One-year mortality after recovery from critical illness: A retrospective cohort study

Lokhandwala, Sharukh; McCague, Ned; Chahin, Abdullah; Escobar, Braiam; Feng, Mengling; Ghassemi, Mohammad M.; Stone, David J.; Celi, Leo Anthony

doi:10.1371/journal.pone.0197226

Cited by 16 publications

(14 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Delirium occurs in up to 50% of critically ill adults, is associated with substantial burden to patients and their families, and is associated with serious ICU and post-ICU complications [1,2]. Mortality after critical illness remains an important concern among ICU survivors and their families [3,4]. Cohort studies evaluating the association either between delirium occurrence or days with delirium and mortality (either during or after ICU discharge) have reported inconsistent results (Additional file 1) [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20].…”

Section: Introductionmentioning

confidence: 99%

Association between incident delirium and 28- and 90-day mortality in critically ill adults: a secondary analysis

Duprey

Boogaard²,

Hoeven³

et al. 2020

Crit Care

View full text Add to dashboard Cite

Background: While delirium prevalence and duration are each associated with increased 30-day, 6-month, and 1-year mortality, the association between incident ICU delirium and mortality remains unclear. We evaluated the association between both incident ICU delirium and days spent with delirium in the 28 days after ICU admission and mortality within 28 and 90 days. Methods: Secondary cohort analysis of a randomized, double-blind, placebo-controlled trial conducted among 1495 delirium-free, critically ill adults in 14 Dutch ICUs with an expected ICU stay ≥2 days where all delirium assessments were completed. In the 28 days after ICU admission, patients were evaluated for delirium and coma 3x daily; each day was coded as a delirium day [≥1 positive Confusion Assessment Method for the ICU (CAM-ICU)], a coma day [no delirium and ≥ 1 Richmond Agitation Sedation Scale (RASS) score ≤ − 4], or neither. Four Cox-regression models were constructed for 28-day mortality and 90-day mortality; each accounted for potential confounders (i.e., age, APACHE-II score, sepsis, use of mechanical ventilation, ICU length of stay, and haloperidol dose) and: 1) delirium occurrence, 2) days spent with delirium, 3) days spent in coma, and 4) days spent with delirium and/or coma. Results: Among the 1495 patients, 28 day mortality was 17% and 90 day mortality was 21%. Neither incident delirium (28 day mortality hazard ratio [HR] = 1.02, 95%CI = 0.75-1.39; 90 day mortality HR = 1.05, 95%CI = 0.79-1.38) nor days spent with delirium (28 day mortality HR = 1.00, 95%CI = 0.95-1.05; 90 day mortality HR = 1.02, 95%CI = 0.98-1.07) were significantly associated with mortality. However, both days spent with coma (28 day mortality HR = 1.05, 95%CI = 1.02-1.08; 90 day mortality HR = 1.05, 95%CI = 1.02-1.08) and days spent with delirium or coma (28 day mortality HR = 1.03, 95%CI = 1.00-1.05; 90 day mortality HR = 1.03, 95%CI = 1.01-1.06) were significantly associated with mortality. Conclusions: This analysis suggests neither incident delirium nor days spent with delirium are associated with short-term mortality after ICU admission. Trial registration: ClinicalTrials.gov, Identifier NCT01785290 Registered

show abstract

Section: Introductionmentioning

confidence: 99%

Association between incident delirium and 28- and 90-day mortality in critically ill adults: a secondary analysis

Duprey

Boogaard²,

Hoeven³

et al. 2020

Crit Care

View full text Add to dashboard Cite

show abstract

“…However, after recovery from critical illness, cirrhotic patients are at higher risk of 1-year mortality (between 27.5% to 68%) compared to the general population. [121][122][123][124] Predictors associated with 1-year mortality after ICU discharge are not clearly reported but could be related to the number and type of OFs that patients experienced in the ICU, as well as the severity of liver disease at admission to and discharge from the ICU. 123 There are currently no reliable tools that identify populations with a very low risk of mortality after initial effective ICU care and discharge from the ICU.…”

Section: Therapiesmentioning

confidence: 99%

Approaches for patients with very high MELD scores

Artru

Samuel

2019

JHEP Reports

View full text Add to dashboard Cite

In the era of the "sickest first" policy, patients with very high model for end-stage liver disease (MELD) scores have been increasingly admitted to the intensive care unit with the expectation that they will receive a liver transplant (LT) in the absence of improvement on supportive therapies. Such patients are often admitted in a context of acute-on-chronic liver failure with extrahepatic failures. Sequential assessment of scores or classification based on organ failures within the first days after admission help to stratify the risk of mortality in this population. Although the prognosis of severely ill cirrhotic patients has recently improved, transplant-free mortality remains high. LT is still the only curative treatment in this population. Yet, the increased relative scarcity of graft resource must be considered alongside the increased risk of losing a graft in the initial postoperative period when performing LT in "too sick to transplant" patients. Variables associated with poor immediate post-LT outcomes have been identified in large studies. Despite this, the performance of scores based on these variables is still insufficient. Consideration of a patient's comorbidities and frailty is an appealing predictive approach in this population that has proven of great value in many other diseases. So far, local expertise remains the last safeguard to LT. Using this expertise, data are accumulating on favourable post-LT outcomes in very high MELD populations, particularly when LT is performed in a situation of stabilization/improvement of organ failures in selected candidates. The absence of "definitive" contraindications and the control of "dynamic" contraindications allow a "transplantation window" to be defined. This window must be identified swiftly after admission given the poor short-term survival of patients with very high MELD scores. In the absence of any prospect of LT, withdrawal of care could be discussed to ensure respect of patient life, dignity and wishes.

show abstract

“…Most of the ICU mortality prediction studies focus on developing powerful mortality prediction models [3][4][5][6][7][8][9][10][11][12][13][14][15][16] in which the higher priority is to provide an accurate label or score about the admitted patients' status. One drawback of such an objective is paying less attention to features' simplicity and interpretability, which is the case with deep learning approaches [7][8][9][10][11][12][13].…”

Section: Introductionmentioning

confidence: 99%

“…Hence, the useful information that can be provided to the medical staff is strictly the prediction output. Moreover, a considerable number of relevant studies focus on investigating the continuously recorded vital signs of ICU patients in order to predict the mortality-risk of those patients [3][4][5][6][7][8][9][10][11][12][13][14][15][16]. A frequently used database in these studies is the medical information mart for intensive care (MIMIC) in its three releases (MIMIC, MIMIC II and III) with different versions [17,18].…”

Section: Introductionmentioning

confidence: 99%

Feature Engineering for ICU Mortality Prediction Based on Hourly to Bi-Hourly Measurements

et al. 2019

View full text Add to dashboard Cite

Mortality prediction for intensive care unit (ICU) patients is a challenging problem that requires extracting discriminative and informative features. This study presents a proof of concept for exploring features that can provide clinical insight. Through a feature engineering approach, it is attempted to improve ICU mortality prediction in field conditions with low frequently measured data (i.e., hourly to bi-hourly). Features are explored by investigating the vital signs measurements of ICU patients, labelled with mortality or survival at discharge. The vital signs of interest in this study are heart and respiration rate, oxygen saturation and blood pressure. The latter comprises systolic, diastolic and mean arterial pressure. In the feature exploration process, it is aimed to extract simple and interpretable features that can provide clinical insight. For this purpose, a classifier is required that maximises the margin between the two classes (i.e., survival and mortality) with minimum tolerance to misclassification errors. Moreover, it preferably has to provide a linear decision surface in the original feature space without mapping to an unlimited dimensionality feature space. Therefore, a linear hard margin support vector machine (SVM) classifier is suggested. The extracted features are grouped in three categories: statistical, dynamic and physiological. Each category plays an important role in enhancing classification error performance. After extracting several features within the three categories, a manual feature fine-tuning is applied to consider only the most efficient features. The final classification, considering mortality as the positive class, resulted in an accuracy of 91.56 % , sensitivity of 90.59 % , precision of 86.52 % and F 1 -score of 88.50 % . The obtained results show that the proposed feature engineering approach and the extracted features are valid to be considered and further enhanced for the mortality prediction purpose. Moreover, the proposed feature engineering approach moved the modelling methodology from black-box modelling to grey-box modelling in combination with the powerful classifier of SVMs.

show abstract

One-year mortality after recovery from critical illness: A retrospective cohort study

Cited by 16 publications

References 25 publications

Association between incident delirium and 28- and 90-day mortality in critically ill adults: a secondary analysis

Association between incident delirium and 28- and 90-day mortality in critically ill adults: a secondary analysis

Approaches for patients with very high MELD scores

Feature Engineering for ICU Mortality Prediction Based on Hourly to Bi-Hourly Measurements

Contact Info

Product

Resources

About