Table 4.
Performance for sequential SOFA.
| Max SOFA | AUC | Brier | H/C-statistics | Comp. | AUC | H/C-statistics | Mort. |
| Pettilä et al (2002) [17], 5 days | 0.792 | LODS | 0.827 | H | |||
| MODS | 0.795 | H | |||||
| Junger et al (2002) [18], ICU stay | 0.922 | I | |||||
| Bota et al (2002) [16], 24 hrs period | 0.898 | MODS | 0.9 | U | |||
| Ferreira et al (2001) [19], ICU stay | 0.9 | I | |||||
| Total Max SOFA | AUC | Brier | H/C-statistics | Comp | AUC | H/C-statistics | Mort. |
| Ho et al (2007) [11], ICU stay | 0.829 | 0.1 | C = 7.4 p = 0.496 | H | |||
| Zygun et al (2005) [14], ICU stay | 0.7 | U = 9.2 p = 0.33 | MODS | 0.65 | 8.07 p = 0.43 | H | |
| Pettilä et al (2002) [17], ICU stay | 0.816 | LODS | 0.839 | H | |||
| MODS | 0.817 | H | |||||
| Zygun et al (2005) [14], ICU stay | 0.69 | U = 7.30 p = 0.50 | MODS | 0.64 | 9.09 p = 0.33 | I | |
| Kajdacsy-Balla Amaral et al (2005) [7], ICU stay | 0.84 | H: p = 0.95 C: p = 0.54 | I | ||||
| Junger et al (2002) [18], ICU stay | 0.921 | I | |||||
| Moreno et al (1999) [21], ICU stay | 0.847 | I | |||||
| Janssens et al (2000) [20], ICU stay | 0.86 | H | |||||
| Delta SOFA | AUC | Brier | H/C-statistics | Comp | AUC | H/C-statistics | Mort. |
| Ho et al (2007) [11], TMS – Adm | 0.635 | 0.12 | C = 20.2 p = 0.001 | H | |||
| Zygun et al (2005) [14], TMS – Adm | 0.54 | U = 53.48 p < 0.01 | MODS | 0.55 | 31.2 p < 0.01 | H | |
| Pettilä et al (2002) [17], day 5 – Adm | 0.6 | LODS | 0.633 | H | |||
| MODS | 0.653 | H | |||||
| Zygun et al (2005) [14], TMS – Adm | 0.51 | U = 98.01 p < 0.01 | MODS | 0.52 | 70.52 p < 0.01 | I | |
| Junger et al (2002) [18], TMS – Adm | 0.828 | I | |||||
| Moreno et al (1999) [21], TMS – Adm | 0.742 | I | |||||
| Ferreira et al (2001) [19], 48 hrs – Adm | 0.69 | I | |||||
| Ferreira et al (2001) [19], 96 hrs – Adm | 0.62 | I | |||||
| Janssens et al (2000) [20], TMS – Adm | 0.62 | H | |||||
| Mean SOFA | AUC | Brier | H/C-statistics | Comp | AUC | H/C-statistics | Mort. |
| Zygun et al (2005) [14], ICU stay | 0.77 | U = 22.66 p < 0.01 | MODS | 0.74 | 46.13 p < 0.01 | H | |
| Zygun et al (2005) [14], ICU stay | 0.79 | U = 28.92 p < 0.01 | MODS | 0.75 | 42.72 p < 0.01 | I | |
| Ferreira et al (2001) [19], ICU stay | 0.88 | I | |||||
| Total SOFA | AUC | Brier | H/C-statistics | Comp | AUC | H/C-statistics | Mort. |
| Ferreira et al (2001) [19], ICU stay | 0.85 | I | |||||
| Modified SOFA | AUC | Brier | H/C-statistics | Comp | AUC | H/C-statistics | Mort. |
| Timsit et al (2002) [15], Adm | 0.729 | U = 11 p = 0.2 | LODS | 0.733 | 11.3 p = 0.19 | H | |
| Timsit et al (2002) [15], day 2 | 0.752 | U = 8.3 p = 0.4 | LODS | 0.748 | H | ||
| Timsit et al (2002) [15], day 3 | 0.773 | U = 11.3 p = 0.19 | LODS | 0.761 | H | ||
| Timsit et al (2002) [15], day 4 | 0.779 | U = 7.3 p = 0.5 | LODS | 0.76 | H | ||
| Timsit et al (2002) [15], day 5 | 0.763 | U = 14.4 p = 0.07 | LODS | 0.749 | H | ||
| Timsit et al (2002) [15], day 6 | 0.784 | U = 11 p = 0.17 | LODS | 0.79 | H | ||
| Timsit et al (2002) [15], day 7 | 0.768 | U = 6.3 p = 0.62 | LODS | 0.746 | H | ||
| Junger et al (2002) [18], Adm | 0.799 | I | |||||
Adm = admission, AUC = Area Under the Receiver Operating Characteristic Curve, Comp. = Compared with, H = Hospital, H/C = H- or C- Hosmer-Lemeshow statistics, hrs = hours, I = ICU = Intensive care unit, LODS = Logistic Organ Dysfunction System, max = maximum, MODS = Multiple Organ Dysfunction Score, Mort. = Mortality, SOFA = Sequential Organ Failure Assessment, TMS = total max SOFA (always measured over entire ICU stay), U = Unspecified (mortality type or H/C statistic).