Informatics infrastructure for syndrome surveillance, decision support, reporting, and modeling of critical illness

Individualized prediction of critical illness in older adults: Validation of an elders risk assessment model

BACKGROUND

The electronic health record (EHR) presents new opportunities for the timely identification of patients at high risk of critical illness and the implementation of preventive strategies. This study aims to externally validate an EHR-based Elders Risk Assessment (ERA) score to identify older patients at high risk of future critical illness during a primary care visit.

METHODS

This historical cohort study included patients aged ≥65 years who had primary care visits at Mayo Clinic Rochester, MN, between July 2019 and December 2021. The ERA score at the time of the primary care visit was used to predict critical illness, defined as death or ICU admission within 1 year of the visit.

RESULTS

A total of 12,885 patients were included in the analysis. The median age at the time of the primary care visit was 75 years, with 44.6% being male. 93.7% of participants were White, and 64.2% were married. The median (25th, 75th percentile) ERA score was 4 (0, 9). 11.3% of study participants were admitted to the ICU or died within 1 year of the visit. The ERA score predicted critical illness within 1 year of a primary care visit with an area under the receiver operating characteristic curve of 0.84 (95% CI 0.83-0.85), which indicates good discrimination. An ERA score of 9 was identified as optimal for implementing and testing potential preventive strategies, with the odds ratio of having the primary outcome in patients with ERA score ≥9 being 11.33 (95%CI 9.98-12.87).

CONCLUSIONS

This simple EHR-based risk assessment model can predict critical illness within 1 year of primary care visits in older patients. The findings of this study can serve as a basis for testing and implementation of preventive strategies to promote the well-being of older adults at risk of critical illness and its consequences.

Building Learning Healthcare Systems for Critical Care Medicine

Learning healthcare systems are an evolving way of integrating informatics, analytics, and continuous improvement into daily practice in healthcare. This article discusses strategies to build learning healthcare systems for critical care medicine.

Temporal trends in critical care utilization and outcomes in allogeneic hematopoietic stem cell transplant recipients

Historically, the prognosis of allogeneic hematopoietic stem cell transplant (allo-HCT) recipients who require intensive care unit (ICU) admission has been poor. We aimed to describe the epidemiological trends of ICU utilization and outcomes in allo-HCT patients. We conducted a retrospective cohort study including adults (≥ 18) undergoing allo-HCT between 01/01/2005 and 31/12/2020 at Mayo Clinic, Rochester. Temporal trends in outcomes were assessed by robust linear regression modelling. Risk factors for hospital mortality were chosen a priori and assessed with multivariable logistic regression modelling. Of 1,249 subjects, there were 486 ICU admissions among 287 individuals. Although older patients underwent allo-HCT (1.64 [95% CI: 1.11 to 2.45] years per year; P = 0.025), there was no change in ICU utilization over time (P = 0.91). The ICU and hospital mortality rates were 19.2% (55/287) and 28.2% (81/287), respectively. There was a decline in ICU mortality (-0.38% [95% CI: -0.70 to -0.06%] per year; P = 0.035). The 1-year post-HCT mortality for those requiring ICU admission was 56.1% (161/287), with no significant difference over time, versus 15.8% (141/891, 71 missing) among those who did not. The frequency and duration of invasive mechanical ventilation (IMV) declined. In multivariable analyses, higher serum lactate, higher sequential organ failure assessment (SOFA) scores, acute respiratory distress (ARDS), and need for IMV were associated with greater odds of hospital mortality. Over time, rates of ICU utilization have remained stable, despite increasing patient age. Several trends suggest improvement in outcomes, notably lower ICU mortality and frequency of IMV. However, long-term survival remains unchanged. Further work is needed to improve long-term outcomes in this population.

DERIVATION OF A MORTALITY PREDICTION MODEL IN CRITICAL CARE PATIENTS WITH CIRRHOSIS AND SEPSIS

ABSTRACT

Objective : The aim of the study is to develop a predictive model for in-hospital mortality in critically ill patients with cirrhosis and sepsis, using clinical and laboratory data. Design : This is a retrospective cohort study. Setting: Medical and mixed intensive care units (ICUs) of a tertiary medical center. Patients : Cirrhotic adults were admitted with sepsis to the ICUs from January of 2007 to May of 2017. Interventions : None. Measurements and Main Results : Of 2,595 ICU admissions of patients with cirrhosis, 277 with first ICU admission for sepsis were included in the analysis, and 37% died in the hospital. Patients who stayed in the ICU for at least 6 h (n = 275) were considered for the multivariate model. Ten-fold cross-validation was used to estimate best parameter values and model performance, and the final model was chosen as the model maximizing area under the receiver-operating characteristic curve. Variables in order of impact were Acute Physiology and Chronic Health Evaluation (APACHE) III score, initial serum lactate, conjugated bilirubin, serum creatinine, model for end-stage liver disease score, age, body mass index, and serum hemoglobin. The final best model from cross-validation presented an area under the receiver operator characteristic curve (AUC) of 0.75, using a cut-point of 50% estimated probability, sensitivity and specificity were 0.46 and 0.90, respectively, with positive predictive value of 0.72 and negative predictive value of 0.74. These results were similar to the APACHE III only model (AUC = 0.74, sensitivity = 0.43, specificity = 0.89, positive predictive value = 0.69, negative predictive value = 0.73). Conclusion : The combination of initial serum lactate level, conjugated bilirubin, initial serum creatinine, model for end-stage liver disease score, age, body mass index, and serum hemoglobin did not yield meaningful improvement in the AUC and did not provide advantage over the APACHE III score for the prediction of in-hospital mortality in critically ill patients with cirrhosis and sepsis.

Serum lactate and mean arterial pressure thresholds in patients with cirrhosis and septic shock

BACKGROUND

The Sepsis-3 guidelines have incorporated serum lactate levels of >2 mmol/L in septic shock definition to account for higher observed mortality. Further evidence is needed to support this threshold in cirrhosis, as well as target mean arterial pressure (MAP) during resuscitation.

METHODS

This observational cohort study investigated the association between initial serum lactate and resuscitation MAP levels on in-hospital mortality in patients with and without cirrhosis. Patients admitted to the intensive care unit for the treatment of septic shock between 2006 and 2021 in a quaternary academic center were included. Patients with cirrhosis documented on imaging and International Classification of Disease codes (n=595) were compared to patients without cirrhosis (n=575). The association of intensive care unit admission lactate levels and median 2-hour MAP with in-hospital mortality and the need for continuous renal replacement therapy was assessed. The association between median 24-hour MAP and in-hospital mortality was analyzed post hoc.

RESULTS

Within the cirrhosis group, admission lactate levels of 2-4 and >4 mmol/L were associated with increased in-hospital mortality compared to lactate <2 mmol/L [adjusted odds ratio (aOR): 1.69, CI: 1.03-2.81, aOR: 4.02, CI: 2.53-6.52]. Median 24-hour MAP 60-65 and <60 mm Hg were also associated with increased in-hospital mortality compared with MAP >65 mm Hg (aOR: 2.84, CI: 1.64-4.92 and aOR: 7.34, CI: 3.17-18.76). In the noncirrhosis group, associations with in-hospital mortality were weaker for lactate 2-4 and >4 mmol/L (aOR: 1.32, CI: 0.77-2.27 and aOR: 2.25, CI: 1.40-3.67) and median 24-hour MAP 60-65 and <60 mm Hg (aOR: 1.70, CI: 0.65-4.14 and aOR: 4.41, CI: 0.79-29.38).

CONCLUSIONS

These findings support utilizing lactate >2 mmol/L in the definition of septic shock, as well as a target MAP of >65 mm Hg during resuscitation in patients with cirrhosis.

Association of Blood Pressure Variability with Delirium in Patients with Critical Illness

BACKGROUND

The objective was to examine the association of blood pressure variability (BPV) during the first 24 h after intensive care unit admission with the likelihood of delirium and depressed alertness without delirium ("depressed alertness").

METHODS

This retrospective, observational, cohort study included all consecutive adult patients admitted to an intensive care unit at Mayo Clinic, Rochester, Minnesota, from July 1, 2004, through October 31, 2015. The primary outcomes were delirium and delirium-free days, and the secondary outcomes included depressed alertness and depressed alertness-free days. Logistic regression was performed to determine the association of BPV with delirium and depressed alertness. Proportional odds regression was used to assess the association of BPV with delirium-free days and depressed alertness-free days.

RESULTS

Among 66,549 intensive care unit admissions, delirium was documented in 20.2% and depressed alertness was documented in 24.4%. Preserved cognition was documented in 55.4% of intensive care unit admissions. Increased systolic and diastolic BPV was associated with an increased odds of delirium and depressed alertness. The magnitude of the association per 5-mm Hg increase in systolic average real variability (the average of absolute value of changes between consecutive systolic blood pressure readings) was greater for delirium (odds ratio 1.34; 95% confidence interval 1.29-1.40; P < 0.001) than for depressed alertness (odds ratio 1.06; 95% confidence interval 1.02-1.10; P = 0.004). Increased systolic and diastolic BPV was associated with fewer delirium-free days but not with depressed alertness-free days.

CONCLUSIONS

BPV in the first 24 h after intensive care unit admission is associated with an increased likelihood of delirium and fewer delirium-free days.

Impact of Various Estimated Glomerular Filtration Rate Equations on the Pharmacokinetics of Meropenem in Critically Ill Adults

IMPORTANCE

Meropenem dosing is typically guided by creatinine-based estimated glomerular filtration rate (eGFR), but creatinine is a suboptimal GFR marker in the critically ill.

OBJECTIVES

This study aimed to develop and qualify a population pharmacokinetic model for meropenem in critically ill adults and to determine which eGFR equation based on creatinine, cystatin C, or both biomarkers best improves model performance.

DESIGN SETTING AND PARTICIPANTS

This single-center study evaluated adults hospitalized in an ICU who received IV meropenem from 2018 to 2022. Patients were excluded if they had acute kidney injury, were on kidney replacement therapy, or were treated with extracorporeal membrane oxygenation. Two cohorts were used for population pharmacokinetic modeling: a richly sampled development cohort (n = 19) and an opportunistically sampled qualification cohort (n = 32).

MAIN OUTCOMES AND MEASURES

A nonlinear mixed-effects model was developed using parametric methods to estimate meropenem serum concentrations.

RESULTS

The best-fit structural model in the richly sampled development cohort was a two-compartment model with first-order elimination. The final model included time-dependent weight normalized to a 70-kg adult as a covariate for volume of distribution (Vd) and time-dependent eGFR for clearance. Among the eGFR equations evaluated, eGFR based on creatinine and cystatin C expressed in mL/min best-predicted meropenem clearance. The mean (se) Vd in the final model was 18.2 (3.5) liters and clearance was 11.5 (1.3) L/hr. Using the development cohort as the Bayesian prior, the opportunistically sampled cohort demonstrated good accuracy and low bias.

CONCLUSIONS AND RELEVANCE

Contemporary eGFR equations that use both creatinine and cystatin C improved meropenem population pharmacokinetic model performance compared with creatinine-only or cystatin C-only eGFR equations in adult critically ill patients.

Population pharmacokinetic model of cefepime for critically ill adults: a comparative assessment of eGFR equations

Cefepime exhibits highly variable pharmacokinetics in critically ill patients. The purpose of this study was to develop and qualify a population pharmacokinetic model for use in the critically ill and investigate the impact of various estimated glomerular filtration rate (eGFR) equations using creatinine, cystatin C, or both on model parameters. This was a prospective study of critically ill adults hospitalized at an academic medical center treated with intravenous cefepime. Individuals with acute kidney injury or on kidney replacement therapy or extracorporeal membrane oxygenation were excluded. A nonlinear mixed-effects population pharmacokinetic model was developed using data collected from 2018 to 2022. The 120 included individuals contributed 379 serum samples for analysis. A two-compartment pharmacokinetic model with first-order elimination best described the data. The population mean parameters (standard error) in the final model were 7.84 (0.24) L/h for CL1 and 15.6 (1.45) L for V1. Q was fixed at 7.09 L/h and V2 was fixed at 10.6 L, due to low observed interindividual variation in these parameters. The final model included weight as a covariate for volume of distribution and the eGFRcr-cysC (mL/min) as a predictor of drug clearance. In summary, a population pharmacokinetic model for cefepime was created for critically ill adults. The study demonstrated the importance of cystatin C to prediction of cefepime clearance. Cefepime dosing models which use an eGFR equation inclusive of cystatin C are likely to exhibit improved accuracy and precision compared to dosing models which incorporate an eGFR equation with only creatinine.

Adequacy of cefepime concentrations in the early phase of critical illness: A case for precision pharmacotherapy

STUDY OBJECTIVE

In critically ill patients, adequacy of early antibiotic exposure has been incompletely evaluated. This study characterized factors associated with inadequate cefepime exposure in the first 24 h of critical illness.

DESIGN

Prospective cohort study.

SETTING

Academic Medical Center.

PATIENTS

Critically ill adults treated with cefepime. Patients with acute kidney injury or treated with kidney replacement therapy or extracorporeal membrane oxygenation were excluded.

INTERVENTION

None.

MEASUREMENTS

A nonlinear mixed-effects pharmacokinetic (PK) model was developed to estimate cefepime concentrations for each patient over time. The percentage of time the free drug concentration exceeded 8 mg/L during the first 24 h of therapy was calculated (%ƒT>8; appropriate for the susceptible breakpoint for Pseudomonas aeruginosa). Factors predictive of low %ƒT>8 were explored with multivariable regression.

MAIN RESULTS

In the 100 included patients, a one-compartment PK model was developed with first-order elimination with covariates for weight and estimated glomerular filtration rate based on creatinine and cystatin C (eGFRSCr-CysC). The median (interquartile range) %ƒT>8 for cefepime in the first 24 h of therapy based on this model was 85% (66%, 100%). Less than 100% ƒT>8 during first 24 h of therapy occurred in 70 (70%) individuals. Lower Sequential Organ Failure Assessment score (p = 0.032) and higher eGFRSCr-CysC (p < 0.001) predicted a lower %ƒT>8. Central nervous system infection source was protective (i.e., associated with a higher %ƒT>8; p = 0.008).

CONCLUSIONS

During early critical illness, cefepime concentrations were inadequate in a significant proportion of patients. Antimicrobial optimization is needed to improve the precision of pharmacotherapy in the critically ill patients.

Extracorporeal membrane oxygenation and acute kidney injury: a single-center retrospective cohort

To assess the relationship between acute kidney injury (AKI) with outcomes among patients requiring extracorporeal membrane oxygenation (ECMO). This is a single-center, retrospective cohort study of adult patients admitted to intensive care units (ICU) at a tertiary referral hospital requiring ECMO from July 1, 2015, to August 30, 2019. We assessed the temporal relationship of AKI and renal replacement therapy with ECMO type (VV vs. VA). The primary outcome was in-hospital mortality rates. We used Kruskal-Wallis or chi-square tests for pairwise comparisons, cause-specific Cox proportional hazards models were utilized for the association between AKI prevalence and in-hospital mortality, and a time-dependent Cox model was used to describe the association between AKI incidence and mortality. After the screening, 190 patients met eligibility criteria [133 (70%) AKI, 81 (43%) required RRT]. The median age was 61 years, and 61% were males. Among AKI patients, 48 (36%) and 85 (64%) patients developed AKI before and after ECMO, respectively. The SOFA Day 1, baseline creatinine, respiratory rate (RR), use of vasopressin, vancomycin, proton pump inhibitor, antibiotics, duration of mechanical ventilation and ECMO, and ICU length of stay were higher in AKI patients compared with those without AKI (P < 0.01). While ICU and in-hospital mortality rates were 46% and 50%, respectively, there were no differences based on the AKI status. The type and characteristics of ECMO support were not associated with AKI risk. Among AKI patients, 77 (58%) were oliguric, and 46 (60%) of them received diuretics. Urine output in the diuretic group was only higher on the first day than in those who did not receive diuretics (P = 0.03). Among ECMO patients, AKI was not associated with increased mortality but was associated with prolonged duration of mechanical ventilation and ICU length of stay.

Clinical Informatics in Critical Care Medicine

Continuous monitoring and treatment of patients in intensive care units generates vast amounts of data. Critical Care Medicine clinicians incorporate this continuously evolving data to make split-second, life or death decisions for management of these patients. Despite the abundance of data, it can be challenging to consider every accessible data point when making the quick decisions necessary at the point of care. Consequently, Clinical Informatics offers a natural partnership to improve the care for critically ill patients. The last two decades have seen a significant evolution in the role of Clinical Informatics in Critical Care Medicine. In this review, we will discuss how Clinical Informatics improves the care of critically ill patients by enhancing not only data collection and visualization but also bedside medical decision making. We will further discuss the evolving role of machine learning algorithms in Clinical Informatics as it pertains to Critical Care Medicine.

Relationship between the rate of fluid resuscitation and acute kidney injury: A retrospective cohort study

Background

Septic shock is the leading cause of acute kidney injury (AKI) in critically ill patients. The foundation of early septic shock management includes early fluid resuscitation, but the association between fluid resuscitation rates and kidney outcomes remains unclear. This investigation examines the association between fluid resuscitation rate and AKI recovery.

Methods

In the medical intensive care unit of Mayo Clinic Rochester, adult patients with AKI and septic shock were retrospectively studied from January 1, 2006 to May 31, 2018. The surviving sepsis campaign recommends an initial fluid bolus of 30 ml/kg for sepsis resuscitation. The cohort of patients was divided into three groups based on the average fluid resuscitation time (<1 h, 1.1-3 h, >3 h) and the corresponding fluid rate ≥0.5, 0.17-0.49, and <0.17 ml/kg/min, respectively. The primary outcome was the recovery of AKI on day 7. To account for potential confounders, multivariable regression analyses were conducted.

Results

After meeting the eligibility, 597 patients were included in the analysis. The AKI recovery was considerably different among the groups (P = 0.006). Patients in group 1 who received fluid resuscitation faster had a higher rate of AKI recovery (53%) compared to group 2 and group 3 (50% and 37.8%).

Conclusion

In septic shock patients with AKI, a higher fluid resuscitation rate of 30 ml/kg IV fluids within the 1st-h sepsis diagnosis (i.e., >0.50 ml/kg/min) lead to higher AKI recovery compared with slower infusion rates.

Echocardiographic parameters and hemodynamic instability at the initiation of continuous kidney replacement therapy

OBJECTIVE

Investigate the association of echocardiographic parameters with hemodynamic instability after initiating continuous kidney replacement therapy (CKRT) in a cohort of intensive care unit (ICU) patients requiring CKRT.

METHODS

Historical cohort study of consecutive adults admitted to the ICU at a tertiary care hospital from December 2006 through November 2015 who underwent CKRT and had an echocardiogram done within seven days before CKRT initiation. The primary outcome was hypotension within one hour of CKRT initiation.

RESULTS

We included 980 patients, 804 (82%) with acute kidney injury (AKI) and 176 (18%) with end-stage kidney disease (ESKD). Median patient age was 63 (± 14) years, and median Sequential Organ Failure Assessment (SOFA) score on the day of CKRT initiation was 12 (IQR 10-14). Multivariable analysis showed that Left (OR 2.01, 95% CI 1.04-3.86), and Right (OR 1.5, 95% CI 1.04-2.25) moderate and severe ventricular enlargement, Vasoactive-Inotropic Score (VIS) one hour before CKRT initiation (OR 1.18 per 10 units increase, 95% CI 1.09-1.28) and high bicarbonate fluid replacement (OR 2.52, 95% CI 1.01-6.2) were associated with hypotension after CKRT initiation.

CONCLUSION

Right and left ventricular enlargement are risk factors associated with hypotension after CKRT initiation.

Validation of an Intensive Care Unit Data Mart for Research and Quality Improvement

Data derived from the electronic health record (EHR) is frequently extracted using undefined approaches that may affect the accuracy of collected variables. Further, efforts to assess data accuracy often suffer from limited collaboration between clinicians and data analysts who perform the extraction. In this manuscript, we describe the methodology behind creation of a structured, rigorously derived intensive care unit (ICU) data mart based on data automatically and routinely derived from the EHR. This ICU data mart includes high-quality data elements commonly used for quality improvement and research purposes. These data elements were identified by physicians working closely with data analysts to iteratively develop and refine algorithmic definitions for complex outcomes and risk factors. We contend that this methodology can be reproduced and applied across other institution or to other clinical domains to create high quality data marts, inclusive of complex outcomes data.

Association of blood pressure variability with short- and long-term cognitive outcomes in patients with critical illness

BACKGROUND

Blood pressure variability (BPV), a modifiable risk factor, can compromise cerebral perfusion in critically ill patients. We studied the association between BPV in the intensive care unit (ICU) and short- and long-term cognitive outcomes.

METHODS

All patients were ≥50 years old. The short-term cognitive end points were delirium and depressed alertness without delirium. The long-term outcome was change in the slope of longitudinal cognitive scores. Primary BPV measure was average real variability (ARV) of systolic blood pressure. Associations were assessed with multivariable multinominal logistic regression and linear mixed effects models.

RESULTS

Of 794 patients (1130 admissions) 185 developed delirium and 274 developed depressed alertness. There was a dose-response association of 24-h systolic ARV with delirium (adjusted OR, 95% CI 2.15 per 5 mm Hg increase, 1.31-3.06, P < 0.017) and with depressed alertness (OR 1.89, 95% CI 1.18-3.03, P < 0.008). For 371 patients with available longitudinal cognitive scores, the decline in cognitive trajectory was accelerated after discharge (annual change OR -0.097, 95% CI -0.122 to -0.073). This acceleration increased with delirium (additional decline -0.132 [-0.233 to 0.030], P = 0.011). We found no significant association between BPV and post-ICU cognitive trajectory.

CONCLUSIONS

BPV was associated with increased likelihood of delirium in the ICU. Delirium, but not BPV, was associated with long-term cognitive decline.

Echocardiographic Characteristics of Cardiogenic Shock Patients with and Without Cardiac Arrest

BACKGROUND

Cardiac arrest (CA) is associated with worse outcomes in patients with cardiogenic shock (CS). To better understand the contribution of CA on CS, we evaluated transthoracic echocardiography (TTE) parameters in CS patients with and without CA. Methods: We retrospectively identified CS patients with a TTE performed near cardiac intensive care unit admission between 2007 to 2018. We compared TTE measurements of left ventricular (LV) and right ventricular (RV) function in patients with and without CA. The primary outcome was all-cause in-hospital mortality, as determined using multivariable logistic regression. Results: We included 1085 patients, 35% of whom had CA. Median age was 70 years and 37% were females. CA patients had higher severity of illness, more invasive mechanical ventilation and greater vasopressor/inotrope use. In-hospital mortality was 31% and was higher in CA patients (45% vs. 23%, p <0.001). Although LV ejection fraction (LVEF) was similar (35% vs. 37%, p = 0.05), CA patients had lower cardiac index, mitral valve E wave peak velocity, E/A ratio and E/e' ratio. TTE variables that were associated with hospital mortality varied, among patients with CA, these included measures of RV pressure and function and among patients without CA, these included parameters reflecting LV systolic function. Conclusions: Doppler assessments of RV systolic dysfunction were the strongest TTE predictors of hospital mortality in CS patients with CA, unlike CS patients without CA in whom LV systolic function was more important. This emphasizes the importance of RV assessment for mortality risk stratification after CA.

Detecting professional interpreter use among patients with limited English proficiency: Derivation and validation study

Objectives

The objective of this derivation and validation study was to develop and validate a search strategy algorithm to detect patients who used professional interpreter services.

Methods

We identified all adults who had at least one intensive care unit admission during their hospital stay across the Mayo Clinic Enterprise between 1 January 2015 and 30 June 2020. Three random subsets of 100 patients were extracted from 60,268 patients to develop the search strategy algorithm. Two physician reviewers conducted gold standard manual chart review and any discrepancies were resolved by a third reviewer. These results were compared with the search strategy algorithm each time it was refined. Sensitivity and specificity were calculated during each phase by comparing the search strategy results to the reference gold standard for both derivation cohorts and the final validation cohort.

Results

The first search strategy resulted in a sensitivity of 100% and a specificity of 89%. The second revised search strategy achieved a sensitivity of 100% and a specificity of 87%. The final version of the search strategy was applied to the validation subset and sensitivity and specificity were 100% and 89%, respectively.

Conclusion

We derived and validated a search strategy algorithm to assess interpreter use among hospitalized patients. Using a search strategy algorithm with high sensitivity and specificity can reduce the time required to abstract data from the electronic medical records compared with manual data abstraction.

Sequential organ failure assessment score improves survival prediction for left ventricular assist device recipients in intensive care

BACKGROUND

Preoperative risk scores facilitate patient selection, but postoperative risk scores may offer valuable information for predicting outcomes. We hypothesized that the postoperative Sequential Organ Failure Assessment (SOFA) score would predict mortality after left ventricular assist device (LVAD) implantation.

METHODS

We retrospectively reviewed data from 294 continuous-flow LVAD implantations performed at Mayo Clinic Rochester during 2007 to 2015. We calculated the EuroSCORE, HeartMate-II Risk Score, and RV Failure Risk Score from preoperative data and the APACHE III and Post Cardiac Surgery (POCAS) risk scores from postoperative data. Daily, maximum, and mean SOFA scores were calculated for the first 5 postoperative days. The area under receiver-operator characteristic curves (AUC) was calculated to compare the scoring systems' ability to predict 30-day, 90-day, and 1-year mortality.

RESULTS

For the entire cohort, mortality was 5% at 30 days, 10% at 90 days, and 19% at 1 year. The Day 1 SOFA score had better discrimination for 30-day mortality (AUC 0.77) than the preoperative risk scores or the APACHE III and POCAS postoperative scores. The maximum SOFA score had the best discrimination for 30-day mortality (AUC 0.86), and the mean SOFA score had the best discrimination for 90-day mortality (AUC 0.82) and 1-year mortality (AUC 0.76).

CONCLUSIONS

We observed that postoperative mean and maximum SOFA scores in LVAD recipients predict short-term and intermediate-term mortality better than preoperative risk scores do. However, because preoperative and postoperative risk scores each contribute unique information, they are best used in concert to predict outcomes after LVAD implantation.

Investigating the cognitive capacity constraints of an ICU care team using a systems engineering approach

Background

ICU operational conditions may contribute to cognitive overload and negatively impact on clinical decision making. We aimed to develop a quantitative model to investigate the association between the operational conditions and the quantity of medication orders as a measurable indicator of the multidisciplinary care team’s cognitive capacity.

Methods

The temporal data of patients at one medical ICU (MICU) of Mayo Clinic in Rochester, MN between February 2016 to March 2018 was used. This dataset includes a total of 4822 unique patients admitted to the MICU and a total of 6240 MICU admissions. Guided by the Systems Engineering Initiative for Patient Safety model, quantifiable measures attainable from electronic medical records were identified and a conceptual framework of distributed cognition in ICU was developed. Univariate piecewise Poisson regression models were built to investigate the relationship between system-level workload indicators, including patient census and patient characteristics (severity of illness, new admission, and mortality risk) and the quantity of medication orders, as the output of the care team’s decision making.

Results

Comparing the coefficients of different line segments obtained from the regression models using a generalized F-test, we identified that, when the ICU was more than 50% occupied (patient census > 18), the number of medication orders per patient per hour was significantly reduced (average = 0.74; standard deviation (SD) = 0.56 vs. average = 0.65; SD = 0.48; p < 0.001). The reduction was more pronounced (average = 0.81; SD = 0.59 vs. average = 0.63; SD = 0.47; p < 0.001), and the breakpoint shifted to a lower patient census (16 patients) when at a higher presence of severely-ill patients requiring invasive mechanical ventilation during their stay, which might be encountered in an ICU treating patients with COVID-19.

Conclusions

Our model suggests that ICU operational factors, such as admission rates and patient severity of illness may impact the critical care team’s cognitive function and result in changes in the production of medication orders. The results of this analysis heighten the importance of increasing situational awareness of the care team to detect and react to changing circumstances in the ICU that may contribute to cognitive overload.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12871-021-01548-7.

Intensive Care Unit Sedation Practices at a Large, Tertiary Academic Center

BACKGROUND

Sedatives are frequently administered in an ICU and are often dependent on patient population and ICU type. These differences may affect patient-centered outcomes.

OBJECTIVE

Our primary objective was to identify differences in sedation practice among three different ICU types at an academic medical center.

METHODS

This was a retrospective cross-sectional study of adult patients (≥18 years) requiring a continuous sedative for ≥6 h and admitted to a medical ICU, surgical ICU, and medical/surgical ICU at a single academic medical center in Rochester Minnesota from June 1, 2018 to May 31, 2020. We extracted baseline characteristics; sedative type, dose, and duration; concomitant therapies; and patient outcomes. Summary statistics are presented.

RESULTS

A total of 2154 patients met our study criteria (1010 from medical ICU, 539 from surgical ICU, 605 from medical/surgical ICU). Propofol was the most frequently used sedative in all ICU settings (74.1% in medical ICU, 53.8% in surgical ICU, 68.9% in medical/surgical ICU, and 67.5% in all ICUs). The mortality rate was highest in the medical/surgical ICU (40.2% in medical ICU, 26.0% in surgical ICU, 40.7% in medical/surgical ICU, and 36.8% in all ICUs). 90.7% of all patients required mechanical ventilation (92.9% in medical ICU, 88.5% in surgical ICU, and 89.1% in medical/surgical ICU). Overall, patients spent more time in light sedation than deep sedation, 75% versus 10.3%, during their ICU admission. Patients in the medical ICU spent a greater proportion of time positive for delirium than the other ICU settings (35.7% in medical ICU, 9.8% in surgical ICU, and 20% in medical/surgical ICU). Similar amounts of opioids (morphine milligram equivalents) were used during the continuous sedative infusion between the three settings.

CONCLUSIONS

We observed that patients in the medical ICU spent more time deeply sedated with multiple agents which was associated with a higher proportion of delirium.

Inclusion of Albumin in the Initial Resuscitation of Adult Patients with Medical Sepsis or Septic Shock: a Propensity Score-Matched Analysis

INTRODUCTION

The impact of albumin resuscitation on sepsis outcomes is debated, particularly in the initial phase of resuscitation. We aimed to investigate the association between albumin use in the initial 6 h of resuscitation and subsequent outcomes in adult septic patients.

METHODS

This single-center, retrospective, propensity score-matched cohort study included adult patients admitted to intensive care units (ICUs) with sepsis or septic shock from January 1, 2006, to May 4, 2018, at a tertiary referral hospital. We compared two groups based on albumin receipt within the initial six resuscitation hours (albumin group vs. non-albumin group). We performed a 1:2 propensity score matching to assess shock-free time in ICU as the primary outcome.

RESULTS

Of 2,732 patients with medical sepsis, 286 cases in the albumin group were matched with 549 individuals in the non-albumin group. Compared to the non-albumin group, the albumin group required more intravenous fluids and had higher net fluid balance, lower mean arterial pressure, and lower serum base excess level in the initial 6 and 24 h of resuscitation. Shock-free time, ICU and hospital length of stay, and 28-day mortality were not different between albumin and non-albumin groups (56 vs. 66 h, P = 0.18; 3.5 days vs. 3.7 days, P = 0.61; 9.1 days vs. 9.5 days, P = 0.27; 36% vs. 32%, P = 0.25, respectively).

CONCLUSIONS

Using albumin during the initial 6 h of resuscitation was not associated with benefits in clinical outcomes of patients with medical sepsis.

Clinical Effect of Systemic Steroids in Patients with Cirrhosis and Septic Shock

PURPOSE

Evidence regarding the utility of systemic steroids in treating patients with cirrhosis and septic shock remains equivocal. This study aimed to evaluate and elucidate the association of steroid use with outcomes and adverse effects in a cohort of patients with cirrhosis and septic shock.

PATIENTS AND METHODS

Retrospective cohort study of patients with cirrhosis and septic shock admitted to a tertiary hospital intensive care unit (ICU) from January 2007 to May 2017, using a validated ICU Datamart. Patients who received vasopressors within 6 h of ICU admission were included in the multivariate analysis. The effect of steroids on outcomes was evaluated using multivariable regression, adjusting for confounding variables.

RESULTS

Out of 179 admissions of patients with cirrhosis and septic shock, 56 received steroids during the ICU admission. Patients who received steroids received a higher total dose of vasopressors (91.2 mg vs. 39.1 mg, P = 0.04) and had a lower initial lactate level (1.8 mmol/L vs. 2.6 mmol/L, P = 0.007). The multivariate analysis included 117 patients and showed no significant differences in mortality, length of ICU admission, or length of hospital stay. Bleeding events, delirium, and renal-replacement therapy requirements were also not associated with the use of steroids.

CONCLUSION

The use of systemic steroids was more prevalent in cirrhotic patients with higher vasopressor requirements. It was not associated with decreased mortality or increased ICU- and hospital-free days, or to adverse effects.

Diamond-Forrester classification using echocardiography haemodynamic assessment in cardiac intensive care unit patients

AIMS

We sought to determine whether the Diamond-Forrester classification using non-invasive haemodynamic measurements by 2-D and Doppler echocardiography would predict hospital mortality in cardiac intensive care unit (CICU) patients.

METHODS AND RESULTS

We retrospectively analysed unique patients admitted to the CICU at Mayo Clinic Rochester from 2007 to 2018. Doppler-derived cardiac index (CI) and ratio of mitral valve E velocity to medial mitral annulus e' velocity (E/e' ratio) were used to classify patients into four profiles: Profile I (warm/dry), Profile II (warm/wet), Profile III (cold/dry), and Profile IV (cold/wet). Logistic regression was used to determine predictors of hospital mortality, and Cox proportional-hazards analysis was used to determine predictors of mortality during one year of follow-up. We included 4563 patients with a mean age of 68.3 ± 14.3 years, including 36.2% female patients. The distribution of each profile was as follows: I, 47.4%; II, 36.2%; III, 7.9%; IV, 8.5%. A total of 5.8% patients died during hospitalization, and 18.1% died by 1 year. Patients with either low CI or elevated E/e' ratio had higher in-hospital and 1 year mortality. Patients with elevated E/e' ratio (i.e. Profiles II and IV) had an increased risk of death during hospitalization and at 1 year after multivariate adjustment (adjusted hazard ratio 1.72 and 2.17 for 1 year mortality, respectively, compared with Profile I, P < 0.01).

CONCLUSIONS

Simple Doppler echocardiographic assessment can be used to identify haemodynamic profiles defined by the Diamond-Forester classification in patients admitted in CICU. These profiles predict outcomes and may be used to guide therapy in critically ill patients.

Continuous Renal Replacement Therapy Liberation and Outcomes of Critically Ill Patients With Acute Kidney Injury

OBJECTIVE

To examine the association between continuous renal replacement therapy (CRRT) liberation and clinical outcomes among patients with acute kidney injury (AKI) requiring CRRT.

METHODS

This single-center, retrospective cohort study included adult patients admitted to intensive care units with AKI and treated with CRRT from January 1, 2007, to May 4, 2018. Based on the survival and renal replacement therapy (RRT) status at 72 hours after the first CRRT liberation, we classified patients into liberated, reinstituted, and those who died. We observed patients for 90 days after CRRT initiation to compare the major adverse kidney events (MAKE₉₀).

RESULTS

Of 1135 patients with AKI, 228 (20%), 437 (39%), and 470 (41%) were assigned to liberated, reinstituted, and nonsurvival groups, respectively. The MAKE₉₀, mortality, and RRT independence rates of the cohort were 62% (707 cases), 59% (674 cases), and 40% (453 cases), respectively. Compared with reinstituted patients, the liberated group had a lower MAKE₉₀ (29% vs 39%; P=.009) and higher RRT independence rate (73% vs 65%; P=.04) on day 90, but without significant difference in 90-day mortality (26% vs 33%; P=.05). After adjustments for confounders, successful CRRT liberation was not associated with lower MAKE₉₀ (odds ratio, 0.71; 95% CI, 0.48 to 1.04; P=.08) but was independently associated with improved kidney recovery at 90-day follow-up (hazard ratio, 1.81; 95% CI, 1.41 to 2.32; P<.001).

CONCLUSION

Our study demonstrated a high occurrence of CRRT liberation failure and poor 90-day outcomes in a cohort of AKI patients treated with CRRT.

FIO2 Trajectory as a Pragmatic Intermediate Marker in Acute Hypoxic Respiratory Failure

BACKGROUND

Several markers of oxygenation are used as prognostic markers in acute hypoxemic respiratory failure. Real-world use is limited by the need for invasive measurements and unreliable availability in the electronic health record. A pragmatic, reliable, and accurate marker of acute hypoxemic respiratory failure is needed to facilitate epidemiologic studies, clinical trials, and shared decision-making with patients. [Formula: see text] is easily obtained at the bedside and from the electronic health record. The [Formula: see text] trajectory may be a valuable marker of recovery in patients with acute hypoxemic respiratory failure.

METHODS

This was a historical cohort study of adult subjects admitted to an ICU with acute hypoxemic respiratory failure secondary to community-acquired pneumonia and/or ARDS.

RESULTS

Our study included 2,670 subjects. [Formula: see text] and [Formula: see text] were consistently more available than was [Formula: see text] in the electronic health record: ([Formula: see text] vs [Formula: see text] vs [Formula: see text] : 100 vs 100 vs 72.8% on day 1, and 100 vs 99 vs 21% on day 5). A worsening [Formula: see text] trajectory was associated with reduced ventilator-free days. From days 2 to 5, every increase in [Formula: see text] by 10% from the previous day was associated with fewer ventilator-free days (on day 2: adjusted mean -1.25 [95% CI -1.45 to -1.05] d, P < .001). The [Formula: see text] trajectory also provided prognostic information. On days 3 - 5, an increase in [Formula: see text] from the previous day was associated with increased ventilator-free days (on day 3: adjusted mean 2.09 (95% CI 1.44-2.74) d; P < .001). [Formula: see text] models did not add predictive information compared with models with [Formula: see text] alone (on day 2: adjusted [Formula: see text] vs [Formula: see text] R2 0.122 vs 0.119; and on day 3: 0.153 vs 0.163).

CONCLUSIONS

[Formula: see text] and [Formula: see text] are pragmatic and readily available intermediate prognostic markers in acute hypoxic respiratory failure. The [Formula: see text] trajectory in the first 5 d of ICU admission provided important prognostic information (ventilator-free days). Although the [Formula: see text] trajectory was also associated with ventilator-free days, it did not provide more information than the [Formula: see text] trajectory alone.

Timing of ARDS Resolution (TARU): A Pragmatic Clinical Assessment of ARDS Resolution in the ICU

Purpose

Lack of a pragmatic outcome measures for acute respiratory distress syndrome (ARDS) resolution is a barrier to meaningful interventional trials of novel treatments. We evaluated a pragmatic, electronic health record (EHR)-based approach toward the clinical assessment of a novel outcome measure: ICU ARDS resolution.

Methods

We conducted a retrospective observational cohort study evaluating adult patients with moderate–severe ARDS admitted to the medical intensive care unit (ICU) at Mayo Clinic in Rochester, MN, from January 2001 through December 2010. We compared the association of ICU ARDS resolution vs non-resolution with mortality. ICU ARDS resolution was defined as improvement in P/F > 200 for at least 48 h or (if arterial blood gas unavailable) SpO₂:FiO₂ (S/F) > 235, or discharge prior to 48 h from first P/F > 200 without subsequent decline in P/F, as documented in EHR.

Results

Of the 254 patients included, ICU ARDS resolution was achieved in 179 (70%). Hospital mortality was lower in patients who met ICU ARDS resolution criteria as compared to those who did not (23% vs. 41%, p < 0.01). After adjusting for age, gender, and illness severity, the patients who met ICU ARDS resolution criteria had lower odds of hospital mortality [odds ratio 0.47, 95% CI 0.25–0.86; p = 0.015].

Conclusion

The electronic health record-based pragmatic measure of ICU ARDS resolution is associated with patient outcomes and may serve as an intermediate outcome assessing novel mechanistic treatments.

The Mayo Cardiac Intensive Care Unit Admission Risk Score is Associated with Medical Resource Utilization During Hospitalization

Objective

To determine whether the Mayo Cardiac Intensive Care Unit (CICU) Admission Risk Score (M-CARS) is associated with CICU resource utilization.

Patients and Methods

Adult patients admitted to our CICU from 2007 to 2018 were retrospectively reviewed, and M-CARS was calculated from admission data. Groups were compared using Wilcoxon test for continuous variables and χ² test for categorical variables.

Results

We included 12,428 patients with a mean age of 67±15 years (37% female patients). The mean M-CARS was 2.1±2.1, including 5890 (47.4%) patients with M-CARS less than 2 and 644 (5.2%) patients with M-CARS greater than 6. Critical care restricted therapies were frequently used, including mechanical ventilation in 28.0%, vasoactive medications in 25.5%, and dialysis in 4.8%. A higher M-CARS was associated with greater use of critical-care therapies and longer CICU and hospital length of stay. The low-risk cohort with M-CARS less than 2 was less likely to require critical-care–restricted therapies, including invasive or noninvasive mechanical ventilation (8.0% vs 46.1%), vasoactive medications (10.1% vs 38.8%), or dialysis (1.0% vs 8.2%), compared with patients with M-CARS greater than or equal to 2 (all P<.001).

Conclusion

Patients with M-CARS less than 2 infrequently require critical-care resources and have extremely low mortality, suggesting that the M-CARS could be used to facilitate the triage of critically ill cardiac patients.

Predicting 1-Year Mortality on Admission Using the Mayo Cardiac Intensive Care Unit Admission Risk Score

OBJECTIVE

To determine whether the Mayo Cardiac Intensive Care Unit (CICU) Admission Risk Score (M-CARS) accurately predicts 1-year mortality.

METHODS

We retrospectively reviewed adult CICU patients admitted from January 1, 2007, through April 30, 2018, and calculated M-CARS using admission data. We examined the association between admission M-CARS, as continuous and categorical variables, and 1-year mortality.

RESULTS

This study included 12,428 unique patients with a mean age of 67.6±15.2 years (4686 [37.7%] female). A total of 2839 patients (22.8%) died within 1 year of admission, including 1149 (9.2%) hospital deaths and 1690 (15.0%) of the 11,279 hospital survivors. The 1-year survival decreased incrementally as a function of increasing M-CARS (P<.001), and all components of M-CARS were significant predictors of 1-year mortality (P<.001). The 1-year survival among hospital survivors decreased incrementally as a function of increasing M-CARS for scores below 3 (all P<.001); however, there was no further decrease in 1-year survival for hospital survivors with M-CARS of 3 or more (P=.99). The M-CARS components associated with 1-year mortality among hospital survivors included blood urea nitrogen, red blood cell distribution width, Braden skin score, and respiratory failure (all P<.001).

CONCLUSION

M-CARS predicted 1-year mortality among CICU admissions, with a plateau effect at high M-CARS of 3 or more for hospital survivors. Significant added predictors of 1-year mortality among hospital survivors included markers of frailty and chronic illness.

Predicting successful continuous renal replacement therapy liberation in critically ill patients with acute kidney injury

PURPOSE

No standardized criteria for continuous renal replacement therapy (CRRT) liberation have been established. We sought to develop and internally validate prediction models for successful CRRT liberation in critically ill patients with acute kidney injury (AKI).

MATERIALS AND METHODS

This single-center, retrospective cohort study included adult patients admitted to intensive care units (ICUs) with AKI and treated with CRRT from January 1, 2007, to May 4, 2018, at a tertiary referral hospital. The cohort was randomly divided into derivation and validation sets. The outcomes were successful CRRT liberation, defined as renal replacement therapy (RRT)-free survival within 72 h after the liberation and hospital discharge. Multivariate logistic regression models were developed and internally validated.

RESULTS

Of 1135 AKI patients requiring CRRT, successful CRRT liberation and RRT-free survival at hospital discharge were observed in 228 (20%) and 395 (35%) individuals, respectively. The independent predictors included mean hourly urine output within 12 h before liberation, mean serum creatinine value within 24 h before liberation, cumulative fluid balance from ICU admission to liberation, CRRT duration before liberation, and the requirement of vasoactive agents within 24 h before liberation. The models demonstrated good discrimination (AUROC, 0.76 and 0.78; positive predictive value, 36% and 48%; negative predictive value, 92% and 94%; respectively) and calibration in the validation set.

CONCLUSIONS

These validated models could assist the decision-making related to the CRRT liberation in critically ill patients with AKI.

Rule-Based Cohort Definitions for Acute Respiratory Distress Syndrome: A Computable Phenotyping Strategy Based on the Berlin Definition

Supplemental Digital Content is available in the text.

OBJECTIVES:

Accurate identification of acute respiratory distress syndrome is essential for understanding its epidemiology, patterns of care, and outcomes. We aimed to design a computable phenotyping strategy to detect acute respiratory distress syndrome in electronic health records of critically ill patients.

DESIGN:

This is a retrospective cohort study. Using a near real-time copy of the electronic health record, we developed a computable phenotyping strategy to detect acute respiratory distress syndrome based on the Berlin definition.

SETTING:

Twenty multidisciplinary ICUs in Mayo Clinic Health System.

SUBJECTS:

The phenotyping strategy was applied to 196,487 consecutive admissions from year 2009 to 2019.

INTERVENTIONS:

The acute respiratory distress syndrome cohort generated by this novel strategy was compared with the acute respiratory distress syndrome cohort documented by clinicians during the same period. The sensitivity and specificity of the phenotyping strategy were calculated in randomly selected patient cohort (50 patients) using the results from manual medical record review as gold standard.

MEASUREMENTS AND MAIN RESULTS:

Among the patients who did not have acute respiratory distress syndrome documented, the computable phenotyping strategy identified 3,169 adult patients who met the Berlin definition, 676 patients (21.3%) were classified to have severe acute respiratory distress syndrome (Pao₂/Fio₂ ratio ≤ 100), 1,535 patients (48.4%) had moderate acute respiratory distress syndrome (100 < Pao₂/Fio₂ ratio ≤ 200), and 958 patients (30.2%) had mild acute respiratory distress syndrome (200 < Pao₂/Fio₂ ratio ≤ 300). The phenotyping strategy achieved a sensitivity of 94.4%, specificity of 96.9%, positive predictive value of 94.4%, and negative predictive value of 96.9% in a randomly selected patient cohort. The clinicians documented acute respiratory distress syndrome in 1,257 adult patients during the study period. The clinician documentation rate of acute respiratory distress syndrome was 28.4%. Compared with the clinicians’ documentation, the phenotyping strategy identified a cohort that had higher acuity and complexity of illness suggested by higher Sequential Organ Failure Assessment score (9 vs 7; p < 0.0001), higher Acute Physiology and Chronic Health Evaluation score (76 vs 63; p < 0.0001), higher rate of requiring invasive mechanical ventilation (99.1% vs 71.8%; p < 0.0001), higher ICU mortality (20.6% vs 16.8%; p < 0.0001), and longer ICU length of stay (5.1 vs 4.2 d; p < 0.0001).

CONCLUSIONS:

Our rule-based computable phenotyping strategy can accurately detect acute respiratory distress syndrome in critically ill patients in the setting of high clinical complexity. This strategy can be applied to enhance early recognition of acute respiratory distress syndrome and to facilitate best-care delivery and clinical research in acute respiratory distress syndrome.

Validation of a Machine Learning Model for Early Shock Detection

OBJECTIVES

The objectives of this study were to test in real time a Trauma Triage, Treatment, and Training Decision Support (4TDS) machine learning (ML) model of shock detection in a prospective silent trial, and to evaluate specificity, sensitivity, and other estimates of diagnostic performance compared to the gold standard of electronic medical records (EMRs) review.

DESIGN

We performed a single-center diagnostic performance study.

PATIENTS AND SETTING

A prospective cohort consisted of consecutive patients aged 18 years and older who were admitted from May 1 through September 30, 2020 to six Mayo Clinic intensive care units (ICUs) and five progressive care units.

MEASUREMENTS AND MAIN RESULTS

During the study time, 5,384 out of 6,630 hospital admissions were eligible. During the same period, the 4TDS shock model sent 825 alerts and 632 were eligible. Among 632 hospital admissions with alerts, 287 were screened positive and 345 were negative. Among 4,752 hospital admissions without alerts, 78 were screened positive and 4,674 were negative. The area under the receiver operating characteristics curve for the 4TDS shock model was 0.86 (95% CI 0.85-0.87%). The 4TDS shock model demonstrated a sensitivity of 78.6% (95% CI 74.1-82.7%) and a specificity of 93.1% (95% CI 92.4-93.8%). The model showed a positive predictive value of 45.4% (95% CI 42.6-48.3%) and a negative predictive value of 98.4% (95% CI 98-98.6%).

CONCLUSIONS

We successfully validated an ML model to detect circulatory shock in a prospective observational study. The model used only vital signs and showed moderate performance compared to the gold standard of clinician EMR review when applied to an ICU patient cohort.

Red blood cell transfusion threshold and mortality in cardiac intensive care unit patients

BACKGROUND

The benefit of red blood cell (RBC) transfusion in anemic critically-ill patients with cardiovascular disease is uncertain, as is the optimal threshold at which RBC transfusion should be considered. We sought to examine the association between RBC transfusion and mortality stratified by nadir Hgb level and admission diagnosis among cardiac intensive care unit (CICU) patients.

METHODS

Retrospective single-center cohort of 11,754 CICU patients admitted between 2007 and 2018. The association between RBC transfusion and hospital mortality at each nadir Hgb (<8 g/dL, 8-9.9 g/dL, ≥10 g/dL) was assessed using multivariable logistic regression adjusted for the propensity to receive RBC transfusion.

RESULTS

The study population had a mean age of 68±15 years, including 38% females; 1,134 (11.4%) received RBC transfusion. Admission diagnoses included: acute coronary syndrome , 42%; heart failure, 50%; cardiac arrest , 12%; and cardiogenic shock , 12%. Patients who received RBC transfusion had higher crude hospital mortality (19% vs. 8%, P<.001). RBC transfusion was associated with lower adjusted hospital mortality in patients with nadir Hgb <8 g/dL after propensity adjustment, including subgroups with acute coronary syndrome, cardiac arrest, or cardiogenic shock (all P <.01). RBC transfusion was not associated with lower adjusted hospital mortality in any subgroup of patients with nadir Hgb ≥8 g/dL.

CONCLUSIONS

These observational data suggest the use of a Hgb threshold <8 g/dL for RBC transfusion in most CICU patients, although we could not exclude a potential benefit of RBC transfusion at a nadir Hgb of 8 to 9.9 g/dL; we did not observe any benefit from RBC transfusion at a nadir Hgb ≥10 g/dL.

Abnormal serum chloride is associated with increased mortality among unselected cardiac intensive care unit patients

Purpose

We sought to describe the association between serum chloride levels and mortality among unselected cardiac intensive care unit (CICU) patients.

Materials and methods

We retrospectively reviewed adult patients admitted to our CICU from 2007 to 2015. The association of dyschloremia and hospital mortality was assessed in a multiple variable model including additional confounders, and the association of dyschloremia and post-discharge mortality were assessed using Cox proportional-hazards analysis.

Results

9,426 patients with a mean age of 67±15 years (37% females) were included. Admission hypochloremia was present in 1,384 (15%) patients, and hyperchloremia was present in 1,606 (17%) patients. There was a U-shaped relationship between admission chloride and unadjusted hospital mortality, with increased hospital mortality among patients with hypochloremia (unadjusted OR 3.0, 95% CI 2.5–3.6, p<0.001) or hyperchloremia (unadjusted OR 1.9, 95% CI 1.6–2.3, p<0.001). After multivariate adjustment, hypochloremia remained associated with higher hospital mortality (adjusted OR 2.1, 95% CI 1.6–2.9, p <0.001). Post-discharge mortality among hospital survivors was higher among patients with admission hypochloremia (adjusted HR 1.3, 95% CI 1.1–1.6; p<0.001).

Conclusion

Abnormal serum chloride on admission to the CICU is associated with increased short- and long-term mortality, with hypochloremia being a strong independent predictor.

Incidence and outcomes of acute kidney injury stratified by cardiogenic shock severity

BACKGROUND

Acute kidney injury (AKI) is common among patients with cardiogenic shock (CS) and it is independently associated with mortality. We sought to assess the prevalence, severity, and prognosis of AKI as a function of cardiogenic shock severity in unselected Cardiac Intensive Care Unit (CICU) patients.

METHODS

We retrospectively reviewed admissions to the Mayo Clinic between 2007 to 2015 and stratified patients by the AKI stage (based on modified Kidney Disease: Improving Global Outcomes criteria) and Society for cardiovascular angiography and interventions (SCAI) shock stage. The association with in-hospital mortality was analyzed using multivariable logistic regression.

RESULTS

We included 9,311 unique patients with a mean age of 67 years and 37% females. SCAI shock stages A, B, C, D, and E were present in 47%, 30%, 15%, 7%, and 1% of patients. The incidence of AKI of any severity was 39% in the CICU and 51% during the hospitalization. Hospital mortality occurred in 8% of all patients, and the risk increased as a function of the rising AKI and SCAI shock stage. Worsening AKI stage was associated with increased adjusted hospital mortality (adjusted OR per AKI stage 1.22, 95% CI 1.10-1.36, p < .001). Higher AKI stages were associated with increased adjusted hospital mortality in SCAI stage A/B (p < .001), but not in SCAI stage C, D, or E (all p > .05).

CONCLUSIONS

Higher AKI stages were independently associated with mortality in CICU patients after accounting for shock severity and may add incremental prognostic utility in patients with lower SCAI stages.

Right Ventricular Pulmonary Artery Coupling and Mortality in Cardiac Intensive Care Unit Patients

Background

Impaired right ventricular (RV) pulmonary artery coupling has been associated with higher mortality in patients with chronic heart disease, but few studies have examined this metric in critically ill patients. We sought to evaluate the association between RV pulmonary artery coupling, defined by the ratio of tricuspid annular peak systolic tissue Doppler velocity (TASV)/estimated RV systolic pressure (RVSP), and mortality in cardiac intensive care unit patients.

Methods and Results

Using a database of unique cardiac intensive care unit admissions from 2007 to 2018, we included patients with TASV/RVSP ratio measured within 1 day of hospitalization. Hospital mortality was analyzed using multivariable logistic regression, and 1‐year mortality was analyzed using multivariable Cox proportional‐hazards analysis. We included 4259 patients with a mean age of 69±15 years (40.1% women). Admission diagnoses included acute coronary syndrome in 56%, heart failure in 52%, respiratory failure in 24%, and cardiogenic shock in 12%. The mean TASV/RVSP ratio was 0.31±0.14, and in‐hospital mortality occurred in 7% of patients. Higher TASV/RVSP ratio was associated with lower in‐hospital mortality (adjusted unit odds ratio, 0.68 per each 0.1‐unit higher ratio; 95% CI, 0.58–0.79; P<0.001) and lower 1‐year mortality among hospital survivors (adjusted unit hazard ratio, 0.83 per each 0.1‐unit higher ratio; 95% CI, 0.77–0.90; P<0.001). Stepwise decreases in hospital and 1‐year mortality were observed in each higher TASV/RVSP quintile. The TASV/RVSP ratio remained associated with mortality after adjusting for left ventricular systolic and diastolic function.

Conclusions

A low TASV/RVSP ratio is associated with increased short‐term and long‐term mortality among cardiac intensive care unit patients, emphasizing importance of impaired RV pulmonary artery coupling as a determinant of poor prognosis. Further study is required to determine whether interventions to optimize RV pulmonary artery coupling can improve outcomes.

Trends in Therapy and Outcomes Associated With Respiratory Failure in Patients Admitted to the Cardiac Intensive Care Unit

PURPOSE

To describe the epidemiology, outcomes, and temporal trends of respiratory failure in the cardiac intensive care unit (CICU).

MATERIALS AND METHODS

Retrospective cohort analysis of 2,986 unique Mayo Clinic CICU patients from 2007 to 2018 with respiratory failure. Temporal trends were analyzed, along with hospital and 1-year mortality. Multivariable logistic regression was used to determine adjusted hospital mortality trends.

RESULTS

The prevalence of respiratory failure in the CICU increased from 15% to 38% during the study period (P < 0.001 for trend). Among patients with respiratory failure, the utilization of invasive ventilation decreased and noninvasive ventilation modalities increased over time. Hospital mortality and 1-year mortality were 24% and 54%, respectively, with variation according to the type of respiratory support (highest among patients receiving invasive ventilation alone: 35% and 46%, respectively). Hospital mortality was highest among patients with concomitant cardiac arrest and/or shock (52% for patients with both). Hospital mortality decreased in the overall population from 35% to 25% (P < 0.001 for trend), but was unchanged among patients receiving positive-pressure ventilation.

CONCLUSIONS

The prevalence of respiratory failure in CICU more than doubled during the last decade. The use of noninvasive respiratory support increased, while overall mortality declined over time. Cardiac arrest and shock accounted for the majority of deaths. Further research is needed to optimize the outcomes of high-risk CICU patients with respiratory failure.

Derivation and Validation of an Automated Search Strategy to Retrospectively Identify Acute Respiratory Distress Patients Per Berlin Definition

Purpose: Acute respiratory distress syndrome (ARDS) is common in critically ill patients and linked with serious consequences. A manual chart review for ARDS diagnosis could be laborious and time-consuming. We developed an automated search strategy to retrospectively identify ARDS patients using the Berlin definition to allow for timely and accurate ARDS detection.

Methods: The automated search strategy was created through sequential steps, with keywords applied to an institutional electronic medical records (EMRs) database. We included all adult patients admitted to the intensive care unit (ICU) at the Mayo Clinic (Rochester, MN) from January 1, 2009 to December 31, 2017. We selected 100 patients at random to be divided into two derivation cohorts and identified 50 patients at random for the validation cohort. The sensitivity and specificity of the automated search strategy were compared with a manual medical record review (gold standard) for data extraction of ARDS patients per Berlin definition.

Results: On the first derivation cohort, the automated search strategy achieved a sensitivity of 91.3%, specificity of 100%, positive predictive value (PPV) of 100%, and negative predictive value (NPV) of 93.1%. On the second derivation cohort, it reached the sensitivity of 90.9%, specificity of 100%, PPV of 100%, and NPV of 93.3%. The strategy performance in the validation cohort had a sensitivity of 94.4%, specificity of 96.9%, PPV of 94.4%, and NPV of 96.9%.

Conclusions: This automated search strategy for ARDS with the Berlin definition is reliable and accurate, and can serve as an efficient alternative to time-consuming manual data review.

Electronic Medical Record-Based Pager Notification Reduces Excess Oxygen Exposure in Mechanically Ventilated Subjects

BACKGROUND

Liberal oxygenation during mechanical ventilation is harmful in critically ill patients and in certain subsets of patients, including those with stroke, acute myocardial infarction, and cardiac arrest. Surveillance through electronic medical records improves safety of mechanical ventilation in the ICU. To date, this practice has not been used for oxygen titration ([Formula: see text]) in adults. We hypothesize that a surveillance system based on the electronic medical record to alert respiratory therapists to titrate [Formula: see text] is feasible, safe, and efficacious.

METHODS

In this pilot study, mechanically ventilated subjects were randomized to respiratory therapist-driven [Formula: see text] titration after an electronic alert versus standard of care (ie, titration based on physician order). An automated surveillance system utilizing a hyperoxemia-detection algorithm generated an electronic alert to a respiratory therapist's pager. Hyperoxemia was defined as [Formula: see text] > 0.5 and [Formula: see text] > 95% for > 30 min. No other aspects of treatment were changed. We assessed feasibility, safety, and preliminary efficacy. Primary outcome was duration of hyperoxemia during mechanical ventilation. An unsafe outcome was identified as hypoxemia ([Formula: see text] < 88%) within 1 h after titration per alert. Feasibility was assessed by a survey of respiratory therapists.

RESULTS

Of 226 randomized subjects, 31 were excluded (eg, programming errors of the electronic alerts, no consent, physician discretion). We included 195 subjects, of whom 86 were in the intervention arm. Alert accuracy was 78%, and respiratory therapists responded to 64% of the alerts. During mechanical ventilation, exposure to hyperoxemia significantly decreased in the intervention group (median 13.5 h [interquartile range 6.2-29.4] vs 18.8 h [interquartile range 9.6-37.4]). No episodes of significant hypoxemia were registered. Most respiratory therapists agreed that the alert was helpful in reducing excessive oxygen exposure.

CONCLUSIONS

Use of an electronic surveillance system to titrate [Formula: see text] was safe and feasible and showed preliminary efficacy in reducing hyperoxemia. Our study serves to justify larger randomized controlled trials for [Formula: see text] titration.

Shock in the cardiac intensive care unit: Changes in epidemiology and prognosis over time

There are few studies documenting the changing epidemiology and outcomes of shock in cardiac intensive care unit (CICU) patients. We sought to describe the changes in shock epidemiology and outcomes over time in a CICU population.

METHODS

We included 1859 unique patients admitted to the Mayo Clinic Rochester CICU from 2007 through 2018 with an admission diagnosis of shock. Temporal trends, including mortality, were assessed across 3-year periods.

RESULTS

Shock comprised 15.1% of CICU admissions during the study period, increasing from 8.8% of CICU admissions in 2007 to 21.6% in 2018 (P < .01 for trend). Mean age was 68 ± 14 years (38% females). Shock was cardiogenic in 65%, septic in 10% and mixed cardiogenic-septic in 15%. Concomitant diagnoses in patients with cardiogenic shock (CS) included acute coronary syndrome (ACS) in 17%, heart failure (HF) in 35% and both in 40%. There was no significant change in the prevalence of individual shock subtypes over time (P > .1). Among patients with CS, the prevalence of ACS decreased and the prevalence of HF increased over time (P < .01). Hospital mortality was highest among patients with mixed shock (39%; P = .05). Among patients with CS, hospital mortality was lower among those with HF compared to those without HF (31% vs. 40%, P < .01). Hospital mortality decreased over time among patients with shock (P < .01) and CS (P = .02).

CONCLUSIONS

The prevalence of shock in the CICU has increased over time, with a substantial prevalence of mixed CS. The etiology of CS has changed over the last decade with HF overtaking ACS as the most common cause of CS in the CICU.

Defining Shock and Preshock for Mortality Risk Stratification in Cardiac Intensive Care Unit Patients

BACKGROUND

Previous studies have defined preshock as isolated hypotension or isolated hypoperfusion, whereas shock has been variably defined as hypoperfusion with or without hypotension. We aimed to evaluate the mortality risk associated with hypotension and hypoperfusion at the time of admission in a cardiac intensive care unit population.

METHODS

We analyzed Mayo Clinic cardiac intensive care unit patients admitted between 2007 and 2015. Hypotension was defined as systolic blood pressure <90 mm Hg or mean arterial pressure <60 mm Hg, and hypoperfusion as admission lactate >2 mmol/L, oliguria, or rising creatinine. Associations between hypotension and hypoperfusion with hospital mortality were estimated using multivariable logistic regression.

RESULTS

Among 10 004 patients with a median age of 69 years, 43.1% had acute coronary syndrome, and 46.1% had heart failure. Isolated hypotension was present in 16.7%, isolated hypoperfusion in 15.3%, and 8.7% had both hypotension and hypoperfusion. Stepwise increases in hospital mortality were observed with hypotension and hypoperfusion compared with neither hypotension nor hypoperfusion (3.3%; all P<0.001): isolated hypotension, 9.3% (adjusted odds ratio, 1.7 [95% CI, 1.4-2.2]); isolated hypoperfusion, 17.2% (adjusted odds ratio, 2.3 [95% CI, 1.9-3.0]); both hypotension and hypoperfusion, 33.8% (adjusted odds ratio, 2.8 [95% CI, 2.1-3.6]). Adjusted hospital mortality in patients with isolated hypoperfusion was higher than in patients with isolated hypotension (P=0.02) and not significant different from patients with both hypotension and hypoperfusion (P=0.18).

CONCLUSIONS

Hypotension and hypoperfusion are both associated with increased mortality in cardiac intensive care unit patients. Hospital mortality is higher with isolated hypoperfusion or concomitant hypotension and hypoperfusion (classic shock). We contend that preshock should refer to isolated hypotension without hypoperfusion, while patients with hypoperfusion can be considered to have shock, irrespective of blood pressure.

Biomarker-Concordant Steroid Use in Critically Ill Patients with Pneumonia

Objectives

To evaluate the frequency and consequences of prescribing corticosteroids for pneumonia in a biomarker-concordant manner.

Patients and Methods

This was a single-center retrospective cohort study of adults with pneumonia admitted to the medical intensive care unit (ICU) at Mayo Clinic in Rochester, Minnesota, between January 1, 2009, and June 30, 2014. Steroid use was “biomarker concordant” if given when C-reactive protein (CRP) was ≥150 mg/L or withheld when CRP was <150 mg/L, and vice versa for biomarker discordant.

Results

Of 3481 ICU admissions with community-acquired pneumonia, 169 (4.9%) had CRPs measured within 48 hours of admission to the ICU. Steroid use in the ICU was biomarker concordant in 88 (52%) patients and biomarker discordant in 81 (48%) patients. Biomarker-concordant steroid use was associated with faster resolution of lung injury: median fraction of inspired oxygen on day 3 (0.4 [0.3, 0.5] vs 0.3 [0.21, 0.4], P=.005), day 4 (0.35 [0.3, 0.5] vs 0.28 [0.21, 0.38], P=<.001), and day 5 (0.30 [0.24, 0.45] vs 0.28 [0.21, 0.40], P=.03), and increased ICU (3.5; 95% CI, 0.5 to 6.4, P=.02), and hospital-free days (3.6; 95% CI, 0.4 to 6.8, P=.03) on multivariate analysis.

Conclusions

In critically ill patients with community-acquired pneumonia, steroid use is rarely biomarker informed and often discordant with inflammatory biomarker levels. Biomarker-concordant steroid use was associated with a faster recovery of hypoxemia and increased ICU- and hospital-free days. Future well-designed prospective studies are justified to test the potential value of biomarker-concordant steroid therapy.

Systemic Inflammatory Response Syndrome Is Associated With Increased Mortality Across the Spectrum of Shock Severity in Cardiac Intensive Care Patients

BACKGROUND

The systemic inflammatory response syndrome (SIRS) frequently occurs in patients with cardiogenic shock and may aggravate shock severity and organ failure. We sought to determine the association of SIRS with illness severity and survival across the spectrum of shock severity in cardiac intensive care unit (CICU) patients.

METHODS

We retrospectively analyzed 8995 unique patients admitted to the Mayo Clinic CICU between 2007 and 2015. Patients with ≥2/4 SIRS criteria based on admission laboratory and vital sign data were considered to have SIRS. Patients were stratified by the 2019 Society for Cardiovascular Angiography and Interventions (SCAI) shock stages using admission data. The association between SIRS and mortality was evaluated across SCAI shock stage using logistic regression and Cox proportional-hazards models for hospital and 1-year mortality, respectively.

RESULTS

The study population had a mean age of 67.5±15.2 years, including 37.2% women. SIRS was present in 33.9% of patients upon CICU admission and was more prevalent in advanced SCAI shock stages. Patients with SIRS had higher illness severity, worse shock, and more organ failure, with an increased risk of mortality during hospitalization (16.8% versus 3.8%; adjusted odds ratio, 2.1 [95% CI, 1.7-2.5]; P<0.001) and at 1 year (adjusted hazard ratio, 1.4 [95% CI, 1.3-1.6]; P<0.001). After multivariable adjustment, SIRS was associated with higher hospital and 1-year mortality among patients in SCAI shock stages A through D (all P<0.01) but not SCAI shock stage E.

CONCLUSIONS

One-third of CICU patients meet clinical criteria for SIRS at the time of admission, and these patients have higher illness severity and worse outcomes across the spectrum of SCAI shock stages. The presence of SIRS identified CICU patients at increased risk of short-term and long-term mortality. Further study is needed to determine whether systemic inflammation truly drives SIRS in this population and whether patients with SIRS respond differently to supportive therapies for shock.

Noninvasive Echocardiographic Left Ventricular Stroke Work Index Predicts Mortality in Cardiac Intensive Care Unit Patients

BACKGROUND

Reduced left ventricular stroke work index (LVSWI) has been associated with adverse outcomes in several populations of patients with chronic heart disease, but no prior studies have examined this metric in cardiac intensive care unit (CICU) patients. We sought to determine whether a low LVSWI, as measured noninvasively using transthoracic echocardiography, is associated with higher mortality in CICU patients.

METHODS

Using a database of unique Mayo Clinic CICU admissions from 2007 to 2018, we identified patients with LVSWI measured by transthoracic echocardiography within 1 day of CICU admission. Hospital mortality was analyzed using multivariable logistic regression, and 1-year mortality was analyzed using multivariable Cox proportional-hazards analysis, adjusted for left ventricular ejection fraction and known predictors of hospital mortality.

RESULTS

We included 4536 patients with a mean age of 68±14 years (36% women). Admission diagnoses (not mutually exclusive) included acute coronary syndrome in 62%, heart failure in 46%, and cardiogenic shock in 11%. The mean LVSWI was 38±14 g×min/m², and in-hospital mortality occurred in 6% of patients. LVSWI had better discrimination for hospital mortality than left ventricular ejection fraction (P<0.001 by De Long test). Higher LVSWI was associated with lower in-hospital mortality (adjusted odds ratio, 0.72 per 10 g×min/m² higher [95% CI, 0.61-0.84]; P<0.001) and lower 1-year mortality (adjusted hazard ratio, 0.812 per 1 g×min/m² higher [95% CI, 0.759-0.868]; P<0.001). Stepwise decreases in hospital and 1-year mortality were observed with higher LVSWI.

CONCLUSIONS

Low LVSWI, reflecting poor left ventricular systolic and diastolic performance, is associated with increased short-term and long-term mortality among CICU patients. This emphasizes the importance of Doppler transthoracic echocardiography as a predictor of outcomes among critically ill patients. Further study is required to determine whether early interventions to optimize LVSWI can improve outcomes in the CICU setting.

Left ventricular systolic dysfunction identification using artificial intelligence-augmented electrocardiogram in cardiac intensive care unit patients

BACKGROUND

An artificial intelligence-augmented electrocardiogram (AI-ECG) can identify left ventricular systolic dysfunction (LVSD). We examined the accuracy of AI ECG for identification of LVSD (defined as LVEF ≤40% by transthoracic echocardiogram [TTE]) in cardiac intensive care unit (CICU) patients.

METHOD

We included unique Mayo Clinic CICU patients admitted from 2007 to 2018 who underwent AI-ECG and TTE within 7 days, at least one of which was during hospitalization. Discrimination of the AI-ECG for LVSD was determined using receiver-operator characteristic curve (AUC) values.

RESULTS

We included 5680 patients with a mean age of 68 ± 15 years (37% females). Acute coronary syndrome (ACS) was present in 55%. LVSD was present in 34% of patients (mean LVEF 48 ± 16%). The AI-ECG had an AUC of 0.83 (95% confidence interval 0.82-0.84) for discrimination of LVSD. Using the optimal cut-off, the AI-ECG had 73%, specificity 78%, negative predictive value 85% and overall accuracy 76% for LVSD. AUC values were higher for patients aged <70 years (0.85 versus 0.80), males (0.84 versus 0.79), patients without ACS (0.86 versus 0.80), and patients who did not undergo revascularization (0.84 versus 0.80).

CONCLUSIONS

The AI-ECG algorithm had very good discrimination for LVSD in this critically-ill CICU cohort with a high prevalence of LVSD. Performance was better in younger male patients and those without ACS, highlighting those CICU patients in whom screening for LVSD using AI ECG may be more effective. The AI-ECG might potentially be useful for identification of LVSD in resource-limited settings when TTE is unavailable.

Analytics with artificial intelligence to advance the treatment of acute respiratory distress syndrome

Artificial intelligence (AI) has found its way into clinical studies in the era of big data. Acute respiratory distress syndrome (ARDS) or acute lung injury (ALI) is a clinical syndrome that encompasses a heterogeneous population. Management of such heterogeneous patient population is a big challenge for clinicians. With accumulating ALI datasets being publicly available, more knowledge could be discovered with sophisticated analytics. We reviewed literatures with big data analytics to understand the role of AI for improving the caring of patients with ALI/ARDS. Many studies have utilized the electronic medical records (EMR) data for the identification and prognostication of ARDS patients. As increasing number of ARDS clinical trials data is open to public, secondary analysis on these combined datasets provide a powerful way of finding solution to clinical questions with a new perspective. AI techniques such as Classification and Regression Tree (CART) and artificial neural networks (ANN) have also been successfully used in the investigation of ARDS problems. Individualized treatment of ARDS could be implemented with a support from AI as we are now able to classify ARDS into many subphenotypes by unsupervised machine learning algorithms. Interestingly, these subphenotypes show different responses to a certain intervention. However, current analytics involving ARDS have not fully incorporated information from omics such as transcriptome, proteomics, daily activities and environmental conditions. AI technology is assisting us to interpret complex data of ARDS patients and enable us to further improve the management of ARDS patients in future with individual treatment plans.

Functional outcome after critical illness in older patients: a population-based study

PURPOSE

To determine the prevalence of disability among ICU survivors one year after admission, and which factors influence functional outcome.

METHODS

We examined consecutive patients enrolled in the population-based Mayo Clinic Olmsted Study of Aging and then admitted to medical or surgical adult ICUs at Mayo Clinic, Rochester between January 1, 2006, and December 31, 2014 to determine one-year functional outcomes.

RESULTS

831cases were included. Mean age was 84 years (IQR 79-88). 569 (68.5%) patients were alive one year after ICU admission. Of them, 546 patients had functional assessment at one year and 367 (67.2%) had good functional outcome. On multivariable analysis, poor one-year functional outcome (death or disability) was more common among women, older patients, and patients with baseline cognitive impairment (mild cognitive impairment or dementia), higher Carlson scores, and longer ICU stay (all P <.01). After excluding deceased patients, these associations remained unchanged. In addition, 120 (32.3%) of 372 patients who had post-ICU cognitive evaluation experienced cognitive decline after the ICU admission.

CONCLUSIONS

On a population-based cohort of older, predominantly elderly patients, approximately two-thirds of survivors maintained or regained good functional status 1 year after ICU hospitalization. However, older age, female sex, greater comorbidities, abnormal baseline cognition, and longer ICU stay were associated with poor functional recovery and cognitive decline was common.

Age and shock severity predict mortality in cardiac intensive care unit patients with and without heart failure

AIMS

Age is an important risk factor for mortality among patients with cardiogenic shock and heart failure (HF). We sought to assess the extent to which age modified the performance of the Society for Cardiovascular Angiography and Interventions (SCAI) shock stage for in-hospital and 1 year mortality in cardiac intensive care unit (CICU) patients with and without HF.

METHODS AND RESULTS

We retrospectively reviewed unique admissions to the Mayo Clinic CICU during 2007-2015 and stratified patients by age and SCAI shock stage. The association between age and in-hospital mortality was analysed using multivariable logistic regression, and 1 year mortality was analysed using Cox proportional hazards analysis, both in the entire cohort and among patients with an admission diagnosis of HF or acute coronary syndrome (ACS). The final study population included 10 004 unique patients with a mean age of 67 ± 15 years, including 46.1% with HF and 43.1% with ACS. Older patients more frequently had HF and had more extensive co-morbidities, higher illness severity, more organ failure, and differential use of critical care therapies. The percentage of patients with SCAI shock stages A, B, C, D, and E were 46%, 30%, 16%, 7%, and 1%, respectively. Patients with HF were older, had greater severity of illness and higher SCAI shock stage, and had higher rates of death at all time points. In-hospital mortality occurred in 908 (9%) patients, including 549 (12%) patients with HF (61% of all hospital deaths). Age was independently associated with hospital mortality (adjusted odds ratio per 10 years 1.3, 95% confidence interval 1.2-1.4, P < 0.001) and 1 year mortality (adjusted hazard ratio per 10 years 1.2, 95% confidence interval 1.2-1.3, P < 0.001) in the overall cohort. The associations of age with both hospital mortality (adjusted odds ratio 1.6 vs. 1.3 per 10 years older) and 1 year mortality (adjusted hazard ratio 1.5 vs. 1.3 per 10 years older) were higher for patients with ACS compared with patients with HF. Older age was associated with higher adjusted hospital mortality and 1 year mortality in each SCAI shock stage (all P < 0.05). Additive increases in both hospital mortality and 1 year mortality were observed with increasing age and SCAI shock stage.

CONCLUSIONS

Age is an independent risk factor for mortality that modifies the relationship between the SCAI shock stage and mortality risk in CICU patients, providing robust risk stratification for in-hospital and 1 year mortality. Although patients with HF had a higher risk of dying, age was more strongly associated with mortality among patients with ACS.

Association between mean arterial pressure during the first 24 hours and hospital mortality in patients with cardiogenic shock

Background

The optimal MAP target for patients with cardiogenic shock (CS) remains unknown. We sought to determine the relationship between mean arterial pressure (MAP) and mortality in the cardiac intensive care unit (CICU) patients with CS.

Methods

Using a single-center database of CICU patients admitted between 2007 and 2015, we identified patients with an admission diagnosis of CS. MAP was measured every 15 min, and the mean of all MAP values during the first 24 h (mMAP₂₄) was recorded. Multivariable logistic regression determined the relationship between mMAP₂₄ and adjusted hospital mortality.

Results

We included 1002 patients with a mean age of 68 ± 13.7 years, including 36% females. Admission diagnoses included acute coronary syndrome in 60%, heart failure in 74%, and cardiac arrest in 38%. Vasoactive drugs were used in 72%. The mMAP₂₄ was higher (75 vs. 71 mmHg, p < 0.001) among hospital survivors (66%) compared with non-survivors (34%). Hospital mortality was inversely associated with mMAP₂₄ (adjusted OR 0.9 per 5 mmHg higher mMAP₂₄, p = 0.01), with a stepwise increase in hospital mortality at lower mMAP₂₄. Patients with mMAP₂₄ < 65 mmHg were at higher risk of hospital mortality (57% vs. 28%, adjusted OR 2.0, 95% CI 1.4–3.0, p < 0.001); no differences were observed between patients with mMAP₂₄ 65–74 vs. ≥ 75 mmHg (p > 0.1).

Conclusion

In patients with CS, we observed an inverse relationship between mMAP₂₄ and hospital mortality. The poor outcomes in patients with mMAP₂₄ < 65 mmHg provide indirect evidence supporting a MAP goal of 65 mmHg for patients with CS.