Acute kidney injury risk prediction score for critically-ill surgical patients.

BACKGROUND: There has been a global increase in the incidence of acute kidney injury (AKI), including among critically-ill surgical patients. AKI prediction score provides an opportunity for early detection of patients who are at risk of AKI; however, most of the AKI prediction scores were derived from cardiothoracic surgery. Therefore, we aimed to develop an AKI prediction score for major non-cardiothoracic surgery patients who were admitted to the intensive care unit (ICU). METHODS: The data of critically-ill patients from non-cardiothoracic operations in the Thai Surgical Intensive Care Unit (THAI-SICU) study were used to develop an AKI prediction score. Independent prognostic factors from regression analysis were included as predictors in the model. The outcome of interest was AKI within 7 days after the ICU admission. The AKI diagnosis was made according to the Kidney Disease Improving Global Outcomes (KDIGO)-2012 serum creatinine criteria. Diagnostic function of the model was determined by area under the Receiver Operating Curve (AuROC). Risk scores were categorized into four risk probability levels: low (0-2.5), moderate (3.0-8.5), high (9.0-11.5), and very high (12.0-16.5) risk. Risk of AKI was presented as likelihood ratios of positive (LH+). RESULTS: A total of 3474 critically-ill surgical patients were included in the model; 333 (9.6%) developed AKI. Using multivariable logistic regression analysis, older age, high Sequential Organ Failure Assessment (SOFA) non-renal score, emergency surgery, large volume of perioperative blood loss, less urine output, and sepsis were identified as independent predictors for AKI. Then AKI prediction score was created from these predictors. The summation of the score was 16.5 and had a discriminative ability for predicting AKI at AuROC = 0.839 (95% CI 0.825-0.852). LH+ for AKI were: low risk = 0.117 (0.063-0.200); moderate risk = 0.927 (0.745-1.148); high risk = 5.190 (3.881-6.910); and very high risk = 9.892 (6.230-15.695), respectively. CONCLUSIONS: The function of AKI prediction score to predict AKI among critically ill patients who underwent non-cardiothoracic surgery was good. It can aid in early recognition of critically-ill surgical patients who are at risk from ICU admission. The scores could guide decision making for aggressive strategies to prevent AKI during the perioperative period or at ICU admission. TRIAL REGISTRATION: TCTR20190408004, registered on April 4, 2019.

Characters of Nutrition Status and Energy-delivery Patterns of the University-based Surgical Intensive Care Units in Thailand (Multi-center THAI-SICU Study).

AIM: The authors aimed to describe nutrition status and energy-delivery characters in multi-center THAI-SICU study. MATERIAL AND METHODS: Eligible patients admitted in SICU were 1,686 after excluding 563 of 2,249 participants owing to very short stay or non-alive within 24 hours after admission and missing data. The study was a posthoc analysis and multicenter descriptive design. The analytic methods described categorical data in percentage and the continuous data in the median with interquartile range. Variables divided into baseline characteristics and nutrition data before SICU admission, and the pattern of energy delivery in SICU. Statistical significance accepted as a p-value less than 0.05. RESULTS: The average age was 64 (52-76) years with 57% male. The median of serum albumin level at admission (interquartile range, IQR) was 2.8 (2.2-3.4). There was 46 -47 percent of nutrition risk patient. Less than 10 percent of the patient had enteral (EN), parenteral (PN) or their combination before admission. History of weight loss and appetite loss was 27-31 percent. However, seventy percent of the patient could not define the duration of the symptom. EN was initiated early, but the tendency of full feeding was 7-10 days. At that period, supplemental PN was added around 30 percent of total calories. The composition of PN was quite low in these study which contains only 15-16 percent of total calories. The average energy delivery was 20 kcal/kg/day (the recommendation is 25-30 kcal/kg/day). CONCLUSION: The patient's nutrition status before SICU admission was at risk of 46-47% and weight loss and appetite loss might unreliable in ICU setting. EN is started early with gradually increase up to 7-10 days. The average total calories requirement is lower than a recommendation.

The Impacts of Surgical Intensive Care Unit Admission Source on Morbidity and Mortality Outcomes: The Results from the THAI-SICU Study.

Objective: The present study aims to examine the association between admission source and outcomes in surgical ICU (SICU) patients. Material and Method: The data in the present report were retrieved from the THAI-SICU database which was designed as a multi-center prospective observational study. The data were collected at 9 university-based surgical ICUs over 22 months from April 2011 to January 2013. Results: The sources of SICU admissions were categorized into operating room (OR) group with 3,238 admissions (69.6%), emergency room (ER) group with 499 admissions (10.7%), ward group with 825 admissions (17.7%), and other ICUs group with 90 admissions (1.9%). In view of transfer from other hospitals, the transfer group included 938 patients (20.2%) while the non-transfer group included 3,714 patients (79.8%). Patients admitted from other ICUs were nearly three-times more likely to die in SICU (adjusted odds ratio (OR) 2.89; 95% confidence interval (CI) 1.52-5.51, p = 0.001) than those who came from operating room. However, the ward group still had a high risk to dying (OR 2.49; 95 % CI 1.88-3.30, p<0.001). In view of outcomes for inter-hospital transfer patients, the transfer group was at greater risk of dying in SICU and had greater risk of 28-day mortality than the non-transfer group. Conclusion: Surgical, critically ill patients, who transferred from other ICUs to SICU, have the highest risk of ICU morbidity and mortality. In addition, ward patients and transfer patients also have high risk of unfavorable outcomes.

The Association between Red Blood Cell Transfusion and Hospital Mortality in Critically Ill Surgical Patients: The Multi-center Thai University-Based Surgical Intensive Care Units Study (THAI-SICU Study).

Objective: Red blood cell transfusion (RBCT) is commonly prescribed to critically ill patients with anemia. Nevertheless, the benefits of RBCT in these patients, particularly critically ill surgical patients, are still controversial. The aim of this study is to explore the association between RBCT and hospital mortality in Thai critically ill surgical patients. Material and Method: This study was a part of the multi-center, prospective, observational study, which included adult patients admitted to the SICUs after surgery. Patients were categorized into transfusion and no transfusion groups according to whether they received RBCT during SICU stay or not. The multiple logistic regression analysis was performed to determine whether RBCT was an independent risk factor for hospital mortality. The patients were also matched between two groups based on the propensity score for RBCT requirement and were then compared. Results: There were 2,531 patients included in this study. The incidence of RBCT in SICU was 40.3%. Overall, there was no association between RBCT in SICU and hospital mortality (adjusted OR 1.33, 95% CI 0.83-2.11) except in the subgroup of patients with age of <65 years old (adjusted OR 1.99, 95% CI 1.03-3.84). However, when the amount of RBCT was more than 1,200 mL, it was independently associated with increased hospital mortality (adjusted OR 2.55, 95% CI 1.35-4.81). In the propensity-score matching cohort, there was no association between RBCT in SICU and hospital mortality (adjusted OR 1.56, 95% CI 0.88-2.77) except when the amount of RBCT was more than 600 mL (601-1,200 mL, adjusted OR 3.14, 95% CI 1.47-6.72 and >1,200 mL, adjusted OR 3.58, 95% CI 1.36-9.48). Conclusion: RBCT should be considered as a life-saving intervention but with potential risks of adverse events. Identifying patients who will likely gain benefit from RBCT and implementing the restrictive transfusion strategy may be the keys to improve outcomes.

Incidence and Risk Factors of Unplanned Extubation in Critically Ill Surgical Patients: The Multi-center Thai University-based Surgical Intensive Care Units Study (THAI-SICU Study).

Objective: Unplanned extubation (UE) is one of the most troubling events in critically ill patients who require endotracheal intubation and mechanical ventilation. The aims of this study are to determine the incidence and to identify the risk factors associated with UE in critically ill surgical patients. Material and Method: This was a multi-center prospective observational cohort study, which included surgical patients admitted to nine university-based surgical intensive care units (SICUs) in Thailand between April 2011 and January 2013. UE was defined as deliberate extubation by patients (self-extubation) or accidental extubation during procedures or transportation. The incidence of UE was calculated, the adjusted logistic regression model was performed to determine the independent risk factors for UE and the outcomes were compared between those with planned extubation and UE. Results: 2,890 patients required endotracheal intubation and mechanical ventilation were included in the analysis. Of these, 54 patients experienced UE and, therefore, the incidence of UE was 1.9%. Five independent risk factors for UE were identified; congestive heart failure (adjusted odds ratio, OR, 3.48; 95% CI, 1.29-9.40), emergency surgery (adjusted OR, 2.18; 95% CI, 1.01-4.74), non-postoperative status (adjusted OR, 2.37; 95% CI, 1.05-5.37), sedation usage (adjusted OR, 3.19; 95% CI, 1.72-5.93) and delirium (adjusted OR, 3.61; 95% CI, 1.71-7.60). ICU length of stay (LOS) was significantly longer in patients with UE than those with planned extubation (adjusted coefficient, 2.76; 95% CI, 1.34-4.19). There was no significant difference between the two groups in terms of hospital LOS as well as ICU and 28-day mortality. Conclusion: The incidence of UE in critically ill surgical patients was 1.9%. Five independent risk factors for UE were: underlying congestive heart failure, emergency surgery, non-postoperative status, sedation usage, and delirium. Patients with UE had significantly longer ICU LOS than those with planned extubation.

The Differences of Nutrition Status, Energy Delivery and Outcomes between Metropolis and Regional University-Based Thai Surgical Intensive Care Units.

Objective: The objective of this study was to compare the differences of nutrition status, nutrition delivery, and the outcomes between the metropolis (MUH) and regional university based hospitals (RUH) in Thailand. Material and Method: The nutrition data were retrieved from the THAI-SICU database. A total of 1,686 patients (MUH 927 patients vs. RUH 759 patients) with completion of nutrition status and nutrition delivery data were included in this analysis. The enrolled patients from study centers located in Bangkok were defined as MUH, and the patients from Chiang Mai were defined as RUH. Patient characteristics, nutrition status using the subjective global assessment (SGA) and nutrition risk screening (NRS), nutrition delivery, and outcomes of treatment were recorded. The outcome associations were analyzed by a multivariable regression model. Results: At admission, there were significant differences of age, gender, body mass index, disease severity, albumin level, and diagnosis. RUH had significantly poorer nutritional status than MUH (RUH vs. MUH: SGA class B and C, 57.7% vs. 37.1%, p<0.001; NRS at risk, 56.3% vs. 38.4%, p<0.001). The tendency of total calories and enteral nutrition delivery per day of RUH was significantly lower than MUH especially in the first three weeks of hospitalization. Carbohydrates were the main resource for parenteral nutrition. Although there was no difference of protein delivery, MUH had a significantly higher prescription of fat emulsion especially in the 1st-2nd weeks. Even though there were higher occurrences of intensive care unit (ICU) mortality, 28-day mortality, sepsis occurrence, ICU length of stay (LOS), and hospital LOS in RUH, the multivariable analysis did not demonstrate the statistical value of these outcomes. Conclusion: RUH had a poorer nutritional status. MUH had more total caloric intake and enteral nutrition delivery per day especially during the first three weeks. However, the treatment outcomes showed no differences in multivariable analysis.

Outcomes of Nutrition Status Assessment by Bhumibol Nutrition Triage/Nutrition Triage (BNT/NT) in Multicenter THAI-SICU Study.

Objective: The objective of the study is to evaluate the nutrition assessment tool used by Bhumibol Nutrition Triage/Nutrition Triage (BNT/NT) for patient outcomes in a surgical intensive care unit (SICU). Material and Method: All data were retrieved from the THAI-SICU database. A total of 1,685 patients from three medical centers were participants in the nutrition project and were enrolled onto this study. The parameters needed for BNT/NT scoring were recorded including body mass index (BMI), weight change, energy delivery, age, and disease severity. The BNT/NT calculation was classified into 4 groups as BNT/NT I to IV. An adjusted odds ratio (OR) with 95% confidence interval (CI) of mortality and sepsis occurrence were reported. Results were classed as being statistically significant at p<0.05. Results: Regarding the nutrition assessment classification, the patients admitted to SICU were classified as BNT/NT class I48.6%, class II 30.0%, class III 9.3%, and class IV 12.1%. There were statistically significant differences between classes in terms of BMI, weight change, energy delivery and disease severity. In addition, the BNT/NT classification was also significantly associated with ICU mortality [OR (95% CI): 1.51 (1.25-1.83); p<0.001], 28 day mortality [1.47 (1.23-1.74); p<0.001], and sepsis occurrence [1.41 (1.25-1.60); p<0.001]. Conclusion: Most of the patients admitted to SICU had a low nutrition risk BNT/NT class I and II. The higher BNT/NT scores were associated with mortality and sepsis occurrence in SICU.

Pain Management in Surgical Intensive Care Units: A Multi-Center Prospective Observational Study (THAI-SICU Study).

Objective: This study is a part of the multi-center Thai university-based Surgical Intensive Care Unit Study (THAI-SICU Study). It aimed to evaluate the patterns of pain management in patients admitted to surgical intensive care units. Material and Method: Case record forms (CRFs) were created by the working group. Data regarding pain management in the ICUs were documented on the daily record form. These included types of analgesics used (opioids and non-opioids), routes of administration (oral, intravenous, intramuscular, epidural and intrathecal) and methods of administration (continuous infusion, regular intermittent, as needed, patient-controlled analgesia and patient-controlled epidural analgesia). Results: Data were gathered from 4,652 patients. The majority of the patients received analgesics (85.2%). The main stay analgesics were morphine (52.3%) and fentanyl (27%). Analgesics were frequently administered via the intravenous route (76.5%) on an as needed basis (48.6%). Conclusion: Analgesics were commonly given to patients in the surgical intensive care units. The analgesics of choice were strong opioids, and the most preferred route and method of administration was the intravenous route and the as needed basis, respectively.

A Multi-center Thai University-based Surgical Intensive Care Unit Study (THAI-SICU Study): Incidence of Acute Myocardial Infarction and Associated Factors.

Introduction: To describe the incidence, characteristics and outcomes of acute myocardial infarction (AMI) and determine risk factor(s) of AMI in THAI-surgical intensive care unit (SICU). Material and Method: This study was multicenter prospective cohorts study that conducted data from 9 university-affiliated SICUs in Thailand between April 2011 and January 2013. We collected and evaluated data of AMI events. The patients were followed-up for up to 28 days after admitted to the SICUs. Results: The overall incidence of AMI in SICU was 1.4% (66 of 4,652 patients). Non-ST elevated MI was the most common electrocardiography (ECG) presentation (75%). The common clinical sign and symptom of AMI included ECG changes (53%) and elevation of cardiac enzymes (48.5%). Patients with AMI had significantly higher 28-day mortality rate (28.8% versus 13.6%, p<0.001) than those with non-MI. The Acute Physiologic and Chronic Health Evaluation (APACHE) II scores (RR 1.04, 95% CI 1.01-1.07, p = 0.003) and age >65 year (RR 2.54, 95% CI 1.36-4.75, p = 0.003) were significant risk factors of AMI. Conclusion: The incidence of AMI in the SICU was uncommon but led to significantly higher mortality rates. The APACHE II score and age ≥65 year were significant predictors of AMI in SICU.

Incidence of Cardiac Arrest and Related Factors in a Multi-Center Thai University-Based Surgical Intensive Care Units Study (THAI-SICU Study)

Objective: To describe the incidences, outcomes and determine the risk factor(s) of cardiac arrest in surgical intensive care unit (SICU). Material and Method: We collected data between April 2011 and January 2013. The case record form (CRF) included the CRF 1 (admission, daily screening and discharge data) and the CRF 2 for cardiac arrest events. The patients were followed-up until discharge from SICU or for up to 28 days after admission in SICU. Results: The incidence of cardiac arrest in SICU was 226 in 4,652 patients (4.9%). The APACHE II score at the day with cardiac arrest were 24.1. Initial monitor rhythm during cardiac was asystole (35.4%), bradycardia (22.6%) and pulseless electrical activity (14.6%). The main cause was poor patient condition before admission (51.3%). Most of the cardiac arrest patients (73.9%) had antecedents within 24 hour and the most common antecedents were hypotension, metabolic disturbances and sepsis and/or septic shock. The overall return of spontaneous circulation rate was 23.5%. At hospital discharge, the mortality rate (91.6%) was statistically different between the cardiac arrest and non-cardiac arrest group (p<0.001). The Acute Physiologic and Chronic Health Evaluation II score (APACHE II score) (Odds ratio, (OR 1.15, 95% CI 1.11-1.19, p<0.001), Sequential Organ Failure Assessment score (SOFA score) (OR 1.12, 95% CI 1.03-1.20, p = 0.005) and American Society of Anesthesiologists physical status physical status (ASA PS) ≥3 (OR 2.32, 95% CI 1.33-4.04, p = 0.003) were significantly risk factors for cardiac arrest. Conclusion: Cardiac arrest in the SICU was uncommon. Initial non-shockable rhythms were common and mostly had antecedents before cardiac arrest. The APACHE II score, SOFA score and ASA PS ≥3 were independent risk factors for cardiac arrest in SICU.

The Impact of Delirium on Clinical Outcomes in Multi-Center Thai Surgical Intensive Care Units: A Prospective Cohort Study.

Objective: Delirium in intensive care units (ICU) increases risks in prolonged mechanical ventilation, hospitalization, and mortality rate. The purpose of this study is to determine if delirium in the surgical intensive care units (SICU) is an independent predictor of clinical outcomes during hospitalization. Material and Method: A multi-center, prospective cohort study was conducted between April 2011 and January 2013. All patients who were admitted to nine university-based SICU were enrolled. Delirium was diagnosed by using the Intensive Care Delirium Screening Checklists. The clinical outcomes of study included length of mechanical ventilation, length of hospital stay, ICU and 28 day mortality. Cox proportional hazard regression model was used to assess the effects of delirium on ICU and 28 day mortality. Results: A total of 4,652 patients were included. One hundred and sixty-three patients were diagnosed delirium (3.5%, 163 of 4,652). Patients who experienced delirium during ICU admission were significantly older (65.0+15.8 years versus 61.6+17.3 years, p = 0.013), had higher American Society of Anesthesiologists physical status (24.3% versus 12.2%, p<0.001), higher Acute Physiology and Chronic Health Evaluation II score (16 (12-23) versus 10 (7-15), p<0.001), and higher Sequential Organ Failure Assessment score (5 (2-8) versus 2 (1-5), p<0.001) compared to non-delirium. Delirious patients also had higher ventilator days (7 (4-17) versus 2 (1-4), p<0.001, longer length of hospital stay (22 (14-34) versus 15 (9-26), p<0.001) and higher ICU mortality (24% versus 9%, p<0.001), and 28-day mortality (28% versus 13%, p<0.001). Patients who developed delirium in the intensive care unit were associated with increased 28-day mortality (adjusted HR = 2.47, 95% CI: 1.13-5.41, p = 0.023). Conclusion: Delirium in an ICU was a major predictor of hospital mortality after adjusted for relevant covariates. Routine monitoring of delirium, early detection, and implementation of preventive strategy are recommended.

Predictors of Prolonged Length of Stay in General Surgical Intensive Care Unit.

Objective: The aims of this study were to explore the incidence, clinical factors, severity scores and outcome associated with prolonged length of intensive care unit (ICU) stay >3 days. Material and Method: This study was a prospective observational study on the noncardiac surgical patients who were admitted to surgical intensive care unit in a tertiary university hospital. The cardiothoracic, neurosurgical and traumatic surgical patients or the patients who died within the first 3 days of ICU admission were excluded. Patient demographic data, preoperative predictors and severity scores (APACHE II, APACHE III, SOFA, SAPS II and MODS scores) at day 1 and day 3 of ICU admission were recorded. Results: A total of 948 patients were observed. The incidence of prolonged ICU stay was 20.1%. Patients with prolonged ICU stay had significantly higher ventilator hours (p<0.001) and ICU length of stay (p<0.001). On the multivariable analysis model of preoperative variables, the significant predictors of prolonged ICU stay were preoperative serum albumin less than 2.6 mg/dL (p = 0.023), preoperative hematocrit less than 34% (p = 0.035), emergency surgery (p = 0.003), having surgical complications (p = 0.017), having anesthetic complications (p = 0.017), admission for respiratory support with or without unstable hemodynamic (p<0.001), and sepsis on ICU admission (p = 0.003). Regarding the multivariable analysis of severity scoring system, the significant severity predictors were found only the preoperative ASA class IV (p<0.001) and emergency ASA status (p<0.001). Conclusion: About one-fifth of the study patients had prolonged ICU stay (>3 days). Low preoperative serum albumin (<2.6 mg/dL), low preoperative hematocrit (<34%), ASA physical status class IV, underwent emergency surgery, having anesthetic complications, surgical complications, sepsis on ICU admission, having respiratory support with or without unstable hemodynamic were significantly associated with prolonged ICU stay.

Incidence of and Risk Factors for Acute Respiratory Distress Syndrome in Patients Admitted to Surgical Intensive Care Units: The Multicenter Thai University- Based Surgical Intensive Care Unit (THAI-SICU) Study.

Objective: The incidence and outcomes of acute lung injury (ALI)/acute respiratory distress syndrome (ARDS) are unclear. We evaluated the cumulative incidence of, risk factors for, and outcomes of ALI/ARDS in surgical ICUs (SICUs). Material and Method: The multicenter Thai University-based Surgical Intensive Care Unit (THAI-SICU) study was a prospective, observational cohort study including nine university-based SICUs throughout Thailand from April 2011 to January 2013. All >18-year-old surgical patients who were admitted to general SICUs were recruited. The primary outcome was the incidence of ALI/ARDS. Results: In total, 4,652 patients were analyzed. ALI/ARDS new developed in 114 patients (2.5%). Patients with ALI/ARDS had higher APACHE II (20.0 vs. 11.4, respectively; p<0.001) and SOFA scores (7.3 vs. 3.1, respectively; p<0.001) and a higher incidence of past or current smoking (48% vs. 36%, respectively; p<0.001) than the non-ARDS patients. The 28-day mortality rate was significantly higher in patients with than without ALI/ARDS (50% vs. 12%; p<0.001). Higher APACHE II and SOFA scores and higher rates of current or past smoking were independent predictors of ALI/ARDS. Conclusion: The incidence of ALI/ARDS in the THAI-SICU study was low, but the mortality rate was high. Higher severity scores and smoking were associated with ALI/ARDS.

A Multi-Center Thai University-Based Surgical Intensive Care Units Study (THAI-SICU Study): Outcome of ICU Care and Adverse Events.

Objective: Surgical intensive care units (SICUs) are special units for critically ill surgical patients both in the pre and postoperative period. There is little aggregated information about surgical patients who are admitted to the Thai surgical ICU. The objective of this report was to describe patient characteristics, outcomes of ICU care, incidence and outcomes of adverse events in the SICU in the participating SICUs. Material and Method: This multi-center, prospective, observational study of nine university-based SICUs was done. All admitted patients with ages >18 years old were included. Information about patient characteristics, underlying medical problems, indication and type of ICU admission, severity score as ASA physical status in operative patients, APACHE II score and SOFA score, adverse events of interest, ventilator days, ICU and 28 days mortality. The association of outcome and predictors was reported by relative risk (RR) with 95% confidence interval (95% CI). Statistical significant difference was defined by p<0.05. Results: During April 2011-January 2013 of total cohort time, a total of 4,652 patients from nine university-based SICUs were included in this study. Mode of patient age was 71-75 year old for both sexes. Median (IQR) of APACHE II scores and SOFA scores were 10 (7-10) and 2 (1-5), respectively. Seventy eight percent of patients were postoperative patients and 50% of them were ASA physical status III. The median of ICU stay was 2 (IQR 1-4) days. Each day of ICU increment was associated with increased 1.4 days of a hospital stay. Three percent of survived at discharge were clinically inappropriate discharge resulting in ICU readmission. Sixty-five percent were discharged home after ICU admission. ICU and 28 days mortality was 9.6% and 13.8%. The seven most common adverse events were sepsis (19.5%), acute kidney injury (AKI) (16.9%), new cardiac arrhythmias (6.2%), acute respiratory distress syndrome (ARDS) (5.8%), cardiac arrest (4.9%), delirium (3.5%) and reintubation within 72 hours (3.0%), respectively. Most of the adverse events occurred in the first five days, significantly less occurred after 15 days of ICU admission. The association between adverse events and 28 days mortality were significant for cardiac arrest (RR, 9.5; 95% CI, 8.6-10.4), respiratory failure [acute respiratory distress syndrome (ARDS) (RR, 4.6; 95% CI, 3.9-5.3), acute lung injury (ALI) (RR, 2.7; 95% CI, 2.1-3.6)], acute kidney injury (AKI) (RR, 4.2; 95% CI, 3.7-4.8), sepsis (RR, 3.6; 95% CI, 3.2-4.2), iatrogenic pneumothorax (RR, 3.2; 95% CI, 2.1-5.1), new seizure (RR, 3.1, 95% CI, 2.2-4.4), upper GI hemorrhage (RR, 3.0, 95% CI, 2.1-4.1), new cardiac arrhythmias (RR, 2.9; 95% CI, 2.4-3.5), delirium (RR, 2.1; 95% CI, 1.7-2.8), acute myocardial infarction (RR, 2.1; 95% CI, 1.4-3.1), unplanned extubation (RR, 2.1; 95% CI, 1.4-3.1), intra-abdominal hypertension (RR, 1.8; 95% CI, 1.2-2.7) and reintubation within 72 hours (RR, 1.5; 95% CI, 1.1-2.1). Conclusion: This is the first large study of surgical critical care in Thailand, which had a systematic patient follow-up program. Most of the patients were elderly. Adverse events were most frequent during the first 5 days of admission and were associated with ICU and 28 days mortality.

Pitfalls in fluid management for critically ill patients in Thailand.

Fluid management is one of the most important treatments for critically ill patients. It has an influence in patients outcomes and is considered one of the most common pitfalls encountered in the management of the critically ill patient. In Thailand, fluid overload (> 10% of fluid accumulation), mismanagement of fluid restrictions and the "bolusing" of colloid solutions are the main pitfalls that may lead to serious complications. These complications can compromise the patient in areas such as decreasing the oxygen index, putting the patient in cardiac failure and possible acute kidney injury (AKI). They can also increase resource utilization and the mortality of critically ill patients. More than 80% of critically ill patients, who are admitted to the intensive care units, are reimbursed from the "universal coverage". Universal coverage does not support the use of albumin solution, which has been reported to improve the function of the endothelial glycocalyx layer vascular permeability and improved outcomes in the hypoalbuminemic patient (serum albumin < or = 2.5 mg/dL) with severe sepsis, in septic shock or undergoing major abdominal surgery. Therefore, primary colloids used for resuscitation the patients are 6% hydroxyethyl starch (HES), 4% gelatin and fresh frozen plasma. AKI and renal replacement therapy (RRT) continue to be a major problem when using these synthetic colloids especially in the high-risk patients who receive large amounts of fluids. Evaluation of the fluid responsiveness for goal directed therapy is another problem in Thailand. This has been predominant in critically ill surgical patients both intra-operatively and postoperatively. To obtain optimal benefits of fluid therapy and for the prevention of complications associated with this treatment, physician need to acquire more knowledge, choose the right strategy, choose the proper type and amount of colloid and assure the correct mode of evaluation.

A multi-center Thai university-based surgical intensive care units study (THAI-SICU study): methodology and ICU characteristics.

OBJECTIVE: Although there were two large intra-operative observational studies on Thai surgical patients (THAI and THAI-AIMS), there has been no available study on critically ill surgical patients regarding their adverse events and outcomes. A THAI-Surgical Intensive Care Unit (SICU) study has been established for monitoring the occurrence of these adverse events and outcomes in the SICU. The objective of this report is to describe the methodology of the THAI-SICU study and participating SICUs' characteristics as well as the early recruitment results on patients enrolled in the present study. MATERIAL AND METHOD: The present study is designed as a multi-center, prospective, observational study. This report describes the method of case record form development and summarizes their collected parameters as well as the adverse event surveillance variables. All of nine SICU characteristics are described regarding their management systems, physicians' and nurses' work patterns. The final group of enrolled patients is reported. RESULTS: A total of nine university-based SICUs were included in the present study. All participating hospitals are residency training centers. Four of the SICUs, fulltime directors are anesthesiologists. Only one hospital's SICU is directed by a surgeon. Two SICUs were closed ICUs, three were mandatory consulting units, one was an elective consultation unit and the remaining three ICUs had no directors. Most of the participating SICUs had heterogeneity of surgical specialty patients. Six SICUs had regular resident rotations and only two of the SICUs had critical care fellowship training. There were significant differences regarding the nursing workload among the ICUs. The patient to registered nurse ratio ranged from 0.9-2.0. After a total of 19.7 months of a recruitment period, the total number of patient admissions was 6,548 (1,894 patients were excluded). A total cohort of 4,654 patients was included for further analytical processes. CONCLUSION: There were differences in ICU management systems, physician and specialist coverage, nurse burdens, nurse sparing, and types of patients admitted in the university based SICUs. This presentation is the pioneer multi-center study on Thai SICUs in which adverse events and outcomes are reported.

Adherence to lung protective mechanical ventilation in patients admitted to a surgical intensive care unit and the associated increased mortality.

Background: The adherence rate to the lung protective ventilation (LPV) strategy, which is generally accepted as a standard practice in mechanically ventilated patients, reported in the literature is approximately 40%. This study aimed to determine the adherence rate to the LPV strategy, factors associated with this adherence, and related clinical outcomes in mechanically ventilated patients admitted to the surgical intensive care unit (SICU). Methods: This prospective observational study was conducted in the SICU of a tertiary university-based hospital between April 2018 and February 2019. Three hundred and six adult patients admitted to the SICU who required mechanical ventilation support for more than 12 h were included. Ventilator parameters at the initiation of mechanical ventilation support in the SICU were recorded. The LPV strategy was defined as ventilation with a tidal volume of equal or less than 8 ml/kg of predicted body weight plus positive end-expiratory pressure of at least 5 cm H2O. Demographic and clinical data were recorded and analyzed. Results: There were 306 patients included in this study. The adherence rate to the LPV strategy was 36.9%. Height was the only factor associated with adherence to the LPV strategy (odds ratio for each cm, 1.10; 95% confidence interval (CI), 1.06-1.15). Cox regression analysis showed that the LPV strategy was associated with increased 90-day mortality (hazard ratio, 1.73; 95% CI, 1.02-2.94). Conclusion: The adherence rate to the LPV strategy among patients admitted to the SICU was modest. Further studies are warranted to explore whether the application of the LPV strategy is simply a marker of disease severity or a causative factor for increased mortality.

Incidence and risk factors of perioperative stroke in Siriraj Hospital.

BACKGROUND: Data concerning perioperative stroke incidence and risk factors are lacking in Thailand. OBJECTIVE: To study incidence and risk factors of perioperative stroke in Siriraj Hospital, Bangkok, Thailand. MATERIAL AND METHOD: The authors conducted a nested case-control study between July 2007 and June 2010. Consecutive perioperative stroke cases were compared with age-matched controls that had undergone surgery without having a stroke at a 1:4 ratio. Patients' characteristics, co-morbidities, clinical manifestation, stroke subtypes, duration, and types of surgery were collected. Multiple logistic regression analysis was performed to identify factors associated with a stroke during perioperative period. RESULTS: Sixty-six patients with perioperative stroke from 99,283 that underwent surgery were included. This resulted in an incidence of 66.5/100,000. Men comprised 65.2%. Mean age of stroke patients was 66.6-years-old. Age-match controls included 264 consecutive patients who underwent surgery without having a stroke. A multivariate analysis showed that male sex (adjusted OR 3.1, p = 0.003), surgical types: open heart surgery (adjusted OR 54.3, p < 0.0001), vascular surgery (adjusted OR 20.6, p < 0.0001) and endovascular procedure (adjusted OR 15.4, p < 0.0001), emergency surgery (adjusted OR 8.3, p < 0.0001), the presence of diabetes mellitus (adjusted OR 2.6, p = 0.018), chronic kidney disease (adjusted OR 2.6, p = 0.027), and coronary artery disease (adjusted OR 0.4, p = 0.039) were associated with perioperative stroke. CONCLUSION: Incidence of perioperative stroke was higher than a previous report. Male, type of surgery, emergency operation, diabetes mellitus, and chronic kidney disease were risk factors of perioperative stroke.

Extracorporeal membrane oxygenation (ECMO) support for acute hypoxemic respiratory failure patients: outcomes and predictive factors.

Background: Extracorporeal membrane oxygenation (ECMO) is an important rescue therapy for patients with refractory respiratory or circulatory failure. High cost and associated complications warrant careful case selection. The aim of this study was to investigate the outcomes and factors associated with mortality in acute hypoxemic respiratory failure patients who received ECMO support, and to externally validate preexisting ECMO survival prediction scoring systems. Methods: This retrospective study enrolled acute hypoxemic respiratory failure patients who received veno-venous (VV) or veno-arterial (VA) ECMO support at Siriraj Hospital (Bangkok, Thailand) from 2010 to 2020. All relevant baseline patient characteristics including ECMO survival prediction scores were recorded. The primary outcome was in-hospital mortality. Multivariate logistic regression analysis was employed to identify independent predictors of in-hospital mortality. Results: Of a total of 65 patients, 34 (52%) were male, the median (IQR) age was 61 years (49-70 years), the median body mass index (BMI) was 22.6 kg/m2 (20.6-28 kg/m2), and the median Sequential Organ Failure Assessment (SOFA) score was 13 [11-16]. Forty-three patients (66%) received VV-ECMO, and 22 (34%) received VA-ECMO support. In-hospital mortality was 69%. Multivariate analysis identified a SOFA score >14, hospitalized >72 hours before ECMO initiation, PaO2/FiO2 ratio <60, and pH <7.2 as independent predictors of in-hospital mortality. These four parameters were combined to create the SHOP (S: SOFA >14, H: hospitalize >72 hours, O: PF ratio <60, and P: pH <7.2) score. Compared with three different preexisting ECMO survival prediction scoring systems, the SHOP score had the highest area under the curve (AUC) for predicting in-hospital mortality (overall: 0.873, VV-EMCO: 0.866, and VA-EMCO: 0.891). Conclusions: In-hospital mortality among ECMO-supported patients was high at 69%. SOFA score >14, hospitalized >72 hours, PaO2/FiO2 ratio <60, and pH <7.2 were found to be independent predictors of in-hospital mortality. A SHOP score of 2 or higher significantly predicts in-hospital mortality in EMCO-supported patients. Trial Registration: www.clinicaltrials.gov (reg. No. NCT04031794).

Protocol-directed vs. physician-directed weaning from ventilator in intra-abdominal surgical patients.

BACKGROUND: Previous studies have demonstrated that protocol-directed weaning is better than physician-directed weaning in terms of shorter duration of mechanical ventilation in general critically ill patients. In this prospective, randomized controlled trial, the authors compared duration of mechanical ventilation between protocol based nurse-directed weaning and physician-directed weaning in patients following intra-abdominal surgery. MATERIAL AND METHOD: One hundred intra-abdominal surgical patients requiring mechanical ventilation for more than 24 hours were randomly assigned to receive either protocol-directed (n=51) or physician-directed (n=49) weaning from mechanical ventilation. Patients assigned to the protocol-directed weaning group underwent daily screening and a spontaneous breathing trial by nursing staff OUTCOMES: The primary outcome was the duration of mechanical ventilation. RESULTS: The median duration of mechanical ventilation was 40 and 72 hrs in protocol-directed and physician-directed groups, respectively (p < 0.001). Two patients in the protocol-directed group and three patients in the physician directed group were re-intubated within the first 72 hours after extubation (p = 0.61). CONCLUSION: Daily screening of respiratory function in intra-abdominal surgical patients followed by trials ofspontaneous breathing performed by nurses resulted in a shorter duration of mechanical ventilation when compared to traditional physician-directed weaning.