Readmission to the Intensive Care Unit: Incidence, Risk Factors, Resource Use, and Outcomes. A Retrospective Cohort Study

Ponzoni, Carolina Rodrigues; Corrêa, Thiago Domingos; Filho, Roberto Rabello; Neto, Ary Serpa; Assunção, Murillo Santucci César de; Pardini, Andréia; Schettino, Guilherme

doi:10.1513/annalsats.201611-851oc

Cited by 93 publications

(110 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The developed model supports several previous studies which identified associations between increased risk of readmission and male gender, older age, admission location and Glasgow coma scale [19,[41][42][43]. Interestingly, previous work found that length of ICU stay was higher among readmitted patients [19,42,43]; however, the present study suggests that if a patient stayed in the ICU one additional day the odds of readmission were decreased by approximately 10%. This finding could be a result of considering a more comprehensive set of administrative and diagnostic variables in a patient's EMR and indicates a potential benefit of prolonging ICU stays.…”

Section: Comparison With Prior Worksupporting

confidence: 89%

“…It may be an indicator of poor or incomplete medical care as transfers between healthcare professionals have been linked to an increased rate of adverse events, higher mortality, and hospital stays [15][16][17][18]. If patients at high risk of readmission could be identified, appropriate interventions, such as careful patient evaluation before discharge, planning for proper placement of the patient after discharge, and a safe and thorough handover of patient care between healthcare providers, could be implemented and patient outcomes could be improved [19]. Readmission to the ICU also represents a major source of avoidable costs for the health care system, as up to 30% of total hospital costs and 1% of the US gross national product are directly linked to ICU expenses [20].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Benchmarking Deep Learning Architectures for Predicting Readmission to the ICU and Describing Patients-at-Risk

Barbieri

Kemp

Perez‐Concha

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Background: Deep learning algorithms achieve high classification accuracy in many applications but their integration into clinical processes remains scarce, partly due to their perceived lack of transparency. Attention layers in deep neural networks increase model interpretability by identifying which components of the input are attended to at any point in time. Objective: To evaluate the feasibility of using an attention-based neural network for predicting the risk of readmission within 30 days of discharge from the intensive care unit (ICU) based on longitudinal electronic medical record (EMR) data and to leverage the interpretability of the model to describe patients-at-risk. Methods: A "time-aware attention" model was trained using publicly available EMR data (MIMIC-III) associated with 45,298 ICU stays for 33,150 patients. The analysed EMR data included static (patient demographics) and timestamped variables (diagnoses, procedures, medications, and vital signs). Bayesian inference was used to compute the posterior distribution of network weights. The prediction accuracy of the proposed model was compared with several baseline models, including recursive neural networks and logistic regression, and evaluated based on average precision, AUROC, and F 1 -Score. Odds ratios (ORs) associated with an increased risk of readmission were computed for static variables. Diagnoses, procedures, and medications were ranked according to the associated risk of readmission. The model was also used to generate reports with predicted risk (and associated uncertainty) justified by specific diagnoses, procedures, medications, and vital signs. Results: A Bayesian ensemble of 10 time-aware attention models could be trained to predict the risk of readmission within 30 days of discharge from the ICU and led to the highest predictive accuracy (average precision: 0.282, AUROC: 0.738, F 1 -Score: 0.353). Male gender, number of recent admissions, age, admission location, insurance type, and ethnicity were all associated with risk of readmission. A longer length of stay in the ICU was found to reduce the risk of readmission (OR: 0.909, 95% credible interval: 0.902, 0.916). Groups of patients at risk included those requiring cardiovascular or ventilatory support, those with poor nutritional state, and those for whom standard medical care was not suitable, e.g. due to contraindications to surgery or medications. Conclusions:The presented deep learning model considers the full clinical history of a patient and can be used to gain insight into the patient population at increased risk of readmission. Ultimately, the development of interpretable machine learning techniques such as proposed here is necessary to allow the integration of predictive models in clinical processes.

show abstract

Section: Comparison With Prior Worksupporting

confidence: 89%

Section: Introductionmentioning

confidence: 99%

Benchmarking Deep Learning Architectures for Predicting Readmission to the ICU and Describing Patients-at-Risk

Barbieri

Kemp

Perez‐Concha

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…They confirmed that unplanned discharges from the ICU during night-time and on weekends were independent risk factors for readmission to the ICU and increased mortality (18,23,24). Others confirmed high age, severity of illness on the day of admission to the ICU and day of discharge from the ICU as well as comorbidities as independent risk factors (8,(25)(26)(27). We did not compare surviving and non-surviving patients after ICU-discharge but non-survivors with or without initial discharge from ICU.…”

Section: Discussionmentioning

confidence: 92%

Mortality of severely injured patients after ICU discharge: An underestimated problem?

Hamsen

Drotleff

Lefering

et al. 2020

Preprint

View full text Add to dashboard Cite

Background Most trauma patients admitted to the hospital alive and die later on, decease during the initial care in the emergency department or the intensive care unit (ICU). However, a number of patients pass away after having been discharged from the ICU during the initial hospital stay. On first sight these cases could be seen as "failure to rescue" of potentially salvageable patients. A low rate of such patients might be a potential indicator of quality for trauma care on ICUs and surgical wards.Methods Retrospective analysis of the TraumaRegister DGU® with data from 2015-2017. Patients that died during the initial ICU stay were compared to those who were discharged from the initial ICU stay for at least 24 hours but died later on.Results A total of 82,313 trauma patients were included in the TraumaRegister DGU®. In total, 6,576 patients (8.0%) died during their hospital stay. Out of those, 5,481 were admitted to the ICU alive and 972 patients (17.7%) were discharged from ICU and died later on. Those were older (mean age: 77 vs. 68 years), less severely injured (mean ISS: 23.1 vs. 30.0 points) and had a longer mean ICU length of stay (10 vs. 6 days). A limitation of life-sustaining therapy due to a documented living will was present in 46.1% of all patients who died during their initial ICU stay and in 59.9% of patients who died after discharge from their initial ICU stay.Conclusions 17.7% of all non-surviving severely injured trauma patients died within the hospital after discharge from their initial ICU treatment. Their death can partially be explained by a limitation of therapy due to a living will. In conclusion, the rate of such late deaths may represent patients that died of potentially avoidable or treatable complications.

show abstract

“…The present study is a post hoc analysis of a retrospective single center cohort study that investigated the impact of ICU readmission on resource use and clinical outcomes [13]. The original study and this post hoc analysis were approved by the Local Ethics Committee at Hospital Israelita Albert Einstein with waiver of informed consent (CAAE:54065716.3.0000.0071).…”

Section: Methodsmentioning

confidence: 99%

“…Multidisciplinary clinical rounds involving ICU physicians, nurses, respiratory therapists, nutritionists, psychologist and clinical pharmacists are performed daily. ICU admissions are made by on-duty intensivists, whereas discharge is a consensus decision-making process involving on-duty intensivists and the primary physician, i.e., the physician who will accept the patient outside the ICU [13].…”

Section: Methodsmentioning

confidence: 99%

Impact of intensive care unit admission during handover on mortality: a propensity matched cohort study

Midega

Filho

Nassar

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

IntroductionHandover is a process of transferring information, responsibility and authority for providing care of critically ill patients from a departing intensivist to an oncoming intensivist. The effect of i admission during a medical handover on clinical outcomes is unknown.ObjectivesOur purpose was to evaluate the impact of ICU admission during a medical handover on clinical outcomes.MethodsPost hoc analysis of a cohort study addressing the effect of ICU admissions during the handover on outcomes. This retrospective, single center, propensity matched cohort study was conducted in a 41-bed open general ICU located in a private tertiary care hospital in São Paulo, Brazil. Based on time of ICU admission, patients were categorized into two cohorts: handover group (ICU admission between 6:30 am to 7:30 or 6:30 pm to 7:30 pm) or control group (admission between 7:31 am to 6:29 pm or 7:31 pm to 6:29 am). Patients in the handover group were propensity matched to patients in the control group at 1:2 ratio. Our primary outcome was hospital mortality.ResultsBetween June 1, 2013 and May 31, 2015, 6,650 adult patients were admitted to the ICU. Following exclusion of ineligible participants, 5,779 patients [389 (6.7%) in handover group and 5390 (93.3%) in control group] were eligible for propensity score matching, of whom 1,166 were successfully matched [389 (33.4%) handover group and 777 (66.6%) in control group]. Before matching, hospital mortality was 14.1% (55/389 patients) in handover group compared to 11.7% (628/5,390) in control group (p=0.142). After propensity-score matching, ICU admission during handover was not associated with increased risk of ICU (OR, 1.40; 95% CI, 0.92 to 2.11; p=0.11) and hospital (OR, 1.23; 95%CI, 0.85 to 1.75; p=0.26) mortality. ICU and hospital length of stay did not differ between the groups.ConclusionIn this propensity-matched single center cohort study, ICU admission during medical handover did not affect clinical outcomes.

show abstract

Readmission to the Intensive Care Unit: Incidence, Risk Factors, Resource Use, and Outcomes. A Retrospective Cohort Study

Abstract: Readmissions to the ICU were frequent and strongly related to poor outcomes. The degree to which ICU readmissions are preventable as well as the main causes of preventable ICU readmissions need to be further determined.

Cited by 93 publications

References 32 publications

Benchmarking Deep Learning Architectures for Predicting Readmission to the ICU and Describing Patients-at-Risk

Benchmarking Deep Learning Architectures for Predicting Readmission to the ICU and Describing Patients-at-Risk

Mortality of severely injured patients after ICU discharge: An underestimated problem?

Impact of intensive care unit admission during handover on mortality: a propensity matched cohort study

Contact Info

Product

Resources

About