Skip to main content

Table 4 Hosmer and Lemeshow and goodness-of-fit test for the multivariable logistic regression model

From: To develop a regional ICU mortality prediction model during the first 24 h of ICU admission utilizing MODS and NEMS with six other independent variables from the Critical Care Information System (CCIS) Ontario, Canada

Partition for the Hosmer and Lemeshow test
  Mortality = 1 Mortality = 0
Group Range of predicted mortality probability Total Observed Expected Observed Expected
1 0–3.58 % 433 11 9.73 422 423.27
2 3.59–6.26 % 432 21 22.03 411 409.97
3 6.27–9.56 % 432 31 33.91 401 398.09
4 9.57–13.06 % 434 45 49.05 389 384.95
5 13.07–18.24 % 432 66 66.83 366 365.17
6 18.25–23.35 % 432 91 90.61 341 341.49
7 23.36–28.25 % 432 108 109.35 324 320.84
8 28.26–39.11 % 438 174 149.02 264 288.98
9 39.12–49.38 % 435 190 190.70 245 244.30
10 49.39–94.08 % 417 249 262.96 168 154.04
Hosmer and Lemeshow goodness-of-fit test
Chi-square 9.3599 p value 0.3128