Evaluation of Acute Physiology and Chronic Health Evaluation II and sequential organ failure assessment scoring systems for prognostication of outcomes among Intensive Care Unit's patients

DEVELOPMENT OF A MACHINE LEARNING MODEL FOR PREDICTING 28-DAY MORTALITY OF SEPTIC PATIENTS WITH ATRIAL FIBRILLATION

ABSTRACT

Introduction: Septic patients with atrial fibrillation (AF) are common in the intensive care unit accompanied by high mortality. The early prediction of prognosis of these patients is critical for clinical intervention. This study aimed to develop a model by using machine learning (ML) algorithms to predict the risk of 28-day mortality in septic patients with AF. Methods: In this retrospective cohort study, we extracted septic patients with AF from the Medical Information Mart for Intensive Care III (MIMIC-III) and IV database. Afterward, only MIMIC-IV cohort was randomly divided into training or internal validation set. External validation set was mainly extracted from MIMIC-III database. Propensity score matching was used to reduce the imbalance between the external validation and internal validation data sets. The predictive factors for 28-day mortality were determined by using multivariate logistic regression. Then, we constructed models by using ML algorithms. Multiple metrics were used for evaluation of performance of the models, including the area under the receiver operating characteristic curve, sensitivity, specificity, recall, and accuracy. Results: A total of 5,317 septic patients with AF were enrolled, with 3,845 in the training set, 960 in the internal testing set, and 512 in the external testing set, respectively. Then, we established four prediction models by using ML algorithms. AdaBoost showed moderate performance and had a higher accuracy than the other three models. Compared with other severity scores, the AdaBoost obtained more net benefit. Conclusion: We established the first ML model for predicting the 28-day mortality of septic patients with AF. Compared with conventional scoring systems, the AdaBoost model performed moderately. The model established will have the potential to improve the level of clinical practice.

Factors affecting 90-day mortality in community and hospital acquired pneumonia patients with or without acute kidney injury

Background

AKI is a significant risk factor for mortality. Inflammatory markers are commonly used in the prediction of prognosis in pneumonia patients. The present study aimed to evaluate the prevalence of AKI in hospitalized CAP and HAP patients and to investigate the role of inexpensive, practical, routinely measured serum biomarkers in predicting 90-day mortality.

Materials and Methods

The retrospective study included 381 patients in CAP patients and HAP patients who were hospitalized in our Chest Diseases clinic or ICU.

Results

Ninety-day mortality occurred in 115 (30.2%) patients (CAP, 28.7%; HAP, 34.7%). AKI was detected in 25.5% of the patients. On multivariate logistic regression analysis, the 90-day mortality risk was 0.931, 1.05, 0.607, and 1.999 times greater in patients with an increased APACHE II score and increased WBC, 1-h creatinine, and 48-h creatinine levels, respectively. In CAP patients, the 90-day mortality risk was 0.296, 0.539, and 1.966 times greater in patients with an increased CURB-65 score and elevated 1-h and 48-h creatinine levels, respectively. In HAP patients, however, the 90-day mortality risk was 3.554 times greater in patients with an increased 48-h creatinine level.

Conclusion

Novel practical scoring systems based on serum creatinine levels are needed for the prediction of long-term prognosis in pneumonia patients.

Prediction of Mortality in the Medical Intensive Care Unit with Serial Full Outline of Unresponsiveness Score in Elderly Patients

Background

Advanced age is one of the key risk factors for mortality and morbidity in intensive care units. The full outline of unresponsiveness (FOUR) score has been developed and introduced to address the limitations of the Glasgow Coma Scale (GCS). The current study aimed to evaluate the ability of the FOUR score in predicting the outcomes (survivors, nonsurvivors).

Materials and methods

This observational study of 168 consecutive elderly patients admitted to medical intensive care during the 14 months carried out prospectively. FOUR score in the 24, 48, and 72 hours of admission, and demographic characteristics of all elderly patients were calculated, then recorded. The receiver operating characteristic (ROC) curve, logistic regression, and Hosmer-Lemeshow test were used (95% confidence interval) for statistical analysis.

Results

FOUR scores in 24, 48, and 72 hours between survivors and nonsurvivors (p <0.0001, p <0.0001, and p <0.0001, respectively) were statistically different. The discrimination power of FOUR score 24 hours of admission was excellent [area under ROC (AUC): 85.7% [standard error (SE)]: 2.8%]; it was acceptable for 48 and 72 hours of admission [AUC: 76.3% (SE: 3.6%), AUC: 75/0% (SE: 3.8%), respectively]. The FOUR score of 24 and 48 hours (x² = 10.06, p = 0.261, x² = 6.82, p = 0.448, respectively) showed acceptable calibration.

Conclusions

The FOUR score is a suitable scoring system for prognostication of outcomes in critically ill elderly patients. The FOUR score 24 hours of admission was superior in terms of discrimination power than 48 and 72 hours, but better calibration power belonged to FOUR score 48 hours.

How to cite this article

Ramazani J, Hosseini M. Prediction of Mortality in the Medical Intensive Care Unit with Serial Full Outline of Unresponsiveness Score in Elderly Patients. Indian J Crit Care Med 2022;26(1):94–99.

Perioperative Risk Stratification: A Need for an Improved Assessment in Surgery and Anesthesia-A Pilot Study

Background and Objectives: Numerous scoring systems have been introduced into modern medicine. None of the scoring systems assessed both anesthetic and surgical risk of the patient, predict the morbidity, mortality, or the need for postoperative intensive care unit admission. The aim of this study was to compare the anesthetic and surgical scores currently used, for a better evaluation of perioperative risks, morbidity, and mortality. Material and Methods: This is a pilot, prospective, observational study. We enrolled 50 patients scheduled for elective surgery. Anesthetic and surgery risk was assessed using American Society of Anesthesiologists (ASA) scale, Physiological and Operative Severity Score for the enumeration of Mortality and morbidity (P-POSSUM), Acute Physiology and Chronic Health Evaluation (APACHE II), and Surgical APGAR Score (SAS) scores. The real and the estimated length of stay (LOS) were registered. Results: We obtained several statistically significant positive correlations: ASA score–P-POSSUM (p < 0.01, r = 0.465); ASA score–SAS, (p < 0.01, r = −0.446); ASA score–APACHE II, (p < 0.01 r = 0.519); predicted LOS and ASA score (p < 0.01, r = 0.676); predicted LOS and p-POSSUM (p < 0.01, r = 0.433); and predicted LOS and APACHE II (p < 0.01, r = 0.454). A significant negative correlation between predicted LOS, real LOS, ASA class, and SAS (p < 0.05) was observed. We found a statistically significant difference between the predicted and actual LOS (p < 001). Conclusions: Anesthetic, surgical, and severity scores, used together, provide clearer information about mortality, morbidity, and LOS. ASA scale, associated with surgical scores and severity scores, presents a better image of the patient’s progress in the perioperative period. In our study, APACHE II is the best predictor of mortality, followed by P-POSSUM and SAS. P-POSSUM score and ASA scale may be complementary in terms of preoperative physiological factors, providing valuable information for postoperative outcomes.

Mortality predictive factors of people living with human immunodeficiency virus and bloodstream infection

INTRODUCTION

Portugal has one of the highest mortality rates for people living with HIV (PLWHIV) in Europe. After antiretroviral therapy introduction, HIV-associated mortality declined, included the one associated with bloodstream infection (BSI). However it is still high, and European data are scarce . Therefore, characterizing BSI and defining prognostic factors may improve our approach.

METHODS

This was a 10-year retrospective study of predictive factors for 30-day and 3-year mortality in PLWHIV with BSI in a tertiary infectious diseases ward.

RESULTS

Of 2134 PLWHIV admissions, 145 (6.8%) had a BSI, mostly respiratory and catheter-related bacteremia and globally community-acquired. Nosocomial infections occurred in 42 (36%) cases, mostly caused by Enterococcus spp, Staphylococcus aureus, and Candida spp. PLWHIV with a BSI had higher 30-day mortality (27%) compared to those without a BSI (14%). APACHE II score, corticotherapy, and current intravenous drug use (IDU) had a prognostic impact on 30-day mortality. Three-year survival was 54% in PLWHIV with a BSI; a CD4 <200 cells, vascular or chronic pulmonary disease, and lymphoma were prognostic factors.

CONCLUSIONS

Patients with a BSI were more likely to present advanced HIV disease, have more comorbidities, a longer length of stay, and higher 30-day mortality. IDU and severity of infection determined the short-term prognosis. Three-year mortality was primarily influenced by lower CD4 cell counts, hematological tumor, and cardiopulmonary comorbidities. Systemic corticotherapy may influence nosocomial BSI and short-term prognosis.

The Association of TLR2, TLR3, and TLR9 Gene Polymorphisms With Susceptibility to Talaromycosis Among Han Chinese AIDS Patients in Guangdong

Background

Talaromycosis (TM) caused by Talaromyces marneffei (T. marneffei) is a growing public health concern. Although Toll-like receptor (TLR) genes play a critical role in the host defense against fungal infection, the influence of polymorphisms in these genes on the susceptibility of acquired immune deficiency syndrome (AIDS) patients to TM remains unknown. This study aims to uncover the associations of single nucleotide polymorphisms (SNPs) in TLR genes with TM susceptibility among patients with AIDS.

Methods

Altogether 200 AIDS patients complicated with TM, 200 matched AIDS patients without TM, and 76 healthy controls (HCs) were enrolled in this case-control study. In total, 23 SNPs in the TLR2, TLR4, and TLR9 genes, which may influence the susceptibility of AIDS patients to TM, were checked by the time of flight mass spectrometry (TOF/MS) method among these Han Chinese subjects.

Results

No significant differences in genotype or allele frequencies of selected SNPs were found among the TM group, Non-TM group, and HC group. Haplotype analysis also demonstrated no correlation of these SNPs with TM. However, subgroup analysis showed that the genotype TT and the T allele in TLR2 SNP rs1339 were more frequent in typical TM cases than controls (50.0 vs. 35.8%, 70.5 vs. 59.7%); the frequency of the GT genotype in TLR2 SNP rs7656411 was markedly higher in severe TM cases compared to controls (57.8 vs. 34.4%).

Conclusion

Our results demonstrate a genetic connection of TLR2 SNPs rs1339 and rs7656411 with an increased susceptibility and severity of TM among Han Chinese populations.

Influence of severity of infection on the effect of appropriate antimicrobial therapy for Acinetobacter baumannii bacteremic pneumonia

Background

The impact of appropriate antimicrobial therapy for A. baumannii bacteremic pneumonia has not been well established due to the inclusion of the three phenotypically indistinguishable Acinetobacter species and confounding factors including underlying diseases and severity of infection. This retrospective study aimed to evaluate the impact of appropriate antimicrobial therapy on 14-day mortality in A. baumannii bacteremic pneumonia patients after adjusting for risk factors.

Methods

This study was conducted at five medical centers in Taiwan between July 2012 and June 2016. A. baumannii species identification was performed using reference molecular methods. Risk factors for 14-day mortality were analyzed via logistic regression. The interaction between the Acute Physiology and Chronic Health Evaluation (APACHE) II score and appropriate antimicrobial therapy was assessed using the logistic model.

Results

A total of 336 patients with monomicrobial A. baumannii bacteremic pneumonia were included in this study. The overall 14-day mortality rate was 47.3%. The crude mortality of appropriate antimicrobial therapy was 35.9% (57 of 151 patients). Appropriate antimicrobial therapy was associated with a lower mortality after multivariate adjustment (odds ratio [OR], 0.57; 95% confidence interval [CI], 0.34–0.97; p = 0.04), and the effect was influenced by APACHE II score (OR for interaction term, 0.0098; 95% CI, 0.0005–0.1885; p = 0.002). Further analysis demonstrated that appropriate antimicrobial therapy significantly reduced 14-day mortality among the patients with an APACHE II score > 35 (OR 0.0098; 95% CI 0.0005–0.1885).

Conclusion

Appropriate antimicrobial therapy decreases 14-day mortality of the most severely ill patients with A. baumannii bacteremic pneumonia.

Prognostic indicators in critically ill poisoned patients: development of a risk-prediction nomogram

OBJECTIVES

The prognosis of acutely poisoned patients is a significant concern for clinical toxicologists. In this study, we sought to determine the clinical and laboratory findings that can contribute to predicting the medical outcomes of poisoned patients admitted to intensive care units (ICUs).

METHODS

This retrospective study was performed from January 2009 to January 2016 in the ICU of Vali-e-Asr Hospital in Birjand, Iran. We included all patients with the diagnosis of acute poisoning admitted to the ICU. Demographic data, laboratory results, the Sequential Organ Failure Assessment (SOFA), and acute physiology score + age points + chronic health points (APACHE) II, and the Simplified Acute Physiology Score (SAPS) II, and outcome were collected. Univariate analysis (Mann-Whitney or t-test), multiple logistic regression, receiver operating characteristics (ROC) curve analysis, and Pearson's correlation test were performed using SPSS, STATA/SE 13.0, and Nomolog software programs.

RESULTS

The multiple logistic regression analysis revealed that five factors were significant for predicting mortality including age (OR 95% CI: 1.1[1.05-1.12], p<0.001), Glasgow Coma Score (GCS) (OR 95% CI: 0.71[0.6-0.84], p<0.001), white blood cell (WBC) count (OR 95% CI: 1.1[1.01-1.12], p=0.04), serum sodium (Na) (OR 95% CI: 1.08[1.01-1.15], p=0.02), and creatinine levels (Cr) (OR 95% CI: 1.86 [1.23-2.81], p=0.003). We generated a five-variable risk-prediction nomogram which could both predict mortality risk and identify high-risk patients.

CONCLUSIONS

Age, GCS, WBC, serum creatinine, and sodium levels are the best prognostic factors for mortality in poisoned patients admitted to the ICU. The APACHE II score can discriminate between non-survivors and survivors. The nomogram developed in the current study can provide a more precise, quick, and simple analysis of risks, thereby enabling the users to predict mortality and identify high-risk patients.

Prognostic indicators in critically ill poisoned patients: development of a risk-prediction nomogram

Objectives The prognosis of acutely poisoned patients is a significant concern for clinical toxicologists. In this study, we sought to determine the clinical and laboratory findings that can contribute to predicting the medical outcomes of poisoned patients admitted to intensive care units (ICUs). Methods This retrospective study was performed from January 2009 to January 2016 in the ICU of Vali-e-Asr Hospital in Birjand, Iran. We included all patients with the diagnosis of acute poisoning admitted to the ICU. Demographic data, laboratory results, the Sequential Organ Failure Assessment (SOFA), and acute physiology score + age points + chronic health points (APACHE) II, and the Simplified Acute Physiology Score (SAPS) II, and outcome were collected. Univariate analysis (Mann-Whitney or t-test), multiple logistic regression, receiver operating characteristics (ROC) curve analysis, and Pearson's correlation test were performed using SPSS, STATA/SE 13.0, and Nomolog software programs. Results The multiple logistic regression analysis revealed that five factors were significant for predicting mortality including age (OR 95% CI: 1.1[1.05-1.12], p<0.001), Glasgow Coma Score (GCS) (OR 95% CI: 0.71[0.6-0.84], p<0.001), white blood cell (WBC) count (OR 95% CI: 1.1[1.01-1.12], p=0.04), serum sodium (Na) (OR 95% CI: 1.08[1.01-1.15], p=0.02), and creatinine levels (Cr) (OR 95% CI: 1.86 [1.23-2.81], p=0.003). We generated a five-variable risk-prediction nomogram which could both predict mortality risk and identify high-risk patients. Conclusions Age, GCS, WBC, serum creatinine, and sodium levels are the best prognostic factors for mortality in poisoned patients admitted to the ICU. The APACHE II score can discriminate between non-survivors and survivors. The nomogram developed in the current study can provide a more precise, quick, and simple analysis of risks, thereby enabling the users to predict mortality and identify high-risk patients.

Improvement of mortality prediction accuracy in critically ill patients through combination of SOFA and APACHE II score with markers of stress haematopoiesis

INTRODUCTION

In critically ill patients nucleated red blood cells (NRBC) and immature granulocytes (IG) appear in the peripheral blood as the consequence of stress haematopoesis. The aim of this retrospective study was to evaluate the diagnostic value of NRBC and IG and to propose a model of improved mortality prediction including these parameters in the assessment of critically ill patients.

METHODS

The study included 338 critically ill adult patients hospitalized at Department of Anaesthesiology and Intensive Medicine, Louis Pasteur University Hospital in Kosice. As NRBC positive patients were considered patients with peripheral NRBC > 0.01 × 10⁹ /L and IG positivity as >0.03 × 10⁹ /L. Apache II index was calculated 24 hours after admission and Systemic Organ Failure Assessment (SOFA) on the day with the worst clinical condition.

RESULTS

NRBC positivity was found in 27.6% of patients. The mortality of NRBC positive patients was 48.38%, significantly higher than 23.7% of NRBC negative patients. IG positivity was 79.0% and their mortality was also higher as compared with that of IG negative patients (69.3% vs 33.8%). Three regression models predicting mortality including stress haematopoiesis markers, APACHE II, SOFA scores and age had sufficient level of sensitivity and specificity.

CONCLUSION

The presence of NRBC in the peripheral blood and the IG increase are available early risk predictors of mortality in critically ill patients. Regression models designed by combination of SOFA, APACHE II, and the new haematological parameters increase the accuracy and effectivity of diagnostic process in predicting prognosis and risk of mortality with high sensitivity and specificity.

Comparison of full outline of unresponsiveness score and Glasgow Coma Scale in Medical Intensive Care Unit

Context:

The Glasgow Coma Scale (GCS) is the most commonly used scale, and Full Outline of Unresponsiveness (FOUR) score is new validated coma scale as an alternative to GCS in the evaluation of the level of consciousness.

Aim:

The aim of the current study was to evaluate FOUR score and GCS ability in predicting the outcomes (Survivors, nonsurvivors) in Medical Intensive Care Unit (MICU).

Setting and Design:

This was an observational and prospective study of 300 consecutive patients admitted to the MICU during a 14 months’ period.

Materials and Methods:

FOUR score, GCS score, and demographic characteristics of all patients were recorded in the first admission 24 h.

Statistical Analysis Used:

A receiver operator characteristic (ROC) curve, Hosmer–Lemeshow test, and Logistic regression were used in the statistical analysis (95% confidence interval).

Results:

Data analysis showed a significant statistical difference in FOUR score and GCS score between survivors and nonsurvivors (P < 0.0001, P < 0.0001; respectively). The discrimination power was good for both FOUR score and GCS (area under ROC curve: 87.3% (standard error [SE]: 2.1%), 82.6% [SE: 2.3%]; respectively). The acceptable calibration was seen just for FOUR score (χ² = 8.059, P = 0.428).

Conclusions:

Both FOUR score and GCS are valuable scales for predicting outcomes in patients are admitted to the MICU; however, the FOUR score showed better discrimination and calibration than GCS, so it is superior to GCS in predicting outcomes in this patients population.

Prediction of ICU mortality in critically ill children : Comparison of SOFA, GCS, and FOUR score

INTRODUCTION

The SOFA (Sequential Organ Failure Assessment), GCS (Glasgow Coma Scale), and FOUR (Full Outline of UnResponsiveness) scores are the most commonly used scoring systems to predict the risk of mortality and morbidity in intensive care units (ICUs). The aim of the current study was to compare the predictive ability of these three models for predicting medical/surgical ICU mortality in critically ill children.

METHODS

In the current observational and prospective study, a total of 90 consecutive patients, age ≤18 years, admitted to medical and surgical ICUs, were enrolled. The SOFA, GCS, FOUR score and demographic characteristics of all children were recorded on the first day of admission. For statistical analyses, a receiver operator characteristic (ROC) curve, the Hosmer-Lemeshow goodness of fit test, and logistic regression were used (95% confidence interval).

RESULTS

The SOFA, GCS, and FOUR scores between survivors and nonsurvivors were statistically different (p = 0.002, p < 0.001, p = 0.004, respectively). The discrimination power for SOFA, GCS, and FOUR score was moderate (area under ROC [AUC] curve: 75.1%; standard error [SE]: 6.0%, 72.9% [SE: 7.2%], 78.7% [SE: 6.6%], respectively). The only well-calibrated model was GCS (x² = 2.76, p = 0.59).

CONCLUSIONS

The performance of the three predictive models SOFA, GCS, and FOUR score for predicting outcomes in children admitted to medical and surgical ICUs was good. The discrimination was moderate for all three models, and calibration was good just for GCS. GCS was superior in predicting outcome in critically ill children; however, further studies are needed to validate these scores in the pediatric population.

The significance of National Early Warning Score for predicting prognosis and evaluating conditions of patients in resuscitation room

Background:

For critical patients in resuscitation room, the early prediction of potential risk and rapid evaluation of disease progression would help physicians with timely treatment, leading to improved outcome. In this study, it focused on the application of National Early Warning Score on predicting prognosis and conditions of patients in resuscitation room. The National Early Warning Score was compared with the Modified Early Warning Score) and the Acute Physiology and Chronic Health Evaluation II.

Objectives:

To assess the significance of NEWS for predicting prognosis and evaluating conditions of patients in resuscitation rooms.

Methods:

A total of 621 consecutive cases from resuscitation room of Xuanwu Hospital, Capital Medical University were included during June 2015 to January 2016. All cases were prospectively evaluated with Modified Early Warning Score, National Early Warning Score, and Acute Physiology and Chronic Health Evaluation II and then followed up for 28 days. For the prognosis prediction, the cases were divided into death group and survival group. The Modified Early Warning Score, National Early Warning Score, and Acute Physiology and Chronic Health Evaluation II results of the two groups were compared. In addition, receiver operating characteristic curves were plotted. The areas under the receiver operating characteristic curves were calculated for assessing and predicting intensive care unit admission and 28-day mortality.

Results:

For the prognosis prediction, in death group, the National Early Warning Score (9.50 ± 3.08), Modified Early Warning Score (4.87 ± 2.49), and Acute Physiology and Chronic Health Evaluation II score (23.29 ± 5.31) were significantly higher than National Early Warning Score (5.29 ± 3.13), Modified Early Warning Score (3.02 ± 1.93), and Acute Physiology and Chronic Health Evaluation II score (13.22 ± 6.39) in survival group ( p < 0.01). For the disease progression evaluation, the areas under the receiver operating characteristic curves of National Early Warning Score, Modified Early Warning Score, and Acute Physiology and Chronic Health Evaluation II were 0.760, 0.729, and 0.817 ( p < 0.05), respectively, for predicting intensive care unit admission; they were 0.827, 0.723, and 0.883, respectively, for predicting 28-day mortality. The comparison of the three systems was significant ( p < 0.05).

Conclusion:

The performance of National Early Warning Score for predicting intensive care unit admission and 28-day mortality was inferior than Acute Physiology and Chronic Health Evaluation II but superior than Modified Early Warning Score. It was able to rapidly predict prognosis and evaluate disease progression of critical patients in resuscitation room.

The evaluation of acute physiology and chronic health evaluation II score, poisoning severity score, sequential organ failure assessment score combine with lactate to assess the prognosis of the patients with acute organophosphate pesticide poisoning

The aim of this study was to assess the ability of acute physiology and chronic health evaluation II (APACHE II) score, poisoning severity score (PSS) as well as sequential organ failure assessment (SOFA) score combining with lactate (Lac) to predict mortality in the Emergency Department (ED) patients who were poisoned with organophosphate.A retrospective review of 59 stands-compliant patients was carried out. Receiver operating characteristic (ROC) curves were constructed based on the APACHE II score, PSS, SOFA score with or without Lac, respectively, and the areas under the ROC curve (AUCs) were determined to assess predictive value. According to SOFA-Lac (a combination of SOFA and Lac) classification standard, acute organophosphate pesticide poisoning (AOPP) patients were divided into low-risk and high-risk groups. Then mortality rates were compared between risk levels.Between survivors and non-survivors, there were significant differences in the APACHE II score, PSS, SOFA score, and Lac (all P < .05). The AUCs of the APACHE II score, PSS, and SOFA score were 0.876, 0.811, and 0.837, respectively. However, after combining with Lac, the AUCs were 0.922, 0.878, and 0.956, respectively. According to SOFA-Lac, the mortality of high-risk group was significantly higher than low-risk group (P < .05) and the patients of the non-survival group were all at high risk.These data suggest the APACHE II score, PSS, SOFA score can all predict the prognosis of AOPP patients. For its simplicity and objectivity, the SOFA score is a superior predictor. Lac significantly improved the predictive abilities of the 3 scoring systems, especially for the SOFA score. The SOFA-Lac system effectively distinguished the high-risk group from the low-risk group. Therefore, the SOFA-Lac system is significantly better at predicting mortality in AOPP patients.