with (18) or without sepsis (31) were included. C-reactive protein and vWF levels were measured on ICU admission and plasma fibrinolysis was assessed by the Euglobulin Clot Lysis Time (ECLT). The sequential organ failure assessment (SOFA) score and the simplified acute physiology score (SAPS) II were calculated on admission.

Results

ECLT was significantly longer in septic than in non-septic patients [1033 min (871–1372) versus 665 min (551–862), p = 0.001]. There were significant correlations between ECLT and C-reactive protein (CRP) concentrations (r = 0.78, p < 0.001) and the Sequential Organ Failure Assessment (SOFA) score (r = 0.39, p = 0.006). The level of vWF was not correlated with the ECLT (r = -0.06, p = 0.65) or the SOFA score (r = -0.02, p = 0.88).

Conclusion

ECLT measurement at admission could be a marker of organ dysfunction and a prognostic indicator in critically ill patients.

[Age as predictive factor of mortality in an intensive and intermediate care unit]

INTRODUCTION

Age by itself is not a criterion of biological prognosis. Scores for physiological variables on admission and multiorgan failure are better predictors of mortality.

PATIENTS AND METHODS

We performed a retrospective/ prospective observational study from September, 2005 to May, 2007. The following variables were analyzed: age, sex, Acute Physiology and Chronic Health Classification System (APACHE) II, modified APACHE II score (without the variable of age), Sequential Organ Failure Assessment (SOFA) score, length of hospital stay, type of disease and mortality, limitation of therapeutic effort (LTE), Katz index on admission, intensive and intermediate care unit (IICU) mortality and in-hospital mortality. Student's t-test was used to analyze continuous variables.

RESULTS

Of the 572 patients admitted to the IICU, we excluded 75 due to transfer to other hospitals, 142 due to direct admission to intermediate care, and 89 due to acute coronary syndrome. Of the 266 remaining patients with medical disease, mortality was higher when the APACHE II score was > 20 (OR = 9.4) and/or the SOFA score was >4 (OR = 15.41) but not when age was 3 76 years (OR = 2.04). Multivariate analysis of these parameters revealed higher mortality in the IICU (P=.01) in patients with a SOFA score > 4 and modified APACHE II score >16, independently of age or the Katz index. In addition to the SOFA and the APACHE II scores, in-hospital mortality was significantly influenced by the Katz index (P=.05). LTE was significantly greater in patients with a Katz index E-G.

CONCLUSIONS

Higher SOFA and APACHE II scores predicted higher IICU mortality, regardless of age. LTE was more frequent in patients with a greater degree of dependence.

Evaluation of SOFA-based models for predicting mortality in the ICU: A systematic review

Introduction

To systematically review studies evaluating the performance of Sequential Organ Failure Assessment (SOFA)-based models for predicting mortality in patients in the intensive care unit (ICU).

Methods

Medline, EMBASE and other databases were searched for English-language articles with the major objective of evaluating the prognostic performance of SOFA-based models in predicting mortality in surgical and/or medical ICU admissions. The quality of each study was assessed based on a quality framework for prognostic models.

Results

Eighteen articles met all inclusion criteria. The studies differed widely in the SOFA derivatives used and in their methods of evaluation. Ten studies reported about developing a probabilistic prognostic model, only five of which used an independent validation data set. The other studies used the SOFA-based score directly to discriminate between survivors and non-survivors without fitting a probabilistic model. In five of the six studies, admission-based models (Acute Physiology and Chronic Health Evaluation (APACHE) II/III) were reported to have a slightly better discrimination ability than SOFA-based models at admission (the receiver operating characteristic curve (AUC) of SOFA-based models ranged between 0.61 and 0.88), and in one study a SOFA model had higher AUC than the Simplified Acute Physiology Score (SAPS) II model. Four of these studies used the Hosmer-Lemeshow tests for calibration, none of which reported a lack of fit for the SOFA models. Models based on sequential SOFA scores were described in 11 studies including maximum SOFA scores and maximum sum of individual components of the SOFA score (AUC range: 0.69 to 0.92) and delta SOFA (AUC range: 0.51 to 0.83). Studies comparing SOFA with other organ failure scores did not consistently show superiority of one scoring system to another. Four studies combined SOFA-based derivatives with admission severity of illness scores, and they all reported on improved predictions for the combination. Quality of studies ranged from 11.5 to 19.5 points on a 20-point scale.

Conclusions

Models based on SOFA scores at admission had only slightly worse performance than APACHE II/III and were competitive with SAPS II models in predicting mortality in patients in the general medical and/or surgical ICU. Models with sequential SOFA scores seem to have a comparable performance with other organ failure scores. The combination of sequential SOFA derivatives with APACHE II/III and SAPS II models clearly improved prognostic performance of either model alone. Due to the heterogeneity of the studies, it is impossible to draw general conclusions on the optimal mathematical model and optimal derivatives of SOFA scores. Future studies should use a standard evaluation methodology with a standard set of outcome measures covering discrimination, calibration and accuracy.

Altered pharmacology in the Intensive Care Unit patient

Critically ill patients, not infrequently present alterations of physiological parameters that determine the success/failure of therapeutic interventions as well as the final outcome. Sepsis and polytrauma are two of the most common and complex syndromes occurring in Intensive Care Unit (ICU) and affect drug absorption, disposition, metabolism and elimination. Pharmacological management of ICU patients requires consideration of the unique pharmacokinetics associated with these clinical conditions and the likely occurrence of drug interaction. Rational adjustment in drug choice and dosing contributes to the appropriateness of treatment of those patients.