Skip to main content

Table 2 Cox proportional hazard and frailty models, post discharge survival

From: Long-term survival and healthcare utilization outcomes attributable to sepsis and pneumonia

Variable Cox proportional hazard model no frailty Weibull model gamma frailty
  Hazard ratio 95 % CI P Value Hazard ratio CI P Value
Infection (reference = No Infection)       
Sepsis 1.36 1.20-1.54 <0.01 1.39 1.22-1.59 <0.01
Sepsis x Year 1 1.21 1.04-1.42 0.02 1.17 1.00-1.37 0.05
Pneumonia 1.55 1.42-1.70 <0.01 1.58 1.42-1.76 <0.01
Pneumonia x Year 1 1.15 1.02-1.30 0.02 1.11 0.98-1.25 0.11
Sepsis & Pneumonia 0.66 0.55-0.80 <0.01 0.66 0.55-0.80 <0.01
Gender (reference = Male)       
Female 0.89 0.85-0.94 <0.01 0.89 0.85-0.94 <0.01
Race (reference = White)       
Black 1.04 0.96-1.13 0.33 1.04 0.96-1.13 0.34
Other 0.96 0.84-1.09 0.49 0.95 0.84-1.08 0.46
Age Group (reference < 50)       
50 to 64 1.76 1.49-2.08 <0.01 1.75 1.48-2.07 <0.01
65 to 74 2.07 1.77-2.41 <0.01 2.06 1.76-2.41 <0.01
75 to 84 3.12 2.67-3.64 <0.01 3.10 2.64-3.64 <0.01
85 to 94 5.25 4.47-6.16 <0.01 5.21 4.36-6.21 <0.01
95 and above 8.38 6.68-10.52 <0.01 8.26 6.47-10.55 <0.01
Insurance (reference = non Medicaid)       
Medicaid 1.22 1.14-1.30 <0.01 1.22 1.14-1.30 <0.01
Number of subjects 17537    17537   
Number of failures 9694    9694   
Log pseudolikelihood −89334.21    −26182.6   
  1. Tests of the Cox proportional hazards assumption showed that the effects of sepsis and pneumonia were not proportional over time. Thus, interactions between sepsis and pneumonia and analysis time (distinguishing the first year after discharge from the rest of the study period) were incorporated into the model. These analyses controlled for healthcare-associated infections and full results are shown in Additional file 2: Table S2.