Validation of a severity of illness score (APACHE II) in a surgical intensive care unit

The purpose of this study was to use natural language processing of physician documentation to predict mortality in patients admitted to the surgical intensive care unit (SICU). The Multiparameter Intelligent Monitoring in Intensive Care III database was used to obtain SICU stays with six different severity of illness scores. Natural language processing was performed on the physician notes. Classifiers for predicting mortality were created. One classifier used only the physician notes, one used only the severity of illness scores, and one used the physician notes with severity of injury scores. There were 3838 SICU stays identified during the study period and 5.4 per cent ended with mortality. The classifier trained with physician notes with severity of injury scores performed with the highest area under the curve (0.88 ± 0.05) and accuracy (94.6 ± 1.1%). The most important variable was the Oxford Acute Severity of Illness Score (16.0%). The most important terms were “dilated” (4.3%) and “hemorrhage” (3.7%). This study demonstrates the novel use of artificial intelligence to process physician documentation to predict mortality in the SICU. The classifiers were able to detect the subtle nuances in physician vernacular that predict mortality. These nuances provided improved performance in predicting mortality over physiologic parameters alone.

Download Full-text

Data Science for Extubation Prediction and Value of Information in Surgical Intensive Care Unit

Journal of Clinical Medicine ◽

10.3390/jcm8101709 ◽

2019 ◽

Vol 8 (10) ◽

pp. 1709 ◽

Cited By ~ 3

Author(s):

Tsung-Lun Tsai ◽

Min-Hsin Huang ◽

Chia-Yen Lee ◽

Wu-Wei Lai

Keyword(s):

Intensive Care Unit ◽

Logistic Regression ◽

Intensive Care ◽

Prediction Model ◽

Value Of Information ◽

Data Science ◽

Apache Ii ◽

Surgical Intensive Care Unit ◽

Surgical Intensive Care ◽

Science Framework

Besides the traditional indices such as biochemistry, arterial blood gas, rapid shallow breathing index (RSBI), acute physiology and chronic health evaluation (APACHE) II score, this study suggests a data science framework for extubation prediction in the surgical intensive care unit (SICU) and investigates the value of the information our prediction model provides. A data science framework including variable selection (e.g., multivariate adaptive regression splines, stepwise logistic regression and random forest), prediction models (e.g., support vector machine, boosting logistic regression and backpropagation neural network (BPN)) and decision analysis (e.g., Bayesian method) is proposed to identify the important variables and support the extubation decision. An empirical study of a leading hospital in Taiwan in 2015–2016 is conducted to validate the proposed framework. The results show that APACHE II and white blood cells (WBC) are the two most critical variables, and then the priority sequence is eye opening, heart rate, glucose, sodium and hematocrit. BPN with selected variables shows better prediction performance (sensitivity: 0.830; specificity: 0.890; accuracy 0.860) than that with APACHE II or RSBI. The value of information is further investigated and shows that the expected value of experimentation (EVE), 0.652 days (patient staying in the ICU), is saved when comparing with current clinical experience. Furthermore, the maximal value of information occurs in a failure rate around 7.1% and it reveals the “best applicable condition” of the proposed prediction model. The results validate the decision quality and useful information provided by our predicted model.

Download Full-text

Evaluation of the consistency of Acute Physiology and Chronic Health Evaluation (APACHE II) scoring in a surgical intensive care unit

Critical Care Medicine ◽

10.1097/00003246-199212000-00014 ◽

1992 ◽

Vol 20 (12) ◽

pp. 1681-1687 ◽

Cited By ~ 39

Author(s):

METTE M. BERGER ◽

ALFIO MARAZZI ◽

JAMES FREEMAN ◽

RENÉ CHIOLÉRO

Keyword(s):

Intensive Care Unit ◽

Intensive Care ◽

Chronic Health Evaluation ◽

Apache Ii ◽

Surgical Intensive Care Unit ◽

Health Evaluation ◽

Surgical Intensive Care ◽

Chronic Health

Download Full-text

Patients’ Characteristics Associated With Readmission to a Surgical Intensive Care Unit

American Journal of Critical Care ◽

10.4037/ajcc2012773 ◽

2012 ◽

Vol 21 (6) ◽

pp. e120-e128 ◽

Cited By ~ 18

Author(s):

T. K. Timmers ◽

M. H. J. Verhofstad ◽

K. G. M. Moons ◽

L. P. H. Leenen

Keyword(s):

Quality Of Life ◽

Intensive Care Unit ◽

Intensive Care ◽

Severity Of Illness ◽

Total Sample ◽

Surgical Intensive Care Unit ◽

Surgical Intensive Care ◽

Term Survival

Background Readmission within 48 hours is a leading performance indicator of the quality of care in an intensive care unit. Objective To investigate variables that might be associated with readmission to a surgical intensive care unit. Methods Demographic characteristics, severity-of-illness scores, and survival rates were collected for all patients admitted to a surgical intensive care unit between 1995 and 2000. Long-term survival and quality of life were determined for patients who were readmitted within 30 days after discharge from the unit. Quality of life was measured with the EuroQol-6D questionnaire. Multivariate logistic analysis was used to calculate the independent association of expected covariates. Results Mean follow-up time was 8 years. Of the 1682 patients alive at discharge, 141 (8%) were readmitted. The main causes of readmission were respiratory decompensation (48%) and cardiac conditions (16%). Compared with the total sample, patients readmitted were older, mostly had vascular (39%) or gastrointestinal (26%) disease, and had significantly higher initial severity of illness (P = .003, .007) and significantly more comorbid conditions (P = .005). For all surgical classifications except general surgery, readmission was independently associated with type of admission and need for mechanical ventilation. Long-term mortality was higher among patients who were readmitted than among the total sample. Nevertheless, quality-of-life scores were the same for patients who were readmitted and patients who were not. Conclusion The adverse effect of readmission to the intensive care unit on survival appears to be long-lasting, and predictors of readmission are scarce.

Download Full-text

AN ICD-9 BASED ILLNESS SEVERITY SCORE (ICISS) OUT PERFORMS APACHE II IN PREDICTING SURVIVAL, HOSPITAL CHARGES AND LENGTH OF STAY IN SURGICAL INTENSIVE CARE UNIT PATIENTS

Critical Care Medicine ◽

10.1097/00003246-199801001-00013 ◽

1998 ◽

Vol 26 (Supplement) ◽

pp. 25A ◽

Cited By ~ 2

Author(s):

Toan Huynh ◽

Jeff Guy ◽

Robert Rutledge

Keyword(s):

Intensive Care Unit ◽

Intensive Care ◽

Length Of Stay ◽

Severity Score ◽

Illness Severity ◽

Apache Ii ◽

Hospital Charges ◽

Surgical Intensive Care Unit ◽

Surgical Intensive Care ◽

Illness Severity Score

Download Full-text

Validation of the APACHE IV model and its comparison with the APACHE II, SAPS 3, and Korean SAPS 3 models for the prediction of hospital mortality in a Korean surgical intensive care unit

Korean Journal of Anesthesiology ◽

10.4097/kjae.2014.67.2.115 ◽

2014 ◽

Vol 67 (2) ◽

pp. 115 ◽

Cited By ~ 17

Author(s):

Hannah Lee ◽

Yoon-Jung Shon ◽

Hyerim Kim ◽

Hyesun Paik ◽

Hee-Pyoung Park

Keyword(s):

Intensive Care Unit ◽

Intensive Care ◽

Hospital Mortality ◽

Apache Ii ◽

Surgical Intensive Care Unit ◽

Surgical Intensive Care ◽

Saps 3 ◽

Iv Model ◽

Apache Iv

Download Full-text

Obesity Does Not Affect Mortality after Trauma

The American Surgeon ◽

10.1177/000313480607201028 ◽

2006 ◽

Vol 72 (10) ◽

pp. 966-969 ◽

Cited By ~ 9

Author(s):

Rodrigo F. Alban ◽

Sergey Lyass ◽

Daniel R. Margulies ◽

M. Michael Shabot

Keyword(s):

Intensive Care Unit ◽

Risk Factor ◽

Intensive Care ◽

Adverse Outcomes ◽

Apache Ii ◽

Surgical Intensive Care Unit ◽

Apache Ii Score ◽

Surgical Intensive Care ◽

Significant Difference ◽

Nonobese Patients

Although obesity has been proposed as a risk factor for adverse outcomes after trauma, numerous studies report conflicting results. The objective of this study was to compare outcomes of obese and nonobese patients after trauma. The study population consisted of all trauma patients admitted to a surgical intensive care unit in a Level I trauma center from January 1999 to December 2002. Admission data, demographics, injury severity score (ISS), severity of illness, hospital course, complications, and outcomes were compared between obese (OB; body mass index [BMI] ≥ 30), and nonobese patients (NOB; BMI ≤ 29). A total of 918 patients was included in the study, 135 OB (14.7%) and 783 NOB (85.3%). There was no significant difference in demographic data, ISS, APACHE II score, and hospital stay. Intensive care unit stay was longer for OB patients (6.8 vs 4.8 days, P = 0.04). Overall mortality was 5.9 per cent for OB and 8.0 per cent for NOB patients (P = 0.48). Mortality by mechanism of injury was 3.4 per cent OB versus 7.4 per cent NOB (P = 0.26) for blunt and 10.6 per cent OB versus 10.2 per cent NOB (P = 0.9) for penetrating injury. The three most common complications associated with death were pulmonary, cardiovascular, and neurological deterioration. Using logistic regression analysis, age and ISS and APACHE II scores were associated with mortality, but BMI was not. We conclude that obesity does not appear to be a risk factor for adverse outcomes after blunt or penetrating trauma. Further research is warranted to uncover the reason for discrepant findings between centers.

Download Full-text