Predicting time-to-intubation after critical care admission using machine learning and cured fraction information.

Intubation for mechanical ventilation (MV) is one of the most common high-risk procedures performed in Intensive Care Units (ICUs). Early prediction of intubation may have a positive impact by providing timely alerts to clinicians and consequently avoiding high-risk late intubations. In this work, we propose a new machine learning method to predict the time to intubation during the first five days of ICU admission, based on the concept of cure survival models. Our approach combines classification and survival analysis, to effectively accommodate the fraction of patients not at risk of intubation, and provide a better estimate of time to intubation, for patients at risk. We tested our approach and compared it to other predictive models on a dataset collected from a secondary care hospital (AZ Groeninge, Kortrijk, Belgium) from 2015 to 2021, consisting of 3425 ICU stays. Furthermore, we utilised SHAP for feature importance analysis, extracting key insights into the relative significance of variables such as vital signs, blood gases, and patient characteristics in predicting intubation in ICU settings. The results corroborate that our approach improves the prediction of time to intubation in critically ill patients, by using routinely collected data within the first hours of admission in the ICU. Early warning of the need for intubation may be used to help clinicians predict the risk of intubation and rank patients according to their expected time to intubation.

Predicting outcomes of acute kidney injury in critically ill patients using machine learning.

Acute Kidney Injury (AKI) is a sudden episode of kidney failure that is frequently seen in critically ill patients. AKI has been linked to chronic kidney disease (CKD) and mortality. We developed machine learning-based prediction models to predict outcomes following AKI stage 3 events in the intensive care unit. We conducted a prospective observational study that used the medical records of ICU patients diagnosed with AKI stage 3. A random forest algorithm was used to develop two models that can predict patients who will progress to CKD after three and six months of experiencing AKI stage 3. To predict mortality, two survival prediction models have been presented using random survival forests and survival XGBoost. We evaluated established CKD prediction models using AUCROC, and AUPR curves and compared them with the baseline logistic regression models. The mortality prediction models were evaluated with an external test set, and the C-indices were compared to baseline COXPH. We included 101 critically ill patients who experienced AKI stage 3. To increase the training set for the mortality prediction task, an unlabeled dataset has been added. The RF (AUPR: 0.895 and 0.848) and XGBoost (c-index: 0.8248) models have a better performance than the baseline models in predicting CKD and mortality, respectively Machine learning-based models can assist clinicians in making clinical decisions regarding critically ill patients with severe AKI who are likely to develop CKD following discharge. Additionally, we have shown better performance when unlabeled data are incorporated into the survival analysis task.

Validated risk prediction models for outcomes of acute kidney injury: a systematic review.

BACKGROUND: Acute Kidney Injury (AKI) is frequently seen in hospitalized and critically ill patients. Studies have shown that AKI is a risk factor for the development of acute kidney disease (AKD), chronic kidney disease (CKD), and mortality. METHODS: A systematic review is performed on validated risk prediction models for developing poor renal outcomes after AKI scenarios. Medline, EMBASE, Cochrane, and Web of Science were searched for articles that developed or validated a prediction model. Moreover, studies that report prediction models for recovery after AKI also have been included. This review was registered with PROSPERO (CRD42022303197). RESULT: We screened 25,812 potentially relevant abstracts. Among the 149 remaining articles in the first selection, eight met the inclusion criteria. All of the included models developed more than one prediction model with different variables. The models included between 3 and 28 independent variables and c-statistics ranged from 0.55 to 1. CONCLUSION: Few validated risk prediction models targeting the development of renal insufficiency after experiencing AKI have been developed, most of which are based on simple statistical or machine learning models. While some of these models have been externally validated, none of these models are available in a way that can be used or evaluated in a clinical setting.

Comparison between Cystatin C- and Creatinine-Based Estimated Glomerular Filtration Rate in the Follow-Up of Patients Recovering from a Stage-3 AKI in ICU.

Background: Acute kidney injury (AKI) in critically ill patients is associated with a significant increase in mortality as well as long-term renal dysfunction and chronic kidney disease (CKD). Serum creatinine (SCr), the most widely used biomarker to evaluate kidney function, does not always accurately predict the glomerular filtration rate (GFR), since it is affected by some non-GFR determinants such as muscle mass and recent meat ingestion. Researchers and clinicians have gained interest in cystatin C (CysC), another biomarker of kidney function. The study objective was to compare GFR estimation using SCr and CysC in detecting CKD over a 1-year follow-up after an AKI stage-3 event in the ICU, as well as to analyze the association between eGFR (using SCr and CysC) and mortality after the AKI event. Method: This prospective observational study used the medical records of ICU patients diagnosed with AKI stage 3. SCr and CysC were measured twice during the ICU stay and four times following diagnosis of AKI. The eGFR was calculated using the EKFC equation for SCr and FAS equation for CysC in order to check the prevalence of CKD (defined as eGFR < 60 mL/min/1.73 m2). Results: The study enrolled 101 patients, 36.6% of whom were female, with a median age of 74 years (30−92), and a median length of stay of 14.5 days in intensive care. A significant difference was observed in the estimation of GFR when comparing formulas based on SCrand CysC, resulting in large differences in the prediction of CKD. Three months after the AKI event, eGFRCysC < 25 mL/min/1.73 m2 was a predictive factor of mortality later on; however, this was not the case for eGFRSCr. Conclusion: The incidence of CKD was highly discrepant with eGFRCysC versus eGFRSCr during the follow-up period. CysC detects more CKD events compared to SCr in the follow-up phase and eGFRCysC is a predictor for mortality in follow-up but not eGFRSCr. Determining the proper marker to estimate GFR in the post-ICU period in AKI stage-3 populations needs further study to improve risk stratification.

Long-term outcome in ICU patients with acute kidney injury treated with renal replacement therapy: a prospective cohort study.

BACKGROUND: In intensive care unit (ICU) patients, acute kidney injury treated with renal replacement therapy (AKI-RRT) is associated with adverse outcomes. The aim of this study was to evaluate variables associated with long-term survival and kidney outcome and to assess the composite endpoint major adverse kidney events (MAKE; defined as death, incomplete kidney recovery, or development of end-stage renal disease treated with RRT) in a cohort of ICU patients with AKI-RRT. METHODS: We conducted a single-center, prospective observational study in a 50-bed ICU tertiary care hospital. During the study period from August 2004 through December 2012, all consecutive adult patients with AKI-RRT were included. Data were prospectively recorded during the patients' hospital stay and were retrieved from the hospital databases. Data on long-term follow-up were gathered during follow-up consultation or, in the absence of this, by consulting the general physician. RESULTS: AKI-RRT was reported in 1292 of 23,665 first ICU admissions (5.5 %). Mortality increased from 59.7 % at hospital discharge to 72.1 % at 3 years. A Cox proportional hazards model demonstrated an association of increasing age, severity of illness, and continuous RRT with long-term mortality. Among hospital survivors with reference creatinine measurements, 1-year renal recovery was complete in 48.4 % and incomplete in 32.6 %. Dialysis dependence was reported in 19.0 % and was associated with age, diabetes, chronic kidney disease (CKD), and oliguria at the time of initiation of RRT. MAKE increased from 83.1 % at hospital discharge to 93.7 % at 3 years. Multivariate regression analysis showed no association of classical determinants of outcome (preexisting CKD, timing of initiation of RRT, and RRT modality) with MAKE at 1 year. CONCLUSIONS: Our study demonstrates poor long-term survival after AKI-RRT that was determined mainly by severity of illness and RRT modality at initiation of RRT. Renal recovery is limited, especially in patients with acute-on-chronic kidney disease, making nephrological follow-up imperative. MAKE is associated mainly with variables determining mortality.

Long-term quality of life in critically ill patients with acute kidney injury treated with renal replacement therapy: a matched cohort study.

INTRODUCTION: Acute kidney injury (AKI) is a common complication in intensive care unit (ICU) patients and is associated with increased morbidity and mortality. We compared long-term outcome and quality of life (QOL) in ICU patients with AKI treated with renal replacement therapy (RRT) with matched non-AKI-RRT patients. METHODS: Over 1 year, consecutive adult ICU patients were included in a prospective cohort study. AKI-RRT patients alive at 1 year and 4 years were matched with non-AKI-RRT survivors from the same cohort in a 1:2 (1 year) and 1:1 (4 years) ratio based on gender, age, Acute Physiology and Chronic Health Evaluation II score, and admission category. QOL was assessed by the EuroQoL-5D and the Short Form-36 survey before ICU admission and at 3 months, 1 and 4 years after ICU discharge. RESULTS: Of 1953 patients, 121 (6.2%) had AKI-RRT. AKI-RRT hospital survivors (44.6%; N = 54) had a 1-year and 4-year survival rate of 87.0% (N = 47) and 64.8% (N = 35), respectively. Forty-seven 1-year AKI-RRT patients were matched with 94 1-year non-AKI-RRT patients. Of 35 4-year survivors, three refused further cooperation, three were lost to follow-up, and one had no control. Finally, 28 4-year AKI-RRT patients were matched with 28 non-AKI-RRT patients. During ICU stay, 1-year and 4-year AKI-RRT patients had more organ dysfunction compared to their respective matches (Sequential Organ Failure Assessment scores 7 versus 5, P < 0.001, and 7 versus 4, P < 0.001). Long-term QOL was, however, comparable between both groups but lower than in the general population. QOL decreased at 3 months, improved after 1 and 4 years but remained under baseline level. One and 4 years after ICU discharge, 19.1% and 28.6% of AKI-RRT survivors remained RRT-dependent, respectively, and 81.8% and 71% of them were willing to undergo ICU admission again if needed. CONCLUSION: In long-term critically ill AKI-RRT survivors, QOL was comparable to matched long-term critically ill non-AKI-RRT survivors, but lower than in the general population. The majority of AKI-RRT patients wanted to be readmitted to the ICU when needed, despite a higher severity of illness compared to matched non-AKI-RRT patients, and despite the fact that one quarter had persistent dialysis dependency.

SILS sigmoidectomy versus multiport laparoscopic sigmoidectomy for diverticulitis.

BACKGROUND AND OBJECTIVES: In this single-institution study, we aimed to compare the safety, feasibility, and outcomes of single-incision laparoscopic sigmoidectomy (SILSS) with multiport laparoscopic sigmoidectomy (MLS) for recurrent diverticulitis. METHODS: Between October 2011 and February 2013, 60 sigmoidectomies were performed by the same surgeon. Forty patients had a MLS and 20 patients had a SILSS. Outcomes were compared. RESULTS: Patient characteristics were similar. There was no difference in morbidity, mortality or readmission rates. The mean operative time was longer in the SILSS group (P=.0012). In a larger proportion of patients from the SILSS group, 2 linear staplers were needed for transection at the rectum (P=.006). The total cost of disposable items was higher in the SILSS group (P<.0001). No additional ports were placed in the SILSS group. Return to bowel function or return to oral intake was faster in the SILSS group (P=.0446 and P=.0137, respectively). Maximum pain scores on postoperative days 1 and 2 were significantly less for the SILSS group (P=.0014 and P=.047, respectively). Hospital stay was borderline statistically shorter in the SILSS group (P=.0053). SILSS was also associated with better cosmesis (P<.0011). CONCLUSION: SILSS is feasible and safe and is associated with earlier recovery of bowel function, a significant reduction in postoperative pain, and better cosmesis.

Contrast-associated AKI in the critically ill: relevant or irrelevant?

PURPOSE OF REVIEW: Iodinated contrast media are frequently administered in ICU patients. Recent studies challenge the relevance of contrast media toxicity in ICU patients and relate occurrence of acute kidney injury to baseline characteristics and severity of illness. RECENT FINDINGS: Various findings in studies with kidney biomarkers indicate the causal relationship between contrast media exposure and kidney damage. Contrast media exposure not only causes direct tubular damage and renal hypoperfusion but also initiates the formation of reactive oxygen species in its turn causing tissue damage. The route of administration determines the incidence of contrast-induced acute kidney injury with a higher incidence when contrast media are administered by intra-arterial route versus intravenous route. The impact of contrast-associated acute kidney injury on hospital length of stay, the need for renal replacement therapy and survival remains a matter of debate because of discrepancies between observational versus case-matched studies and limitations of the individual studies. SUMMARY: There are diverse pathophysiologic mechanisms explaining the causal relationship between the administration of contrast media and the development of acute kidney injury. Some studies challenge the relevance of contrast media toxicity in ICU patients. However, limitations of the available studies in ICU patients preclude firm conclusions. A precautionary approach in the administration of contrast media is justified.

Severe lactic acidosis in critically ill patients with acute kidney injury treated with renal replacement therapy.

PURPOSE: Severe lactic acidosis (SLA) is frequent in intensive care unit (ICU) patients with acute kidney injury (AKI) treated with renal replacement therapy (RRT). The aim of the study is to describe the epidemiology of SLA in this setting. MATERIALS AND METHODS: An observational single-center cohort analysis was performed on AKI patients treated with RRT. At initiation of RRT, SLA patients (serum lactate concentration>5 mmol/L and pH<7.35) were compared with non-SLA patients. RESULTS: Of the 454 patients dialyzed during the study period, 342 patients matched inclusion criteria (116 with and 226 patients without SLA). In SLA patients, lactate stabilized/decreased in 69.7% at 4 hours (P=.001) and in 81.8% during the period of 4 to 24 hours (P<.001) after initiation of RRT. Mortality during this 24-hour period was 31.0%. Intensive care unit mortality was 83.6% compared with 47.3% in non-SLA patients. Initial lactate concentration was not related to ICU mortality in SLA patients. CONCLUSIONS: Severe lactic acidosis was frequent in AKI patients treated with RRT. Severe lactic acidosis patients were more severely ill and had higher mortality compared with patients without. During the first 24 hours of RRT, a correction of lactate concentration and acidosis was observed. In SLA patients, lactate concentration at initiation of RRT was not able to discriminate between survivors and nonsurvivors.

Implementing the Kidney Disease: Improving Global Outcomes/acute kidney injury guidelines in ICU patients.

PURPOSE OF REVIEW: Acute kidney injury (AKI) is a frequent finding in critically ill patients and is associated with adverse outcomes. With the purpose of improving outcome of AKI, the Kidney Disease: Improving Global Outcomes (KDIGO) group, a group of experts in critical care nephrology, has presented a set of guidelines in 2012, based on the evidence gathered until mid 2011. This review will update these guidelines with recent evidence. RECENT FINDINGS: Early application of a set of therapeutic measures - a bundle - is advised for the prevention and therapy of AKI. Hemodynamic optimization remains the cornerstone of prevention and treatment of AKI. Fluid resuscitation should be with isotonic crystalloids. Recent evidence demonstrated a higher risk for renal replacement therapy (RRT) and mortality in hydroxyethyl starch-exposed patients. Further, blood pressure should be maintained by the use of vasopressors in vasomotor shock. Nephrotoxic drugs should be avoided or stopped when possible. Contrast-associated AKI should be prevented by prehydration with either NaCl 0.9% or a bicarbonate solution. Other therapies, including intravenous N-acetylcysteine and hemofiltration are not recommended. Optimal timing of RRT remains controversial. Fluid overload remains an important determinant for the initiation of RRT. Continuous therapies are preferred in hemodynamically unstable patients; otherwise, choice of modality does not impact on outcomes. SUMMARY: The KDIGO guidelines as presented in 2012 provide guidelines on the domain of definition of AKI, prevention and treatment, contrast-induced AKI and dialysis interventions for AKI. Especially, early application of a set of measures, the AKI bundle, may prevent AKI and improve outcome.

AKI patients have worse long-term outcomes, especially in the immediate post-ICU period.

Acute kidney injury (AKI) is associated with worse outcome in the acute phase of acute illness but also in the chronic phase. In a large Danish study in this issue of Critical Care, 1-year mortality was higher in patients with AKI than in patients without AKI. Mortality was most important during the first 50 days after admission to the intensive care unit (ICU), whereas after 2 months the survival curves of patients with AKI and those of patients without AKI were similar. The reasons for this observation are not clear, but protracted critical illness and fragility after acute critical illness probably play important roles. Because we see more and more of these patients, they should be the focus of ICU research. Consequently, ICU and post-ICU care for these patients requires focus and a more integrated approach to the specific problems of these survivors of acute critical illness.

Clinical consequences of acute kidney injury.

Acute kidney injury (AKI) can no longer be considered a surrogate marker for severity of illness. Recent epidemiologic data demonstrate the association of AKI and mortality. Even small decreases of kidney function are associated with increased mortality. Several clinical consequences of AKI may explain the association of AKI and mortality. Decreased free water clearance leading to volume overload contributes to morbidity and mortality, but also to deterioration of kidney function. Acid-base disorders and electrolyte abnormalities interfere with normal functioning of many processes in the body. Critically ill patients have an increased prevalence of infection. Infection and antimicrobial therapy can be the cause of AKI, but infection can also be a consequence of AKI. Finally, inadequate antimicrobial dosing probably plays an important role in the morbidity and mortality of AKI. These findings have led to a paradigm shift: patients die because of AKI rather than with AKI.

Serum urea concentration is probably not related to outcome in ICU patients with AKI and renal replacement therapy.

BACKGROUND: Acute kidney injury (AKI) is a common complication in patients admitted to the intensive care unit (ICU). Among other variables, serum urea concentrations are recommended for timing of initiation of renal replacement therapy (RRT). The aim of this study was to evaluate whether serum urea concentration or different serum urea concentration cutoffs as recommended in the literature were associated with in-hospital mortality at time of initiation of RRT for AKI. METHODS: This is a retrospective single- centre study during a 3-year period (2004-07), in a 44-bed tertiary care centre ICU of adult AKI patients who were treated with RRT. RESULTS: Three hundred and two patients were included: 68.9% male, median age 65 years and an APACHE II score of 21. The overall in-hospital mortality was 57.9%. Non-survivors were older (67 versus 64 years, P = 0.016) and had a higher APACHE II score (22 versus 20, P < 0.001). At time of initiation of RRT, they were more severely ill and had a lower serum urea concentration compared to survivors (130 versus 141 mg/dL, P = 0.038). Serum urea concentration, as well as the different historical serum urea concentration cut-offs had low area under the curves for the receiver operating characteristic curve for prediction of mortality. In multivariate analysis, age, and at time of initiation of RRT, potassium, SOFA score with exclusion of points for AKI and RIFLE class were associated with mortality, but serum urea concentration and the different cut-offs were not. CONCLUSIONS: This retrospective study suggests that serum urea concentration and serum urea concentration cut-offs at time of initiation of RRT have no predictive value for in-hospital mortality in ICU patients with AKI.