Skip to main content
. 2016 Dec 1;11(12):e0167227. doi: 10.1371/journal.pone.0167227

Table 2. Factors of multiple myeloma inpatient mortality by logistic regression in Generalized Estimating Equation (GEE) model.

Adjusted OR (95%CI) P
Hemodialysis
    Without Reference
    With 3.044 (2.651–3.494) <0.001***
Gender
    Male 1.162 (1.034–1.305) 0.012*
    Female Reference
Age group (years)
    ≦55 Reference
    >55 1.511 (1.272–1.796) <0.001***
Low-income household
    Without Reference
    With 1.613 (0.827–3.146) 0.161
Catastrophic illness#
    Without Reference
    With 1.347 (1.171–1.549) <0.001***
CCI_R 0.998 (0.942–1.057) 0.944
Inpatient season
    Spring (March-May) Reference
    Summer (June-August) 0.898 (0.766–1.053) 0.187
    Autumn (September-November) 0.706 (0.601–0.828) <0.001***
    Winter (December-Februrary) 0.827 (0.703–0.974) 0.023*
Urbanization level
    High 1.443 (1.134–1.811) 0.003**
    Middle 1.726 (1.389–2.143) <0.001***
    Low Reference
Level of care
    Medical center 0.864 (0.655–1.140) 0.302
    Regional hospital 0.897 (0.686–1.172) 0.426
    Local hospital Reference
Department of hematology & oncology
    Without Reference
    With 0.884 (0.782–0.999) 0.049*
Surgical operation
    Without Reference
    With 1.050 (0.911–1.210) 0.497
Length of days 1.005 (1.005–1.006) <0.001***

Adjusted OR = Adjusted Odds Ratio, adjusted for all variables in the table. CI = confidence interval. Nagelkerke R-square of model = 0.138. Outpatient location had multicollinearity with urbanization level. Medical cost had multicollinearity with length of d.

*P<0.05

**P<0.01

***P<0.001

# Catastrophic illness: Illness other than multiple myeloma contained the the Catastrophic Illnesses Patient Database.