Critically ill patients admitted to the intensive care unit (ICU) frequently suffer from circulatory shock or respiratory distress, with high morbidity and mortality up to 40%. After initial fluid resuscitation other complications associated with either treatment or disease may arise. A consequence of treatment might be fluid overload or overfilling. Multiple studies have shown the possible negative effects of - too much - fluid administration, such as venous congestion. Venous congestion entails venous fluid overload, manifested by for example an increased central venous pressure (CVP) or peripheral oedema. This venous congestion may contribute to the occurrence of short-term organ failure by causing a high ''afterload'' in the venous tracts of organs. There is no consensus on how to measure venous congestion. It is important to identify variables that reflect the development of venous congestion in order to investigate whether venous congestion is associated with short-term organ failure. Variables that indicate venous congestion may be obtained with clinical examination and biochemical analyses, supplemented by hemodynamic variables derived from critical care ultrasonography (CCUS) with information about organ perfusion, and both arterial and venous function. The development of short-term organ failure can be assessed by collecting clinical, biochemical and hemodynamic variables at multiple moments. Using repeated measurements is likely to add dynamic information about the diagnostic and prognostic value of these variables. The dynamics of variables, in any direction, over time might improve the diagnostic accuracy and prognostic value of clinical, biochemical and hemodynamic variables that can be collected at the beside of the critically ill patient. Aim and hypotheses This study aims to investigate the association between dynamic variables that reflect venous congestion and the development of short-term organ failure and mortality in the critically ill. The primary objective of this study is to identify the combination of variables at different time points that indicate venous congestion and predict patient outcome. Secondary objectives are to identify a combination of CCUS variables that precede serum creatine rises in patients who develop acute kidney injury (AKI) after an acute ICU admission {diagnostic}; to identify a combination of variables per organ system or subset of populations to predict short-term organ deterioration and 7-day mortality {prognostic}; to identify a combination of variables over 48 hours of ICU admission that predict long-term (90 day) morbidity and mortality {prognostic} and; to validate multiple prognostic risk scores developed for critically ill ICU patients.
Registry procedures: This study is the follow-up study of the Simple Intensive Care Studies I (SICS-I, NCT02912624). All eligible patients will first be included in the Simple Observational Critical Care Studies (SOCCS, NCT03553069) within 3 hours after ICU admission. The SOCCS includes all acutely admitted ICU patients by means of a standardized onetime physical examination and registration of observational standard care data to predict patient outcome. The SICS-II will screen these included patients within 24 hours of ICU admission and exclude patients with a non-traumatic neurological reason for admission. All eligible patients will undergo clinical examination and critical care ultrasonography on day one, three and five. Monitoring: Monitoring will be performed by independent researchers of the University Medical Center Groningen (UMCG). Audits are planned to take place once a year. The first audit is planned in August 2018. Recruitment: Inclusion of patients and measurements of variables will be performed by the study coordinator or a co-researcher under supervision and responsibility of the principle investigator. Due to its observational nature, informed consent is not deemed necessary for this study. However, we will obtain informed consent to also cover possible advances within this study. Source data verification At inclusion all conventional hemodynamic variables are derived by physical examination and recording data from the basic hemodynamic monitoring (Philips ImageVue monitor with tracing of heart rate, electrocardiogram (ECG), arterial oxygen saturation (SpO2), arterial pressure from arterial line pressure measurement and/or from non-invasively blood pressure monitoring. CCUS will be used to envision cardiac function, pulmonary edema, venal cave overload and renal perfusion. All variables are predefined (see data dictionary) to standardize all measurements conducted by the student researchers. A CCUS protocol was written before start of the study to train researchers in CCUS during a pilot phase. Researchers were trained by an expert cardiologist-intensivist. Measurements will later be validated by independent specialists blinded for all other measurements. General patient characteristics and laboratory measurements were recorded from patient's electronic charts and the Acute Physiology and Chronic Health Evaluation (APACHE) II and IV score, Simplified Acute Physiology Score II (SAPS) scores are extracted from our local National Intensive Care Evaluation database. Follow-up of all-cause mortality will be acquired using the municipal personal records database. Data collection: Data are collected on different moments as follows: T=0, patient admission, extracted from electronic patient file Date and time of admission and patient history are recorded. T=1, within 3 hours of inclusion, variables obtained through physical examination (SOCCS) Heart rate (HR): will be recorded from the bedside electrocardiographic monitor. In case of an irregular rhythm (i.e. atrial fibrillation) the investigators will use the mean heart rate over a minute. Presence of atrial fibrillation will be recorded. Systolic blood pressure (SBP), diastolic blood pressure (DBP) and mean arterial pressure (MAP): will be recorded from the bedside electrocardiographic monitor, measured using an arterial line Central venous pressure (CVP): will be recorded if a central venous line is situated in the internal jugular or subclavian vene. Respiratory rate: will be recorded from the bedside electrocardiographic monitor. If a patient is on mechanical ventilation, see below. Capillary refill time (CRT): will be measured after exerting firm pressure for 10 seconds, preferably on the sternum and on the central part of the knee. The original upper limit of a normal CRT was considered to be 2 seconds by Champions' Trauma score (Champion HR, Sacco WJ, Hannan DS, et al. Assessment of injury severity: the triage index. Crit Care Med 1980). CRT in a healthy population is age and temperature dependent with an upper limit for healthy older adults of 4.5 seconds (Schriger DL, Baraff L. Defining normal capillary refill: variation with age, sex, and temperature. Ann Emerg Med 1988). In 2011 Ait-Oufella et al found that an index CRT upper limit of 2.4 seconds is predictive of 14-day mortality in septic shock patients (Ait-Oufella et al. Capillary refill time exploration during septic shock. Intensive Care Med 2014). The investigators will therefore both use a cut-off value of 2.5, 4.5 seconds and a continuous measure of CRT. Central temperature (Tcentral): will be measured using the already available temperature sensor which is attached through the bladder catheter. If this is not available, rectal or aural temperature will be recorded if available. Skin temperature (Tskin): will be measured subjectively and objectively. The subjective measure will be conducted by palpating the patient's extremities. A distinction between either 'warm' or 'cold' will be made using the dorsal surface of the hands of the examiner. Patients will be considered to have 'cold' skin extremities if all examined extremities are considered cool, or if only the lower extremities are cool despite warm upper extremities. The objective measurements will be performed with a skin probe, placed on the dorsum of the foot. Mottling score: this score was described by Ait-Oufella et al in 2011 (Ait-Oufella et al. Mottling score predicts survival in septic shock. Intensive Care Med 2011). The score ranges from 0 -5, depending on the extensiveness of the mottled area. A score of 0-1 is regarded mild, 2-3 moderate and 4-5 severe. Peripheral edema: Pit depth was estimated visually and according to Brodovicz et al: 0=no clinical edema, 1=slight pitting (2 mm depth) with no visible distortion, 2=somewhat deeper pit (4 mm) with no readily detectable distortion, 3=noticeably deep pit (6 mm) with the dependent extremity full and swollen, and 4=very deep pit (8 mm) with the dependent extremity grossly distorted. (Brodovicz KG, McNaughton K, Uemura N, Meininger G, Girman CJ, Yale SH. Reliability and feasibility of methods to quantitatively assess peripheral edema. Clin Med Res. 2009;7(1-2):21-31. doi:10.3121/cmr.2009.819). Urine output (ml/kg/h): measured as part of regular care. The investigators will use both the urine output over the hour before examination and the mean urine output per hour, calculated using the six hours prior to the physical examination. If these data are unavailable, the mean urine output of the previous hour(s) will be calculated depending on the available data. In patients with pre-existing renal failure, urine output will be recorded but not used for primary analysis of kidney function. Other variables: EMV score: this can be used to obtain a quick impression of a patient's state of consciousness. EMV score will be corrected in case of sedated patients. Serum lactate, creatinine and hemoglobin: are determined as part of regular care. For study purposes the investigators will use the value closest to our examination. Other biochemical values will also be recorded. Mechanical ventilation: data on the presence and type of mechanical ventilation will be collected, as well as basic information on respiratory conditions (i.e. PEEP and respiratory rate). Note: in case of mechanical ventilation the value 'respiratory rate' will be filled in twice in the CRF. If both values are the same, it will be assumed that the patient breathes at a machine-set respiratory rate. If they differ, spontaneous breathing will be assumed. Non-invasive ventilation: data on the type of ventilation (e.g. Ventimask, nasal cannula) will be gathered, as well as FiO2 and SpO2. Inotrope-, vasopressor and sedative use: any inotrope or vasopressin requirement, type, dose and speed will be recorded. Estimations of pump function and patient outcome: an estimation will be made, either by a member of the treating team, or by the researcher. T=2, within 24 hours of admission, physical examination, CCUS and biochemical variables All variables mentioned in T=1 will be collected again. Furthermore, a CCUS will be performed. Using CCUS, the following variables will be obtained by trained researchers. Exact methods are described in the CCUS SICS-II protocol. Cardiac output: will be measured by transthoracic echocardiography. Both cardiac output and cardiac index (i.e. cardiac output corrected for body surface area) will be calculated. Tricuspid Annular Plane Systolic Excursion (TAPSE): will be measured in the apical 4 chamber view using M-Mode, as a variable that reflects right ventricular function. Mitral Annular Plane Systolic Excursion (MAPSE): will be measured in the apical 4 chamber view using M-Mode, as a variable that reflects left ventricular function. Right Ventricular Systolic Excursion (RV S'): will be measured in the apical 4 chamber view, using Tissue Velocity Imaging, as a variable that reflects right ventricular function. Strain: Color Tissue Dopper Imaging is performed on the RV free wall, septum and LV free wall to study regional and global systolic function by means of strain. Kerly B-Lines: B-lines will be investigated using pulmonary ultrasound following the BLUE protocol (Lichtenstein D.A. BLUE-protocol and FALLS-protocol: two applications of lung ultrasound in the critically ill. Chest 2015). More than 3 B-lines per view point will be considered abnormal and as a possible sign of pulmonary edema. Inferior vena cava: the investigators will measure both diameter, obtained using CCUS just below the xiphoid process, and the Inferior Vena Cava Collapsibility Index (IVCCI), allowing analysis of fluid status. Renal ultrasound: the kidney will be envisioned using the 4C abdominal probe. The kidney size will be measured in centimeters. Doppler will be used to analyze the renal resistive index (RI), venous impedance index (VII) and the intrarenal venous flow pattern (IRVF). T=3 and T=4, respectively on day 3 and day 5 after admission, include the same variables from T=2 Data management: Data will be recorded using OpenClinica and transferred for analysis. After export from OpenClinica, all data will be managed with Stata version 15 (StataCorp, College Station, TX) All study subjects will receive a study subject ID in line with the SOCCS. This study subject ID will be used in both OpenClinica and Stata. Only a researcher with 'study director' account properties in OpenClinica will be able to link study subject ID to patient number. Images will be saved anonymously and will be coded in a systematic fashion, using the study subject ID, session number, and image contents. Sample size assessment: As this is an observational and explorative study evaluating unknown associations, no exact sample size could be calculated. It is however estimated that around 900 subjects will be included based on admittance numbers and previous experience with SICS-I. Roughly 3000 patients are admitted to the ICU in the UMCG each year, of which 50% (1500) is acutely admitted. It is expected that 70% (1050) of these unplanned admissions will fulfill the inclusion criteria. However, it will probably not be possible to include all eligible patients due to logistic and practical reasons. Therefore, the aim is to include at least 300 patients each year, in total an estimated 900 patients during the entire study period. We have described overal statistical plan in the statistical analysis plan (SAP) of SICS-I. Plan for missing data: We will follow the steps proposed by Jakobsen et al and will first identify the mechanisms causing missing data: missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR) \[Jakobsen JC, Gluud C, Wetterslev J, Winkel P. When and how should multiple imputation be used for handling missing data in randomised clinical trials - a practical guide with flowcharts. BMC Med Res Methodol, 2017\]. Missing values are MCAR when there is no correlation between the missing values and other observed data: a statistically insignificant Little's test (P\>0.05) confirms that the missing values are MCAR, and a significant test confirms a MAR or MNAR pattern in our missing values. From our experiences with the SICS-I, we expect the data to be MAR. Therefore, we will conduct our primary analyses with imputation for missing data using multiple imputations (MI). A threshold of up to 50% missing data will be considered acceptable for use of MI. Robustness of conclusions will be checked by secondary sensitivity analyses including available data and imputation of worst-best and best-worst case scenarios covering also missing not at random (MNAR) scenarios. If our missing values are MCAR or missingness is only confined to the outcome variable, we will use complete case analysis for our primary analyses. Statistical analysis plan: The investigators will use the general characteristics to create a baseline table. Statistical analyses will be performed using the Stata 15 (StataCorp, College Station, TX). Data will be presented as means with standard deviation if normally distributed or as medians with ranges in case of skewed data. Categorical or dichotomous data will be presented as proportions with confidence intervals. Univariate analyses will be conducted and all variables with p\<0.25 will be included in the multivariate models. Multivariate analyses will be conducted using forward stepwise regression by adding blocks of variables. All analyses will be adjusted for age and gender; other general characteristics will not be added to the model standardly. All analyses will be tested two-sided and p-values of less than 0.05 will be considered as statistical significant. Multiplicity issues will be addressed in our detailed SAP. Furthermore, we will assess the possibilities of Machine Learning. Machine Learning (ML) is a branch of Artificial Intelligence which allows data scientists to design supervised or unsupervised algorithms to "learn" from generally large data samples by means of inference. ML was mostly used as method in gene analysis but since then possibilities have increased. It has been proven that ML may boost clinically-oriented research and the analysis of big data. Furthermore, the use of ML-based frameworks in critical care has been reported in several studies, using different types of data. This data can be obtained in different ways, with the most common being through bedside measurements, which is how most data within this research will be gathered. The primary outcome of this study is mainly explorative because there is currently no consensus about which (combination of) variables to use to diagnose venous congestion. The aim is to define venous congestion after analysis of normal values and patterns within patients of the variables that reflect venous congestion. It is expected that then an algorithm can be generated to establish a cut-off for venous congestion. Longitudinal data analyses will be used to assess the differences in variables between different moments. Mortality will be modelled using multivariable logistic regression and survival analyses. A statistician will be consulted to further review the statistical analysis plan.
Study Type
OBSERVATIONAL
Enrollment
1,010
University Medical Center Groningen
Groningen, Netherlands
The combination of variables at different time points that indicate venous congestion and predict patient outcome in the critically ill.
Venous congestion is not yet well defined. We measure multiple variables which, if combined, may indicate venous congestion. These variables will be analysed and used to compute a cut-offs for venous congestion. The primary outcome is the incidence of venous congestion, the course/trajectory of these variables throughout multiple measurements and its association with patient outcome. The variables that will be measured are obtained using critical care ultrasonography and include amongst others inferior caval vein collapsibility, kerly B-lines as sign of pulmonary edema and right ventricular function. All other variables are listed in the protocol and detailed description.
Time frame: 7 days
Short term organ failure
All short term organ failure within the first seven days such as Acute Kidney Injury (defined by the KDIGO criteria) will be recorded.
Time frame: 7 days
Short-term mortality
Short term mortality and in hospital mortality will be recorded, including mortality reason, extracted from the electronic patient file
Time frame: 7 days
Long term mortality
Long term mortality will be acquired using the municipal personal records database 90 days after inclusion.
Time frame: 90 days
Ultrasonography variables that predict AKI
To identify a combination of CCUS variables that precede serum creatine rises in patients who develop AKI after acute ICU admission {diagnostic}.
Time frame: 7 days
Predictors for short term organ failure
To identify a combination of variables per organ system or subset of populations to predict short-term organ deterioration and mortality
Time frame: 7 days
Predictors for long term mortality
To identify a combination of variables over 48 hours of ICU admission that predict long-term outcome {prognostic}.
Time frame: 90 days
Risk scores
To validate multiple prognostic risk scores developed for critically ill ICU patients within our cohort. The risk scores that will be validated for our cohort include for example the SAPS, SOFA and APACHE scores.
Time frame: 90 days
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.