The Development of a Machine Learning Inpatient Acute Kidney Injury Prediction Model*

The Development of a Machine Learning Inpatient Acute Kidney Injury Prediction Model*

 by Koyner, Jay L.; Carey, Kyle A.; Edelson, Dana P.; Churpek, Matthew M.

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1070–1077

Objectives: To develop an acute kidney injury risk prediction model using electronic health record data for longitudinal use in hospitalized patients. Design: Observational cohort study. Setting: Tertiary, urban, academic medical center from November 2008 to January 2016. Patients: All adult inpatients without pre-existing renal failure at admission, defined as first serum creatinine greater than or equal to 3.0 mg/dL, International Classification of Diseases, 9th Edition, code for chronic kidney disease stage 4 or higher or having received renal replacement therapy within 48 hours of first serum creatinine measurement. Interventions: None. Measurements and Main Results: Demographics, vital signs, diagnostics, and interventions were used in a Gradient Boosting Machine algorithm to predict serum creatinine–based Kidney Disease Improving Global Outcomes stage 2 acute kidney injury, with 60% of the data used for derivation and 40% for validation. Area under the receiver operator characteristic curve (AUC) was calculated in the validation cohort, and subgroup analyses were conducted across admission serum creatinine, acute kidney injury severity, and hospital location. Among the 121,158 included patients, 17,482 (14.4%) developed any Kidney Disease Improving Global Outcomes acute kidney injury, with 4,251 (3.5%) developing stage 2. The AUC (95% CI) was 0.90 (0.90–0.90) for predicting stage 2 acute kidney injury within 24 hours and 0.87 (0.87–0.87) within 48 hours. The AUC was 0.96 (0.96–0.96) for receipt of renal replacement therapy (n = 821) in the next 48 hours. Accuracy was similar across hospital settings (ICU, wards, and emergency department) and admitting serum creatinine groupings. At a probability threshold of greater than or equal to 0.022, the algorithm had a sensitivity of 84% and a specificity of 85% for stage 2 acute kidney injury and predicted the development of stage 2 a median of 41 hours (interquartile range, 12–141 hr) prior to the development of stage 2 acute kidney injury. Conclusions: Readily available electronic health record data can be used to predict impending acute kidney injury prior to changes in serum creatinine with excellent accuracy across different patient locations and admission serum creatinine. Real-time use of this model would allow early interventions for those at high risk of acute kidney injury.

Increased Heat Generation in Postcardiac Arrest Patients During Targeted Temperature Management Is Associated With Better Outcomes*

Increased Heat Generation in Postcardiac Arrest Patients During Targeted Temperature Management Is Associated With Better Outcomes*

by Uber, Amy J.; Perman, Sarah M.; Cocchi, Michael N.; Patel, Parth V.; Ganley, Sarah E.; Portmann, Jocelyn M.; Donnino, Michael W.; Grossestreuer, Anne V

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1133–1138

Objectives: Assess if amount of heat generated by postcardiac arrest patients to reach target temperature (Ttarget) during targeted temperature management is associated with outcomes by serving as a proxy for thermoregulatory ability, and whether it modifies the relationship between time to Ttarget and outcomes. Design: Retrospective cohort study. Setting: Urban tertiary-care hospital. Patients: Successfully resuscitated targeted temperature management–treated adult postarrest patients between 2008 and 2015 with serial temperature data and Ttarget less than or equal to 34°C. Interventions: None. Measurements and Main Results: Time to Ttarget was defined as time from targeted temperature management initiation to first recorded patient temperature less than or equal to 34°C. Patient heat generation (“heat units”) was calculated as inverse of average water temperature × hours between initiation and Ttarget × 100. Primary outcome was neurologic status measured by Cerebral Performance Category score; secondary outcome was survival, both at hospital discharge. Univariate analyses were performed using Wilcoxon rank-sum tests; multivariate analyses used logistic regression. Of 203 patients included, those with Cerebral Performance Category score 3–5 generated less heat before reaching Ttarget (median, 8.1 heat units [interquartile range, 3.6–21.6 heat units] vs median, 20.0 heat units [interquartile range, 9.0–33.5 heat units]; p = 0.001) and reached Ttarget quicker (median, 2.3 hr [interquartile range, 1.5–4.0 hr] vs median, 3.6 hr [interquartile range, 2.0–5.0 hr]; p = 0.01) than patients with Cerebral Performance Category score 1–2. Nonsurvivors generated less heat than survivors (median, 8.1 heat units [interquartile range, 3.6–20.8 heat units] vs median, 19.0 heat units [interquartile range, 6.5–33.5 heat units]; p = 0.001) and reached Ttarget quicker (median, 2.2 hr [interquartile range, 1.5–3.8 hr] vs median, 3.6 hr [interquartile range, 2.0–5.0 hr]; p = 0.01). Controlling for average water temperature between initiation and Ttarget, the relationship between outcomes and time to Ttarget was no longer significant. Controlling for location, witnessed arrest, age, initial rhythm, and neuromuscular blockade use, increased heat generation was associated with better neurologic (adjusted odds ratio, 1.01 [95% CI, 1.00–1.03]; p = 0.039) and survival (adjusted odds ratio, 1.01 [95% CI, 1.00–1.03]; p = 0.045) outcomes. Conclusions: Increased heat generation during targeted temperature management initiation is associated with better outcomes at hospital discharge and may affect the relationship between time to Ttarget and outcomes.

The gastro-renal effects of intra-abdominal hypertension: Implications for critical care nurses

The gastro-renal effects of intra-abdominal hypertension: Implications for critical care nurses

by Sherree Gray, Martin Christensen, Judy Craft  

Intensive and Critical Care Nursing

Intra-abdominal hypertension is classified as either primary or secondary – primary occurs due to intra-abdominal or retro-peritoneal pathophysiology, whereas secondary results in alterations in capillary fluid dynamics due to factors, such as massive fluid resuscitation and generalised inflammation. The renal and gastro-intestinal effects occur early in the progression of intra-abdominal hypertension, and may lead to poor patient outcomes if not identified. As a direct response to intra-abdominal hypertension, renal function is reduced with remarkable impairment from pressures of around 10 mmHg, oliguria developing at 15 mmHg and anuria developing at 30 mmHg.

Combined Biomarkers Predict Acute Mortality Among Critically Ill Patients With Suspected Sepsis*

Combined Biomarkers Predict Acute Mortality Among Critically Ill Patients With Suspected Sepsis*

by Kelly, Brendan J.; Lautenbach, Ebbing; Nachamkin, Irving; Coffin, Susan E.; Gerber, Jeffrey S.; Fuchs, Barry D.; Garrigan, Charles; Han, Xiaoyan; Bilker, Warren B.; Wise, Jacqueleen; Tolomeo, Pam; Han, Jennifer H.; for the Centers for Disease Control and Prevention (CDC) Prevention Epicenters Program  

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1106–1113

Objectives: Sepsis is associated with high early and total in-hospital mortality. Despite recent revisions in the diagnostic criteria for sepsis that sought to improve predictive validity for mortality, it remains difficult to identify patients at greatest risk of death. We compared the utility of nine biomarkers to predict mortality in subjects with clinically suspected bacterial sepsis. Design: Cohort study. Setting: The medical and surgical ICUs at an academic medical center. Subjects: We enrolled 139 subjects who met two or more systemic inflammatory response syndrome (systemic inflammatory response syndrome) criteria and received new broad-spectrum antibacterial therapy. Interventions: We assayed nine biomarkers (α-2 macroglobulin, C-reactive protein, ferritin, fibrinogen, haptoglobin, procalcitonin, serum amyloid A, serum amyloid P, and tissue plasminogen activator) at onset of suspected sepsis and 24, 48, and 72 hours thereafter. We compared biomarkers between groups based on both 14-day and total in-hospital mortality and evaluated the predictive validity of single and paired biomarkers via area under the receiver operating characteristic curve. Measurements and Main Results: Fourteen-day mortality was 12.9%, and total in-hospital mortality was 29.5%. Serum amyloid P was significantly lower (4/4 timepoints) and tissue plasminogen activator significantly higher (3/4 timepoints) in the 14-day mortality group, and the same pattern held for total in-hospital mortality (Wilcoxon p ≤ 0.046 for all timepoints). Serum amyloid P and tissue plasminogen activator demonstrated the best individual predictive performance for mortality, and combinations of biomarkers including serum amyloid P and tissue plasminogen activator achieved greater predictive performance (area under the receiver operating characteristic curve > 0.76 for 14-d and 0.74 for total mortality). Conclusions: Combined biomarkers predict risk for 14-day and total mortality among subjects with suspected sepsis. Serum amyloid P and tissue plasminogen activator demonstrated the best discriminatory ability in this cohort.

Ventilator Bundle and Its Effects on Mortality Among ICU Patients: A Meta-Analysis*

Ventilator Bundle and Its Effects on Mortality Among ICU Patients: A Meta-Analysis*

by Pileggi, Claudia; Mascaro, Valentina; Bianco, Aida; Nobile, Carmelo G. A.; Pavia, Maria  

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1167–1174

Objectives: To assess the effectiveness of the ventilator bundle in the reduction of mortality in ICU patients. Data Sources: PubMed, Scopus, Web of Science, Cochrane Library for studies published until June 2017. Study Selection: Included studies: randomized controlled trials or any kind of nonrandomized intervention studies, made reference to a ventilator bundle approach, assessed mortality in ICU-ventilated adult patients. Data Extraction: Items extracted: study characteristics, description of the bundle approach, number of patients in the comparison groups, hospital/ICU mortality, ventilator-associated pneumonia–related mortality, assessment of compliance to ventilator bundle and its score. Data Synthesis: Thirteen articles were included. The implementation of a ventilator bundle significantly reduced mortality (odds ratio, 0.90; 95% CI, 0.84–0.97), with a stronger effect with a restriction to studies that reported mortality in ventilator-associated pneumonia patients (odds ratio, 0.71; 95% CI, 0.52–0.97), to studies that provided active educational activities was analyzed (odds ratio, 0.88; 95% CI, 0.78–0.99), and when the role of care procedures within the bundle (odds ratio, 0.87; 95% CI, 0.77–0.99). No survival benefit was associated with compliance to ventilator bundles. However, these results may have been confounded by the differential implementation of evidence-based procedures at baseline, which showed improved survival in the study subgroup that did not report implementation of these procedures at baseline (odds ratio, 0.82; 95% CI, 0.70–0.96). Conclusions: Simple interventions in common clinical practice applied in a coordinated way as a part of a bundle care are effective in reducing mortality in ventilated ICU patients. More prospective controlled studies are needed to define the effect of ventilator bundles on survival outcomes.

Early Enteral Nutrition Provided Within 24 Hours of ICU Admission: A Meta-Analysis of Randomized Controlled Trials*

Early Enteral NutritionProvided Within 24 Hours of ICU Admission: A Meta-Analysis of RandomizedControlled Trials*

 by Tian, Feng; Heighes, Philippa T.; Allingstrup, Matilde J.; Doig, Gordon S.  

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1049–1056

Objectives: To identify, appraise, and synthesize the most current evidence to determine whether early enteral nutrition alters patient outcomes from critical illness. Data Sources: Medline and Embase were searched. The close out date was November 20, 2017. Study Selection: Early enteral nutrition was defined as a standard formula commenced within 24 hours of ICU admission. Comparators included any form of nutrition support “except” early enteral nutrition. Only randomized controlled trials conducted in adult patients requiring treatment in an ICU were eligible for inclusion. Data Extraction: The primary outcome was mortality. Secondary outcomes included pneumonia, duration of mechanical ventilation, and ICU and hospital stay. Data Synthesis: Six-hundred ninety-nine full-text articles were retrieved and screened. Sixteen randomized controlled trials enrolling 3,225 critically ill participants were included. Compared with all other types of nutrition support, commencing enteral nutrition within 24 hours of ICU admission did not result in a reduction in mortality (odds ratio, 1.01; 95% CI, 0.86–1.18; p = 0.91; I2 = 32%). However, there was a differential treatment effect between a priori identified subgroups (p = 0.032): early enteral nutrition reduced mortality compared with delayed enteral intake (odds ratio, 0.45; 95% CI, 0.21–0.95; p = 0.038; I2 = 0%), whereas a mortality difference was not detected between early enteral nutrition and parenteral nutrition (odds ratio, 1.04; 95% CI, 0.89–1.22; p = 0.58; I2 = 30%). Overall, patients who were randomized to receive early enteral nutrition were less likely to develop pneumonia (odds ratio, 0.75; 95% CI, 0.60–0.94; p = 0.012; I2 = 48%). Conclusions: Overall, there was no difference between early enteral nutrition and all other forms of nutrition support. A priori planned subgroup analysis revealed early enteral nutrition reduced mortality and pneumonia compared with delayed enteral intake; however, there were no clear clinical advantages of early enteral nutrition over parenteral nutrition.

Effect of Administration of Ramelteon, a Melatonin Receptor Agonist, on the Duration of Stay in the ICU: A Single-Center Randomized Placebo-Controlled Trial*

Effect of Administration of Ramelteon, a Melatonin Receptor Agonist, on the Duration of Stay in the ICU: A Single-Center Randomized Placebo-Controlled Trial*

 by Nishikimi, Mitsuaki; Numaguchi, Atsushi; Takahashi, Kunihiko; Miyagawa, Yasuhiro; Matsui, Kota; Higashi, Michiko; Makishi, Go; Matsui, Shigeyuki; Matsuda, Naoyuki  / 

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1099–1105

Objectives: Occurrence of delirium in the ICU is associated with a longer stay in the ICU. To examine whether the use of ramelteon, a melatonin agonist, can prevent delirium and shorten the duration of ICU stay of critically ill patients. Design: A single-center, triple-blinded, randomized placebo-controlled trial. Setting: ICU of an academic hospital. Patients: Eligible patients were ICU patients who could take medicines orally or through a nasogastric tube during the first 48 hours of admission. Interventions: The intervention group received ramelteon (8 mg/d), and the control group received placebo (1 g/d of lactose powder) at 20:00 hours every day until discharge from the ICU. Measurements and Main Results: A total of 88 subjects were randomized to the ramelteon group (45 subjects) or the placebo group (43 subjects). As the primary endpoint, there was a trend toward decrease in the duration of ICU stay (4.56 d) in the ramelteon group compared with the placebo group (5.86 d) (p = 0.082 and p = 0.028 before and after adjustments). As the secondary endpoints, statistically significant decreases in the occurrence rate (24.4% vs 46.5%; p = 0.044) and duration (0.78 vs 1.40 d; p = 0.048) of delirium were observed in the ramelteon group. The nonintubated patients of the ramelteon group showed statistically significantly fewer awakenings per night and a higher proportion of nights without awakenings. Conclusions: Ramelteon tended to decrease the duration of ICU stay as well as decreased the occurrence rate and duration of delirium statistically significantly.

What Faces Reveal: A Novel Method to Identify Patients at Risk of Deterioration Using Facial Expressions*

What Faces Reveal: A NovelMethod to Identify Patients at Risk of Deterioration Using Facial Expressions*

by Madrigal-Garcia, Maria Isabel; Rodrigues, Marcos; Shenfield, Alex; Singer, Mervyn; Moreno-Cuesta, Jeronimo  

Critical Care Medicine: July 2018 - Volume 46 - Issue 7 - p 1057–1062

Objectives: To identify facial expressions occurring in patients at risk of deterioration in hospital wards. Design: Prospective observational feasibility study. Setting: General ward patients in a London Community Hospital, United Kingdom. Patients: Thirty-four patients at risk of clinical deterioration. Interventions: A 5-minute video (25 frames/s; 7,500 images) was recorded, encrypted, and subsequently analyzed for action units by a trained facial action coding system psychologist blinded to outcome. Measurements and Main Results: Action units of the upper face, head position, eyes position, lips and jaw position, and lower face were analyzed in conjunction with clinical measures collected within the National Early Warning Score. The most frequently detected action units were action unit 43 (73%) for upper face, action unit 51 (11.7%) for head position, action unit 62 (5.8%) for eyes position, action unit 25 (44.1%) for lips and jaw, and action unit 15 (67.6%) for lower face. The presence of certain combined face displays was increased in patients requiring admission to intensive care, namely, action units 43 + 15 + 25 (face display 1, p < 0.013), action units 43 + 15 + 51/52 (face display 2, p < 0.003), and action units 43 + 15 + 51 + 25 (face display 3, p < 0.002). Having face display 1, face display 2, and face display 3 increased the risk of being admitted to intensive care eight-fold, 18-fold, and as a sure event, respectively. A logistic regression model with face display 1, face display 2, face display 3, and National Early Warning Score as independent covariates described admission to intensive care with an average concordance statistic (C-index) of 0.71 (p = 0.009). Conclusions: Patterned facial expressions can be identified in deteriorating general ward patients. This tool may potentially augment risk prediction of current scoring systems.

Revisited: A Systematic Review of Therapeutic Hypothermia for Adult Patients Following Traumatic Brain Injury

Revisited: A Systematic Review of TherapeuticHypothermia for Adult Patients Following Traumatic Brain Injury

by Watson, Hannah I.; Shepherd, Andrew A.; Rhodes, Jonathan K. J.; Andrews, Peter J. D. 

Objectives: Therapeutic hypothermia has been of topical interest for many years and with the publication of two international, multicenter randomized controlled trials, the evidence base now needs updating. The aim of this systematic review of randomized controlled trials is to assess the efficacy of therapeutic hypothermia in adult traumatic brain injury focusing on mortality, poor outcomes, and new pneumonia. Data Sources: The following databases were searched from January 1, 2011, to January 26, 2018: Cochrane Central Register of Controlled Trial, MEDLINE, PubMed, and EMBASE. Study Selection: Only foreign articles published in the English language were included. Only articles that were randomized controlled trials investigating adult traumatic brain injury sustained following an acute, closed head injury were included. Two authors independently assessed at each stage. Data Extraction: Quality was assessed using the Cochrane Collaboration’s tool for assessing the risk of bias. All extracted data were combined using the Mantel-Haenszel estimator for pooled risk ratio with 95% CIs. p value of less than 0.05 was considered statistically significant. All statistical analyses were conducted using RevMan 5 (Cochrane Collaboration, Version 5.3, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Data Synthesis: Twenty-two studies with 2,346 patients are included. Randomized controlled trials with a low risk of bias show significantly more mortality in the therapeutic hypothermia group (risk ratio, 1.37; 95% CI, 1.04–1.79; p = 0.02), whereas randomized controlled trials with a high risk of bias show the opposite with a higher mortality in the control group (risk ratio, 0.70; 95% CI, 0.60–0.82; p < 0.00001). Conclusions: Overall, this review is in-keeping with the conclusions published by the most recent randomized controlled trials. High-quality studies show no significant difference in mortality, poor outcomes, or new pneumonia. In addition, this review shows a place for fever control in the management of traumatic brain injury.