Skip to main content

Table 1 Demographic characteristics of patients in model training and testing data sets

From: Predicting the cumulative risk of death during hospitalization by modeling weekend, weekday and diurnal mortality risks

  Training data (10,745,181 admissions) Test data (987,079 admissions)
% of training data Death rates % of test data Death rates
Age distribution     
[0–5] 10.2% 0.3% 8.3% 0.2%
[615] 4.4% 0.1% 3.6% 0.1%
[16–35] 21.9% 0.1% 18.6% 0.1%
[36–55] 25.2% 0.5% 23.3% 0.4%
[56–65] 13.5% 1.4% 14.8% 1.0%
[66–75] 12.8% 2.8% 14.8% 1.7%
[76–85] 9.9% 6.1% 13.0% 3.5%
> = 85 2.2% 19.6% 3.7% 9.5%
Charlson score of comorbidity     
Zero to mild (0) 68.6% 0.2% 65.5% 0.2%
Mild [1, 2] 20.0% 2.7% 20.2% 1.7%
Moderate [3, 4] 6.8% 5.7% 8.5% 3.5%
Severe [> = 5] 4.5% 14.7% 5.8% 10.1%
Gender     
Female 54.7% 1.5% 53.6% 1.1%
Male 45.3% 2.1% 46.4% 1.6%
Overall death rates 1.8%   1.3%  
LOS     
Mean/Std. deviation 3.21/11.42   2.75/5.134  
Median 1.00   1.00  
Range 599   171  
Interquartile range 2   1