Research Article
Predicting Outcomes for Interhospital Transferred Patients of Emergency General Surgery
Table 5
Multivariable ordinal logistic regression measuring association of clinical factors with patients’ outcomes of HLOS and disposition. The order for HLOS was 0 (short, ≤5 days), 1 (medium, >5 days to ≤12 days), and 2 (long, >12 days). The order for disposition was 1 (home), 2 (rehab or subacute facility), 3 (skilled nursing facility), and 4 (discharge to hospice or death). All independent variables selected a priori were included in the models; only variables that were statistically significant were reported.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶Higher score indicates more severe disease. Goodness-of-fit test: †Deviance test. Degrees of freedom: 1267, χ2: 1159, . ‡Deviance test; Degrees of freedom: 2091, χ2: 1117, P = 0.99. (+) correlation coefficient is associated with the lowest rank of the ordinal regression and the next one after it; (+) value is associated with lower HLOS and discharge home. (−) correlation coefficient is associated with the highest rank of the ordinal regression and the next one below it; (−) value is associated with longer HLOS and discharge to hospice or death. CI: confidence interval; ED: emergency department; HLOS: hospital length of stay; OR: odds ratio; SOFA: Sequential Organ Failure Assessment. |