The hospital frailty risk score is of limited value in intensive care unit patients

The hospital frailty risk score is of limited value in intensive care unit patients

by Raphael Romano Bruno, Bernhard Wernly, Hans Flaatten, Fabian Schölzel, Malte Kelm and Christian Jung 

The identification of patients with frailty is of utmost importance, in particular during intensive care treatment of very old intensive care patients (VOPs). It is quite obvious that tools for this triage process should differ from younger patients. Frailty—not necessarily age—is associated with a negative impact on outcome, especially in critically ill patients [1]. This problem is of great importance as VOPs are one of the fastest growing subgroups in intensive care medicine. We expect an increase in the proportion of the world population older than 60 years from 12% in 2013 to 21% in 2050 [2]…

Improve short-term survival in postcardiotomy cardiogenic shock by simultaneous use of intra-aortic balloon pumping with veno-arterial extracorporeal membrane oxygenation: Beware of confounders!

Improve short-term survival in postcardiotomy cardiogenic shock by simultaneous use of intra-aortic balloon pumping with veno-arterial extracorporeal membrane oxygenation: Beware of confounders!

by Patrick M. Honore, David De Bels, Sebastien Redant and Kianoush Kashani 

We enthusiastically read the recently published retrospective study by Chen et al. [1] who demonstrated that simultaneous use of intra-aortic balloon pumping (IABP) together with veno-arterial extracorporeal membrane oxygenation (VA-ECMO) in postcardiotomy cardiogenic shock (PCS) patients improved short-term survival and reduced peripheral perfusion complications. In this study, 42 (28%) patients were on concomitant IABP and VA-ECMO [1]. While the study adds substantial value to the current knowledge, the current literature about the concomitant use of VA-ECMO and IABP remains controversial [2]. A more mechanical and pragmatic approach would be to state that VA-ECMO increases left ventricular (LV) afterload and decreases the blood flow in the coronary arteries due to retrograde blood flow, which can potentially deteriorate cardiac function while IABP could reduce these effects. Reduced LV afterload and increased blood flow in the coronary arteries by IABP theoretically promote myocardial recovery and could potentially improve survival (although improved survival was never shown) [2, 3]. In the baseline characteristics of patients before VA-ECMO implantation among the non-survivors, they were significantly more hypertension (35 vs. 15%; P < .004), secondary thoracotomies (39 vs. 19%; P < .007), cardiac arrests (34 vs. 11%; P < .001), bedside implantations 42 vs. 11%; P < .0001) and significantly less concomitant insertions of VA-ECMO and IABP (22 vs. 41%; P < .025) when compared with the study survivors [1]. All mentioned variables are well-described risk factors for increased mortality [3]. It is also reported that brain and kidney blood flow improves with concurrent initiation of IABP with ECMO [1]. Therefore, the question would be to find the mechanism by which concurrent initiation could reduce the need for continuous renal replacement therapy (CRRT) and decrease neurological complications [1]. Strategies aiming to prevent acute kidney injury (AKI) by increasing global blood flow to the kidneys have failed [4] as increasing blood flow mostly impacts the cortex while medulla remains hypoperfused. Therefore, it remains unclear why the use of IABP added to VA-ECMO in order to improve renal blood flow could significantly reduce the need for CRRT [1, 3, 4]. In order to decrease the chances of bias in the reported findings, the traditional AKI risk factors like diabetes mellitus, contrast exposure, the presence of shock and need for inotropes should be included in the comparison of these groups (VA-ECMO alone vs. VA-ECMO plus IABP) [1]. Adding IABP to VA-ECMO was not reported to increase limb ischemia [1]. This is in contradiction with a recent study by Yang et al. [5] which stated major vascular complications (MVCs) to be common and associated with higher in-hospital mortality among adult PCS patients receiving peripheral VA-ECMO support. Previously, obesity, concomitant IABP/ECMO, SOFA score at 24 h post-ECMO, and bleeding disorders were reported as independent risk factors for MVCs [5]. In conclusion and according to our interpretation, this very interesting study does not definitively show that adding IABP is improving short-term survival as many confounders could explain the observed difference in mortality.

Focus on sepsis

Focus on sepsis

Intensive Care Medicine

Authors: Morten Hylander Møller, Waleed Alhazzani, Manu Shankar-Hari

Sepsis continues to be an important clinical and research problem within critical care, as highlighted in the most recent literature.

The Surviving Sepsis Campaign bundle was updated in 2018 [1]. It was emphasised that within 1 h of presentation with sepsis, clinicians should: measure lactate, obtain blood cultures, administer broad-spectrum antimicrobials, begin fluid resuscitation with 30 ml/kg crystalloids, and apply vasopressors in case of fluid refractory shock. It was recommended that this new sepsis 1-h bundle should be used systematically in emergency departments, wards, and ICUs to reduce the global burden of sepsis [1]. While the 1-h bundle is welcomed and reasonable from a patient perspective, the quality of evidence supporting some individual elements of the bundle is low. A group of international experts representing the European Society of Intensive Care Medicine and the Society of Critical Care Medicine recently highlighted research priorities in the recent Surviving Sepsis Campaign guideline [2]. The top-six research priorities were use of personalised medicine in sepsis, fluid resuscitation, rapid diagnostic tests, empirical antibiotic combination therapy, long-term outcomes, and predictors of organ dysfunction…

Reporting of Organ Support Outcomes in Septic Shock Randomized Controlled Trials: A Methodologic Review—The Sepsis Organ Support Study

Reporting of Organ Support Outcomes in Septic Shock Randomized Controlled Trials: A Methodologic Review—The Sepsis Organ Support Study

by Bourcier, Simon; Hindlet, Patrick; Guidet, Bertrand; Dechartres, Agnès 

Objectives: Many recent randomized controlled trials in the field of septic shock failed to demonstrate a benefit on mortality. Randomized controlled trials increasingly report organ support duration and organ support-free days as primary or secondary outcomes. We conducted a methodologic systematic review to assess how organ support outcomes were defined and reported in septic shock randomized controlled trials.

Data Sources: MEDLINE via PubMed, Embase, Cochrane Central Register of Controlled Trials, and Web of Science.

Study Selection: We included randomized controlled trials published between January 2004 and March 2018 that involved septic shock adults and assessed organ support duration and/or organ support-free days for hemodynamic support, respiratory support, or renal replacement therapy. 

Data Extraction: For each randomized controlled trial, we extracted the definitions of organ support duration and organ support-free days. We particularly evaluated how nonsurvivors were accounted for. Study authors were contacted to provide any missing information regarding these definitions. Data Synthesis: We included 28 randomized controlled trials. Organ support duration and organ support-free days outcomes were reported in 17 and 15 randomized controlled trials, respectively, for hemodynamic support, 15 and 15 for respiratory support, and five and nine for renal replacement therapy. Nonsurvivors were included in the organ support duration calculation in 13 of 14 randomized controlled trials (93%) for hemodynamic support and nine of 10 (90%) for respiratory support. The organ support-free days definition for hemodynamic support, respiratory support, and renal replacement therapy was reported in six of 15 randomized controlled trials (40%), eight of 15 randomized controlled trials (53%), and six of nine randomized controlled trials (67%) reporting an organ support-free days outcome, respectively. Of these, one half assigned “0” to nonsurvivors, and the other half attributed one point per day alive free of organ support up to a predefined time point. Conclusions: This study highlights the heterogeneity and infrequency of organ support duration/organ support-free days outcome reporting in septic shock trials. When reported, the definitions of these outcome measures and methods of calculation are also infrequently reported, in particular how nonsurvivors were accounted for, which may have an important impact on interpretation.

Heparin-binding protein in sepsis: player! predictor! positioning?

Heparin-binding protein in sepsis: player! predictor! positioning?

by Patrick M. Honore, David De Bels, Leonel Barreto Gutierrez, Sebastien Redant and Herbert D. Spapen 

In a post hoc study of the multicenter FINNAKI trial, Tverring et al. recently reported that measuring heparin-binding protein (HBP) on admission in the intensive care unit (ICU) improved prediction of sepsis-induced acute kidney injury (AKI). In addition, high plasma HBP levels were associated with a significantly higher fluid balance within 24 h, more organ failure within the first week, and increased 28-day mortality [1]. These observations support HBP as a novel prominent pawn on the already well-stuffed AKI biomarker chessboard…

When less is more in the active management of elevated body temperature of ICU patients

When less is more in the active management of elevated body temperature of ICU patients

By Paul J. Young, Hallie C. Prescott

Abstract

Fever is a pathophysiological response in which the body’s normal thermoregulatory set-point is adjusted upwards leading to an increase in body temperature. In contrast, hyperthermia occurs from excessive heat production or insufficient thermoregulation (e.g., heat stroke or drug reactions). Although temperature elevation is common in Intensive Care Unit (ICU) patients, a newly elevated body temperature should prompt consideration of a diagnostic evaluation. It is always prudent to consider the possibility of infection; however, for critically patients with acute brain pathologies in particular, elevated body temperature is common, even in the absence of infection. Body temperature may be elevated due to drugs, particularly antipsychotic, serotonergic, sympathomimetic, anesthetic, and anticholinergics drugs [1]. Thyrotoxicosis and pheochromocytoma should also be considered in the differential diagnosis. Often elevated temperature is multifactorial and, in many patients, particularly after major surgery, a specific cause is not found.

Although body temperature is recorded assiduously in the ICU [2], it is often unclear when or how to intervene when a patient’s body temperature is elevated. A recent individual patient data meta-analysis reported that more active fever management did not increase survival compared with less active fever management in an all-comers population of critically ill adults [3]. Survival by treatment group was similar in a range of subgroups defined by age, illness severity, receipt of specific organ supports, and the presence versus absence of high fever at baseline. These data suggest that, in general, when it comes to active management of fever in ICU patients, although less may not be more, doing less to treat fever results in similar outcomes to doing more.

The association between nutritional adequacy and 28-day mortality in the critically ill is not modified by their baseline nutritional status and disease severity

The association between nutritional adequacy and 28-day mortality in the critically ill is not modified by their baseline nutritional status and disease severity

by Charles Chin Han Lew, Gabriel Jun Yung Wong, Ka Po Cheung, Robert J. L. Fraser, Ai Ping Chua, Mary Foong Fong Chong and Michelle Miller

Background

During the initial phase of critical illness, the association between the dose of nutrition support and mortality risk may vary among patients in the intensive care unit (ICU) because the prevalence of malnutrition varies widely (28 to 78%), and not all ICU patients are severely ill. Therefore, we hypothesized that a prognostic model that integrates nutritional status and disease severity could accurately predict mortality risk and classify critically ill patients into low- and high-risk groups. Additionally, in critically ill patients placed on exclusive nutritional support (ENS), we hypothesized that their risk categories could modify the association between dose of nutrition support and mortality risk.

Methods

A prognostic model that predicts 28-day mortality was built from a prospective cohort study of 440 patients. The association between dose of nutrition support and mortality risk was evaluated in a subgroup of 252 mechanically ventilated patients via logistic regressions, stratified by low- and high-risk groups, and days of exclusive nutritional support (ENS) [short-term (≤ 6 days) vs. longer-term (≥ 7 days)]. Only the first 6 days of ENS was evaluated for a fair comparison.

Results

The prognostic model demonstrated good discrimination [AUC 0.78 (95% CI 0.73–0.82), and a bias-corrected calibration curve suggested fair accuracy. In high-risk patients with short-term ENS (≤ 6 days), each 10% increase in goal energy and protein intake was associated with an increased adjusted odds (95% CI) of 28-day mortality [1.60 (1.19–2.15) and 1.47 (1.12–1.86), respectively]. In contrast, each 10% increase in goal protein intake during the first 6 days of ENS in high-risk patients with longer-term ENS (≥ 7 days) was associated with a lower adjusted odds of 28-day mortality [0.75 (0.57–0.99)]. Despite the opposing associations, the mean predicted mortality risks and prevalence of malnutrition between short- and longer-term ENS patients were similar.

Conclusions

Combining baseline nutritional status and disease severity in a prognostic model could accurately predict 28-day mortality. However, the association between the dose of nutrition support during the first 6 days of ENS and 28-day mortality was independent of baseline disease severity and nutritional status.

The Restrictive IV Fluid Trial in Severe Sepsis and Septic Shock (RIFTS): A Randomized Pilot Study*

The Restrictive IV Fluid Trial in Severe Sepsis and Septic Shock (RIFTS): A Randomized Pilot Study*

by Corl, Keith A.; Prodromou, Michael; Merchant, Roland C.; Gareen, Ilana; Marks, Sarah; Banerjee, Debasree; Amass, Timothy; Abbasi, Adeel; Delcompare, Cesar; Palmisciano, Amy; Aliotta, Jason; Jay, Gregory; Levy, Mitchell M.

Objectives: It is unclear if a low- or high-volume IV fluid resuscitation strategy is better for patients with severe sepsis and septic shock.

Design: Prospective randomized controlled trial.

Setting: Two adult acute care hospitals within a single academic system.

Patients: Patients with severe sepsis and septic shock admitted from the emergency department to the ICU from November 2016 to February 2018.

Interventions: Patients were randomly assigned to a restrictive IV fluid resuscitation strategy (≤ 60 mL/kg of IV fluid) or usual care for the first 72 hours of care.

Measurements and Main Results: We enrolled 109 patients, of whom 55 were assigned to the restrictive resuscitation group and 54 to the usual care group. The restrictive group received significantly less resuscitative IV fluid than the usual care group (47.1 vs 61.1 mL/kg; p = 0.01) over 72 hours. By 30 days, there were 12 deaths (21.8%) in the restrictive group and 12 deaths (22.2%) in the usual care group (odds ratio, 1.02; 95% CI, 0.41–2.53). There were no differences between groups in the rate of new organ failure, hospital or ICU length of stay, or serious adverse events.

Conclusions: This pilot study demonstrates that a restrictive resuscitation strategy can successfully reduce the amount of IV fluid administered to patients with severe sepsis and septic shock compared with usual care. Although limited by the sample size, we observed no increase in mortality, organ failure, or adverse events. These findings further support that a restrictive IV fluid strategy should be explored in a larger multicenter trial.

Sleep and Work in ICU Physicians During a Randomized Trial of Nighttime Intensivist Staffing*

Sleep and Work in ICU Physicians During a Randomized Trial of Nighttime Intensivist Staffing*

by Bakhru, Rita N.; Basner, Mathias; Kerlin, Meeta Prasad; Halpern, Scott D.; Hansen-Flaschen, John; Rosen, Ilene M.; Dinges, David F.; Schweickert, William D

Objectives: To compare sleep, work hours, and behavioral alertness in faculty and fellows during a randomized trial of nighttime in-hospital intensivist staffing compared with a standard daytime intensivist model.

Design: Prospective observational study. Setting: Medical ICU of a tertiary care academic medical center during a randomized controlled trial of in-hospital nighttime intensivist staffing. Patients: Twenty faculty and 13 fellows assigned to rotations in the medical ICU during 2012. Interventions: As part of the parent study, there was weekly randomization of staffing model, stratified by 2-week faculty rotation. During the standard staffing model, there were in-hospital residents, with a fellow and faculty member available at nighttime by phone. In the intervention, there were in-hospital residents with an in-hospital nighttime intensivist. Fellows and faculty completed diaries detailing their sleep, work, and well-being; wore actigraphs; and performed psychomotor vigilance testing daily.

Measurements and Main Results: Daily sleep time (mean hours [sd]) was increased for fellows and faculty in the intervention versus control (6.7 [0.3] vs 6.0 [0.2]; p < 0.001 and 6.7 [0.1] vs 6.4 [0.2]; p < 0.001, respectively). In-hospital work duration did not differ between the models for fellows or faculty. Total hours of work done at home was different for both fellows and faculty (0.1 [< 0.1] intervention vs 1.0 [0.1] control; p < 0.001 and 0.2 [< 0.1] intervention vs 0.6 [0.1] control; p < 0.001, respectively). Psychomotor vigilance testing did not demonstrate any differences. Measures of well-being including physical exhaustion and alertness were improved in faculty and fellows in the intervention staffing model.

Conclusions: Although no differences were measured in patient outcomes between the two staffing models, in-hospital nighttime intensivist staffing was associated with small increases in total sleep duration for faculty and fellows, reductions in total work hours for fellows only, and improvements in subjective well-being for both groups. Staffing models should consider how work duration, sleep, and well-being may impact burnout and sustainability.

Preserving the quality of life: nutrition in the ICU

Preserving the quality of life: nutrition in the ICU

by Pierre Singer 

Abstract

Critically ill patients require adequate nutritional support to meet energy requirements both during and after intensive care unit (ICU) stay to protect against severe catabolism and prevent significant deconditioning. ICU patients often suffer from chronic critical illness causing an increase in energy expenditure, leading to proteolysis and related muscle loss. Careful supplementation and modulation of caloric and protein intake can avoid under- or overfeeding, both associated with poorer outcomes. Indirect calorimetry is the preferred method for assessing resting energy expenditure and the appropriate caloric and protein intake to counter energy and muscle loss. Physical exercise may have favorable effects on muscle preservation and should be considered even early in the hospital course of a critically ill patient. After liberation from the ventilator or during non-invasive ventilation, oral intake should be carefully evaluated and, in case of severe dysphagia, should be avoided and replaced by enteral of parenteral nutrition. Upon transfer from the ICU to the ward, adequate nutrition remains essential for long-term rehabilitation success and continued emphasis on sufficient nutritional supplementation in the ward is necessary to avoid a suboptimal nutritional state.

Formal guidelines: management of acute respiratory distress syndrome

Formal guidelines: management of acute respiratory distress syndrome

by Laurent Papazian, Cécile Aubron, Laurent Brochard, Jean-Daniel Chiche, Alain Combes, Didier Dreyfuss, Jean-Marie Forel, Claude Guérin, Samir Jaber, Armand Mekontso-Dessap, Alain Mercat, Jean-Christophe Richard, Damien Roux, Antoine Vieillard-Baron and Henri Faure

Abstract

Fifteen recommendations and a therapeutic algorithm regarding the management of acute respiratory distress syndrome (ARDS) at the early phase in adults are proposed. The Grade of Recommendation Assessment, Development and Evaluation (GRADE) methodology has been followed. Four recommendations (low tidal volume, plateau pressure limitation, no oscillatory ventilation, and prone position) had a high level of proof (GRADE 1 + or 1 −); four (high positive end-expiratory pressure [PEEP] in moderate and severe ARDS, muscle relaxants, recruitment maneuvers, and venovenous extracorporeal membrane oxygenation [ECMO]) a low level of proof (GRADE 2 + or 2 −); seven (surveillance, tidal volume for non ARDS mechanically ventilated patients, tidal volume limitation in the presence of low plateau pressure, PEEP > 5 cmH2O, high PEEP in the absence of deleterious effect, pressure mode allowing spontaneous ventilation after the acute phase, and nitric oxide) corresponded to a level of proof that did not allow use of the GRADE classification and were expert opinions. Lastly, for three aspects of ARDS management (driving pressure, early spontaneous ventilation, and extracorporeal carbon dioxide removal), the experts concluded that no sound recommendation was possible given current knowledge. The recommendations and the therapeutic algorithm were approved by the experts with strong agreement.

High-density lipoprotein (HDL) particle size and concentration changes in septic shock patients

High-density lipoprotein (HDL) particle size and concentration changes in septic shock patients

by Sébastien Tanaka, Dévy Diallo, Sandrine Delbosc, Claire Genève, Nathalie Zappella, Jennyfer Yong-Sang, Jessica Patche, Anatole Harrois, Sophie Hamada, Erick Denamur, Philippe Montravers, Jacques Duranteau and Olivier Meilhac

Background

Sepsis is associated with systemic inflammation that may impact lipoprotein function. In particular, high-density lipoproteins (HDLs) that display pleiotropic protective roles may be dysfunctional in septic conditions. The aim of this study was to evaluate the HDL profile and the inflammatory context in septic shock patients admitted to our intensive care unit (ICU).

Methods

In this study, 20 septic shock patients and 20 controls (ICU patients without septic shock) were included. Plasma samples were collected on days 1, 2 and 7. Total cholesterol and lipoprotein concentrations were determined. HDL profiles were obtained using the Lipoprint® System (non-denaturing electrophoresis). Quantification of pro-inflammatory cytokines (interleukin 1b, 6 and 8), cell-free DNA and lipopolysaccharide-binding protein was also performed.

Results

HDL concentration was statistically lower in septic shock patients than in controls. At days 1 and 2, septic patients had significantly more large-sized HDL than control patients. Patients recovered a normal lipid profile at day 7.

Conclusions

Our results emphasize that HDL levels are dramatically decreased in the acute phase of septic shock and that there is a shift toward large HDL particles, which may reflect a major dysfunction of these lipoproteins. Further mechanistic studies are required to explore this shift observed during sepsis.

Prevalence of post-traumatic stress disorder symptoms in adult critical care survivors: a systematic review and meta-analysis

Prevalence of post-traumatic stress disorder symptoms in adult critical care survivors: a systematic review and meta-analysis

by Cássia Righy, Regis Goulart Rosa, Rodrigo Teixeira Amancio da Silva, Renata Kochhann, Celina Borges Migliavaca, Caroline Cabral Robinson, Stefania Pigatto Teche, Cassiano Teixeira, Fernando Augusto Bozza and Maicon Falavigna

Background: As more patients are surviving intensive care, mental health concerns in survivors have become a research priority. Among these, post-traumatic stress disorder (PTSD) can have an important impact on the quality of life of critical care survivors. However, data on its burden are conflicting. Therefore, this systematic review and meta-analysis aimed to evaluate the prevalence of PTSD symptoms in adult critical care patients after intensive care unit (ICU) discharge.

Methods: We searched MEDLINE, EMBASE, LILACS, Web of Science, PsycNET, and Scopus databases from inception to September 2018. We included observational studies assessing the prevalence of PTSD symptoms in adult critical care survivors. Two reviewers independently screened studies and extracted data. Studies were meta-analyzed using a random-effects model to estimate PTSD symptom prevalence at different time points, also estimating confidence and prediction intervals. Subgroup and meta-regression analyses were performed to explore heterogeneity. Risk of bias was assessed using the Joanna Briggs Institute tool and the GRADE approach.

Results: Of 13,267 studies retrieved, 48 were included in this review. Overall prevalence of PTSD symptoms was 19.83% (95% confidence interval [CI], 16.72–23.13; I2 = 90%, low quality of evidence). Prevalence varied widely across studies, with a wide range of expected prevalence (from 3.70 to 43.73% in 95% of settings). Point prevalence estimates were 15.93% (95% CI, 11.15–21.35; I2 = 90%; 17 studies), 16.80% (95% CI, 13.74–20.09; I2 = 66%; 13 studies), 18.96% (95% CI, 14.28–24.12; I2 = 92%; 13 studies), and 20.21% (95% CI, 13.79–27.44; I2 = 58%; 7 studies) at 3, 6, 12, and > 12 months after discharge, respectively.

Conclusion: PTSD symptoms may affect 1 in every 5 adult critical care survivors, with a high expected prevalence 12 months after discharge. ICU survivors should be screened for PTSD symptoms and cared for accordingly, given the potential negative impact of PTSD on quality of life. In addition, action should be taken to further explore the causal relationship between ICU stay and PTSD, as well as to propose early measures to prevent PTSD in this population.

In-hospital mortality associated with the misdiagnosis or unidentified site of infection at admission

In-hospital mortality associated with the misdiagnosis or unidentified site of infection at admission

by Toshikazu Abe, et al.

Background

Rapid detection, early resuscitation, and appropriate antibiotic use are crucial for sepsis care. Accurate identification of the site of infection may facilitate a timely provision of appropriate care. We aimed to investigate the relationship between misdiagnosis of the site of infection at initial examination and in-hospital mortality.

Methods

This was a secondary-multicenter prospective cohort study involving 37 emergency departments. Consecutive adult patients with infection from December 2017 to February 2018 were included. Misdiagnosis of the site of infection was defined as a discrepancy between the suspected site of infection at initial examination and that at final diagnosis, including those infections remaining unidentified during hospital admission, whereas correct diagnosis was defined as site concordance. In-hospital mortality was compared between those misdiagnosed and those correctly diagnosed.

Results

Of 974 patients included in the analysis, 11.6% were misdiagnosed. Patients diagnosed with lung, intra-abdominal, urinary, soft tissue, and CNS infection at the initial examination, 4.2%, 3.8%, 13.6%, 10.9%, and 58.3% respectively, turned out to have an infection at a different site. In-hospital mortality occurred in 15%. In both generalized estimating equation (GEE) and propensity score-matched models, misdiagnosed patients exhibited higher mortality despite adjustment for patient background, site infection, and severity. The adjusted odds ratios (misdiagnosis vs. correct diagnosis) for in-hospital mortality were 2.66 (95% CI, 1.45–4.89) in the GEE model and 3.03 (95% CI, 1.24–7.38) in the propensity score-matched model. The difference in the absolute risk in the GEE model was 0.11 (0.04–0.18).

Conclusions

Among patients with infection, misdiagnosed site of infection is associated with a > 10% increase in in-hospital mortality.