Development of a daily mortality probability prediction model from Intensive Care Unit patients using a discrete-time event history analysis

Huang, Ying; Chang, Kuang Yi; Lin, Shih Pin; Chen, Kung; Chan, Kwok-Hon; Chang, Polun

doi:10.1016/j.cmpb.2013.03.018

Cited by 9 publications

(21 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Characteristics of the mortality prediction models and underlying derivation cohorts are presented in Table 1. In all, 19 mortality prediction models (44%) were developed using prospectively collected data specifically gathered for the development of the prediction model, 6,[13][14][15][16][17][18][19][20][21][22][23][24][25][26][27] whereas 24 (56%) were developed using either retrospective data [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44] or prospective data previously collected for other purposes. [45][46][47][48][49] The start of data collection for the development cohorts spanned 36 years (1979-2015), and the duration of the cohort studies varying from 2 months up to 10 years for each cohort.…”

Section: Characteristics Of the Included Mortality Prediction Modelsmentioning

confidence: 99%

“…Two mortality prediction models (4.7%) did not report the timespan during which their development cohort was assembled. 22,33 In all, 31 mortality prediction models (74%) were developed in a single country, 14,[18][19][20][21][22][23][24][25][26][27]29,31,[33][34][35][36][37][38][39][40][41][42][43][44][45]47,49 six (14%) in neighbouring countries (two or more) 6,13,28,30,32,46 and five (12%) were developed in multiple countries worldwide. [15][16][17]48 The number of patients included in the development databases ranged from 232 to 731 611 patients with a median of 4,895 (IQR 528-35 878).…”

Section: Characteristics Of the Included Mortality Prediction Modelsmentioning

confidence: 99%

“…Hospital mortality was the most frequently used primary outcome in 29 (67%) mortality prediction models. 6,[13][14][15][16][17][18][19]21,22,24,27,28,[30][31][32][33]35,36,38,[41][42][43]45,46 Other primary outcome variables were ICU mortality (7%), 23,26,34 28-day mortality (4.7%), 39,44 90-day mortality (4.7%), 48,49 3-to 28-day mortality (4.7%), 40 30-day mortality (2.3%), 47 180-day mortality (2.3%), 20 6-month mortality (2.3%), 25 15-year mortality (2.3%), 37 and 6-and 12-month mortality (2.3%). 29 Secondary outcomes were 1-month mortality after ICU admission (4.7%), 24,31 hospital mortality (4.7%), 29,34 ICU mortality (2.3%), 45 3-month mortality after ICU admission (2.3%), 31 6-month mortality after ICU admission (2.3%), 31 9-month mortality (2.3%), 47 1-year mortality (2.3%)…”

Section: Outcome Measuresmentioning

confidence: 99%

See 2 more Smart Citations

Mortality prediction models in the adult critically ill: A scoping review

Keuning

Kaufmann

Wiersema

et al. 2019

Acta Anaesthesiol Scand

View full text Add to dashboard Cite

show abstract

Section: Characteristics Of the Included Mortality Prediction Modelsmentioning

confidence: 99%

Section: Characteristics Of the Included Mortality Prediction Modelsmentioning

confidence: 99%

Section: Outcome Measuresmentioning

confidence: 99%

See 1 more Smart Citation

Mortality prediction models in the adult critically ill: A scoping review

Keuning

Kaufmann

Wiersema

et al. 2019

Acta Anaesthesiol Scand

View full text Add to dashboard Cite

show abstract

“…There are several scales to quantify severity of critical illness; most of them were validated in different populations, frequently in the context of clinical trials [1]. Although there is continuous progress in prediction model performance, their use in the individual patient should be considered with caution given its poor sensitivity [2, 3]. In order to achieve a better estimate of mortality, scales based on biomarkers related to presence of oxidative stress or systemic inflammatory response have been developed; among the biomarkers most used in the ICU are C-reactive protein, glucose, and lactate [4–7].…”

Section: Introductionmentioning

confidence: 99%

Eosinophil count at intensive care unit admission was not predictor of hospital mortality: results of a case control study

et al. 2015

View full text Add to dashboard Cite

BackgroundPredicting mortality in the intensive care unit (ICU) is one of the biggest challenges in critical care medicine. Several studies have linked the presence of eosinopenia with adverse outcomes in different populations.MethodsWe performed a case control study to determine whether the eosinophil count at ICU admission was a predictor of hospital mortality. We included data from patients 18 years or older admitted to the medical or surgical ICU in a university hospital in northern of Mexico. Medical records of 86 non-survivors (cases) and 99 discharged alive patients (controls) were randomly reviewed; clinical records of patients with an ICU stay of less than 24 h and those whose information was incomplete were excluded.ResultsMedian of eosinophil count at ICU admission was 0.013 (interquartile range (IQR) 0.00 to 0.57) K/μL. There was no significant statistical difference in eosinophils at admission between survivors and non-survivors (0.014 [IQR 0.00 to 0.36] vs. 0.010 [IQR 0.00 to 0.57] K/μL, P = 0.35). In the multivariate analysis, APACHE II score at ICU admission and discharge were the only mortality predictors. Survivors had a significantly greater increase in eosinophil count during the first 7 days of ICU stay (0.104 [IQR −0.64 to 0.41] vs. 0.005 [IQR −1.79 to 0.43] K/μL, P = 0.004).ConclusionsIn our study, eosinophil count at ICU admission was not associated with increased hospital mortality. The larger increase in number of eosinophils observed during the first week of ICU stay in surviving patients deserves to be investigated further.

show abstract

“…However, such statistical approaches (APACHE, SAPS, MODS, etc. ), have been shown by several studies to generalize sub-optimally [3,4]. [4] show that over time, fixed scores' performance tends to deteriorate (e.g.…”

Section: Introductionmentioning

confidence: 99%

ISeeU: Visually interpretable deep learning for mortality prediction inside the ICU

Caicedo-Torres

2019

Journal of Biomedical Informatics

View full text Add to dashboard Cite

To improve the performance of Intensive Care Units (ICUs), the field of biostatistics has developed scores which try to predict the likelihood of negative outcomes. These help evaluate the effectiveness of treatments and clinical practice, and also help to identify patients with unexpected outcomes. However, they have been shown by several studies to offer sub-optimal performance. Alternatively, Deep Learning offers state of the art capabilities in certain prediction tasks and research suggests deep neural networks are able to outperform traditional techniques. Nevertheless, a main impediment for the adoption of Deep Learning in healthcare is its reduced interpretability, for in this field it is crucial to gain insight on the why of predictions, to assure that models are actually learning relevant features instead of spurious correlations. To address this, we propose a deep multi-scale convolutional architecture trained on the Medical Information Mart for Intensive Care III (MIMIC-III) for mortality prediction, and the use of concepts from coalitional game theory to construct visual explanations aimed to show how important these inputs are deemed by the network. Our results show our model attains state of the art performance while remaining interpretable. Supporting code can be found at https://github.com/williamcaicedo/ISeeU.

show abstract

Development of a daily mortality probability prediction model from Intensive Care Unit patients using a discrete-time event history analysis

Cited by 9 publications

References 49 publications

Mortality prediction models in the adult critically ill: A scoping review

Mortality prediction models in the adult critically ill: A scoping review

Eosinophil count at intensive care unit admission was not predictor of hospital mortality: results of a case control study

ISeeU: Visually interpretable deep learning for mortality prediction inside the ICU

Contact Info

Product

Resources

About